File: tf-help

package info (click to toggle)
tf5 5.0beta8-7
links: PTS, VCS
area: main
in suites: buster
size: 3,800 kB
sloc: ansic: 25,492; perl: 103; makefile: 82; sh: 79
file content (10788 lines) | stat: -rw-r--r-- 482,218 bytes
parent folder | download | duplicates (7)
&/addworld
&addworld()

addworld()

  [1mFunction[22;0m usage: 

  [1mADDWORLD[22;0m(<[4mname[24m>, <[4mtype[24m>, [<[4mhost[24m>, <[4mport[24m> [, <[4mchar[24m>, <[4mpass[24m> [, <[4mfile[24m> [, 
  <[4mflags[24m> [, <[4msrchost[24m>]]]]])

  Command usage: 

  [1m/ADDWORLD[22;0m [-pxe] [-T<[4mtype[24m>] [-s<[4msrchost[24m>] <[4mname[24m> [<[4mchar[24m> <[4mpass[24m>] <[4mhost[24m> 
  <[4mport[24m> [<[4mfile[24m>]
  [1m/ADDWORLD[22;0m [-T<[4mtype[24m>] [-s<[4msrchost[24m>] <[4mname[24m>
  [1m/ADDWORLD[22;0m [-T<[4mtype[24m>] DEFAULT [<[4mchar[24m> <[4mpass[24m> [<[4mfile[24m>]]
  ____________________________________________________________________________

  Defines a new [1mworld[22;0m or redefines an existing [1mworld[22;0m with the name <[4mname[24m>.  
  <[4mName[24m> may not contain spaces; addtionally, when defining a new world, 
  <[4mname[24m> may not begin with "(".  

  <[4mHost[24m> is a server's internet hostname, IPv4 address, or (if your platform 
  supports it) IPv6 address.  <[4mPort[24m> is the number or name of a TCP port on 
  the host.  If <[4mhost[24m> and <[4mport[24m> are blank, then "connecting" to the world 
  will only create a tf window for the world, it will not open an actual 
  network connection; this is called a "connectionless" socket.  

  There may be a special [1mworld[22;0m named "default" which does not have a <[4mhost[24m> or 
  <[4mport[24m>.  If a normal [1mworld[22;0m is defined without a <[4mcharacter[24m>, <[4mpass[24m>, <[4mtype[24m>, 
  or <[4mmfile[24m>, then that [1mworld[22;0m will use the corresponding field of the 
  "default" [1mworld[22;0m if there is one.  If the "default" [1mworld[22;0m is redefined, 
  [1mworlds[22;0m with omitted fields will use the new default values.  

  In function form, <[4mflags[24m> is a string of 0 or more letters that modify the 
  behavior of the function.  For compatability with older versions of TF, an 
  "f" or "0" in <[4mflags[24m> has the same effect as "p", and an "n" or "1" in 
  <[4mflags[24m> has no effect.  

  [1mOptions[22;0m: 
  command: -p 
  function: <[4mflags[24m> contains "p" 
          [1m%{proxy_host}[22;0m will be ignored, and all connections to the world will 
          be direct.  By default, worlds use [1m%{proxy_host}[22;0m if it is set.  
  command: -x 
  function: <[4mflags[24m> contains "x" 
          TF will use the SSL protocol to make connections to this world.  
  command: -e 
  function: <[4mflags[24m> contains "e" 
          all text sent to the world will be echoed right back as if it were 
          received from the world (in addition to being sent to the server).  
          This is most useful with connectionless sockets.  
  command: -s<[4msrchost[24m> 
  function: <[4msrchost[24m> 
          defines the host name or IP address to use for the local (tf) side 
          of the connection.  This is useful if the host has multiple network 
          interfaces and you need to override the default choice of the OS.  
  command: -T<[4mtype[24m> 
  function: <[4mtype[24m> 
          The optional <[4mtype[24m> is used in hooks and triggers, and for automatic 
          [1mlogin[22;0m and flag setting.  (See below.) 

  The library pre-defines WORLD and LOGIN hooks for types that match these 
  [1mglob patterns[22;0m: 

  (none)  TinyMud [1mlogin[22;0m format ("connect <[4mchar[24m> <[4mpass[24m>"), the value of [1mlp[22;0m is 
          not changed.  

  tiny, tiny.* 
          TinyMud [1mlogin[22;0m format ("connect <[4mchar[24m> <[4mpass[24m>"), [1mlp[22;0m=off.  

  lp, lp.* 
  diku, diku.* 
  aber, aber.* 
          LP/Diku [1mlogin[22;0m format (sends <[4mchar[24m> and <[4mpass[24m> on separate lines), 
          [1mlp[22;0m=on.  For servers that send unterminated prompts.  

  lpp, lpp.* 
          LP/Diku [1mlogin[22;0m format, [1mlp[22;0m=off.  For muds that use GOAHEAD or EOR 
          [1mprompt protocol[22;0m.  

  telnet, telnet.* 
          Telnet [1mlogin[22;0m format (sends <[4mchar[24m> and <[4mpass[24m> when "login:" and 
          "password:" prompts are received), [1mlp[22;0m=on, [1m/localecho[22;0m on.  For any 
          line-by-line telnet service.  

  You can define your own world types for use in other triggers or hooks.  If 
  you use names that match the [1mglob[22;0m patterns above, the standard library hooks 
  will still work.  For example, if you did: 

    [1m/test[22;0m [1maddworld[22;0m("Cave", "tiny.muck.",     "cave.tcp.com", 2283, <[4mchar[24m>, <[4mpass[24m>)
    [1m/test[22;0m [1maddworld[22;0m("Foo",  "tiny.muck.msp.", "foo.com",      9999, <[4mchar[24m>, <[4mpass[24m>)
    [1m/test[22;0m [1maddworld[22;0m("Cow",  "tiny.moo.",      "cow.com",      8267, <[4mchar[24m>, <[4mpass[24m>)
    [1m/test[22;0m [1maddworld[22;0m("Buzz", "tiny.moo.msp.",  "buzz.org",     8267, <[4mchar[24m>, <[4mpass[24m>)

  then tiny-style [1mautologin[22;0m would still work (using the library hooks), and 
  you could also define your own [1mtriggers[22;0m and [1mhooks[22;0m specific to TinyMUCKs or 
  TinyMOOs (e.g., "[1m/def[22;0m [1m-T[22;0mtiny.muck.*") or to worlds that support MSP 
  regardless of their server type (e.g., "[1m/def[22;0m [1m-T[22;0m*.msp.*"), etc.  Note the 
  trailing period on the world type defintions, which make it easier to write 
  matching triggers.  

  Any <[4mtype[24m> is valid, but is only useful if it is matched by a "[1m-T<[4mtype[24m>[22;0m" 
  option of a [1mhook[22;0m or [1mtrigger[22;0m.  

  If [1maddworld()[22;0m with a password is executed from a file that has permissions 
  making it readable by others, it will produce a warning.  You should change 
  the file permissions to prevent other people from reading your password.  

  See: [1mworlds[22;0m, [1m/connect[22;0m, [1m/fg[22;0m, [1m/unworld[22;0m, [1m/edworld[22;0m, [1m/telnet[22;0m 

&/addtiny
&/addlp
&/addlpp
&/adddiku
&/addtelnet

/add<[4mworldtype[24m>

  The comamnds [1m/addtiny[22;0m, [1m/addlp[22;0m, [1m/addlpp[22;0m, [1m/adddiku[22;0m, and [1m/addtelnet[22;0m take the 
  same arguments as [1m/addworld[22;0m, and also give that world a type.  A world's 
  type determines the format for automatic login and flag settings.  

  See: [1m/addworld[22;0m 

&/alias
&/unalias

/alias

  Usage: 

  [1m/REQUIRE[22;0m alias.tf

  [1m/ALIAS[22;0m [<[4mname[24m> [<[4mcommand[24m>]]
  [1m/UNALIAS[22;0m <[4mname[24m>
  [1m/PURGEALIAS[22;0m
  ____________________________________________________________________________

  With no arguments, [1m/alias[22;0m lists all aliases.  With a <[4mname[24m> argument, [1m/alias[22;0m 
  lists the alias with names that match the glob pattern <[4mname[24m>.  Otherwise, 
  [1m/alias[22;0m defines <[4mname[24m> as an alias for <[4mcommand[24m>.  

  [1m/Unalias[22;0m undefines an alias for <[4mname[24m> that was defined with [1m/alias[22;0m.  

  [1m/Purgealias[22;0m undefines all aliases defined with [1m/alias[22;0m.  Note that 
  [1m/purgealias[22;0m does not take a pattern argument.  

  To use an alias, just type its name followed by any optional arguments.  
  Unlike [1mmacros[22;0m defined with [1m/def[22;0m, you do not type '/' before <[4mname[24m> to 
  execute an alias.  [1mArgument substitution[22;0m in aliases works the same as in 
  [1mmacros[22;0m.  

  As of 3.5 alpha 11, aliases can be called from other aliases or [1mmacros[22;0m.  To 
  send a line of text to the server without alias calls, use [1msend()[22;0m.  If an 
  old alias that used to work now results in "Too many recursions", you need 
  to rewrite the alias to use [1msend()[22;0m.  

  Using [1m/def[22;0m instead of [1m/alias[22;0m is recommended.  

  See: [1m/def[22;0m, [1mmacros[22;0m, [1msubstitution[22;0m, [1mtfrc[22;0m 

&/at

/at

  Usage: 

  [1m/AT[22;0m [-v] [<[4mdate[24m>] <[4mtime[24m> <[4mcommands[24m>
  ____________________________________________________________________________

  <[4mCommands[24m> will be executed at <[4mdate[24m> and <[4mtime[24m>.  <[4mDate[24m> must be of the 
  form "<[4myear[24m>-<[4mmonth[24m>-<[4mday[24m>" or "<[4mmonth[24m>-<[4mday[24m>", where <[4myear[24m> may be 2 or 4 
  digits.  <[4mTime[24m> must be of the form "<[4mhours[24m>:<[4mminutes[24m>" or 
  "<[4mhours[24m>:<[4mminutes[24m>:<[4mseconds[24m>", where <[4mhours[24m> is between 0 and 23, and 
  <[4mseconds[24m> may be specified to the nearest microsecond.  If any part of the 
  date is omitted, it defaults to the nearest value for which <[4mdate[24m> and 
  <[4mtime[24m> are in the future.  For example, if the current time is 16:00, then 
  an argument of "15:00" means 15:00 tomorrow, and "17:00" means 17:00 today.  

  [1mOptions[22;0m: 
  -v      verbose: prints full date and time 

  Examples: 
  [1m/at[22;0m 04-01 00:00:00 [1m/echo[22;0m Happy April Fools Day!
  [1m/def[22;0m lunch_reminder = [1m/at[22;0m 12:00 [1m/echo[22;0m Lunchtime!%%; /lunch_reminder

  See: [1mprocesses[22;0m, [1m/repeat[22;0m, [1m/quote[22;0m 

&/bamf

/bamf

  Usage: 

  [1m/BAMF[22;0m [OFF|ON|OLD]
  ____________________________________________________________________________

  Sets the flag [1m%{bamf}[22;0m.  This flag controls whether TF will cooperate with 
  portals.  A portal allows a mud character to move from one server to another 
  transparently, by simply going through a seemingly normal mud exit.  

  How it works: A "portal" is text sent by a server of the form: 


    #### Please reconnect to <[4mname[24m>@<[4maddr[24m> (<[4mhost[24m>) port <[4mport[24m> #### 

  For example: 


    #### Please reconnect to Islandia@128.100.102.51 (hawkwind.utcs.toronto.edu) port 2323 #### 

  If [1m%{bamf}[22;0m is off, lines in this format have no effect.  If [1m%{bamf}[22;0m is on, 
  Fugue will attempt to use the portal as an UnterMUD portal: it will 
  disconnect from the [1mcurrent[22;0m world, and attempt to connect to the new world; 
  if the [1m%{login}[22;0m flag is also on, TF will try to log in to the new world 
  using the name and password from the [1mcurrent[22;0m world.  If bamf is "old", Fugue 
  will connect to the new world without disconnecting from the [1mcurrent[22;0m world.  
  If [1m%{login}[22;0m is also on, and the new world has been defined with a name and 
  password in an [1m/addworld[22;0m command, Fugue will attempt to log in 
  automatically.  

  Note that on many servers, arbitrary users can spoof the portal text, 
  redirecting your tf against your will if you have bamfing enabled.  

  The flag [1m%{bamf}[22;0m defaults to 0 (off).  

  See: [1mworlds[22;0m, [1msockets[22;0m, [1m%bamf[22;0m, [1m%login[22;0m 

&/beep

/beep

  Usage: 

  [1m/BEEP[22;0m [<[4mnumber[24m>|ON|OFF]
  ____________________________________________________________________________

  [1m/beep[22;0m causes Fugue to emit <[4mnumber[24m> beeps (ASCII 7).  [1m/beep[22;0m with no 
  arguments will emit three beeps.  [1m/beep[22;0m OFF causes Fugue to ignore further 
  calls to [1m/beep[22;0m until a [1m/beep[22;0m ON is performed.  

  Note that on many terminals, multiple immediate beeps are indistinguishable. 
  You can use [1m/repeat[22;0m to put a delay between beeps: 

    [1m/repeat[22;0m -0.2 5 [1m/beep[22;0m

&/bind

/bind

  Usage: 

  [1m/BIND[22;0m <[4msequence[24m> = <[4mcommand[24m>
  ____________________________________________________________________________

  Creates a [1mmacro[22;0m that will be executed when <[4msequence[24m> is typed at the 
  keyboard.  The <[4msequence[24m> may use ^<[4mkey[24m> notation for a control key, and 
  \<[4mnumber[24m> for an ascii character code in octal, hexadecimal, or decimal.  
  For example, the escape character can be given by any of these forms: ^[, 
  \033, \0x1B, or \27.  

  When the key sequence <[4msequence[24m> is typed at the keyboard, <[4mcommand[24m> is 
  executed.  The command is actually a [1mmacro[22;0m body, so all the substitutions 
  described under "[1mevaluation[22;0m" will be performed.  The most common command 
  used with a key binding is [1m/dokey[22;0m.  

  At [1mstartup[22;0m, TF defines bindings for [1m/dokey[22;0m BSPC, BWORD, DLINE, REFRESH, 
  LNEXT, UP, DOWN, RIGHT, and LEFT based on your terminal settings.  Also, the 
  standard [1mmacro[22;0m library defines a set of (invisible) default bindings, one 
  for each of the [1m/dokey[22;0m functions.  

  If [1m/bind[22;0m fails for any reason, it returns 0.  Otherwise, it returns the 
  number of the new [1mmacro[22;0m (useful in [1m/undefn[22;0m and [1m/edit[22;0m).  

  As of version 3.5, the NUL character (^@) is allowed in keybindings.  

  The command
  [1m/bind[22;0m <[4msequence[24m> = <[4mcommand[24m>
  is equivalent to
  [1m/def[22;0m [1m-b[22;0m"<[4msequence[24m>" = <[4mcommand[24m>. 

  Examples: 

    [1m/bind[22;0m ^Xtw = :jumps to the left%;:steps to the right!
    [1m/bind[22;0m ^[q = [1m/set[22;0m [1mmore[22;0m off
    [1m/bind[22;0m ~ky = [1m/input[22;0m Kyosuke

  See: [1mkeys[22;0m, [1m/dokey[22;0m, [1m/unbind[22;0m, [1m/input[22;0m, [1minterface[22;0m 

&/break

/break

  Usage: 

  [1m/BREAK[22;0m [<[4mn[24m>]
  ____________________________________________________________________________

  During [1mmacro[22;0m evaluation, [1m/BREAK[22;0m unconditionally terminates the nearest 
  enclosing [1m/WHILE[22;0m loop.  If <[4mn[24m> is specified, it will break out of <[4mn[24m> 
  enclosing [1m/WHILE[22;0m loops.  If used outside a [1m/WHILE[22;0m loop, the [1mmacro[22;0m evaluation 
  is terminated.  

  See: [1m/while[22;0m, [1m/return[22;0m, [1m/exit[22;0m, [1mevaluation[22;0m 

&/cat

/cat

  Usage: 

  [1m/CAT[22;0m [%]
  ____________________________________________________________________________

  Concatenates (puts together) all subsequent lines until a line containing a 
  single "." is typed.  If the argument "%" is given, a "%;" sequence is 
  appended to each intermediate line.  The concatenated result is then 
  executed as a single line.  

  The concatenated result is stored in the input [1mhistory[22;0m as a single line, so 
  intermediate lines can not be recalled separately.  

  Example: 

    [1m/cat[22;0m %
    :foo
    :bar
    :baz
    . 

  This produces: 

    :foo%;:bar%;:baz

  If the [1m%{sub}[22;0m flag is set on, this will [1mexpand[22;0m to three lines ":foo", ":bar" 
  and ":baz" and be sent to the [1msocket[22;0m.  

  See: [1m/paste[22;0m, [1m/sub[22;0m, [1mgeneral[22;0m, [1mhistory[22;0m 

&/changes

/changes

  Usage: 

  [1m/CHANGES[22;0m [<[4mversion[24m>]
  ____________________________________________________________________________

  List the changes in a <[4mversion[24m> of TinyFugue; if omitted, <[4mversion[24m> defaults 
  to the current version.  <[4mVersion[24m> can be a full version name (e.g., "5.0 
  beta 7") or just the major and minor numbers (e.g., "5.0").  The information 
  is kept in the file [1m%TFLIBDIR/CHANGES[22;0m.  

  A list of changes in the latest version of tf can be found at 
  [1mhttp://tinyfugue.sourceforge.net/CHANGES[22;0m.  

  See: [1m/version[22;0m 

&completion
&/complete

/complete

  Usage: 

  [1m/COMPLETE[22;0m [<[4mtype[24m>]
  ____________________________________________________________________________

  When a part of a word is typed, and then [1m/complete[22;0m is called (from a 
  [1mkeybinding[22;0m), it will attempt to fill in the rest of the word.  The possible 
  words it chooses from depend on <[4mtype[24m>.  If no <[4mtype[24m> is given, it completes 
  from context: it will choose the type of completion based on earlier parts 
  of the line being typed, plus previous [1minput history[22;0m.  For example, if the 
  line begins with "/connect", it will use worldname completion; if the word 
  begins with "%" or "%{", it will use variable name completion; etc.  

  The following table lists the meanings and the default [1mkeybindings[22;0m for each 
  type.  

      Keys                Type            Meaning
      ----                ----            -------
      ^[^I (ESC TAB)                      complete word depending on context
      ^[^W                worldname       complete tf world name
      ^[$                 macroname       complete tf macro name
      ^[%                 variable        complete tf variable name
      ^[/                 filename        complete file name (unix only)
      ^[;                 user_defined    complete from [1m%{completion_list}[22;0m
      ^[i                 input_history   complete from previously typed words
                          sockname        complete name of open tf socket

  The "ESC TAB" and "ESC ;" bindings will use the [1m%{completion_list}[22;0m [1mvariable[22;0m, 
  in which you can store a list of any words you want to be able to complete.  

  You can also define your own types of completion.  See the 
  [1m%{TFLIBDIR}[22;0m/complete.tf file for more information.  

  See: [1mkeybindings[22;0m, [1minterface[22;0m 

&/connect

/connect

  Usage: 

  [1m/CONNECT[22;0m [-lqxbf] [<[4mworld[24m>]
  [1m/CONNECT[22;0m <[4mhost[24m> <[4mport[24m>
  ____________________________________________________________________________

  In the first form, [1m/connect[22;0m attempts to open a [1msocket[22;0m connected to <[4mworld[24m>.  
  <[4mWorld[24m> must be defined by the [1m/addworld[22;0m command and not already open.  If 
  <[4mworld[24m> is omitted, the first defined world will be used.  If <[4mworld[24m> does 
  not have a host and port, [1m/connect[22;0m will create a "connectionless" [1msocket[22;0m.  
  In the form "[1m/connect[22;0m <[4mhost[24m> <[4mport[24m>", it will define a temporary world named 
  "(unnamed<[4mN[24m>)" with the given address, and try to connect to it.  <[4mHost[24m> may 
  be an internet hostname, an [1mIPv4[22;0m address, or (if your platform supports it) 
  an [1mIPv6[22;0m address.  A temporary world will be undefined when it is no longer 
  in use.  

  [1mOptions:[22;0m 
  -l      No [1mautomatic login[22;0m (i.e., don't call the [1mLOGIN[22;0m [1mhook[22;0m).  
  -q      Quiet login (overrides [1m%{quiet}[22;0m flag).  
  -x      Connect using SSL (not necessary if [1mworld[22;0m was defined with the "x" 
          flag).  
  -f      Connect in the foreground 
  -b      Connect in the background 

  The first thing [1m/connect[22;0m does is create a new [1msocket[22;0m.  If the -f option was 
  given, or [1m/connect[22;0m was called from the foreground (e.g., from the command 
  line), the new [1msocket[22;0m is immediately brought into the [1mforeground[22;0m.  If the -b 
  option was given, or [1m/connect[22;0m was called from the background (e.g., from a 
  DISCONNECT [1mhook[22;0m in a [1mbackground[22;0m world), the connection proceeds in the 
  background.  

  If a hostname was given, TF must look it up to find one or more [1mIPv4[22;0m or (if 
  your platform supports it) [1mIPv6[22;0m addresses.  If [1m%{gethostbyname}[22;0m is 
  "nonblocking" (the default), and this process takes more than a fraction of 
  a second, TF will print "Hostname resolution for <[4mworld[24m> in progress" (the 
  PENDING [1mhook[22;0m), and TF will continue running normally while the lookup 
  proceeds.  But if [1m%{gethostbyname}[22;0m is "blocking", TF will freeze until the 
  lookup is finished.  Either way, if the lookup succeeds, TF will try to 
  connect; if it fails, you will be notified.  

  Next, TF tries to open a network connection to the IP address, and prints 
  "Trying to connect to <[4mworld[24m>: <[4maddress[24m> <[4mport[24m>" (the PENDING hook).  On 
  most platforms, if [1m%{connect}[22;0m is "nonblocking" (the default), TF continues 
  running normally while the network connection proceeds.  But if [1m%{connect}[22;0m 
  is "blocking", TF will freeze until the network connection is finished.  If 
  the connection succeeds, a message is printed, but (unlike previous versions 
  of TF) the [1msocket[22;0m is not automatically brought to the [1mforeground[22;0m.  However, 
  if you had run [1m/connect[22;0m in the foreground (e.g.  from the command line), the 
  [1msocket[22;0m would already be in the [1mforeground[22;0m, unless it was nonblocking and had 
  taken a long time and you [1mforegrounded[22;0m another [1msocket[22;0m while you were 
  waiting, in which case you probably wouldn't want to automatically 
  [1mforeground[22;0m the new [1msocket[22;0m.  If you prefer automatic [1mforegrounding[22;0m upon 
  successful connection, you can [1mdefine[22;0m a CONNECT [1mhook[22;0m that calls "[1m/fg[22;0m [1m%{1}[22;0m".  

  Even if [1m%{gethostbyname}[22;0m and/or [1m%{connect}[22;0m are "blocking", they can be 
  interrupted with the SIGINT [1msignal[22;0m (^C).  

  If the connection fails, TF normally prints "Connection to <[4mworld[24m> failed: 
  <[4maddress[24m> <[4mport[24m>: <[4mreason[24m>" (the CONFAIL [1mhook[22;0m).  But, if the failure was in 
  the specific address, and there is more than one address associated with the 
  [1mworld[22;0m's hostname, the message will instead say "Intermediate connection to 
  <[4mworld[24m> failed: ..." (the ICONFAIL [1mhook[22;0m), and TF will try to connect to the 
  next address.  So, a failed [1m/connect[22;0m will always result in a series of zero 
  or more ICONFAIL [1mhooks[22;0m followed by exactly one CONFAIL [1mhook[22;0m.  But an 
  ICONFAIL may also be followed by a successful connection to an alternate 
  address.  

  If the network connection is successful, or the [1msocket[22;0m is "connectionless", 
  these events occur: 

    * If the [1mworld[22;0m was defined with an <[4mmfile[24m>, that file will be loaded 
      (and the LOAD [1mhook[22;0m will be called); 
    * The CONNECT [1mhook[22;0m is called (unless the socket is connectionless or 
      the connection is via a [1mproxy[22;0m).  
    * If [1m%{login}[22;0m is on, and a character and password is defined for the 
      [1mworld[22;0m, the LOGIN [1mhook[22;0m is called (unless the socket is connectionless or 
      the connection is via a [1mproxy[22;0m).  The default LOGIN [1mhooks[22;0m sends the 
      character name and password in a format corresponding to the world type. 
      To automatically login to a world that expects a different login format, 
      define your own LOGIN [1mhook[22;0m.  

  If you have trouble connecting (especially if you use SOCKS), try "[1m/set[22;0m 
  [1mconnect[22;0m=blocking".  

  If your host has multiple network interfaces, the OS will choose one of them 
  for the client end of the connection according to its own rules.  To 
  override the system's choice, set the [1mtfhost[22;0m variable or define the [1mworld[22;0m 
  with a <[4msrchost[24m> parameter to [1maddworld[22;0m.  

  [1m/connect[22;0m returns 0 on error or failure, 1 for immediate success, or 2 if the 
  name lookup or network connection is pending.  

  See: [1mworlds[22;0m, [1msockets[22;0m, [1mproxy[22;0m, [1m/world[22;0m, [1m/addworld[22;0m, [1m/fg[22;0m, [1m/retry[22;0m, [1m%login[22;0m, 
  [1m%gethostbyname[22;0m, [1m%connect[22;0m, [1mhooks[22;0m [1mprocotols[22;0m 

&disconnect
&close
&/dc

/dc

  Usage: 

  [1m/DC[22;0m [<[4mworld[24m>|-ALL]
  ____________________________________________________________________________

  Disconnects from the named world, or the [1mcurrent[22;0m world if no world is given, 
  or all worlds if "-all" is given.  If the flag [1m%{quitdone}[22;0m is on, and [1m/dc[22;0m 
  disconnects the last [1msocket[22;0m, TF will exit.  

  Disconnecting with [1m/dc[22;0m does not invoke the [1mDISCONNECT[22;0m [1mhook[22;0m.  

  See: [1msockets[22;0m, [1m%quitdone[22;0m, [1m/quit[22;0m 

&/def

/def

  Usage: 

  [1m/DEF[22;0m [<[4moptions[24m>] [<[4mname[24m>] [= <[4mbody[24m>]
  ____________________________________________________________________________

  Defines a [1mmacro[22;0m with an optional [1mkeybinding[22;0m, [1mtrigger[22;0m and/or [1mhook[22;0m associated 
  with it.  The [1moptions[22;0m and their meanings are: 

#-msimple
#-mglob
#-mregexp
#/def -m
#-m
  -m<[4mmatching[24m> 
          Determines which matching style should be used for [1m-t[22;0m, [1m-h[22;0m, or [1m-T[22;0m 
          options.  Valid values are "[1msimple[22;0m", "[1mglob[22;0m", and "[1mregexp[22;0m" (see also: 
          [1mpatterns[22;0m).  If omitted, the value of [1m%{matching}[22;0m ("[1mglob[22;0m" by default) 
          is used, unless [1m-P[22;0m is also given, in which case "[1mregexp[22;0m" is used.  

#/def -n
#-n
  -n<[4mshots[24m> 
          The [1mmacro[22;0m is a multi-shot, that is, it will be deleted after it is 
          [1mtrigger[22;0med or [1mhook[22;0med <[4mshots[24m> times.  A value of 0 makes the [1mmacro[22;0m 
          permanent.  Default: 0.  

#/def -E
#-E
  -E<[4mexpression[24m> 
          Before this [1mmacro[22;0m is tested for a [1mtrigger[22;0m ([1m-t[22;0m) or [1mhook[22;0m ([1m-h[22;0m) match, 
          <[4mexpression[24m> is evaluated; if its value is 0, the macro will not be 
          considered a match, so no [1mattributes[22;0m (-a) will be applied, and this 
          macro will not prevent matches of lower [1mpriority[22;0m (-p), and its body 
          will not be executed.  If the value of <[4mexpression[24m> is non-zero, the 
          comparison proceedes as usual.  Note: 
          * [1mpositional parameters[22;0m ([1m%n[22;0m) and [1msubexpression matches[22;0m ([1m%Pn[22;0m) are not 
          available in <[4mexpression[24m>.  
          * Remember that for every macro with a trigger and an -E expression, 
          its <[4mexpression[24m> must be evaluated for every line received.  So, you 
          should keep it simple (e.g., "enable_foo" or "[1m${world_name}[22;0m =~ 
          [1mfg_world[22;0m()").  More complex expressions should be put in the body of 
          the macro.  
          * The body of a high [1mpriority[22;0m [1mmacro[22;0m is not necessarily executed 
          before the -E expression of a lower [1mpriority[22;0m [1mmacro[22;0m is tested, so 
          <[4mexpression[24m> should not rely on values that may be changed by other 
          macros that match the same [1mtrigger[22;0m or [1mhook[22;0m.  
          Default: no [1mexpression[22;0m (i.e., always match if the [1mtrigger[22;0m or [1mhook[22;0m 
          matches).  See: [1mexpressions[22;0m.  

#/def -t
#-t
  -t<[4mpattern[24m> 
          Defines a [1mtrigger[22;0m pattern which will cause the [1mmacro[22;0m to be called 
          when it is matched by a line of text from a socket.  <[4mPattern[24m> may 
          be enclosed in quotes (", ', or `); if so, all occurances of quotes 
          and '\' within the pattern must be preceded with a '\'.  The [1mpattern[22;0m 
          matching style is determined by the [1m-m[22;0m option, or defaults to the 
          value of [1m%{matching}[22;0m.  Default: no [1mtrigger[22;0m.  See: [1mtriggers[22;0m.  

#/def -h
#-h
  -h"<[4mevent[24m>[ <[4mpattern[24m>]" 
          Specifies that the [1mmacro[22;0m will be called automatically whenever 
          <[4mevent[24m> occurs and its arguments match <[4mpattern[24m>.  <[4mEvent[24m> may be a 
          single event name or a list separated by '|'.  If <[4mpattern[24m> is 
          omitted, it will match any arguments, and the quotes may also be 
          omitted.  If quotes are used, then all occurances of quotes and '\' 
          within the option argument must be preceded with a '\'.  The [1mpattern[22;0m 
          matching style is determined by the [1m-m[22;0m option, or defaults to the 
          value of [1m%{matching}[22;0m.  Default: no [1mhook[22;0m.  See: [1mhooks[22;0m.  

#/def -b
#-b
  -b<[4mbind[24m> 
          The [1mmacro[22;0m will be called when the string <[4mbind[24m> is typed at the 
          keyboard.  Default: no binding.  The <[4mbind[24m> string may contain the 
          special codes described under "[1mbind[22;0m".  See: [1mkeys[22;0m.  

#/def -B
#-B
  -B<[4mkeyname[24m> 
          Deprecated.  The [1mmacro[22;0m will be called when the key named <[4mkeyname[24m> 
          (according to the termcap database) is typed at the keyboard.  
          Default: none.  See "[1mkeys[22;0m".  

#/def -p
#-p
  -p<[4mpri[24m> 
          Sets the [1mpriority[22;0m of the [1mmacro[22;0m's [1mtrigger[22;0m or [1mhook[22;0m to <[4mpri[24m>.  As in 
          all [1mnumeric options[22;0m, the argument to -p may be an [1mexpression[22;0m that 
          has a numeric value.  E.g.  "[1m/def[22;0m -pmaxpri ..." will set the macro's 
          priority to the value of the variable maxpri.  The [1mexpression[22;0m is 
          evaluated only once, when the macro is defined.  Default: 1.  See 
          also: [1mfall-thru[22;0m.  See: [1mpriority[22;0m, [1m/def -F[22;0m.  

#/def -c
#-c
  -c<[4mchance[24m> 
          Sets the percent probability of executing the body of a matched 
          [1mtrigger[22;0m or [1mhook[22;0m.  (The macro still counts as a match for attributes 
          and priority even if it does not execute.) Default: 100%.  

#/def -w
#-w
  -w<[4mworld[24m> 
          If the [1mmacro[22;0m has a [1mtrigger[22;0m or [1mhook[22;0m, it can be matched only by text 
          or events from <[4mworld[24m>.  Default: any world.  

#/def -T
#-T
  -T<[4mtype[24m> 
          If the [1mmacro[22;0m has a [1mtrigger[22;0m or [1mhook[22;0m, it can be matched only by text 
          or events from worlds of type <[4mtype[24m>.  (See: [1m/addworld[22;0m).  The 
          [1mpattern[22;0m matching style is determined by the [1m-m[22;0m option, or defaults 
          to the value of [1m%{matching}[22;0m.  Default: any type.  

#/def -F
#-F
  -F      [1mFall-thru[22;0m: on a [1mtrigger[22;0m or [1mhook[22;0m, allows additional matches of lower 
          [1mpriority[22;0m to be run.  Default: not [1mfall-thru[22;0m.  See: [1mpriority[22;0m 

#/def -a
#-a
  -a[ngGLAurBbhC] 
          Set [1mattribute[22;0m(s) (normal, [1mgag[22;0m, nohistory, nolog, noactivity, 
          underline, reverse, bold, bell, [1mhilite[22;0m, Color) used to display text 
          matched by the [1mtrigger[22;0m or to display the default message of a [1mhook[22;0m.  
          Default: normal.  See: [1mattributes[22;0m.  

#/def -P
#-P
  -P[<[4mpart[24m>]<[4mattr[24m>[;[<[4mpart[24m>]<[4mattr[24m>]...  
          Define a "partial [1mhilite[22;0m".  The argument consists of a list of pairs 
          of parts (<[4mpart[24m>) and attributes (<[4mattr[24m>), separated by ';'.  When a 
          line matches the [1mregexp[22;0m [1mtrigger[22;0m of this macro, each <[4mattr[24m> is 
          applied to the corresponding <[4mpart[24m> of the line.  <[4mAttr[24m> can contain 
          any of the [1mattribute[22;0m codes "nxurBhC".  (normal, exclusive, 
          underline, reverse, bold, [1mhilite[22;0m, Color).  The value of <[4mpart[24m> 
          determines which part of the text is affected: 
          L       text to the left of the [1mregexp[22;0m match 
          R       text to the right of the [1mregexp[22;0m match 
          0       text matched by the entire [1mregexp[22;0m 
          <[4mnumber[24m>
                  text matched by the the <[4mnumber[24m>th parenthesized 
                  subexpression of the [1mregexp[22;0m.  
          If <[4mpart[24m> is omitted it defaults to 0.  If <[4mpart[24m> is a number and 
          there are multiple matches in the text, the <[4mattr[24m> will be applied 
          to all of the matches.  Implies [1m-m[22;0mregexp.  Only one [1m-P[22;0m option is 
          allowed.  See: [1mattributes[22;0m.  

#/def -f
#-f
  -f      Same as [1m-a[22;0m, for backward compatibility.  

#/def -I
#-I
#/def -i
#-i
  -i 
  -I      Makes the [1mmacro[22;0m "invisible".  Invisible [1mmacros[22;0m are not processed by 
          [1m/list[22;0m, [1m/save[22;0m, or [1m/purge[22;0m unless forced.  Default: not invisible.  

#/def -q
#-q
  -q      Makes the [1mmacro[22;0m "quiet".  If called as a [1mtrigger[22;0m, the [1mmacro[22;0m will not 
          count toward the [1mBACKGROUND[22;0m [1mhook[22;0m or the return value of [1m/trigger[22;0m.  
          If called as a [1mSEND[22;0m [1mhook[22;0m, the [1mmacro[22;0m will not prevent the sending of 
          the original input.  If called as a [1mPROMPT[22;0m [1mhook[22;0m, the [1mmacro[22;0m will not 
          remove the text from the data stream.  

#-1
  -1      Defines a one-shot.  Equivalent to "[1m-n[22;0m1".  

#
  <[4mname[24m>  The name of the [1mmacro[22;0m.  Default: no name.  Names should begin with a 
          letter, and contain letters, numbers, or '_' characters.  This is 
          not enforced, but other characters (especially '$', '/', and '%') 
          may cause unwanted interpretations during [1mexpansion[22;0m.  

  = <[4mbody[24m> 
          Text to be executed when [1mmacro[22;0m is called.  Default: no body.  

  If [1m/def[22;0m could not create a new [1mmacro[22;0m, it returns 0.  Otherwise, it returns 
  the number of the new [1mmacro[22;0m (useful with [1m/undefn[22;0m and [1m/edit[22;0m).  
  ____________________________________________________________________________

##follow
  Example: 

    [1m/def[22;0m follow = \
        [1m/def[22;0m [1m-T^tiny[22;0m [1m-mregexp[22;0m [1m-p2[22;0m [1m-t[22;0m"^[1m%{1}[22;0m goes ([a-z]*)\\\\.$$" do_follow = \
            go %%P1

  This will create a [1mmacro[22;0m named "follow".  When it is called like "/follow 
  Joe", it will execute the command 

    [1m/def[22;0m [1m-T^tiny[22;0m [1m-mregexp[22;0m [1m-p2[22;0m [1m-t[22;0m"^Joe goes ([a-z]*)\\.$" do_follow = go [1m%P1[22;0m

  Note the [1msubstitutions[22;0m that occurred: "[1m%{1}[22;0m" was replaced with the first 
  (and only) argument; each "[1m\\[22;0m" was replaced with "\"; "[1m$$[22;0m" was replaced with 
  "$"; and "[1m%%[22;0m" was replaced with "%".  

  That command, in turn, defines another [1mmacro[22;0m called "do_follow", with a 
  [1mregexp[22;0m [1mtrigger[22;0m 

    ^Joe goes ([a-z]*)\.$

  which will only match on worlds whose type matches the [1mregexp pattern[22;0m 
  "^tiny".  

  Thereafter, when a line like "Joe goes north." is received, it will match 
  the [1mtrigger[22;0m, and cause this command to be executed: 

    go north

  Note how "[1m%P1[22;0m" was substituted with the text matched by the first set of 
  parentheses (in this case, "north").  

  When writing nested [1mmacros[22;0m like this, it is usually easiest to think 
  backwards.  In this example, you would first figure out how /do_follow 
  should be defined, and then figure out how to define /follow in such a way 
  that it will define /do_follow.  
#
  ____________________________________________________________________________

  [1m/def[22;0m is sufficient to perform all the functions of the [1m/trig[22;0m, [1m/trigp[22;0m, 
  [1m/trigc[22;0m, [1m/trigpc[22;0m, [1m/gag[22;0m, [1m/hilite[22;0m, [1m/partial[22;0m, [1m/hook[22;0m, and [1m/bind[22;0m commands.  

  See: [1mmacros[22;0m, [1mtriggers[22;0m, [1mpatterns[22;0m, [1mhooks[22;0m, [1mpriority[22;0m, [1mevaluation[22;0m, [1mattributes[22;0m, 
  [1m/undef[22;0m, [1m/undefn[22;0m, [1m/purge[22;0m, [1m/list[22;0m, [1m/save[22;0m, [1m/load[22;0m 

&/dokey

/dokey

  Usage: 

  [1m/DOKEY[22;0m <[4mname[24m>
  ____________________________________________________________________________

  Performs an action that is intended to be invoked from a [1mkeybinding[22;0m created 
  with [1m/bind[22;0m or [1m/def -b[22;0m.  Most of the actions not meaningful or useful when 
  the [1m/dokey[22;0m command is executed from the command line.  


    Name          Default binding   Action
    ----          ---------------   --------
#bs
#backspace
#bspc
    BSPC          (stty), ^H, ^?    Backspace
#bword
    BWORD         (stty), ^W        Delete previous word
#dline
    DLINE         (stty), ^U        Delete entire line
#refresh
    REFRESH       (stty), ^R        Refresh line
#lnext
    LNEXT         (stty), ^V        Ignore any binding next key might have
#

#up
    UP            (none)            Cursor up
#down
    DOWN          (none)            Cursor down
#right
    RIGHT         [1mkey_right[22;0m         Cursor right
#left
    LEFT          [1mkey_left[22;0m          Cursor left
#

#newline
    NEWLINE       ^J, ^M            Execute current line
#recallb
    RECALLB       ^P                Recall previous input line
#recallf
    RECALLF       ^N                Recall next input line
#recallbeg
    RECALLBEG     ^[<               Recall first input line
#recallend
    RECALLEND     ^[>               Recall last input line
#searchb
    SEARCHB       ^[p               Search backward in input history
#searchf
    SEARCHF       ^[n               Search forward in input history
#socketb
    SOCKETB       ^[b               Switch to previous [1msocket[22;0m
#socketf
    SOCKETF       ^[f               Switch to next [1msocket[22;0m
#dword
    DWORD         ^[d               Delete word
#del
#delete
#dch
    DCH           ^D                Delete character under cursor
#redraw
    REDRAW        ^L                Redraw screen
#clear
    CLEAR         ^[^L              Clear screen
#home
    HOME          ^A                Go to beginning of line
#end
    END           ^E                Go to end of line
#wleft
    WLEFT         ^B                Go left, to beginning of word
#wright
    WRIGHT        ^F                Go right, to end of word
#deol
    DEOL          ^K                Delete from cursor to end of line
#pause
    PAUSE         ^S                Pause screen
#page
    PAGE          [1mkey_tab[22;0m           Scroll 1 page forward ("[1mmore[22;0m")
#pageback
    PAGEBACK      (none)            Scroll 1 page backward ("[1mmore[22;0m")
#hpage
    HPAGE         ^X]               Scroll half page forward ("[1mmore[22;0m")
#hpageback
    HPAGEBACK     ^X[               Scroll half page backward ("[1mmore[22;0m")
#pgup
    PGDN          [1mkey_pgdn[22;0m          [1m/dokey_hpage[22;0m
#pgup
    PGUP          [1mkey_pgup[22;0m          [1m/dokey_hpageback[22;0m
#line
    LINE          ^[^N              Scroll forward 1 line ("[1mmore[22;0m")
#lineback
    LINEBACK      ^[^P              Scroll backward 1 line ("[1mmore[22;0m")
#flush
    FLUSH         ^[j               Jump to end of scroll buffer
#selflush
    SELFLUSH      ^[J               Show lines with [1mattributes[22;0m,
                                        and jump to end of buffer

#

  A default of "(stty)" means the key sequence is that used by your terminal 
  driver.  A default of the form "key_<[4mname[24m>" means the key named <[4mname[24m> (see 
  [1mkeybindings[22;0m).  

  The return value of [1m/dokey[22;0m depends on the action.  The movement and deletion 
  actions return the new position of the cursor; the scrolling actions return 
  the number of lines scrolled.  The return values of other actions aren't 
  very useful.  

  See "[1mkeybindings[22;0m" for a complete list of keybindings.  

  Example: 

    [1m/bind[22;0m ^B = [1m/dokey[22;0m RECALLB
    [1m/bind[22;0m ^F = [1m/dokey[22;0m RECALLF

  Then, ^B and ^F could be used to recall input backwards and forwards.  

  See: [1mkeybindings[22;0m, [1m/bind[22;0m, [1msockets[22;0m, [1mhistory[22;0m, [1m/more[22;0m 

&/echo
&/_echo
&echo()

echo()

  [1mFunction[22;0m usage: 

  [1mECHO[22;0m(<[4mtext[24m> [, <[4mattrs[24m> [, <[4minline[24m> [, <[4mdest[24m>]]])

  Command usage: 

  [1m/ECHO[22;0m [-peA] [-a<[4mattrs[24m>] [-w[<[4mworld[24m>]] <[4mtext[24m>
  [1m/_ECHO[22;0m <[4mtext[24m>
  ____________________________________________________________________________

  Displays <[4mtext[24m> on the [1mtfout stream[22;0m (i.e., the screen, usually), unless 
  otherwise redirected by options.  

  [1mOptions[22;0m and arguments: 
  command: -a<[4mattrs[24m> 
  function: <[4mattrs[24m> 
          Echo <[4mtext[24m> with the [1mattributes[22;0m given by <[4mattrs[24m>.  
  command: -p 
  function: <[4minline[24m> = "on" or 1 
          Interpet "@{<[4mattr[24m>}" strings as commands to set [1mattributes[22;0m inline.  
          "@@" strings are interpreted as "@".  "@{n}" or "@{x}" will turn 
          attributes off.  See also: [1mdecode_attr()[22;0m.  
  command: -w<[4mworld[24m> 
  function: <[4mdest[24m> = "w<[4mworld[24m>" 
          Echo <[4mtext[24m> to the <[4mworld[24m> [1mstream[22;0m instead of the default [1mtfout 
          stream[22;0m (see [1mtfio[22;0m).  If <[4mworld[24m> is blank, the [1mcurrent[22;0m world is 
          assumed.  
  command: -e 
  function: <[4mdest[24m> = "e" 
          Echo <[4mtext[24m> to the [1mtferr stream[22;0m, instead of the default [1mtfout stream[22;0m 
          (see [1mtfio[22;0m).  
  function: <[4mdest[24m> = "o" 
          Echo <[4mtext[24m> to the [1mtfout stream[22;0m (the default).  
  command: -A 
  function: <[4mdest[24m> = "a" 
          Echo <[4mtext[24m> to the [1malert stream[22;0m, instead of the default [1mtfout stream[22;0m 
          (see [1mtfio[22;0m).  

  The command form is usually more convenient, but the function form is the 
  only way to echo text with leading or trailing spaces.  Remember that "-" by 
  itself can be used to mark the end of command [1moptions[22;0m, in case <[4mtext[24m> begins 
  with "-".  

  [1m/_echo[22;0m is more efficient than [1m/echo[22;0m, so it is better for use in heavily used 
  macros that don't need all the options of [1m/echo[22;0m.  

  When echoing to the [1mtferr stream[22;0m, if no <[4mattrs[24m> are specified, text will be 
  echoed with the "E" [1mattribute[22;0m.  

  Example: Both of these commands 

    [1m/test[22;0m [1mecho[22;0m("@{u}Hello@{n}, world!", "BCred", 1)
    [1m/echo[22;0m -aBCred -p @{u}Hello@{n}, world!

  echo the following line, with "Hello" underlined, and the whole line bold 
  red: 

    [31m[1m[4mHello[24m, world![22m[0m

  Echoed text is not matched against [1mtriggers[22;0m.  To do that, use [1m/trigger[22;0m.  

  See: [1mattributes[22;0m, [1mworlds[22;0m, [1mfwrite()[22;0m, [1mpad()[22;0m, [1mtfio[22;0m 

&/edit

/edit

  Usage: 

  [1m/EDIT[22;0m [<[4moptions[24m>] [<[4mname[24m>] [= <[4mbody[24m>]
  ____________________________________________________________________________

  Edits a currently existing [1mmacro[22;0m or the [1mtrigger[22;0m associated with a [1mmacro[22;0m.  
  Options are described under "[1mdef[22;0m".  The name of the [1mmacro[22;0m must be specified 
  and cannot be changed, with the following two exceptions: 

  1.  The [1mmacro[22;0m name can be specified as "#<[4mnum[24m>" where <[4mnum[24m> is the number of 
  the [1mmacro[22;0m instead of the name.  A [1mmacro[22;0m number can be determined by listing 
  the [1mmacro[22;0m with [1m/list[22;0m, or from the return value of [1m/def[22;0m or [1m/edit[22;0m.  

  2.  The [1mmacro[22;0m name can be specified as "$<[4mpattern[24m>" where <[4mpattern[24m> is the 
  [1mtrigger[22;0m pattern.  You may still change the pattern if this is used to locate 
  the [1mmacro[22;0m.  

  In either case, the name cannot be changed.  It is possible to create a 
  [1mmacro[22;0m which changes the name of a [1mmacro[22;0m, if it does not have any options 
  other than a name and a body: 

  [1m/def[22;0m rename = [1m/def[22;0m [1m%2[22;0m = $[1m%1[22;0m%; [1m/undef[22;0m [1m%1[22;0m 

  How this works is discussed in the help section "[1mexpansion[22;0m".  

  Also, the [1m/edmac[22;0m command will allow you to edit an existing macro definition 
  on the command line.  

  The [1m-i[22;0m flag will be cleared automatically from the [1mmacro[22;0m if it is not 
  explicitly given to [1m/edit[22;0m.  The body may be cleared by specifiying "=" with 
  nothing after it; if "=" is not present at all, the macro's body will be 
  unchanged.  It is not possible to clear the [1m-F[22;0m option.  The [1m-w[22;0m, [1m-T[22;0m [1m-t[22;0m, and 
  [1m-h[22;0m options also can not be cleared, but their arguments can be changed.  The 
  [1m-T[22;0m, [1m-t[22;0m, and [1m-h[22;0m options will use the [1mpattern matching style[22;0m specified by the 
  [1m-m[22;0m option to the [1m/edit[22;0m command; they will [4mnot[24m inherit [1m-m[22;0m from the original 
  definition.  Any other options that are not specified with [1m/edit[22;0m will remain 
  unchanged from the original definition.  

  As of version 5.0, [1m/edit[22;0m does not renumber the macro being edited.  

  Example: 

  [1m/def[22;0m [1m-p2[22;0m [1m-t[22;0m"* has arrived." [1m-ah[22;0m greet = :greets [1m%1[22;0m
  [1m/edit[22;0m -c0 greet 

  The second command will change the probability of /greet's [1mtrigger[22;0m from 100% 
  to 0%, effectively disabling it without actually [1mundefining[22;0m it (however, 
  because it is not [1mfall-through[22;0m, it will still block other triggers of lower 
  [1mpriority[22;0m).  

  See: [1mmacros[22;0m, [1mtriggers[22;0m, [1mpatterns[22;0m, [1mevaluation[22;0m, [1mattributes[22;0m, [1m/def[22;0m, [1m/list[22;0m, [1m/edmac[22;0m 

&/escape

/escape

  [1mFunction[22;0m usage: 

  [1mESCAPE[22;0m(<[4mmetacharacters[24m>, <[4mstring[24m>)

  Command usage: 

  [1m/ESCAPE[22;0m <[4mmetacharacters[24m> <[4mstring[24m>
  ____________________________________________________________________________

  Echoes (in command form) or returns (in [1mfunction[22;0m form) <[4mstring[24m>, with any 
  <[4mmetacharacters[24m> or '\' characters contained in <[4mstring[24m> preceded by a '\' 
  character.  

  Example: 

  [1m/def[22;0m blue = [1m/def[22;0m [1m-a[22;0mCblue [1m-t[22;0m"$([1m/escape[22;0m " [1m%*[22;0m)"
  /blue * pages, "*" 

  When the second command executes, it will [1mexpand[22;0m to: 

  [1m/def[22;0m [1m-a[22;0mCblue [1m-t[22;0m"* pages, \"*\"" 

  See: [1mevaluation[22;0m 

&/not
&/eval
&eval
&eval()

eval()

  Function usage: 

  [1meval[22;0m(<[4mtext[24m> [, <[4mlevel[24m>])

  Command usage: 

  [1m/EVAL[22;0m [-s<[4mlevel[24m>] <[4mtext[24m>
  [1m/NOT[22;0m [-s<[4mlevel[24m>] <[4mtext[24m>
  ____________________________________________________________________________

  <[4mText[24m> is [1mevaluated[22;0m as a [1mmacro[22;0m body: it goes through [1msubstitution[22;0m, and is 
  executed in a new [1mscope[22;0m.  The return value of [1meval()[22;0m and [1m/eval[22;0m is that of 
  the last command in <[4mtext[24m>; the return value of [1m/not[22;0m is the logical negation 
  of return value of the last command in <[4mtext[24m>.  

  Positional parameters ([1m%1[22;0m, etc) are inherited from the caller.  

  [1mOptions[22;0m and arguments: 
  command: -s<[4mlevel[24m> 
  function: <[4mlevel[24m> 
          Expands the <[4mtext[24m> as if [1m%{sub}[22;0m were set to <[4mlevel[24m>.  By default, 
          [1meval[22;0m expands the <[4mtext[24m> as if [1m%{sub}[22;0m were "full", and echoes it if 
          [1m%{mecho}[22;0m is not "off".  

  Note: calling [1m/eval[22;0m with arguments from a [1mtrigger[22;0m could be dangerous.  If 
  not written carefully, such a [1mtrigger[22;0m could allow anyone with access to the 
  server to gain access to your tf or shell account (if they have not been 
  [1m/restrict[22;0med).  

  Example:
  command: [1m/def[22;0m showvar = [1m/eval[22;0m [1m/echo[22;0m [1m%{1}[22;0m is %%{[1m%{1}[22;0m}. 
  command: /showvar borg
  output: borg is on. 

  "[1m/Eval[22;0m -s0" can be useful when the argument is generated by an expansion.  
  For example, if you defined "[1m/def[22;0m do = [1m%{*}[22;0m, and then called "/do /echo 
  test", it would send "/echo test" to the server instead of executing it as a 
  tf command.  But if you defined "[1m/def[22;0m do = [1m/eval[22;0m -s0 [1m%{*}[22;0m", then "/do /echo 
  test" would execute "/echo test" as a tf command.  

  Note: Instead of [1m/not[22;0m, you should normally use the "/!<[4mcommand[24m>" syntax to 
  execute "/<[4mcommand[24m>" and negate its result.  [1m/not[22;0m evaluates its arguments, 
  which may be undesirable.  

  See: [1mevaluation[22;0m 

&/exit

/exit

  Usage: 

  [1m/EXIT[22;0m [<[4mn[24m>]
  ____________________________________________________________________________

  When called directly or indirectly during a [1m/load[22;0m, [1m/exit[22;0m aborts execution of 
  all enclosing macro bodies, and aborts <[4mn[24m> (default 1) enclosing [1m/load[22;0m's.  

  When called outside of a [1m/load[22;0m, [1m/exit[22;0m has no effect.  

  Example: one way to prevent a file from being loaded more than once is to 
  put commands like these at the beginning of the file: 

    [1m/if[22;0m (<[4mvariable[24m>) [1m/exit[22;0m%; [1m/endif[22;0m
    [1m/set[22;0m <[4mvariable[24m>=1

  ...where <[4mvariable[24m> is the name of the file or some other unique name.  

  See: [1m/load[22;0m, [1m/return[22;0m, [1m/break[22;0m, [1m/loaded[22;0m 

&/export

/export

  Usage: 

  [1m/EXPORT[22;0m <[4mvariable[24m>
  ____________________________________________________________________________

  If <[4mvariable[24m> is a global [1mvariable[22;0m, it becomes an environment [1mvariable[22;0m.  
  This makes <[4mvariable[24m> available to the environment for "[1m/sh[22;0m" and "[1m/quote[22;0m !". 

  Local [1mvariables[22;0m may not be exported.  

  See: [1menvironment[22;0m, [1mvariables[22;0m, [1m/setenv[22;0m 

&/expr

/expr

  Usage: 

  [1m/EXPR[22;0m <[4m[1mexpression[22;0m[24m>
  ____________________________________________________________________________

  Evaluates <[4m[1mexpression[22;0m[24m> and prints its value.  This almost the same as "[1m/eval[22;0m 
  [1m/echo[22;0m -- $$[<[4m[1mexpression[22;0m[24m>]", except that [1m{#}[22;0m and [1mpositional parameters[22;0m ({1}, 
  etc) are not defined.  If you neet to print a value of an expression that 
  uses positional parameters, use [1m/result[22;0m or [1mecho()[22;0m.  

  Example: 

  command: [1m/set[22;0m x=4
  command: [1m/expr[22;0m x * 2
  output: 8

  See: [1mexpressions[22;0m 

&/features

/features

  Usage: 

  [1m/FEATURES[22;0m [<[4mname[24m>]
  ____________________________________________________________________________

  With no arguments, [1m/features[22;0m prints a list of optional TF features, each 
  prefixed with "+" or "-" to indicate that it is enabled or disabled, 
  respectively.  

  With a <[4mname[24m> argument, [1m/features[22;0m returns 0 or 1 if the feature <[4mname[24m> is 
  disabled or enabled, respectively, in this instance of tf.  Case is 
  insignificant in <[4mname[24m>.  

    Feature           Meaning
    -------           -------
    256colors         256 color support
    core              If tf crashes, it can dump a core file
    float             Floating point arithmetic and functions
    ftime             [1mftime[22;0m() accepts % formatting
    history           /recall and /quote #
    IPv6              Internet Protocol version 6
    locale            allow alternate character sets and date formats
                      (see: [1mlocale[22;0m)
    MCCPv1            Mud Client Compression Protocol version 1 (see: [1mmccp[22;0m)
    MCCPv2            Mud Client Compression Protocol version 2 (see: [1mmccp[22;0m)
    process           /repeat and /quote
    SOCKS             SOCKS proxy
    ssl               Secure Sockets Layer
    subsecond         time is measured with subsecond accuracy
    TZ                honors the [1mTZ[22;0m variable

  Example: 

    [1m/if[22;0m (!features("ssl")) [1m/echo[22;0m -e warning: socket is not secure%; /endif
    

&/bg
&/fg

/fg

  Usage: 

  [1m/FG[22;0m [-nsq<>l] [-c<[4mN[24m>] [<[4mworld[24m>]
  [1m/BG[22;0m
  ____________________________________________________________________________

  Bring the [1msocket[22;0m associated with <[4mworld[24m> into the [1mforeground[22;0m.  The <[4mworld[24m> 
  must already be connected with the [1m/connect[22;0m command.  Any lines that arrived 
  while the [1msocket[22;0m was in the background will be displayed or counted in the 
  [1mmore[22;0m prompt, unless the -q option is given.  

  /fg [1mOptions:[22;0m 
  -n      no [1msocket[22;0m: put all [1msockets[22;0m in the [1mbackground[22;0m.  
  -s      suppress error messages.  
  -<      previous [1msocket[22;0m in cycle.  
  ->      next [1msocket[22;0m in cycle.  
  -c<[4mN[24m>   Repeat the -< or -> option <[4mN[24m> times.  
  -l      ignored.  
  -q      quiet: jump to the last screenful of text, instead of starting at 
          the same location you were at the last time the [1msocket[22;0m was in the 
          [1mforeground[22;0m.  

  If successful, [1m/fg[22;0m returns nonzero and invokes the WORLD [1mhook[22;0m; otherwise, it 
  returns 0.  

  By default, [1m/fg[22;0m draws a dividing line between old and new text.  If you 
  would prefer no dividing line, or clearing old text, this can be configured 
  with [1m%textdiv[22;0m.  

  [1m/bg[22;0m puts all [1msockets[22;0m in the [1mbackground[22;0m, and is equivalent to [1m/fg[22;0m -n.  By 
  default, [1m/bg[22;0m is bound to the ^] [1mkey[22;0m (not ESC, which is ^[) 

  See: [1m/connect[22;0m, [1mworlds[22;0m, [1msockets[22;0m, [1m%textdiv[22;0m, [1m%textdiv_str[22;0m.  

&finger.tf
&/finger

/finger

  Usage: 

  [1m/REQUIRE[22;0m finger.tf

  [1m/FINGER[22;0m [<[4muser[24m>][@<[4mhost[24m>]
  ____________________________________________________________________________

  Like unix finger, [1m/finger[22;0m reports information about <[4muser[24m> (default: all 
  users) on <[4mhost[24m> (default: localhost), assuming that <[4mhost[24m> is running a 
  standard finger daemon.  

  See: [1m/require[22;0m, [1mworlds[22;0m, [1msockets[22;0m 

&/for

/for

  Usage: 

  [1m/FOR[22;0m <[4mvariable[24m> <[4mstart[24m> <[4mend[24m> <[4mcommands[24m>
  ____________________________________________________________________________

  The <[4mvariable[24m> will take on all numeric values between <[4mstart[24m> and <[4mend[24m>, 
  inclusive.  The <[4mcommands[24m> will be executed once for each of the values.  If 
  <[4mend[24m> is less then <[4mstart[24m>, <[4mcommands[24m> will not be executed.  

  <[4mCommands[24m> are executed in a new [1mevaluation scope[22;0m.  This means, for example, 
  that a [1m/for[22;0m called from a [1mmacro[22;0m must use "%%{...}" and "%%;" instead of 
  "%{...}" and "%;" to have the [1msubstitutions[22;0m performed when the [1m/for[22;0m is 
  [1mexpanded[22;0m instead of when the calling [1mmacro[22;0m is [1mexpanded[22;0m.  

  Example: 

  Given the definition 

    [1m/def[22;0m countdown = [1m/for[22;0m i 0 %{1} say $$[%{1} - i]
    

  then the command "/countdown 10" would cause you to execute the commands 
  "say 10", "say 9", ...  "say 0".  Note that the "%{1}" is [1msubstituted[22;0m when 
  /countdown is [1mexpanded[22;0m, and the "$$" is replaced with "$".  The resulting 
  "$[10 - i]" is [1msubstituted[22;0m when [1m/for[22;0m is [1mexpanded[22;0m.  If /countdown used 
  "$[...]" instead of "$$[...]" in the <[4mcommands[24m>, it would be [1msubstituted[22;0m 
  when /countdown is [1mexpanded[22;0m, and you would repeat "10" 11 times.  If 
  /countdown used "%%{1}" or "{1}" instead of "%{1}" inside the [1mexpression[22;0m, it 
  would not be [1msubstituted[22;0m until [1m/for[22;0m was [1mexpanded[22;0m, so it would have the value 
  of [1m/for[22;0m's first argument (the string "i", which has numeric value 0), and 
  you would end up counting down from 0 to -10.  

  See: [1m/while[22;0m 

&ftime
&ftime()

ftime()

  [1mFunction[22;0m usage: 

  [1mftime[22;0m([<[4mformat[24m> [, <[4mtime[24m>]])
  ____________________________________________________________________________

  Returns a string formatted from an absolute system time <[4mtime[24m> (obtained 
  from [1mtime()[22;0m or [1mmktime()[22;0m) according to <[4mformat[24m>.  If <[4mtime[24m> is omitted, it 
  defaults to the current time.  If <[4mtime[24m> is out of range, ftime() returns an 
  empty string and prints an error message.  If <[4mformat[24m> is omitted, it 
  defaults to [1m%time_format[22;0m.  If <[4mformat[24m> is "@", a raw system time (e.g., 
  seconds since 1970-01-01 00:00:00 UTC) will be displayed.  Otherwise, each 
  "%" in <[4mformat[24m> describes a conversion: 
  %@      raw system time, in seconds, to the nearest microsecond 
          (nonstandard) 
  %.      microseconds since last whole second (nonstandard) 
  %a      abbreviated weekday name 
  %A      full weekday name 
  %b      abbreviated month name 
  %B      full month name 
  %c      [1mlocal[22;0m time and date representation 
  %d      day of month (01-31) 
  %F      ISO 8601 date format (equivalent to "%Y-%m-%d") 
  %H      hour on 24-hour clock (00-23) 
  %I      hour on 12-hour clock (01-12) 
  %j      day of year (001-366) 
  %m      month (01-12) 
  %M      minute (00-59) 
  %p      [1mlocal[22;0m equivalent of "AM" or "PM" 
  %s      raw system time, rounded down to the nearest whole second 
          (nonstandard) 
  %S      second (00-61) 
  %T      ISO 8601 time format (equivalent to "%H:%M:%S") 
  %U      week number of year, Sunday is first day of week (00-53) 
  %w      weekeday (0-6, Sunday is 0) 
  %W      week number of year, Monday is first day of week (00-53) 
  %x      [1mlocal[22;0m date representation 
  %X      [1mlocal[22;0m time representation 
  %y      year without century (00-99) 
  %Y      year with century 
  %Z      time zone name, if any 
  %%      "%" 
  Names and conversions labeled "local" may be affected by the setting of the 
  LC_TIME [1mlocale[22;0m category.  Additional "%" conversions may be supported by 
  your system, including 3-character conversions starting with "%E" and "%O"; 
  see your system's strftime() documentation for details.  All other 
  characters in <[4mformat[24m> are copied unmodified to the result.  

  The formats "%@" and "%s.%." do not give the same results if <[4mtime[24m> is 
  negative.  

  Example:
  command: [1m/expr[22;0m [1mftime[22;0m("Today is %a %b %d", [1mtime[22;0m())
  output: Today is Thu Jul 02

  See: [1mfunctions[22;0m, [1mtime()[22;0m, [1mlocale[22;0m, [1m%TZ[22;0m, [1m%time_format[22;0m, [1m%clock_format[22;0m.  

&/gag

/gag

  Usage: 

  [1m/GAG[22;0m [<[4mpattern[24m> [=<[4mresponse[24m>]]
  ____________________________________________________________________________

  Creates a [1mmacro[22;0m which will [1mtrigger[22;0m on text matching <[4m[1mpattern[22;0m[24m> and prevent it 
  from being displayed, optionally executing <[4mresponse[24m>.  

  With no arguments, [1m/gag[22;0m sets the flag [1m%{gag}[22;0m to 1 (on).  This flag enables 
  the [1mgag[22;0m [1mattribute[22;0m on [1mtriggers[22;0m.  It is on by default.  

  The matching style of the [1mgag[22;0m [1mpattern[22;0m is determined by [1m%{matching}[22;0m.  The 
  [1mpriority[22;0m of the [1mgag[22;0m is determined by [1m%{gpri}[22;0m.  These variables are examined 
  when the [1mgag[22;0m is defined, not when it is executed.  

  [1mGagged[22;0m lines from [1mbackground[22;0m worlds will not set the activity indicator on 
  the [1mstatus line[22;0m or call the activity [1mhook[22;0m.  

  If [1m/gag[22;0m does not create a new [1mmacro[22;0m, it returns 0.  Otherwise, it returns 
  the number of the new [1mmacro[22;0m (useful in [1m/undefn[22;0m and [1m/edit[22;0m).  

  [1m/gag[22;0m <[4mpattern[24m> [= <[4mresponse[24m>]
  is equivalent to
  [1m/def[22;0m [1m-ag[22;0m [1m-t[22;0m"<[4mpattern[24m>" [= <[4mresponse[24m>]. 

  See: [1mtriggers[22;0m, [1mpatterns[22;0m, [1mevaluation[22;0m, [1m%gag[22;0m, [1m/def[22;0m, [1m/nogag[22;0m 

&download
&/getfile_MUCK
&/getfile_LP
&/getfile_UNIX
&/getfile

/getfile

  Usage: 

  [1m/REQUIRE[22;0m filexfer.tf

  [1m/GETFILE_MUCK[22;0m <[4mfile[24m> [<[4mremote-file[24m>]
  [1m/GETFILE_LP[22;0m <[4mfile[24m> [<[4mremote-file[24m>]
  [1m/GETFILE_UNIX[22;0m <[4mfile[24m> [<[4mremote-file[24m>]
  ____________________________________________________________________________

  Downloads text <[4mremote-file[24m> from a MUCK, LP, or remote UNIX shell to <[4mfile[24m> 
  on the local host.  If <[4mremote-file[24m> is omitted, <[4mfile[24m> is used as the name 
  on both ends.  Do not use "wildcard" globbing characters in the file names.  

  When using [1m/getfile_UNIX[22;0m, an extra line of garbage may appear at the 
  beginning of the downloaded file unless you first disable remote echo with 
  "stty -echo".  

  Bug: if there is a log open for the [1mcurrent[22;0m world, it will be closed by 
  [1m/getfile[22;0m.  

  See: [1m/putfile[22;0m, [1m/log[22;0m 

&/grab

/grab

  Usage: 

  [1m/GRAB[22;0m <[4mtext[24m>
  ____________________________________________________________________________

  This command puts <[4mtext[24m> into the input buffer.  It is not really useful 
  from the normal command line, but is quite useful when called from a [1mmacro[22;0m 
  to redefine [1mmacros[22;0m, or perhaps when bound to a key to speed up part of a 
  line ([1mmacros[22;0m allow you to largely do what this would allow, however).  Any 
  text already in the input buffer is discarded.  

  Example: 

    [1m/def[22;0m reedit = [1m/grab[22;0m [1m/edit[22;0m [1m%1[22;0m = $[1m%1[22;0m

  If you had previously done "[1m/def[22;0m flail = :flails at his keyboard", the 
  command "/reedit flail" would place "[1m/edit[22;0m flail = :flails at his keyboard" 
  in the input buffer and allow you to edit it using the editing keys.  See 
  "[1mevaluation[22;0m" for details on how [1mmacros[22;0m like this work.  

  See: [1m/input[22;0m, [1mgeneral[22;0m 

&oldgrep
&grep.tf

/grep

  Usage: 

  [1m/REQUIRE[22;0m grep.tf

  [1m/FGREP[22;0m <[4mpattern[24m> <[4mcommand[24m>
  [1m/GREP[22;0m <[4mpattern[24m> <[4mcommand[24m>
  [1m/EGREP[22;0m <[4mpattern[24m> <[4mcommand[24m>
  ____________________________________________________________________________

  Executes <[4mcommand[24m> and prints only the output that matches <[4mpattern[24m> (which 
  must not contain spaces).  [1m/fgrep[22;0m prints lines that [4mcontain[24m the string 
  <[4mpattern[24m>; [1m/grep[22;0m prints lines that match the [1mglob[22;0m <[4mpattern[24m>; [1m/egrep[22;0m prints 
  lines that match the [1mregexp[22;0m <[4mpattern[24m>.  

  Remember to use "*" at each end of <[4mpattern[24m> to make [1m/grep[22;0m match lines that 
  [4mcontain[24m a piece that matches the [1mglob[22;0m <[4mpattern[24m>; without the "*"s, the 
  entire line must match.  

  Example: "/fgrep T'tiny.muck' /listworlds" lists all the worlds defined with 
  the -T'tiny.muck' option.  

  See: [1mtextutil.tf[22;0m, [1m/require[22;0m, [1mpatterns[22;0m, [1mexpressions[22;0m, [1mfunctions[22;0m 

&/man
&/help

/help

  Usage: 

  [1m/HELP[22;0m [<[4mtopic[24m>]
  ____________________________________________________________________________

  Displays help on the topic specified, or displays a quick summary of 
  available topics if no topic is given.  

  In the documentation, words or phrases in [1mthis format[22;0m are references to 
  other topics.  That is, a hyperlink in HTML, or something that can be used 
  as an argument to [1m/help[22;0m in [1mtf[22m.  

  Commands are described with the format "/COMMAND arguments".  Words in all 
  caps must be spelled exactly as shown (but do not need to be capitalized).  
  Arguments in <[4mthis format[24m> (underlined angle brackets in /help, or italics 
  in HTML) can be given any value.  Arguments in [square brackets] may be 
  omitted.  The character | means "or".  For example, "[OFF|ON]" means you may 
  type "off", "on", or nothing.  

  Some help topics have punctuation in their names: variables begin with "%", 
  commands begin with "/", and functions end with "()".  A name with omitted 
  punctuation will usually match the same topic (e.g., "[1m/def[22;0m" and "[1mdef[22;0m" both 
  match the /def command topic), but sometime will match a different topic 
  (e.g., "[1m%MAIL[22;0m" matches the MAIL variable topic, but "[1mMAIL[22;0m" matches the MAIL 
  hook topic).  There are also (sub)topics for various tf syntax constructions 
  such as "[1m%{}[22;0m" and "[1m$()[22;0m".  

  For [1m/help[22;0m to work, the [1mvariable[22;0m [1m%TFHELP[22;0m must contain the name of the 
  helpfile.  It is set when TF is installed, and should not normally be 
  changed.  If the helpfile or the help index is not found, [1m/help[22;0m will not 
  function.  The help file is in ASCII with embedded ANSI display codes, so 
  can be read or printed by any program that can handle ANSI codes.  

#html
  The help documents are also available on the web at 
  [1mhttp://tinyfugue.sourceforge.net/help/[22;0m.  
#

  See: [1mindex[22;0m, [1mintro[22;0m, [1moptions[22;0m 

&/highlight
&/hilite

/hilite

  Usage: 

  [1m/HILITE[22;0m [<[4mpattern[24m> [= <[4mresponse[24m>]]
  ____________________________________________________________________________

  Creates a [1mmacro[22;0m which will [1mtrigger[22;0m on text matching <[4m[1mpattern[22;0m[24m> and display it 
  with the [1mhilite[22;0m [1mattribute[22;0m, optionally executing <[4mresponse[24m>.  

  With no arguments, [1m/hilite[22;0m sets the flag [1m%{hilite}[22;0m to 1 (on).  This flag 
  enables [1mhilite[22;0m and other [1mattributes[22;0m on [1mtriggers[22;0m.  It is on by default.  

  The [1mattribute[22;0m(s) for [1mhilite[22;0md text are determined by the [1m%{hiliteattr}[22;0m 
  [1mvariable[22;0m.  The default is bold ([1mhiliteattr[22;0m=B).  Colors are also available 
  (e.g., [1mhiliteattr[22;0m=Cgreen); see "[1mattributes[22;0m" and "[1mcolor[22;0m" for more 
  information.  

  The matching style of the [1mhilite[22;0m [1mpattern[22;0m is determined by [1m%{matching}[22;0m.  The 
  [1mpriority[22;0m of the [1mhilite[22;0m is determined by [1m%{hpri}[22;0m.  These [1mvariables[22;0m are 
  examined when the [1mhilite[22;0m is defined, not when it is executed.  

  If [1m/hilite[22;0m does not create a new [1mmacro[22;0m, it returns 0.  Otherwise, it returns 
  the number of the new [1mmacro[22;0m (useful in [1m/undefn[22;0m and [1m/edit[22;0m).  

  The [1mstandard library[22;0m also defines [1m/hilite_page[22;0m and [1m/hilite_whisper[22;0m which 
  [1mhilite[22;0m several different commonly used page and whisper formats.  

  [1m/hilite[22;0m <[4mpattern[24m> [=<[4mresponse[24m>]
  is equivalent to
  [1m/def[22;0m [1m-ah[22;0m [1m-t[22;0m"<[4mpattern[24m>" [=<[4mresponse[24m>]. 

  Example: 

    [1m/hilite[22;0m {*} tried to kill you!

  With the default settings, any line matching that pattern will appear bold.  

  To hilite messages generated by tf, see [1mhooks[22;0m.  

  See: [1mtriggers[22;0m, [1mpatterns[22;0m, [1mattributes[22;0m, [1m/def[22;0m, [1m/nohilite[22;0m, [1m/partial[22;0m 

&/histsize

/histsize

  Usage: 

  [1m/HISTSIZE[22;0m [-lig] [-w[<[4mworld[24m>]] [<[4msize[24m>]
  ____________________________________________________________________________

  [1mOptions:[22;0m 
  -l      local history 
  -i      input history 
  -g      global [1mhistory[22;0m (default) 
  -w<[4mworld[24m> 
          world history 

  If <[4msize[24m> is not given, [1m/histsize[22;0m reports the maximum number of lines that 
  can be stored in the specified [1mhistory[22;0m.  

  If <[4msize[24m> is given, [1m/histsize[22;0m changes the maximum size of the specified 
  [1mhistory[22;0m to <[4msize[24m>.  If the new size is less than the old size, the oldest 
  lines will be lost immediately.  If the new size is greater than the old 
  size, no more old lines will be lost until enough new lines are added to 
  reach the new size.  

  [1m/histsize[22;0m returns 0 for failure, and the size of the [1mhistory[22;0m otherwise.  

  The [1m%{histsize}[22;0m [1mvariable[22;0m can be used to set the default size of world 
  histories before they are created.  

  See: [1mhistory[22;0m, [1m%histsize[22;0m 

&/hook

/hook

  Usage: 

  [1m/HOOK[22;0m <[4mevent[24m>[ <[4mpattern[24m>] [= <[4mbody[24m>]
  [1m/HOOK[22;0m [OFF|ON]
  ____________________________________________________________________________

  Creates a [1mmacro[22;0m which will execute <[4mbody[24m> when <[4mevent[24m> occurs and the 
  event's arguments match the optional <[4mpattern[24m>.  The <[4mevent[24m> may be a single 
  event or a list of events separated by '|'.  If omitted, <[4mpattern[24m> will 
  default to "*".  

  [1m/hook[22;0m with no arguments displays the state of the [1m%{hook}[22;0m flag.  [1m/hook[22;0m with 
  an argument of ON or OFF sets the [1m%{hook}[22;0m flag, which determines if [1mhooks[22;0m 
  will execute their associated [1mmacros[22;0m.  

  The matching style of the [1mhook[22;0m pattern is determined by [1m%{matching}[22;0m.  This 
  [1mvariable[22;0m is examined when the [1mhook[22;0m is defined, not when it is executed.  

  Defining a [1mhook[22;0m will not replace an existing [1mhook[22;0m on the same event, but 
  rather creates an additional [1mhook[22;0m [1mmacro[22;0m on the event.  The [1mmacro[22;0m or [1mmacros[22;0m 
  to be executed are chosen by the normal [1mpriority[22;0m rules.  

  See the section "[1mhooks[22;0m" for details on [1mhook[22;0m operation, a list of event 
  names, and examples.  

  If [1m/hook[22;0m does not create a new [1mmacro[22;0m, it returns 0.  Otherwise, it returns 
  the number of the new [1mmacro[22;0m (useful in [1m/undefn[22;0m and [1m/edit[22;0m).  

  [1m/hook[22;0m <[4mevent[24m>[ <[4mpattern[24m>] [=<[4mresponse[24m>]
  is equivalent to
  [1m/def[22;0m [1m-h[22;0m"<[4mevent[24m>[ <[4mpattern[24m>]" [=<[4mresponse[24m>]. 

  Example: 

    [1m/hook[22;0m MAIL = [1m/sh[22;0m mutt

  will automatically invoke "mutt" to read mail when it arrives.  

  See: [1mhooks[22;0m, [1mmacros[22;0m, [1mevaluation[22;0m, [1mpatterns[22;0m, [1m/def[22;0m, [1m/unhook[22;0m 

&/if
&/then
&/elseif
&/else
&/endif
&/if

/if

  Usage: 

  [1m/IF[22;0m ([4mexpr[24m) [4mlist[24m [ [1m/ELSEIF[22;0m ([4mexpr[24m) [4mlist[24m ]...  [ [1m/ELSE[22;0m [4mlist[24m ] [1m/ENDIF[22;0m
  [1m/IF[22;0m [4mlist[24m [1m/THEN[22;0m [4mlist[24m [ [1m/ELSEIF[22;0m [4mlist[24m [1m/THEN[22;0m [4mlist[24m ]...  [ [1m/ELSE[22;0m [4mlist[24m ] [1m/ENDIF[22;0m
  ____________________________________________________________________________

  <[4mList[24m> is any list of commands.  The return value of a <[4mlist[24m> is the return 
  value of the last command executed in the <[4mlist[24m>.  Note that each <[4mlist[24m> 
  must be terminated by "[1m%;[22;0m".  

  <[4mexpr[24m> is any [1mexpression[22;0m, and must be surrounded by parentheses.  

  The <[4mlist[24m> or <[4mexpr[24m> following the [1m/IF[22;0m is executed or evaluated.  If the 
  result is non-zero, the next <[4mlist[24m> is executed.  Otherwise, this is 
  repeated for each [1m/ELSEIF[22;0m.  If none of the [1m/IF[22;0m or [1m/ELSEIF[22;0m <[4mlist[24m>s or <[4mexpr[24m>s 
  return non-zero, the [1m/ELSE[22;0m <[4mlist[24m> is executed if there is one.  

  The return value of the [1m/IF[22;0m...[1m/ENDIF[22;0m statement is undefined.  

  [1m/IF[22;0m (expr) body%; [1m/ENDIF[22;0m 
  is equivalent to
  [1m/IF[22;0m [1m/TEST[22;0m expr%; [1m/THEN[22;0m body%; [1m/ENDIF[22;0m 
  except that in the former, <[4mexpr[24m> does not undergo macro body [1msubstitution[22;0m.  

  When [1m/IF[22;0m is used on the command line, "[1m%;[22;0m" command separation is done even 
  if [1m%sub[22;0m=off.  Of course, full substitution will be done if [1m%sub[22;0m=full.  

  If <[4mlist[24m> is a server (mud) command, the condition being tested is whether 
  the command is sent successfully; that is, whether there is a [1mcurrent 
  socket[22;0m.  TF has no way of knowing how the server deals with the command or 
  what is considered "success" for a server command, and tf does not wait for 
  a server response which will be delayed by network latency.  So, doing 
  something like "[1m/if[22;0m rob corpse%; /then ..." will not have the effect you 
  probably want.  To achieve that effect, you should define a [1mtrigger[22;0m on each 
  of the possible server responses, before you send your command.  

  Example: 

    [1m/if[22;0m ([1mTERM[22;0m !~ "dumb") [1m/visual[22;0m on%; [1m/endif[22;0m

  will do "[1m/visual[22;0m on" if your [1m%{TERM}[22;0m is not "dumb".  

  See: [1mevaluation[22;0m, [1mexpressions[22;0m, [1m/test[22;0m, [1m/def -E[22;0m, 

&builtins
&commands
&index

index

  Commands marked with '+' are new in the current version.  Commands marked 
  with '*' have changed significantly in the current version.  

  *[1mADDWORLD[22;0m      *[1mFG[22;0m             [1mLISTVAR[22;0m        [1mREPLACE[22;0m        [1mTOGGLE[22;0m        
  *[1mAT[22;0m             [1mFINGER[22;0m         [1mLISTWORLDS[22;0m    *[1mRESTRICT[22;0m       [1mTR[22;0m            
   [1mBAMF[22;0m           [1mFOR[22;0m            [1mLOAD[22;0m           [1mRETURN[22;0m         [1mTRIG[22;0m          
   [1mBEEP[22;0m           [1mGAG[22;0m            [1mLOCALECHO[22;0m     +[1mRUNTIME[22;0m       *[1mTRIGGER[22;0m       
  *[1mBIND[22;0m           [1mGETFILE[22;0m        [1mLOG[22;0m            [1mSAVE[22;0m           [1mUNBIND[22;0m        
   [1mBREAK[22;0m          [1mGRAB[22;0m           [1mmapping[22;0m        [1mSAVEWORLD[22;0m      [1mUNDEF[22;0m         
   [1mCAT[22;0m            [1mHELP[22;0m          *[1mMORE[22;0m          *[1mSEND[22;0m           [1mUNDEFN[22;0m        
   [1mCHANGES[22;0m        [1mHILITE[22;0m         [1mNOHILITE[22;0m       [1mSET[22;0m            [1mUNDEFT[22;0m        
  *[1mCONNECT[22;0m        [1mHISTSIZE[22;0m       [1mPARTIAL[22;0m        [1mSETENV[22;0m         [1mUNHOOK[22;0m        
   [1mDC[22;0m             [1mHOOK[22;0m          *[1mPASTE[22;0m          [1mSH[22;0m             [1mUNSET[22;0m         
  *[1mDEF[22;0m            [1mIF[22;0m            *[1mPS[22;0m             [1mSHIFT[22;0m          [1mUNTRIG[22;0m        
  *[1mDOKEY[22;0m          [1mINPUT[22;0m          [1mPURGE[22;0m          [1mspelling[22;0m       [1mUNWORLD[22;0m       
  *[1mECHO[22;0m           [1mKILL[22;0m           [1mPURGEWORLD[22;0m     [1mSUB[22;0m            [1mVERSION[22;0m       
  *[1mEDIT[22;0m           [1mLCD[22;0m            [1mPUTFILE[22;0m        [1mSUBSTITUTE[22;0m     [1mWATCHDOG[22;0m      
   [1mESCAPE[22;0m         [1mLET[22;0m           *[1mQUIT[22;0m           [1mSUSPEND[22;0m        [1mWATCHNAME[22;0m     
  *[1mEVAL/NOT[22;0m      +[1mLIMIT[22;0m         *[1mQUOTE[22;0m          [1mTELNET[22;0m         [1mWHILE[22;0m         
   [1mEXIT[22;0m           [1mlist commands[22;0m  [1mquoter.tf[22;0m      [1mTEST[22;0m           [1mWORLD[22;0m         
   [1mEXPORT[22;0m         [1mLIST[22;0m          *[1mRECALL[22;0m        *[1mtextutil.tf[22;0m                  
   [1mEXPR[22;0m          *[1mLISTSOCKETS[22;0m    [1mRECORDLINE[22;0m     [1mTICK[22;0m                         
  +[1mFEATURES[22;0m       [1mLISTSTREAMS[22;0m   *[1mREPEAT[22;0m         [1mTIME[22;0m                         

  See also: [1mintro[22;0m, [1mtopics[22;0m 

&/input

/input

  Usage: 

  [1m/INPUT[22;0m <[4mtext[24m>
  ____________________________________________________________________________

  Enters <[4mtext[24m> into the input buffer as if it had been typed at the keyboard, 
  without deleting the current contents of the input buffer.  

  [1m/Input[22;0m is perhaps most useful in combination with [1m/bind[22;0m, to create short key 
  sequences that expand to longer text.  For example, if you have this 
  binding: 

  [1m/bind[22;0m ^[oj = [1m/input[22;0m OliverJones 

  and then type "page ^[oj = snausages!" at the keyboard, it will appear in 
  the input window as "page OliverJones = snausages!".  

  See: [1m/bind[22;0m, [1m/grab[22;0m 

&/ismacro

/ismacro

  Usage: 

  [1m/ISMACRO[22;0m <[4mmacro-options[24m>
  ____________________________________________________________________________

  If <[4mmacro-options[24m> matches one or more existing [1mmacros[22;0m, [1m/ismacro[22;0m returns the 
  number of the last matching [1mmacro[22;0m; otherwise, [1m/ismacro[22;0m returns 0.  
  <[4mMacro-options[24m> may include any of the options accepted by [1m/list[22;0m.  If -m is 
  not specified, [1m%{matching}[22;0m is used.  

  Example: 

    [1m/if[22;0m /!ismacro -b"^X*"%; [1m/then[22;0m [1m/bind[22;0m ^X = [1m/foobar[22;0m%; [1m/endif[22;0m

  See: [1m/list[22;0m, [1mmacros[22;0m 

&/isvar

/isvar

  Usage: 

  [1m/ISVAR[22;0m <[4mname[24m>
  ____________________________________________________________________________

  Returns 1 if [1mvariable[22;0m <[4mname[24m> is set, 0 otherwise.  

  Example: 

    [1m/if[22;0m (!isvar('LANG')) [1m/set[22;0m LANG=en_US%; [1m/endif[22;0m

  See: [1m/listvar[22;0m, [1mvariables[22;0m 

&/kill

/kill

  Usage: 

  [1m/KILL[22;0m <[4mpid[24m>... 
  ____________________________________________________________________________

  For each <[4mpid[24m> given, [1m/kill[22;0m terminates the corresponding [1mprocess[22;0m ([1m/quote[22;0m or 
  [1m/repeat[22;0m command).  The pid of a [1mprocess[22;0m can be determined from the return 
  value of the [1m/quote[22;0m or [1m/repeat[22;0m, the [1m/ps[22;0m command, or a PROCESS [1mhook[22;0m.  

  Bug: [1m/kill[22;0m on a pending [1m/quote[22;0m ! will block until the shell process exits.  
  The block can be broken with an interrupt.  

  See: [1mprocesses[22;0m, [1m/quote[22;0m, [1m/repeat[22;0m, [1m/ps[22;0m 

&/cd
&/pwd
&/lcd

/lcd

  Usage: 

  [1m/LCD[22;0m [<[4mdir[24m>]
  [1m/CD[22;0m [<[4mdir[24m>]
  [1m/PWD[22;0m
  ____________________________________________________________________________

  [1m/lcd[22;0m and [1m/cd[22;0m change to a new working directory.  If <[4mdir[24m> is omitted with 
  [1m/lcd[22;0m, the current directory is displayed (if supported on your system).  If 
  <[4mdir[24m> is omitted with [1m/cd[22;0m, [1m%{HOME}[22;0m is assumed.  

  The <[4mdir[24m> name is expanded as described under "[1mfilenames[22;0m".  

  [1m/pwd[22;0m displays the current working directory (if supported on your system).  

&/let

/let

  Usage: 

  [1m/LET[22;0m <[4mname[24m>=<[4mvalue[24m>
  [1m/LET[22;0m <[4mname[24m> <[4mvalue[24m>
  ____________________________________________________________________________

  Assigns <[4mvalue[24m> to [1mvariable[22;0m <[4mname[24m> in the current local [1mscope[22;0m.  Can only be 
  used during [1mmacro[22;0m [1mexpansion[22;0m.  The [1mvariable[22;0m will be destroyed when the [1mscope[22;0m. 
  in which it was created exits.  

  Note to lisp users: this is nothing like lisp's let.  

  See: [1m/set[22;0m, [1mvariables[22;0m 

&/limit
&/relimit
&/unlimit

/limit

  Usage: 

  [1m/LIMIT[22;0m [-v] [-a] [-m<[4mstyle[24m>] [<[4mpattern[24m>]
  [1m/RELIMIT[22;0m
  [1m/UNLIMIT[22;0m
  ____________________________________________________________________________

  /Limit redraws the window, showing only lines that match <[4m[1mpattern[22;0m[24m>.  It is 
  then possible to scroll forward and backward within the "limited" window.  
  The limit affects only the current screen, and stays in effect until 
  [1m/unlimit[22;0m is called.  

  /Limit [1moptions:[22;0m 
  -v      show only lines that [4mdon't[24m match <[4m[1mpattern[22;0m[24m> 
  -a      show only lines that have attributes 
  -m<[4mstyle[24m> 
          use matching style ([1msimple[22;0m, [1mglob[22;0m, or [1mregexp[22;0m), instead of the default 
          [1m%{matching}[22;0m.  

  If <[4m[1mpattern[22;0m[24m> is given, only lines in the given range that match <[4m[1mpattern[22;0m[24m> 
  will be recalled.  The matching style is determined by the -m option if 
  given, [1m%{matching}[22;0m otherwise.  By default, the @more [1mstatus field[22;0m does not 
  count lines that are omitted by [1m/limit[22;0m.  

  With no options or arguments, [1m/limit[22;0m returns 1 if a limit is in effect, 0 if 
  not.  

  [1m/unlimit[22;0m disables the [1m/limit[22;0m so all lines are displayed.  

  During [1m/limit[22;0m, scrolling to any point, including the bottom, results in a 
  More prompt that shows the number of lines (possibly 0) below the status 
  line.  In this state, [1m/unlimit[22;0m will leave the bottom visible line where it 
  is, and redraw the unlimited lines above it.  Thus, you can use [1m/limit[22;0m to 
  find a line you are interested in, use the scrolling keys to position that 
  line at the bottom of the window, then [1m/unlimit[22;0m to see the context of that 
  line.  But if you attempt to scroll [4mpast[24m the bottom during [1m/limit[22;0m, the More 
  prompt changes to "LIMIT ON"; in this state, [1m/unlimit[22;0m will redraw with the 
  previously invisible last line at the bottom of the screen.  

  [1m/relimit[22;0m repeats the last [1m/limit[22;0m.  

  The default keybinding ^[L toggles the last limit off and on.  

  See: [1m/recall[22;0m 

&/listbind
&/listdef
&/listgag
&/listhilite
&/listhook
&/listtrig
&/list

/list

  Usage: 

  [1m/LIST[22;0m [-s] [<[4mmacro-options[24m>] [<[4mname[24m>] [= <[4mbody[24m>]
  ____________________________________________________________________________

  Lists [1mmacros[22;0m having all the specified options.  Except for "-s", each option 
  is compared against a [1mmacro[22;0m's option, and the [1mmacro[22;0m selected only if the 
  options match.  Omitted options are "don't care", and will not be used in 
  the comparison.  Thus, with no arguments, [1m/list[22;0m will list all non-[1minvisible[22;0m 
  [1mmacros[22;0m.  

#list options
  [1mOptions:[22;0m 
  -s      List [1mmacros[22;0m in short format.  
  -S      Sort [1mmacros[22;0m by name.  
  -m<[4mmatching[24m> 
          Determines matching style used for comparison of string fields 
          ([1mtrigger[22;0m, keybinding, keyname, [1mhook[22;0m, worldtype, name, and body).  
          This is [4mnot[24m compared against the -m options of [1mmacros[22;0m.  If omitted, 
          the style is determined by [1m%{matching}[22;0m.  
  -t<[4mpattern[24m> 
  -b<[4mpattern[24m> 
  -B<[4mpattern[24m> 
  -E<[4mpattern[24m> 
  -T<[4mpattern[24m> 
          Matches [1mmacros[22;0m with a corresponding [1m/def[22;0m option whose 
          option-argument matches <[4mpattern[24m>.  <[4mpattern[24m>.  An option with no 
          pattern matches all [1mmacros[22;0m that have that option, regardless of the 
          value of the option-argument.  A "{}" [1mglob[22;0m pattern or "^$" [1mregexp[22;0m 
          can be used to match [1mmacros[22;0m that [4mdon't[24m have that option, 
  -h["<[4mevent[24m>[ <[4mpattern[24m>]"] 
          Matches [1mmacros[22;0m with [1mhooks[22;0m matching <[4mevent[24m> and <[4mpattern[24m>.  "-h" by 
          itself matches all non-empty [1mhooks[22;0m; "-h0" matches only [1mmacros[22;0m 
          without [1mhooks[22;0m.  
  -a<[4mattrs[24m> 
          Matches [1mmacros[22;0m having one or more of the display [1mattributes[22;0m in 
          <[4mattrs[24m>.  
  -P<[4mpart[24m><[4mattrs[24m> 
          Matches [1mmacros[22;0m having a [1m-P[22;0m<[4mpart[24m> with one or more of the display 
          [1mattributes[22;0m in <[4mattrs[24m>.  
  -i      Matches invisible [1mmacros[22;0m as well as normal [1mmacros[22;0m.  
  -I      Matches only invisible [1mmacros[22;0m.  
  <[4mname[24m>  A pattern that [1mmacro[22;0m names must match.  The [1mglob[22;0m pattern "{}" or 
          [1mregexp[22;0m "^$" will match only [1mmacros[22;0m without names.  If <[4mname[24m> starts 
          with "#", it is compared against macro numbers, instead of as a 
          pattern against macro names.  
  = <[4mbody[24m> 
          <[4mbody[24m> is a pattern that macro bodies must match.  The [1mglob[22;0m pattern 
          "{}", or the [1mregexp[22;0m "^$" or the [1msimple pattern[22;0m "" will match 
          bodyless [1mmacros[22;0m only.  
#

  Other options allowed by [1m/def[22;0m may be used with [1m/list[22;0m, and are compared 
  directly to macros.  

  The return value of [1m/list[22;0m is the number of the last [1mmacro[22;0m listed, or 0 if no 
  [1mmacros[22;0m were listed (because of error or none matched the specified options). 

  The standard library also defines the [1mmacros[22;0m [1m/listbind[22;0m, [1m/listdef[22;0m, [1m/listgag[22;0m, 
  [1m/listhilite[22;0m, [1m/listfullhilite[22;0m, [1m/listpartial[22;0m, [1m/listhook[22;0m, and [1m/listtrig[22;0m, which 
  list [1mmacros[22;0m of the appropriate type.  

  Example: 

      [1m/list[22;0m -mregexp -n0 -t -aurh ^foo =

  will list all [1mmacros[22;0m whose names begin with "foo"; have a [1mtrigger[22;0m; are not 
  multi-shots; have any of the underline, reverse, or [1mhilite[22;0m [1mattributes[22;0m; and 
  have an empty body.  

  To list functions for named keys, try "[1m/list[22;0m -i key_*".  

  See: [1mmacros[22;0m, [1mtriggers[22;0m, [1mpatterns[22;0m, [1mattributes[22;0m, [1mlibrary[22;0m, [1m/def[22;0m 

&/car
&/cdr
&/cadr
&/cddr
&/caddr
&/cdddr
&/length
&/reverse
&/mapcar
&/maplist
&/remove
&/unique
&lisp
&lisp.tf
&list
&list commands

list commands

  Usage: 

  [1m/REQUIRE[22;0m lisp.tf
  ____________________________________________________________________________

  These commands operate on lists of words, and are similar to those in lisp.  
  They all give their results with [1m/echo[22;0m, and are intended to be used in 
  [1m$(...) command substitution[22;0m to capture the result.  

  [1m/car[22;0m <[4mlist[24m> 
          Echo first word.  (Same as [1m/first[22;0m).  
  [1m/cdr[22;0m <[4mlist[24m> 
          Echo all words after first.  (Same as [1m/rest[22;0m).  
  [1m/cadr[22;0m <[4mlist[24m> 
          Echo second word.  
  [1m/cddr[22;0m <[4mlist[24m> 
          Echo all words after second.  
  [1m/caddr[22;0m <[4mlist[24m> 
          Echo third word.  
  [1m/cdddr[22;0m <[4mlist[24m> 
          Echo all words after third.  

  [1m/length[22;0m <[4mlist[24m> 
          Echo number of words in <[4mlist[24m>.  

  [1m/reverse[22;0m <[4mlist[24m> 
          Reverse the order of the words in <[4mlist[24m>.  

  [1m/mapcar[22;0m <[4mcmd[24m> <[4mlist[24m> 
          Execute "<[4mcmd[24m> <[4mword[24m>" for each word in <[4mlist[24m>.  
  [1m/maplist[22;0m <[4mcmd[24m> <[4mlist[24m> 
          Execute "<[4mcmd[24m> <[4mlist[24m>" repeatedly, removing the first word from 
          <[4mlist[24m> each time, until <[4mlist[24m> is empty.  

  [1m/remove[22;0m <[4mword[24m> <[4mlist[24m> 
          Echo <[4mlist[24m> with all occurrences of <[4mword[24m> removed.  

  [1m/unique[22;0m <[4mlist[24m> 
          Remove all duplicate words from <[4mlist[24m>.  Note: [1m/unique[22;0m is very slow 
          on long lists.  

  See: [1m/nth[22;0m 

&/listsockets

/listsockets

  Usage: 

  [1m/LISTSOCKETS[22;0m [-sn] [-m<[4mstyle[24m>] [-S<[4mfield[24m>] [-T<[4mtype[24m>] [<[4mname[24m>]
  ____________________________________________________________________________

  Lists the [1msockets[22;0m to which TinyFugue is connected.  

  [1mOptions[22;0m and arguments: 
  -s      short form, list only world names 
  -n      print host and port in numeric form 
  -m<[4mstyle[24m> 
          Use <[4mstyle[24m> for [1mpattern matching[22;0m in other options (default: 
          [1m%{matching}[22;0m).  
  -S<[4mfield[24m> 
          Sort sockets by <[4mfield[24m>.  <[4mField[24m> may be "name", "type", 
          "character", "host", "port", "lines", "idle", or "-" (don't sort; 
          this is the default).  Only the first character is necessary.  
  -T<[4mtype[24m> 
          list only worlds with a type matching the [1mpattern[22;0m <[4mtype[24m>.  
  <[4mname[24m>  list only worlds with a name matching the [1mpattern[22;0m <[4mname[24m>.  

  The output will look something like this (unless the -s option is given): 

       LINES IDLE TYPE      NAME            HOST                       PORT
     10+  48  13h tiny.muck Cave            tcp.com                    2283
  *  foregnd   1m tiny.mush DeepSeas        muds.okstate.edu           6250
           0   7s telnet    whitehouse.gov, whitehouse.gov             smtp
   ?       0  15s tiny      SlowMUD         slow.machine.com           4201

  The columns and their meanings are: 
  unlabeled first column 
          "*" marks the [1mcurrent[22;0m socket.  
  unlabeled second column 
          the state of the [1msocket[22;0m is one of: 
          !       dead 
          ?       hostname lookup or network connection is incomplete 
          C/c     an established normal connection 
          S/s     an established connection currently in telnet subnegotiation 
          X/x     an established [1mSSL[22;0m connection 
          O       an open [1mconnectionless[22;0m socket 
          A lowercase state character indicates the connection is using [1mMCCP[22;0m.  
  unlabeled third column 
          "P" if the connection is [1mproxied[22;0m 
  LINES   for a [1mbackground[22;0m [1msocket[22;0m, the number of old (seen) and new (unseen) 
          lines past the bottom of the [1msocket[22;0m's window (ignoring any [1mlimit[22;0m 
          that may be in effect on that window); or, "foregnd" for a 
          [1mforeground[22;0m [1msocket[22;0m.  
  IDLE    how long since the last text was received on the [1msocket[22;0m.  
  TYPE    the type of the world (set with [1m/addworld[22;0m -T).  
  NAME    the name of the world associated with the [1msocket[22;0m.  
  HOST    the host to which the [1msocket[22;0m is connected.  
  PORT    the port to which the [1msocket[22;0m is connected.  

  The return value of [1m/listsockets[22;0m is the number of sockets listed.  

  See: [1msockets[22;0m, [1m%background[22;0m, [1m/connect[22;0m, [1m/fg[22;0m, [1mnactive()[22;0m, [1midle()[22;0m 

&/liststreams

/liststreams

  Usage: 

  [1m/LISTSTREAMS[22;0m
  ____________________________________________________________________________

  Lists [1mtfio streams[22;0m opened by [1mtfopen()[22;0m.  The [1mtfin[22;0m, [1mtfout[22;0m, and [1mtferr[22;0m [1mstreams[22;0m 
  are not included.  

  The columns and their meanings are: 
  HANDLE  The handle returned by [1mtfopen()[22;0m.  
  MODE    The mode argument given to [1mtfopen()[22;0m.  
  FLUSH   Whether automatic flushing is enabled.  See [1mtfflush()[22;0m.  
  NAME    The name argument, if any, given to [1mtfopen()[22;0m.  Files of mode "q" do 
          not need a name, but you may wish to give them one anyway so it 
          appears here.  

  The return value of [1m/liststreams[22;0m is the number of open streams listed.  

  See: [1mtfio[22;0m 

&/listvar

/listvar

  Usage: 

  [1m/LISTVAR[22;0m [-m<[4mmatching[24m>] [-gxsv] [<[4mname[24m> [<[4mvalue[24m>]]
  ____________________________________________________________________________

  [1mOptions:[22;0m 
  -m<[4mmatching[24m> 
          Determines matching style used for comparison of <[4mname[24m> and <[4mvalue[24m>. 
          If omitted, the style is determined by [1m%{matching}[22;0m.  
  -g      List only global (unexported) variables.  
  -x      List only variables that are exported to the environment.  
  -s      Short format: list variable names only.  
  -v      List values only.  

  [1m/Listvar[22;0m lists values of [1mvariables[22;0m whose name and value match <[4mname[24m> and 
  <[4mvalue[24m> according to <[4mmatching[24m>, sorted by name.  If neither -g nor -x is 
  given, global and environment variables are listed.  

  The return value of [1m/listvar[22;0m is the number of variables listed.  

  See: [1mvariables[22;0m, [1m/set[22;0m, [1m/setenv[22;0m, [1m/export[22;0m, [1m/let[22;0m, [1m/unset[22;0m 

&/listworlds

/listworlds

  Usage: 

  [1m/LISTWORLDS[22;0m [-cus] [-m<[4mstyle[24m>] [-S<[4mfield[24m>] [-T<[4mtype[24m>] [<[4mname[24m>]
  ____________________________________________________________________________

  Lists world definitions.  

  [1mOptions[22;0m and arguments: 
  -m<[4mstyle[24m> 
          Use <[4mstyle[24m> for [1mpattern matching[22;0m of <[4mtype[24m> and <[4mname[24m> patterns.  
          (default: [1m%{matching}[22;0m).  
  -s      Display short format (world names only).  
  -c      Display command format (including passwords).  
  -S<[4mfield[24m> 
          Sort worlds by <[4mfield[24m>.  <[4mField[24m> may be "name" (the default), 
          "type", "character", "host", "port", or "-" (don't sort).  Only the 
          first character is necessary.  
  -u      Include unnamed temporary worlds in the listing.  
  -T<[4mtype[24m> 
          List only worlds with a type matching the [1mpattern[22;0m <[4mtype[24m>.  
  <[4mname[24m>  List only worlds with a name matching the [1mpattern[22;0m <[4mname[24m>.  

  If neither -s nor -c are given, a table format is used, and passwords are 
  not shown.  The return value of [1m/listworlds[22;0m is the number of worlds listed.  

  See: [1mworlds[22;0m, [1mpatterns[22;0m 

&/loadbind
&/loaddef
&/loadgag
&/loadhilite
&/loadtrig
&/loadhook
&/loadworld
&/require
&/loaded
&/load

/load

  Usage: 

  [1m/LOAD[22;0m [-q] <[4mfile[24m>
  [1m/REQUIRE[22;0m [-q] <[4mfile[24m>

  [1m/LOADED[22;0m <[4mtoken[24m>
  ____________________________________________________________________________

  [1m/Load[22;0m and [1m/require[22;0m both read and execute commands from <[4mfile[24m>.  They are 
  identical, except that if <[4mfile[24m> calls [1m/loaded[22;0m and has already been read 
  once, [1m/require[22;0m will not read it again (but the LOAD message/[1mhook[22;0m will still 
  be displayed/called).  

  "[1m/Loaded[22;0m <[4mtoken[24m>" should be the first command in a file that is designed to 
  be loaded only once with [1m/require[22;0m.  <[4mToken[24m> should be a string that does not 
  contain space or [1mglob[22;0m metacharacters, and is different than the token used 
  by any other [1m/loaded[22;0m call.  The file's full name is usually a good choice 
  for <[4mtoken[24m>.  

  [1mOptions:[22;0m 
  -q      Do not echo the "% Loading commands from <[4mfile[24m>" message in this 
          [1m/load[22;0m call or any [1m/load[22;0m calls in <[4mfile[24m>.  (but the [1mLOAD hook[22;0m will 
          still be called).  

  The file may contain any legal TinyFugue commands.  Blank lines and lines 
  beginning with ';' or '#' are ignored.  Any leading whitespace on a line is 
  stripped.  Any line ending in '\' will have the following line joined to it 
  (after leading spaces are stripped).  A '%' preceding a '\' eliminates its 
  special meaning.  

  The <[4mfile[24m> name is expanded as described under "[1mfilenames[22;0m".  

  If the [1mCOMPRESS_SUFFIX[22;0m and [1mCOMPRESS_READ[22;0m [1mmacros[22;0m are defined, the file will 
  be automatically uncompressed if needed.  

  If the expanded filename is not an absolute path name, TF will search first 
  in the current directory (which can be changed with [1m/lcd[22;0m), and then in the 
  list of directories named by [1m%{TFPATH}[22;0m.  If [1m%{TFPATH}[22;0m is blank or unset, the 
  single directory named by [1m%{TFLIBDIR}[22;0m is used.  

  A [1m/load[22;0m may be aborted early with the [1m/exit[22;0m command in the file.  

  Loaded files may be given any name, but names ending in ".tf" are 
  recommended.  

  [1m/Load[22;0m and [1m/require[22;0m return 1 if successful (for [1m/require[22;0m, this includes not 
  needing to read the file), or 0 if not successful.  [1m/Loaded[22;0m does not return 
  if the file that calls it has already been loaded.  

  The standard [1mmacro[22;0m library also defines the commands [1m/loaddef[22;0m, [1m/loadbind[22;0m, 
  [1m/loadhilite[22;0m, [1m/loadgag[22;0m, [1m/loadtrig[22;0m, [1m/loadhook[22;0m, and [1m/loadworld[22;0m.  These [1mmacros[22;0m 
  will load from a [1mdefault file[22;0m if no file is specified.  

  See: [1mmacros[22;0m, [1mlibrary[22;0m, [1m/exit[22;0m, [1m/def[22;0m, [1m/save[22;0m, [1m/lcd[22;0m, [1mfilenames[22;0m, [1mcompression[22;0m 

&%always_echo
&always_echo
&/localecho

/localecho

  Usage: 

  [1m/LOCALECHO[22;0m [ON|OFF]
  ____________________________________________________________________________

  [1m/Localecho[22;0m with no arguments returns 1 if local echoing is enabled for the 
  [1mcurrent socket[22;0m, 0 otherwise.  TF echoes its input by default, unless the 
  server has negotiated otherwise.  

  [1m/Localecho[22;0m with an argument attempts to enable or disable echoing for the 
  [1mcurrent socket[22;0m.  If the server is not known to support TELNET protocol, 
  "[1m/localecho[22;0m [ON|OFF]" does nothing, and returns 0.  ON tells the server DONT 
  ECHO; if the server acknowledges (as it must according to TELNET protocol), 
  tf will echo its own input.  OFF tells the server to DO ECHO; if the server 
  acknowledges, tf will not echo its own input, expecting the server to do it. 
  The actual change of state takes place after the server agrees, which may be 
  delayed by network latency ("netlag").  

  Note that tf does not transmit input until a newline is pressed, and the 
  server can not echo it until it is received; so, with [1m/localecho[22;0m off, your 
  typing will not be visible until you hit return, at which time the server 
  may echo back the entire line.  

  Some mud servers use the ECHO option to disable local echo during password 
  entry.  Telnet servers, however, try to disable local echo for the entire 
  session, which would interfere with many useful tf features.  Hooks defined 
  in the [1mstandard library[22;0m use [1m/localecho[22;0m to override the telnet server 
  automatically.  

  [1m/Localecho[22;0m is intended to be called by library macros, and should not need 
  to be called by the user.  [1m/Localecho[22;0m obsoletes %{always_echo}.  

  The TELNET ECHO option is defined in RFC 857.  

  See: [1mprompts[22;0m, [1m%telopt[22;0m, [1m/telnet[22;0m 

&/log

/log

  Usage: 

  [1m/LOG[22;0m [-ligw[<[4mworld[24m>]] [OFF|ON|<[4mfile[24m>]
  ____________________________________________________________________________

  Enables or disables logging, or lists currently open log files.  An [-ligw] 
  option specifies which [1mhistory[22;0m is used (only one can be used).  The 
  [OFF|ON|<[4mfile[24m>] argument specifies what action is taken on that [1mhistory[22;0m.  

  [1mOptions:[22;0m 
  -w<[4mworld[24m> 
          Output from <[4mworld[24m> only.  
  -w      Output from the [1mcurrent[22;0m world.  
  -l      Local output (i.e., output generated by TF).  
  -i      Keyboard input.  
  -g      Global output (all worlds and local TF output).  

  Arguments: 
  OFF     Disable specified log, or all logs if unspecified.  
  ON      Log to [1m${LOGFILE}[22;0m; -g is assumed if -ligw not given.  
  <[4mfile[24m>  Log to <[4mfile[24m>; -g is assumed if -ligw not given.  
  (none)  With no option, lists all open logs.  
  (none)  With an -ligw option, same as "ON".  

  When logging is enabled for a [1mhistory[22;0m, lines that are normally recorded in 
  that [1mhistory[22;0m are also appended to the log file (unless the line has the "L" 
  nolog [1mattribute[22;0m).  The previously existing contents of the file, if any, are 
  not affected.  

  It is possible to have multiple log files open simultaneously.  It is also 
  possible to have several types of output go to the same log file, by using 
  several [1m/log[22;0m commands.  For example, 

    [1m/log[22;0m -i tt.log
    [1m/log[22;0m -wTT tt.log
    [1m/log[22;0m -g on

  will send input from the keyboard and output from the world TT to the file 
  "tt.log", and also send all (global) output to the file named by the [1mLOGFILE[22;0m 
  macro.  

  This example logs the [1mcurrent[22;0m world's output to a file whose name contains 
  the world's name and today's date: 

    [1m/eval[22;0m [1m/log[22;0m -w [1m${world_name}[22;0m.$[[1mftime[22;0m("%F")]
    

  The functions of the [1m/logme[22;0m command in older versions of TF can be performed 
  with [1m/log[22;0m -i.  

  Wrapping will be done in the log file only if the [1m%{wraplog}[22;0m [1mvariable[22;0m is 
  "on".  

  Logging is disabled by default.  The default value of [1m${LOGFILE}[22;0m is 
  "tiny.log".  

  Note: the natural logarithm function was renamed from log() to ln() in 
  version 5.0, to avoid confusion with /log.  

  See: [1m%wraplog[22;0m, [1mhistory[22;0m, [1mnlog()[22;0m [1mfwrite()[22;0m 

&/logme

/logme

  Obsolete.  See "[1mlog[22;0m".  

&/map
&/mark
&/path
&/revert
&/savepath
&/unpath
&/unmark
&/dopath
&mapping

mapping

  Usage: 

  [1m/REQUIRE[22;0m map.tf

  [1m/MARK[22;0m <[4mdir[24m>
  [1m/UNMARK[22;0m
  [1m/PATH[22;0m
  [1m/RETURN[22;0m
  [1m/MAP[22;0m
  [1m/UNPATH[22;0m
  [1m/SAVEPATH[22;0m <[4mname[24m>
  [1m/DOPATH[22;0m <[4mpath[24m>
  ____________________________________________________________________________

  These commands, similar to those in tintin, help keep track of sequences of 
  directions between two locations on a mud.  When mapping is enabled with 
  [1m/mark[22;0m, all mud movement commands (n, s, e, w, ne, sw, nw, se, u, d) that you 
  type are recorded in the "current path".  

  [1m/mark[22;0m clears the current path and starts recording your movement.  

  [1m/unmark[22;0m disables map recording (but does not clear the current path).  

  [1m/path[22;0m prints the current recorded path.  

  [1m/revert[22;0m "undoes" the last movement by deleting it from the path and 
  executing the opposite movement command.  (This was called "/return" prior 
  to version 4.0).  

  [1m/map[22;0m adds <[4mdir[24m> to the current path as if you had actually gone in that 
  direction.  

  [1m/unpath[22;0m deletes the last movement from the path (but does not move you to 
  your previous position) 

  [1m/savepath[22;0m [1mdefines[22;0m a [1mmacro[22;0m named <[4mname[24m> that will execute the movements in 
  the currently defined path.  (To save this [1mmacro[22;0m to a file, use "[1m/save[22;0m [-a] 
  <[4mfile[24m> <[4mname[24m>").  

  [1m/dopath[22;0m executes a <[4mpath[24m>.  <[4mPath[24m> must be a space-separated list of 
  movement commands with optional repeat counts.  For example, "[1m/dopath[22;0m 10 n e 
  d 2 w" will execute "n" 10 times, "e" once, "d" once, and "w" twice.  

  See: [1m/require[22;0m, [1mspeedwalk[22;0m 

&scroll
&pager
&--more--
&--More--
&/more

/more

  Usage: 

  [1m/MORE[22;0m [OFF|ON]
  ____________________________________________________________________________

  Sets the value of the [1m%{more}[22;0m flag.  If the [1m%{more}[22;0m flag is ON when the 
  screen or output window fills up, output will stop, and a "More" prompt will 
  be displayed.  With the default keybindings, TAB will scroll one screenful, 
  PgDn and PgUp will scroll a half screen forward or backward, ^[^N and ^[^P 
  will scroll one line forward or backward, and ^[j will Jump to the last 
  screenful.  

  Regardless of the setting of the [1m%more[22;0m flag, you can use "[1m/dokey[22;0m pause" (^S) 
  at any time to pause the screen immediately, or use any of the scrolling 
  commands to scroll backward and forward.  After doing so, the "more" prompt 
  will remain until you reach the bottom line again; after that point, newly 
  displayed lines will obey the [1m%more[22;0m flag normally.  

  In [1mvisual mode[22;0m, with the default [1mstatus bar[22;0m settings, the More prompt 
  displays the number of old lines (i.e., how far you have scrolled backwards) 
  and the number of new lines you haven't had a chance to see yet (i.e.  lines 
  that arrived since the More prompt first appeared).  If you have not 
  scrolled backwards, only the count of new lines is shown, so the More prompt 
  looks the same as it would have in version 4.0.  If either count would not 
  fit in the space allotted to it in the More prompt, they may be displayed in 
  units of thousands (e.g., "17523" would be shown as "17k").  

  Each [1msocket[22;0m and open world [1mworld[22;0m has its own window with its own "more" 
  state.  

  If your terminal can't scroll in [1mvisual mode[22;0m, TF will start over at the top 
  of the output window instead.  

  See: [1m/dokey[22;0m, [1mvisual[22;0m, [1m%more[22;0m, [1mmorescroll()[22;0m, [1mmoresize()[22;0m, [1mstatus_fields[22;0m 

&/nogag

/nogag

  Usage: 

  [1m/NOGAG[22;0m [<[4mpattern[24m>]
  ____________________________________________________________________________

  Eliminates a [1mmacro[22;0m that is [1mtrigger[22;0med by <[4mpattern[24m> and has the [1mgag[22;0m [1mattribute[22;0m. 
  [1m/nogag[22;0m with no arguments turns off the flag [1m%{gag}[22;0m, disabling all [1mgag[22;0m 
  [1mattributes[22;0m.  <[4mPattern[24m> is matched against existing patterns using simple 
  comparison.  

  The flag [1m%{gag}[22;0m defaults to 1 (on).  

  See: [1mtriggers[22;0m, [1m/gag[22;0m, [1m%gag[22;0m 

&/nohilite

/nohilite

  Usage: 

  [1m/NOHILITE[22;0m [<[4mpattern[24m>]
  ____________________________________________________________________________

  With a <[4mpattern[24m> argument, [1m/nohilite[22;0m undefines a [1mmacro[22;0m that is [1mtrigger[22;0med by 
  <[4mpattern[24m> and has the [1mhilite[22;0m [1mattribute[22;0m.  <[4mPattern[24m> is matched against 
  existing patterns using simple comparison.  

  With no argument, [1m/nohilite[22;0m turns off the flag [1m%{hilite}[22;0m, disabling all 
  display [1mattributes[22;0m.  

  The flag [1m%{hilite}[22;0m defaults to 1 (on).  

  See: [1mtriggers[22;0m, [1m/hilite[22;0m, [1m%hilite[22;0m 

&/first
&/last
&/nth

/nth

  Usage: 

  [1m/FIRST[22;0m <[4mtext[24m>
  [1m/LAST[22;0m <[4mtext[24m>
  [1m/NTH[22;0m <[4mn[24m> <[4mtext[24m>
  ____________________________________________________________________________

  Echoes the first, last, or <[4mn[24m>th word from text.  `[1m/first[22;0m <[4mtext[24m>' is 
  equivalent to `[1m/nth[22;0m 1 <[4mtext[24m>'.  

  These commands can be useful in command substitutions.  For example, to make 
  "ctrl-O 1" input the first word of the most recent mud output, you could do 
  this: 

  [1m/bind[22;0m ^O1 = [1m/input[22;0m $([1m/first[22;0m $([1m/recall[22;0m 1)) 

  See: [1mparameters[22;0m, [1mcommand substitution[22;0m 

&/partial

/partial

  Usage: 

  [1m/PARTIAL[22;0m <[4mregexp[24m>
  ____________________________________________________________________________

  Creates a [1mmacro[22;0m which will [1mhilite[22;0m the part of a line containing text matched 
  by the [1mregular expression[22;0m <[4mregexp[24m>.  Remember that [1mregular expressions[22;0m are 
  case sensitive.  The new [1mmacro[22;0m is a [1mfall-thru[22;0m, so multiple [1m/partial[22;0ms (and 
  other [1mtriggers[22;0m) can match the same text.  

  The [1mattribute[22;0m(s) for [1mhilited[22;0m text are determined by the [1m%{hiliteattr}[22;0m 
  [1mvariable[22;0m.  The default is bold ([1mhiliteattr[22;0m=B).  [1mColors[22;0m are also available.  

  For example, "[1m/partial[22;0m [Hh]awkeye" will [1mhilite[22;0m any occurrence of "Hawkeye" 
  or "hawkeye".  

  Unlike version 3.2, a partial [1mhilite[22;0m will be applied to every match on a 
  line, not just the first match.  

  [1m/partial[22;0m <[4mregexp[24m>
  is equivalent to
  [1m/def[22;0m [1m-Ph[22;0m [1m-F[22;0m [1m-t[22;0m<[4mregexp[24m>

  See: [1mattributes[22;0m, [1mpatterns[22;0m, [1m/hilite[22;0m, [1m/def[22;0m 

&paste_prefix
&%paste_prefix
&/endpaste
&/paste

/paste

  Usage: 

  [1m/PASTE[22;0m [-pnx] [<[4mprefix[24m>]
  [1m/ENDPASTE[22;0m
  ____________________________________________________________________________

  After executing [1m/paste[22;0m, every line of input (including lines that begin with 
  "/") will have <[4mprefix[24m> prepended to it and then get sent to the [1mcurrent 
  socket[22;0m.  If <[4mprefix[24m> is omitted and -n is not specified, the prefix defaults 
  to the value of %paste_prefix; if %paste_prefix is empty or unset, it 
  defaults to ":|".  Typing "[1m/endpaste[22;0m" or "." on a line by itself ends the 
  pasting; "/abort" on a line by itself aborts the pasting.  [1m/Paste[22;0m can be 
  very useful when using the cut-and-paste mechanism of many windowing 
  systems.  

  [1mOptions:[22;0m 
  -p      "paragraph mode": adjacent non-blank lines are joined, and leading 
          spaces are stripped (this is particularly useful when pasting text 
          cut from a web browser or a window of different width).  
  -n      Don't prepend any prefix.  
  -x      After prepending the prefix (if any), execute the resulting line as 
          a command (without [1msubstitution[22;0m) instead of sending it.  
  -w<[4mworld[24m> 
          Send the text to <[4mworld[24m>.  
  -e<[4mend[24m> 
          End when the user types <[4mend[24m> (default: "/endpaste").  With or 
          without this option, "." will always work.  
  -a<[4mabort[24m> 
          Abort when the user types <[4mabort[24m> (default: "/abort").  With or 
          without this option, interrupt (^C) will always work.  
  -q      quiet: do not print "Entering paste mode" message.  
  -s      strip trailing spaces from each pasted line 
  -h      invoke matching [1mSEND hooks[22;0m for each line sent by [1m/paste[22;0m.  

  Note that [1m/endpaste[22;0m is not actually a command, but a "magic cookie" 
  recognized by [1m/paste[22;0m.  "[1m/Endpaste[22;0m", ".", and SIGINT (^C) are the only ways 
  to end [1m/paste[22;0m.  

  Lines sent by [1m/paste[22;0m will invoke matching [1mSEND hooks[22;0m.  

  Examples: 
  Prepare to paste text from a web page to a mud: 
          [1m/paste[22;0m -p 
  Prepare to paste a bunch of lines to be recorded in your input [1mhistory[22;0m: 
          [1m/paste[22;0m -x [1m/recordline[22;0m -i - 

  See: [1m/quote[22;0m 

&/prompt

/prompt

  [1mFunction[22;0m usage: 

  [1mPROMPT[22;0m(<[4mtext[24m>)

  Command usage: 

  [1m/PROMPT[22;0m [-a<[4mattrs[24m>] [-p] <[4mtext[24m>
  ____________________________________________________________________________

  Sets the [1mprompt[22;0m for the [1mcurrent[22;0m [1msocket[22;0m to <[4mtext[24m>, replacing any existing 
  [1mprompt[22;0m.  

  Command [1moptions[22;0m: 
  -a<[4mattrs[24m> 
          Apply the [1mattributes[22;0m given by <[4mattrs[24m> to <[4mtext[24m>.  
  -p      Interpet "@{<[4mattr[24m>}" strings within <[4mtext[24m> as commands to set 
          [1mattributes[22;0m inline.  See [1mdecode_attr()[22;0m.  

  [1m/prompt[22;0m is most useful when called from a PROMPT [1mhook[22;0m, like this: 

      [1m/def[22;0m [1m-h[22;0m"PROMPT *> " catch_prompt = [1m/test[22;0m [1mprompt[22;0m([1m{*}[22;0m)

  Then, any text that ends in ">" without a newline will be made the prompt.  

  For a more sophisticated example, see "[1mstatus line[22;0m".  

  See: [1mprompts[22;0m, [1mhooks[22;0m (PROMPT) 

&/ps

/ps

  Usage: 

  [1m/PS[22;0m [-srq] [-w<[4mworld[24m>] [<[4mpid[24m>]
  ____________________________________________________________________________

  [1mOptions[22;0m: 
  -s      short form, lists only PIDs.  
  -r      list /repeats only.  
  -q      list /quotes only.  
  -w[<[4mworld[24m>] 
          list only processes for <[4mworld[24m>.  

  Lists information about [1mprocess[22;0m <[4mpid[24m>, or all currently running [1m/quote[22;0m and 
  [1m/repeat[22;0m [1mprocesses[22;0m: 

  PID     unique [1mprocess[22;0m identification number.  
  NEXT    time remaining until next execution of [1mprocess[22;0m, or "pending" if 
          [1mprocess[22;0m is waiting for output from a shell command.  
  T       the type of the command: "q" for 
          [1mquote[22;0m or "r" for [1mrepeat[22;0m.  
  D       disposition of [1m/quote[22;0m lines: "e" for echo, "s" for send, or "x" for 
          exec.  
  WORLD   world to which output is sent, if not the [1mcurrent[22;0m world.  
  PTIME   delay between executions.  
  COUNT   number of [1m/repeat[22;0m executions remaining.  
  COMMAND 
          the command to be executed.  

  See: [1mprocesses[22;0m 

&/purgebind
&/purgedef
&/purgedeft
&/purgegag
&/purgehilite
&/purgehook
&/purgetrig
&/purge

/purge

  Usage: 

  [1m/PURGE[22;0m [<[4mmacro-options[24m>] [<[4mname[24m>] [= <[4mbody[24m>]
  ____________________________________________________________________________

  Removes all [1mmacros[22;0m matching the specified restrictions.  The [1m<[4mmacro-options[24m>[22;0m 
  are the same as those in the [1m/list[22;0m command; see "[1m/list[22;0m" for details.  
  Invisible [1mmacros[22;0m will not be purged unless "-i" is specified.  Remember that 
  "macros" includes keybindings, [1mhilite[22;0ms, [1mgag[22;0ms, [1mtriggers[22;0m, and [1mhooks[22;0m.  

  The standard [1mmacro[22;0m library also defines the commands [1m/purgedef[22;0m, [1m/purgebind[22;0m, 
  [1m/purgehilite[22;0m, [1m/purgegag[22;0m, [1m/purgetrig[22;0m, [1m/purgedeft[22;0m, and [1m/purgehook[22;0m, which purge 
  [1mmacros[22;0m of the appropriate type.  These always use [1mglob[22;0m matching.  

  See: [1mmacros[22;0m, [1mtriggers[22;0m, [1mpatterns[22;0m, [1mattributes[22;0m, [1mlibrary[22;0m, [1m/def[22;0m, [1m/list[22;0m, 
  [1m/purgeworld[22;0m 

&/purgeworld

/purgeworld

  Usage: 

  [1m/PURGEWORLD[22;0m [-m<[4mstyle[24m>] [-T<[4mtype[24m>] [<[4mname[24m>]
  ____________________________________________________________________________

  Removes [1mworld[22;0m definitions.  

  [1mOptions[22;0m and arguments: 
  -m<[4mstyle[24m> 
          Use <[4mstyle[24m> for [1mpattern matching[22;0m of <[4mtype[24m> and <[4mname[24m> patterns.  
          (default: [1m%{matching}[22;0m).  
  -T<[4mtype[24m> 
          Remove only worlds with a type matching the [1mpattern[22;0m <[4mtype[24m>.  
  <[4mname[24m>  Remove only worlds with a name matching the [1mpattern[22;0m <[4mname[24m>.  

  The return value of [1m/purgeworld[22;0m is the number of world definitions that were 
  removed.  

  See: [1mworlds[22;0m, [1m/listworlds[22;0m, [1mpatterns[22;0m 

&upload
&/putfile_MUCK
&/putfile_UNIX
&/putfile_LP
&/putfile

/putfile

  Usage: 

  [1m/REQUIRE[22;0m filexfer.tf

  [1m/PUTFILE_MUCK[22;0m <[4mfile[24m> [<[4mremote-file[24m>]
  [1m/PUTFILE_LP[22;0m <[4mfile[24m> [<[4mremote-file[24m>]
  [1m/PUTFILE_UNIX[22;0m <[4mfile[24m> [<[4mremote-file[24m>]
  ____________________________________________________________________________

  Uploads text <[4mfile[24m> from the local system to <[4mremote-file[24m> on a MUCK, LP, or 
  UNIX server, using an editor on the remote system.  If <[4mremote-file[24m> is 
  omitted, <[4mfile[24m> is used as the name of the remote file.  

  [1m/Putfile_LP[22;0m assumes the LPmud has an "ed" editor similar to that in UNIX.  

  For backward compatibility, [1m/putfile[22;0m is the same as [1m/putfile_MUCK[22;0m.  

  See: [1m/getfile[22;0m, [1m/quote[22;0m 

&/quit

/quit

  Usage: 

  [1m/QUIT[22;0m [-y]
  ____________________________________________________________________________

  Exits TF.  If TF is [1minteractive[22;0m, and there are any [1mworlds[22;0m with unseen text, 
  /quit first asks you to confirm the exit; if you type anything other than 
  "Y" or "y", TF does not exit.  

  [1mOptions:[22;0m 
  -y      exit unconditionally, without prompting.  
  When TF exits, all [1msocket[22;0m connections will be disconnected; all logfiles 
  will be closed; all [1m/quote[22;0ms and [1m/repeat[22;0ms will be killed; and all [1mhistory[22;0m, 
  unsaved [1mmacros[22;0m, and [1mvariables[22;0m will be lost.  

  If you prefer to never be prompted by /quit, you can redefine it like this: 

    [1m/def[22;0m quit = /@quit -y

  See also: [1m/dc[22;0m, [1m%quitdone[22;0m 

&/quote

/quote

  Usage: 

  [1m/QUOTE[22;0m [<[4moptions[24m>] [<[4mpre[24m>] '"<[4mfile[24m>"[<[4msuf[24m>]
  [1m/QUOTE[22;0m [<[4moptions[24m>] [<[4mpre[24m>] #"<[4mrecall_args[24m>"[<[4msuf[24m>]
  [1m/QUOTE[22;0m [<[4moptions[24m>] [<[4mpre[24m>] !"<[4mshell_cmd[24m>"[<[4msuf[24m>]
  [1m/QUOTE[22;0m [<[4moptions[24m>] [<[4mpre[24m>] `"<[4mTF_cmd[24m>"[<[4msuf[24m>]
  ____________________________________________________________________________

  [1m/Quote[22;0m generates lines of text, one for each line quoted from a file, shell 
  command, [1mhistory[22;0m, or TF command.  Each generated line is then echoed, sent 
  to a socket, or executed as a command.  Lines will be generated at a rate 
  described in the section "[1mprocesses[22;0m".  

  [1mOptions[22;0m and arguments: 
  -d<[4mdisp[24m> 
          disposition of generated text.  <[4mDisp[24m> is one of: "echo" (echo to 
          the screen), "send" (send directly to the [1msocket[22;0m), or "exec" 
          (execute text as a tf command).  The default <[4mdisp[24m> is "send" if 
          there is no <[4mpre[24m>, and "exec" if there is a <[4mpre[24m>.  
  -w<[4mworld[24m> 
          Generated commands will be executed with <[4mworld[24m> as the [1mcurrent[22;0m 
          world.  If <[4mworld[24m> is blank, it uses the world that was [1mcurrent[22;0m when 
          the [1m/quote[22;0m started.  If -w is omitted, each command's [1mcurrent[22;0m world 
          will be whatever happens to be in the [1mforeground[22;0m when each command 
          occurs.  (See "[1msockets[22;0m").  
  -<[4mtime[24m> 
          The delay between each generated line.  It can have the format 
          "<[4mhours[24m>:<[4mminutes[24m>:<[4mseconds[24m>", "<[4mhours[24m>:<[4mminutes[24m>", or "<[4mseconds[24m>", 
          and <[4mseconds[24m> may be specified to the nearest microsecond.  If 
          -<[4mtime[24m> is omitted, the [1mvariable[22;0m [1m%{ptime}[22;0m is used.  If <[4mtime[24m> is 
          given as the letter "S", the quote will run synchronously, with no 
          delay.  If a slow shell command is used with [1m/quote[22;0m -S !, tf will 
          hang until the command produces some output or exits.  A synchronous 
          [1m/quote[22;0m may be used inside another [1m/quote[22;0m.  If <[4mtime[24m> is given as the 
          letter "P", the quote will run whenever a [1mprompt[22;0m is received.  See 
          "[1mprocesses[22;0m" for more information on process timing.  
  -s<[4msub[24m> 
          [1mExpand[22;0m <[4mTF_cmd[24m> as if [1m%{sub}[22;0m were set to <[4msub[24m>.  By default, [1m/quote[22;0m 
          [1mexpands[22;0m <[4mTF_cmd[24m> as if [1m%{sub}[22;0m were "full".  
  <[4mpre[24m>   <[4mpre[24m> is prefixed to each generated line.  If <[4mpre[24m> contains any of 
          the command characters ('!`#), they must be preceded with '\' to 
          remove their special meaning.  
  '<[4mfile[24m> 
          Get text from <[4mfile[24m>.  The <[4mfile[24m> name is expanded as described 
          under [1m/help[22;0m filenames.  
  !<[4mshell_cmd[24m> 
          Get text from the standard output and standard error of executing 
          <[4mshell_cmd[24m> in the shell.  
  `<[4mTF_cmd[24m> 
          Get text from the output of executing <[4mTF_cmd[24m> in tf.  
  #<[4mrecall_args[24m> 
          Get text from executing [1m/recall[22;0m <[4mrecall_args[24m>.  (See "[1mrecall[22;0m" for 
          the exact syntax).  
  <[4msuf[24m>   Append <[4msuf[24m> to each generated line.  If omitted, the double quotes 
          around the <[4mfile[24m> or <[4mcommand[24m> may also be omitted.  

  An asynchronous (background) [1m/quote[22;0m (i.e., a [1m/quote[22;0m without -S) returns the 
  pid of the new [1mprocess[22;0m, or 0 if an error occurred.  A synchronous (-S) shell 
  (!) or command (`) quote returns the return value of the command.  A 
  synchronous file (') quote returns 0 on error, nonzero otherwise.  

  The [1mlibrary[22;0m file [1mquoter.tf[22;0m defines some useful [1mquoter commands[22;0m that are 
  shortcuts for some common uses of [1mquote[22;0m.  

  The following is a list of some nearly equivalent pairs of commands: 
  [1m/quote[22;0m -S -dexec '<[4mfile[24m> 
          [1m/load[22;0m <[4mfile[24m> 
  [1m/quote[22;0m -S /echo -aG - #<[4margs[24m> 
          [1m/recall[22;0m <[4margs[24m> 
  [1m/quote[22;0m <[4mopts[24m> `[1m/recall[22;0m <[4margs[24m> 
          [1m/quote[22;0m <[4mopts[24m> #<[4margs[24m> 

  ____________________________________________________________________________

  Examples: 


    [1m/quote[22;0m -1 :reads about '"/usr/dict/words" in the dictionary.

  This sends off lines like:

    :reads about aardvark in the dictionary.
    :reads about aardvore in the dictionary.

  with one-second delays between lines.  


    [1m/quote[22;0m -S [1m/echo[22;0m !ps -gux

  This displays the output of the system command "ps -gux" by echoing it 
  locally, immediately.  


    [1m/quote[22;0m -0 :heard: #-wCave /2 *pages*

  This sends off quickly: 

  :heard: [the last 2 lines from Cave that contain "pages"] 


    [1m/quote[22;0m :is using `[1m/version[22;0m

  will tell everybody in the room what version of TF you're running.  


    [1m/quote[22;0m -wlpmud -dsend 'prog.c

  will send the file "prog.c" to the world "lpmud", without any interpretation 
  of leading spaces or slashes (in lines like "/* comment */"), etc.) 

  ____________________________________________________________________________

  See: [1mprocesses[22;0m, [1m%ptime[22;0m, [1m%lpquote[22;0m, [1mquoter.tf[22;0m, [1mhistory[22;0m, [1mcommand subs[22;0m, [1m/load[22;0m, 
  [1m/recall[22;0m, [1m/sh[22;0m, [1m/sys[22;0m, [1m/paste[22;0m 

&/qdef
&/qmac
&/qworld
&/qtf
&/qsh
&/qmud
&quoter
&quoter.tf

Quoter Commands

  [1m/REQUIRE[22;0m quoter.tf
  ____________________________________________________________________________

  After doing "[1m/REQUIRE[22;0m quoter.tf", the quoting commands can be used to take 
  the output of various sources and execute them as commands, typically 
  quoting them to a mud server.  These are all just shortcuts for things you 
  can already do with [1m/quote[22;0m -S.  The default <[4mprefix[24m> is ":|", which will 
  perform a pose on Tiny-style muds.  The default prefix can be changed by 
  setting the appropriate variable: qdef_prefix, qmac_prefix, qworld_prefix, 
  qtf_prefix, qsh_prefix, or qmud_prefix.  An alternate <[4mprefix[24m> can be given 
  on the command line for [1m/qdef[22;0m, [1m/qmac[22;0m, [1m/qworld[22;0m, and [1m/qfile[22;0m.  Also, before any 
  output is generated, the command used to generate the output is quoted.  

  [1m/QDEF[22;0m [<[4mprefix[24m>] <[4mname[24m> 
          Prepends <[4mprefix[24m> to each line generated by "[1m/list[22;0m <[4mname[24m>", and 
          executes each resulting line as a command.  

  [1m/QMAC[22;0m [<[4mprefix[24m>] <[4mname[24m> 
          Searches for the definition of macro <[4mname[24m> in a group of tf files, 
          prepends <[4mprefix[24m> to each line found, "[1m/quote[22;0m <[4mname[24m>", and executes 
          each resulting line as a command.  

  [1m/QWORLD[22;0m [<[4mprefix[24m>] <[4mname[24m> 
          Prepends <[4mprefix[24m> to each line generated by "[1m/listworlds[22;0m <[4mname[24m>", 
          and executes each resulting line as a command.  

  [1m/QFILE[22;0m [<[4mprefix[24m>] <[4mname[24m> 
          Prepends <[4mprefix[24m> to each line of file <[4mname[24m>, and executes each 
          resulting line as a command.  

  [1m/QTF[22;0m <[4mcmd[24m> 
          Prepends <[4mprefix[24m> to each line generated by executing <[4mcmd[24m> in tf, 
          and executes each resulting line as a command.  

  [1m/QSH[22;0m <[4mcmd[24m> 
          Prepends <[4mprefix[24m> to each line generated by executing <[4mcmd[24m> in the 
          shell, and executes each resulting line as a command.  

  [1m/QMUD[22;0m [-w<[4mworld[24m>] <[4mcmd[24m> 
          Prepends <[4mprefix[24m> to each line generated by executing <[4mcmd[24m> on world 
          <[4mworld[24m> (default: the [1mcurrent[22;0m [1mworld[22;0m), and executes each resulting 
          line as a command.  [1m/Qmud[22;0m requires that the mud supports the 
          OUTPUTPREFIX and OUTPUTSUFFIX commands.  

  Examples: 

  The command 

    [1m/qsh[22;0m finger

  would generate a series of commands something like this: 

    :! finger
    :| Login       Name              TTY Idle    When    Site Info
    :| hawkeye  Ken Keys              p3       Fri 19:32 
    :| hawkeye  Ken Keys              p4       Sat 17:37 

  And, on a Tiny-style mud named "Cave", the command 

    [1m/qmud[22;0m score

  would generate a series of commands something like this: 

    :| Cave> score
    :| You have 8704 pennies.

  ____________________________________________________________________________

  See: [1m/quote[22;0m, [1mprocesses[22;0m, [1m/paste[22;0m 

&/recall

/recall

  Usage: 

  [1m/RECALL[22;0m [-w<[4mworld[24m>] [-ligv] [-t[<[4mformat[24m>]] [-a<[4mattrs[24m>] [-m<[4mstyle[24m>] [-A<[4mn[24m>] 
  [-B<[4mn[24m>] [-C<[4mn[24m>] [#]<[4mrange[24m> [<[4mpattern[24m>]
  ____________________________________________________________________________

  Recalls lines from a [1mhistory[22;0m buffer.  Only one of the [-ligw] options can be 
  used, to specify the history from which to recall.  

  [1mOptions:[22;0m 
  -w      recall from [1mcurrent[22;0m world's [1mhistory[22;0m (default) 
  -w<[4mworld[24m> 
          recall from <[4mworld[24m>'s [1mhistory[22;0m 
  -l      recall from local [1mhistory[22;0m (i.e., TF output) 
  -g      recall from global [1mhistory[22;0m (all worlds, and local) 
  -i      recall from input history 
  -t[<[4mformat[24m>] 
          display timestamps on each line, using <[4mformat[24m>.  If <[4mformat[24m> is 
          omitted, "[[1m%{time_format}[22;0m]" will be used.  The format is described 
          in [1mftime()[22;0m.  
  -v      recall lines that [4mdon't[24m match the [1mpattern[22;0m 
  -q      quiet: suppress the header and footer lines 
  -a<[4mattr[24m> 
          suppress specified [1mattributes[22;0m (e.g., -ag shows [1mgag[22;0mged lines) 
  -m<[4mstyle[24m> 
          matching style ([1msimple[22;0m, [1mglob[22;0m, or [1mregexp[22;0m).  
  -A<[4mn[24m>   Print <[4mn[24m> lines of context after each matching line.  
  -B<[4mn[24m>   Print <[4mn[24m> lines of context before each matching line.  
  -C<[4mn[24m>   Equivalent to -A<[4mn[24m> -B<[4mn[24m>.  
  #       display line numbers (must be last option, before <[4mrange[24m>) 

  <[4mrange[24m> can have one of the formats below.  If <[4mx[24m> and <[4my[24m> are plain 
  integers, they are interpreted as line numbers or counts.  If they have the 
  form "<[4mhours[24m>:<[4mminutes[24m>" or "<[4mhours[24m>:<[4mminutes[24m>:<[4mseconds[24m>", they are 
  interpreted as time values (either a period of time, or a clock time within 
  the last 24 hours).  If they are real numbers (with up to 6 decimal places), 
  they are interpreted as absolute system times.  
  /[4mx[24m      Recall the last <[4mx[24m> matching lines.  
  [4mx[24m       Recall from the last <[4mx[24m> lines, or lines within the last time period 
          <[4mx[24m>.  
  [4mx[24m-[4my[24m     Recall lines starting with <[4mx[24m> and ending with <[4my[24m>.  
  -[4my[24m      If <[4my[24m> is a line number, recall the <[4my[24m>th previous line; if <[4my[24m> is a 
          time, recall lines earlier than <[4my[24m>.  Remember to use "[1m-[22;0m" before 
          "-[4my[24m" so it isn't interpreted as an [1moption[22;0m.  
  [4mx[24m-      Recall lines after <[4mx[24m>.  

  If <[4mrange[24m> is prefixed with "#", line numbers will be displayed.  

  If <[4m[1mpattern[22;0m[24m> is given, only lines in the given range that match <[4m[1mpattern[22;0m[24m> 
  will be recalled.  The matching style is determined by the -m option if 
  given, [1m%{matching}[22;0m otherwise.  

  If the output of [1m/recall[22;0m is being sent to the screen, it will be preceded by 
  "================ Recall start ================" and follwed by 
  "================= Recall end =================" unless -q is used.  These 
  lines will not be produced if the output is redirected, for example with 
  [1m$(...)[22;0m [1mcommand substitution[22;0m or "[1m/quote `[22;0m[1m/recall[22;0m".  When -A, -B, or -C is 
  used, groups of lines that are not adjacent in history will be separated by 
  "--".  

  If lines are received while tf is suspended (by [1m^Z[22;0m or [1m/suspend[22;0m) or in a 
  subshell (by [1m/sh[22;0m), the timestamps on the lines will correspond to the time 
  tf resumed control, not the time they actually arrived.  

  The return value of [1m/recall[22;0m is the number of lines that were actually 
  recalled.  

  Because the output of [1m/recall[22;0m may clutter the current window, you may wish 
  to use [1m/limit[22;0m instead.  

  Examples These examples assume that [1mmatching[22;0m=glob (the default).  
  Recall every line beginning with "Kite whispers" that arrived in the last 
  hour:   [1m/recall[22;0m 1:00 Kite whispers* 
  Recall every line that arrived between 11 am and 1 pm: 
          [1m/recall[22;0m 11:00-13:00 
  Recall the last 5 lines containing "spam": 
          [1m/recall[22;0m /5 *spam* 
  Recall the last 4th most recent line: 
          [1m/recall[22;0m - -4 

  See: [1mhistory[22;0m, [1mattributes[22;0m, [1m/limit[22;0m, [1m/quote[22;0m, [1m%time_format[22;0m 

&/recordline

/recordline

  Usage: 

  [1m/RECORDLINE[22;0m [-lig] [-w[<[4mworld[24m>]] [-t<[4mtime[24m>] <[4mtext[24m>
  ____________________________________________________________________________

  Records <[4mtext[24m> into a [1mhistory[22;0m buffer.  

  [1mOptions:[22;0m 
  -w      record to [1mcurrent[22;0m world's [1mhistory[22;0m 
  -w<[4mworld[24m> 
          record to <[4mworld[24m>'s [1mhistory[22;0m 
  -l      record to local [1mhistory[22;0m 
  -g      record to global [1mhistory[22;0m (default) 
  -i      record to input history 
  -t<[4mtime[24m> 
          record the line with the system time <[4mtime[24m> (as displayed by [1m/recall[22;0m 
          -t@) instead of the current time 
  -a<[4mattrs[24m> 
          Record <[4mtext[24m> with the [1mattributes[22;0m given by <[4mattrs[24m>.  
  -p      Interpet "@{<[4mattr[24m>}" strings as commands to set [1mattributes[22;0m inline.  
          "@@" strings are interpreted as "@".  "@{n}" or "@{x}" will turn 
          attributes off.  See also: [1mdecode_attr()[22;0m.  

  The <[4mtext[24m> will not be echoed to the screen or saved in any log.  

  [1m/Recordline[22;0m can be combined with [1m/quote[22;0m to read a log file back into 
  [1mhistory[22;0m.  For example, if you had created a log with "[1m/log[22;0m -i input.log" in 
  an earlier tf session, you could start a new tf session and use 

  [1m/quote[22;0m -S -dexec [1m/recordline[22;0m -i [1m-[22;0m 'input.log 

  to restore that input [1mhistory[22;0m.  That way, you could use the RECALLB, 
  RECALLF, RECALLBEG, RECALLEND, SEARCHB, and SEARCHF (^P, ^N, ^[<, ^[>, ^[P, 
  and ^[N) keys to recall lines you typed in the earlier session.  

  Note that [1m/recordline[22;0m always appends to the end of a [1mhistory[22;0m.  [1m/Recordline[22;0m 
  -t<[4mtime[24m> makes it possible to insert lines that are not in chronological 
  order, which may produce strange results with [1m/recall[22;0m.  

  See: [1m/recall[22;0m, [1m/quote[22;0m, [1mhistory[22;0m 

&delay
&/repeat

/repeat

  Usage: 

  [1m/REPEAT[22;0m [-w[<[4mworld[24m>]] [-n] {[-<[4mtime[24m>]|-S|-P} <[4mcount[24m> <[4mcommand[24m>
  ____________________________________________________________________________

  Repeats <[4mcommand[24m>, <[4mcount[24m> times.  <[4mCommand[24m> may be any legal [1mmacro[22;0m body.  
  If <[4mcount[24m> is "i", the <[4mcommand[24m> repeats indefinitely.  This works through a 
  [1mprocess[22;0m, which runs concurrently with normal operations.  

  [1mOptions:[22;0m 
  -w[<[4mworld[24m>] 
          <[4mCommand[24m> will execute with <[4mworld[24m> as the [1mcurrent[22;0m world.  If 
          <[4mworld[24m> is omitted, it is assumed to be the world that was [1mcurrent[22;0m 
          for /repeat.  If this option is omitted entirely, the <[4mcommand[24m>'s 
          [1mcurrent[22;0m world will be whatever world happens to be in the [1mforeground[22;0m 
          when it's time for <[4mcommand[24m> to run.  
  -<[4mtime[24m> 
          <[4mTime[24m> is the delay between each execution of <[4mcommand[24m>.  <[4mTime[24m> may 
          be specified in the format "<[4mhours[24m>:<[4mminutes[24m>:<[4mseconds[24m>", 
          "<[4mhours[24m>:<[4mminutes[24m>", or "<[4mseconds[24m>" (<[4mseconds[24m> may be specified to 
          the nearest microsecond).  
  -S      The repeat will run synchronously.  
  -P      The repeat will run whenever a [1mprompt[22;0m is received.  
  -n      When combined with the -<[4mtime[24m> option, this makes the first 
          execution of <[4mcommand[24m> happen with no delay.  
  At most one of the -S, -P, and -<[4mtime[24m> options should be specified.  If none 
  are specified, the delay between each execution of <[4mcommand[24m> is determined 
  by the [1mvariable[22;0m [1m%{ptime}[22;0m.  See "[1mprocesses[22;0m" for more information on process 
  timing.  

  The <[4mcommand[24m> undergoes [1mmacro[22;0m body [1msubstitution[22;0m when it is executed.  

  An asynchronous [1m/repeat[22;0m (without -S) returns the pid of the new [1mprocess[22;0m, or 
  0 if an error occurred.  A synchronous [1m/repeat[22;0m returns the return value of 
  the last command.  

  Since the first run is not done until after the first interval (for [1m/repeat[22;0m 
  without -S or -n), a useful trick is to use "[1m/repeat[22;0m -<[4mtime[24m> 1 <[4mcommand[24m>" to 
  delay the execution of a single command.  

  Example: [1m/repeat[22;0m -0:30 1 [1m/echo[22;0m -ab Dinner's ready 
#sleep

  There is no good way to directly "sleep" within a [1mmacro[22;0m body.  Any attempt 
  to write your own /sleep macro will, at best, "freeze" tf for the duration 
  of the sleep, or even worse hog the machine's CPU time in a busy wait.  The 
  best way to achieve the effect a sleep in a [1m/while[22;0m loop is probably to use a 
  [1m/repeat[22;0m where each execution of the [1m/repeat[22;0m body corresponds to an iteration 
  of the desired [1m/while[22;0m loop.  That is, if you want to write 

      [1m/def[22;0m foo = \
          /before_stuff%; \
          [1m/while[22;0m (condition) \
              /do_stuff%; \
              /sleep 5%; \
          [1m/done[22;0m%; \
          /after_stuff

  you must instead write: 

      [1m/def[22;0m foo = \
          /before_stuff%; \
          /foo_loop

      [1m/def[22;0m foo_loop = \
          [1m/if[22;0m (condition) \
              /do_stuff%; \
              [1m/repeat[22;0m -5 1 /foo_loop%; \
          [1m/else[22;0m
              /after_stuff%; \
          [1m/endif[22;0m

  Of course, local [1mvariables[22;0m will not survive between calls of /do_stuff in 
  the second version as they would in the first (if it were possible), so any 
  [1mvariables[22;0m you need to share between iterations must be global.  

  But, if the reason you want to sleep is to wait for a response from a 
  server, then you really don't want to sleep at all: you want a [1mtrigger[22;0m.  
  First, set up [1mtriggers[22;0m on the possible responses, then send the command.  If 
  one of the possible responses is no response at all, then a [1m/repeat[22;0m can be 
  useful to wait for some maximum timeout and then handle the no-reponse case 
  and delete the response [1mtriggers[22;0m.  This is in general the best way to write 
  [1mmacros[22;0m that interact with a server.  
#

  See: [1mprocesses[22;0m, [1m%ptime[22;0m, [1m/at[22;0m, [1mkbnum[22;0m 

&/replace

/replace

  [1mFunction[22;0m usage: 

  [1mREPLACE[22;0m(<[4mold[24m>, <[4mnew[24m>, <[4mstring[24m>)

  Command usage: 

  [1m/REPLACE[22;0m <[4mold[24m> <[4mnew[24m> <[4mstring[24m>
  ____________________________________________________________________________

  Echoes (in command form) or returns (in [1mfunction[22;0m form) <[4mstring[24m>, with any 
  occurrences of <[4mold[24m> in <[4mstring[24m> replaced by <[4mnew[24m>.  

#replace-ex
  Example: 

  This example replaces "TF" with "TinyFugue" in every line sent by the 
  server.  

    [1m/def[22;0m [1m-m[22;0mregexp [1m-t[22;0m"TF" replace_tf = \
        [1m/test[22;0m [1msubstitute[22;0m([1mreplace[22;0m("TF", "TinyFugue", {P0}))

  See: [1mevaluation[22;0m, [1m/tr[22;0m 

&security
&/restrict

/restrict

  Usage: 

  [1m/RESTRICT[22;0m [SHELL|FILE|WORLD]
  ____________________________________________________________________________

  With no arguments, [1m/restrict[22;0m reports the current restriction level.  

  With an argument, [1m/restrict[22;0m sets the restriction level.  Once restriction 
  has been set to a particular level, it can not be lowered.  
  level 0: NONE 
          No restrictions.  
  level 1: SHELL 
          Prevents all access to shell or external commands.  Disables TF 
          builtins "[1m/sh[22;0m" and "[1m/quote[22;0m !", and uncompression during [1m/load[22;0m and 
          [1m/help[22;0m.  
  level 2: FILE 
          Prevents reading and writing of files.  Disables TF builtins 
          "[1m/load[22;0m", "[1m/save[22;0m", "[1m/saveworld[22;0m", "[1m/lcd[22;0m", "[1m/log[22;0m", and "[1m/quote[22;0m '", 
          "[1mtfopen()[22;0m", the "[1msockmload[22;0m feature.  Implies [1m/restrict[22;0m shell.  
  level 3: WORLD 
          Disallows all new user-defined connections.  The TF builtins 
          [1m/addworld[22;0m and the "[1m/connect[22;0m <[4mhost[24m> <[4mport[24m>" semantics are disabled.  
          Implies [1m/restrict[22;0m file.  

  [1m/Restrict[22;0m is typically placed in [1m%{TFLIBDIR}[22;0m/[1mlocal.tf[22;0m by an administrator of 
  a public copy of TF who wishes to restrict users' access.  

  Note that while I believe these options to be secure, I provide no warranty 
  to that effect.  

  See: [1mwarranty[22;0m 

&/result
&/return

/return and /result

  Usage: 

  [1m/RETURN[22;0m [<[4m[1mexpression[22;0m[24m>]
  [1m/RESULT[22;0m [<[4m[1mexpression[22;0m[24m>]
  ____________________________________________________________________________

  [1m/return[22;0m stops execution of the [1mmacro[22;0m body that called it, and causes the 
  macro to return the string value of <[4m[1mexpression[22;0m[24m>.  If the <[4m[1mexpression[22;0m[24m> is 
  omitted, the return value of the [1mmacro[22;0m is the empty string.  

  When a macro that calls [1m/result[22;0m was called as a [1mfunction[22;0m, [1m/result[22;0m is 
  identical to [1m/return[22;0m.  When a macro that calls [1m/result[22;0m was called as a 
  [1mcommand[22;0m, [1m/result[22;0m has the additional effect of echoing the value of 
  <[4m[1mexpression[22;0m[24m> to the [1mtfout stream[22;0m.  [1m/Result[22;0m thus allows the same macro to be 
  called usefully as either a [1mcommand[22;0m or a [1mfunction[22;0m.  

  Note that [1m/return[22;0m and [1m/result[22;0m take the [4mstring[24m value of <[4mexpression[24m>.  This 
  is not a problem for integer- or float-valued expressions, since they 
  convert freely to strings and back without loss of information.  But if the 
  expression is an [1menumerated special variable[22;0m (e.g., borg), the value 
  returned will be its string value (e.g., "on"), not its integer value (e.g., 
  1).  To force it to use the integer value, you can use the unary plus 
  operator (e.g., +borg).  

  The return value of the last command (builtin or macro) is stored in [1m%{?}[22;0m.  
  The return value of a function (builtin or macro) is just the value of the 
  function.  

  These examples define several [1mmacros[22;0m intended to be called as a [1mfunctions[22;0m: 

    [1m/def[22;0m square = [1m/return[22;0m [1mpow[22;0m({1}, 2)

    [1m/def[22;0m hypot = [1m/return[22;0m [1msqrt[22;0m(square({1}) + square({2}))

    [1m/def[22;0m strrev = \
        /let len=$[[1mstrlen[22;0m({*})]%; \
        [1m/return[22;0m (len <= 1) ? {*} : \
            [1mstrcat[22;0m(strrev([1msubstr[22;0m({*},len/2)), strrev([1msubstr[22;0m({*},0,len/2)))

  If those examples had used [1m/result[22;0m instead of [1m/return[22;0m, they could also be 
  used as commands when echoing is more convenient.  For example, 

      [1m/eval[22;0m say My name backwards is [1m$([22;0m/strrev ${world_character}).

  See: [1m/if[22;0m, [1m/while[22;0m, [1m/test[22;0m, [1m/break[22;0m, [1m/exit[22;0m, [1mexpressions[22;0m, [1mevaluation[22;0m, [1mvariables[22;0m 

&/runtime

/runtime

  Usage: 

  [1m/runtime[22;0m <[4mcommand[24m>
  ____________________________________________________________________________

  Executes <[4mcommand[24m>, and prints the real time and cpu time used.  <[4mCommand[24m> 
  is not put through any additional [1msubstitution[22;0m before being executed.  The 
  return value of [1m/runtime[22;0m is that of <[4mcommand[24m>.  

  See: [1mcputime()[22;0m, [1mdebugging[22;0m.  

&mudwho
&rwho.tf
&/rwho

/rwho

  Usage: 

  [1m/REQUIRE[22;0m rwho.tf

  [1m/RWHO[22;0m
  [1m/RWHO[22;0m name=<[4mplayer[24m>
  [1m/RWHO[22;0m mud=<[4mmud[24m>
  ____________________________________________________________________________

  Gets a remote WHO list from a mudwho server.  The first form gives a 
  complete list, the other forms give partial lists.  Due to the short timeout 
  of the mudwho server, sometimes the complete list is sent even if the second 
  or third format is used (send complaints to the author or maintainer of the 
  mudwho server, not to me).  

  Make sure you [1m/load[22;0m rwho.tf _after_ you define your worlds, or rwho will be 
  the default world.  

&/savebind
&/savedef
&/savegag
&/savehilite
&/savehook
&/savetrig
&/save

/save

  Usage: 

  [1m/SAVE[22;0m [-a] <[4mfile[24m> [<[4mlist-options[24m>]
  ____________________________________________________________________________

  Saves specified [1mmacros[22;0m to <[4mfile[24m>.  The [1m<[4mlist-options[24m>[22;0m are the same as those 
  in the [1m/list[22;0m command; see "[1m/list[22;0m" for details.  Invisible [1mmacros[22;0m will not be 
  saved unless "-i" is specified.  

  If "-a" is specified, [1mmacros[22;0m will be appended to <[4mfile[24m>.  Otherwise, the 
  [1mmacros[22;0m will overwrite any existing contents of <[4mfile[24m>.  

  The return value of [1m/save[22;0m is the number of the last [1mmacro[22;0m listed, or 0 if no 
  [1mmacros[22;0m were listed (because of error or none matched the specified options). 

  The standard [1mmacro[22;0m library also defines commands that save macros of a 
  particular type: 
  [1m/savedef[22;0m 
          macros with names, but no [1mtriggers[22;0m, [1mhooks[22;0m, or [1mkeybindings[22;0m 
  [1m/savebind[22;0m 
          macros with [1mkeybindings[22;0m 
  [1m/savehilite[22;0m 
          macros with [1mtriggers[22;0m and [1mattributes[22;0m other than -ag 
  [1m/savegag[22;0m 
          macros with [1mtriggers[22;0m and the -ag [1mattribute[22;0m 
  [1m/savetrig[22;0m 
          macros with [1mtriggers[22;0m and no [1mattributes[22;0m 
  [1m/savehook[22;0m 
          macros with hooks 
  These commands take a filename argument; if it is omitted, a default file 
  name will be used.  No -a (append) option is allowed.  

  The [1m/save*[22;0m commands are useful if your [1mmacros[22;0m are few and simple, but if you 
  have many and/or complex [1mmacros[22;0m, you will probably find it easier to write 
  them with an editor and then [1m/load[22;0m them in tf, instead of writing them in tf 
  and [1m/save[22;0m'ing them to a file.  Avoiding [1m/save[22;0m allows you to keep the file(s) 
  nicely formatted, use comments, and organize them better.  Use whatever 
  works best for you.  

  Note that when tf starts, it does not automatically read files created with 
  any of the [1m/save[22;0m commands.  To make it do so, add the corresponding [1m/load[22;0m 
  command to your [1m.tfrc[22;0m file.  

  Except for its return value,
  [1m/save[22;0m [-a] <[4mfile[24m> [<[4mlist-options[24m>]
  is equivalent to
  [1m/eval[22;0m [1m/list[22;0m [<[4mlist-options[24m>] [1m%|[22;0m [1m/writefile[22;0m [-a] <[4mfile[24m> 

  See: [1mmacros[22;0m, [1mpatterns[22;0m, [1mattributes[22;0m, [1mlibrary[22;0m, [1m/def[22;0m, [1m/list[22;0m, [1m/load[22;0m, [1m/saveworld[22;0m 

&/saveworld

/saveworld

  Usage: 

  [1m/SAVEWORLD[22;0m [-a] [<[4mfile[24m>]
  ____________________________________________________________________________

  Saves world definitions to <[4mfile[24m> if specified, otherwise from the file 
  named in the body of the [1mWORLDFILE[22;0m macro.  

  If "-a" is given, world definitions will be appended to <[4mfile[24m>; otherwise, 
  the world definitions will replace any original contents of <[4mfile[24m>.  

  Note that when tf starts, it does not automatically read files created with 
  [1m/saveworld[22;0m.  To make it do so, add the [1m/loadworld[22;0m command to your [1m.tfrc[22;0m 
  file.  

  See: [1mworlds[22;0m, [1mlibrary[22;0m, [1m/addworld[22;0m, [1m/load[22;0m 

&send()
&/send

/send

  Function usage: 

  [1mSEND[22;0m(<[4mtext[24m>[, <[4mworld[24m>[, <[4mflags[24m>]])

  Command Usage: 

  [1m/SEND[22;0m [-W] [-T<[4mtype[24m>] [-w[<[4mworld[24m>]] [-n] <[4mtext[24m>
  ____________________________________________________________________________

  Sends <[4mtext[24m> to a world.  If no world is specified, the current world is 
  used.  By default, [1msend[22;0m does not execute SEND [1mhooks[22;0m.  

  In the function form, the optional <[4mflags[24m> is a string containing letters 
  that modify the function's behavior: 
  "h"     test for and invoke matching SEND [1mhooks[22;0m.  
  "u"     send <[4mtext[24m> unterminaed (i.e., without a CR LF end-of-line marker).  
  For backwards compatibility, the flags "o", "n", and "1" are ignored, and 
  the flags "0" and "f" are equivalent to "u".  

  Command [1moptions:[22;0m 
  -w<[4mworld[24m> 
          sends <[4mtext[24m> to <[4mworld[24m>.  
  -T<[4mtype[24m> 
          sends <[4mtext[24m> to all connected worlds with a type that matches the 
          pattern <[4mtype[24m>.  
  -W      sends <[4mtext[24m> to all connected worlds.  
  -n      send <[4mtext[24m> without an end-of-line marker (CR LF).  
  -h      test for and invoke matching SEND [1mhooks[22;0m.  

  The return value of [1msend[22;0m is 0 if the text is not successfully sent, nonzero 
  if it is.  

  See: [1mfunctions[22;0m.  

&/set

/set

  Usage: 

  [1m/SET[22;0m <[4mname[24m>=<[4mvalue[24m>
  [1m/SET[22;0m [<[4mname[24m> [<[4mvalue[24m>]]
  ____________________________________________________________________________

  In the first form, or with two arguments, [1m/set[22;0m will set the value of 
  [1mvariable[22;0m <[4mname[24m> to <[4mvalue[24m>.  With one argument, [1m/set[22;0m will display the value 
  of [1mvariable[22;0m <[4mname[24m>.  With no arguments, [1m/set[22;0m will display the value of all 
  internal [1mvariables[22;0m.  If the first form is used, there should be no spaces on 
  either side of the '='.  

  [1mVariable[22;0m <[4mname[24m> will be an internal [1mvariable[22;0m unless it has already been 
  defined as an environment [1mvariable[22;0m.  

  Note: The [1mvariables[22;0m 'L' and 'R' are reserved.  You should not assign values 
  to them.  

  When setting a variable, [1m/set[22;0m returns 1 if successful, 0 if not.  When 
  listing variables, [1m/set[22;0m returns the number of variables listed.  

  See: [1mvariables[22;0m, [1m/listvar[22;0m, [1m/setenv[22;0m, [1m/export[22;0m, [1m/let[22;0m, [1m/unset[22;0m, [1m/edvar[22;0m 

&/setenv

/setenv

  Usage: 

  [1m/SETENV[22;0m [<[4mname[24m> [<[4mvalue[24m>]]
  [1m/SETENV[22;0m <[4mname[24m>=<[4mvalue[24m>

  With two arguments, [1m/setenv[22;0m will set the value of <[4mname[24m> to <[4mvalue[24m> in the 
  environment.  With one argument, [1m/setenv[22;0m will display the value of <[4mname[24m>.  
  With no arguments, [1m/setenv[22;0m will display the value of all environment 
  [1mvariables[22;0m.  If the second form is used, spaces around the '=' will not be 
  stripped.  

  If <[4mname[24m> was already defined as an internal [1mvariable[22;0m, it will become an 
  environment [1mvariable[22;0m.  

  When setting a variable, [1m/setenv[22;0m returns 1 if successful, 0 if not.  When 
  listing variables, [1m/setenv[22;0m returns the number of variables listed.  

  See: [1mvariables[22;0m, [1m/listvar[22;0m, [1m/set[22;0m, [1m/export[22;0m 

&/sh

/sh

  Usage: 

  [1m/SH[22;0m [-q] [<[4mcommand[24m>]
  [1m/PSH[22;0m [<[4mcommand[24m>]
  ____________________________________________________________________________

  If no command is given, [1m/sh[22;0m and [1m/psh[22;0m execute an interactive shell named by 
  [1m%{SHELL}[22;0m.  With a <[4mcommand[24m>, [1m/sh[22;0m will execute <[4mcommand[24m> in the default shell 
  (/bin/sh on unix), and [1m/psh[22;0m will execute <[4mcommand[24m> in the shell named by 
  [1m%{SHELL}[22;0m.  <[4mCommand[24m> is executed interactively, so it may accept input and 
  may produce any output.  

  In [1mvisual mode[22;0m, [1m/sh[22;0m and [1m/psh[22;0m will fix the screen first, and restore it after 
  executing the shell.  [1m/Sys[22;0m does not.  

  If the -q option is given, /sh will be quiet: the [1mSHELL[22;0m [1mhook[22;0m will not be 
  called, and the "Executing" line will not be printed.  

  If the [1m%{shpause}[22;0m and [1m%{interactive}[22;0m flags are on, TF will wait for a 
  keypress before returning.  

  Note: calling [1m/sh[22;0m or [1m/psh[22;0m with arguments from a [1mtrigger[22;0m is very dangerous.  
  If not written carefully, such a [1mtrigger[22;0m could allow anyone connected to the 
  server to gain access to your shell account.  

  The return value of [1m/sh[22;0m and [1m/psh[22;0m is the exit status of the shell if it 
  exited normally, -1 otherwise.  Note that UNIX shell commands usually return 
  0 for success and nonzero for failure.  

  See: [1m/quote[22;0m, [1m/sys[22;0m, [1mutilities[22;0m ([1m/psh[22;0m) 

&/shift

/shift

  Usage: 

  [1m/SHIFT[22;0m [[4mn[24m]
  ____________________________________________________________________________

  Shifts the positional parameters left by <[4mn[24m>.  That is, the positional 
  parameters %([4mn[24m+1) ...  [1m%#[22;0m are renamed to [1m%1[22;0m ...  %(#-[4mn[24m).  If <[4mn[24m> is omitted, 
  1 is assumed.  

  [1m/shift[22;0m is useful only during [1mmacro[22;0m [1mexpansion[22;0m.  

  Example: 

    [1m/def[22;0m worlds = [1m/while[22;0m ({#}) [1m/world[22;0m [1m%1[22;0m%; [1m/shift[22;0m%; [1m/done[22;0m

  Then, the command "[1m/worlds[22;0m foo bar baz" would execute the commands "[1m/world[22;0m 
  foo", "[1m/world[22;0m bar", and "[1m/world[22;0m baz".  

  See: [1mvariables[22;0m, [1mevaluation[22;0m, [1mlist commands[22;0m 

&/signal

/signal

  Usage: 

  [1m/SIGNAL[22;0m [<[4msig[24m>]
  ____________________________________________________________________________

  Sends signal <[4msig[24m> to the tf process, or with no arguments, [1m/signal[22;0m lists 
  all valid signal names.  Valid signals usually include: HUP, INT, QUIT, 
  KILL, SEGV, TERM, USR1, USR2, and TSTP.  The complete list varies from 
  system to system.  

  See: [1msignals[22;0m, [1m/suspend[22;0m, [1mgetpid()[22;0m, [1mhooks[22;0m (SIGHUP, SIGTERM, SIGUSR1, SIGUSR2) 

&spell
&spelling
&/spell_line

spelling checker

  Usage: 

  [1m/REQUIRE[22;0m spell.tf

  [1m/SPELL_LINE[22;0m
  Keybinding: ^[s
  ____________________________________________________________________________

  After executing "[1m/require[22;0m spell.tf", typing "^[s" will call [1m/spell_line[22;0m to 
  report any misspellings in the current input line.  [1m/Spell_line[22;0m can of 
  course be bound to other keys with "[1m/def[22;0m [1m-b[22;0m".  

  [1m/Spell_line[22;0m assumes your system has a program called "spell" that reports 
  misspellings in its standard input.  

  See: [1minterface[22;0m, [1mkeys[22;0m 

&/split

/split

  Usage: 

  [1m/split[22;0m <[4margs[24m>
  ____________________________________________________________________________

  Sets [1m%{P1}[22;0m to the substring of <[4margs[24m> before the first '=', and sets [1m%{P2}[22;0m 
  to the substring of <[4margs[24m> after the first '='.  If there is no '=' in 
  <[4margs[24m>, [1m%{P1}[22;0m will contain the entire string and [1m%{P2}[22;0m will be empty.  [1m%{P0}[22;0m 
  will contain the entire string.  

  Spaces surrounding the '=' are stripped.  

  See: [1mgetopts()[22;0m 

&/sub

/sub

  Usage: 

  [1m/SUB[22;0m [OFF|ON|FULL]
  ____________________________________________________________________________

  Sets the flag [1m%{sub}[22;0m.  

  If the flag [1m%{sub}[22;0m is OFF (0), all lines except for [1mhistory[22;0m substitutions 
  (line beginning with '^') and commands (/) are sent as-is to the [1msocket[22;0m.  

  If the flag [1m%{sub}[22;0m is ON (1), the sequences "[1m%;[22;0m" and "%\" are substituted 
  with newlines, and the sequence "[1m%%[22;0m" is substituted with "%", and the 
  sequence "[1m\<[4mn[24m>[22;0m" is substituted with the character with decimal ASCII code 
  <[4mn[24m>.  

  If the flag [1m%{sub}[22;0m is FULL, text is processed just as if it were the body of 
  a [1mmacro[22;0m (see "[1mevaluation[22;0m") called without any arguments.  This allows you to 
  have in-line [1mmacros[22;0m in regular input.  

  The flag [1m%{sub}[22;0m defaults to 0 (off).  

  See: [1mgeneral[22;0m, [1mevaluation[22;0m 

&/substitute
&substitute()

/substitute

  [1mFunction[22;0m usage: 

  [1mSUBSTITUTE[22;0m(<[4mtext[24m> [, <[4mattrs[24m> [, <[4minline[24m>]])

  Command usage: 

  [1m/SUBSTITUTE[22;0m [-a<[4mattrs[24m>] [-p] <[4mtext[24m>
  ____________________________________________________________________________

  When called from a [1mtrigger[22;0m (directly or indirectly), the entire [1mtrigger[22;0ming 
  line is replaced with <[4mtext[24m>.  After a [1m/substitute[22;0m, it will appear as if 
  <[4mtext[24m> is what caused the [1mtrigger[22;0m; the original line is lost.  In 
  particular, this means when [1m/substitute[22;0m is called from a [1mfall-thru[22;0m [1mtrigger[22;0m, 
  [1mtriggers[22;0m of lower [1mpriority[22;0m will be compared against <[4mtext[24m> instead of the 
  original line.  

  [1mOptions[22;0m and arguments: 
  command: -a<[4mattrs[24m> 
  function: <[4mattrs[24m> 
          Give <[4mtext[24m> the [1mattributes[22;0m described by <[4mattrs[24m>.  These are added to 
          the original line's [1mattributes[22;0m unless <[4mattrs[24m> include the "x" 
          [1mattribute[22;0m.  
  command: -p 
  function: <[4minline[24m> = "on" or 1 
          Interpet @{<[4mattr[24m>} strings as commands to set [1mattributes[22;0m inline, as 
          in [1m/echo[22;0m.  (See [1m/echo[22;0m).  

  Example: 

  On a mud that uses MUFpage, you could set your #prepend string to "##page>", 
  and define a [1mtrigger[22;0m like: 

    [1m/def[22;0m [1m-ah[22;0m [1m-t[22;0m"##page> *" [1mhilite[22;0m_mufpage = [1m/substitute[22;0m [1m%-1[22;0m

  This will match no matter what page format the sender uses, and strip off 
  the "##page>" so you never see it.  

  For another example, see [1m/replace[22;0m.  

  See: [1mtriggers[22;0m 

&/suspend

/suspend

  Usage: 

  [1m/SUSPEND[22;0m
  ____________________________________________________________________________

  Suspends the TF process, if your system and shell support job control.  This 
  has the same effect as typing ^Z on most UNIX-like systems.  When TF is 
  resumed, it redraws the screen and processes all [1m/repeat[22;0m and [1m/quote[22;0m commands 
  that were scheduled to run while TF was suspended and processes all text 
  that was received while TF was suspended.  

  See: [1msignals[22;0m, [1m/signal[22;0m.  

&/sys

/sys

  Usage: 

  [1m/SYS[22;0m <[4mshell-command[24m>
  ____________________________________________________________________________

  Executes <[4mshell-command[24m>.  The command is executed without a tty, so it 
  should have no input, and its output, if any, should be plain text.  The 
  command's stdout and stderr are echoed to tf's output window.  [1m/sys[22;0m differs 
  from [1m/sh[22;0m in that [1m/sys[22;0m can not do an interactive shell command, but does not 
  redraw the screen or produce any extra messages.  

  Note: calling [1m/sys[22;0m with arguments from a [1mtrigger[22;0m is dangerous.  If not 
  written carefully, such a [1mtrigger[22;0m could allow anyone with access to the 
  server to gain access to your shell account.  

  The return value of [1m/sys[22;0m is the exit status of the shell if it exited 
  normally, -1 otherwise.  Note that UNIX shell commands usually return 0 for 
  success and nonzero for failure, which is the opposite of the TF convention. 

  [1m/sys[22;0m executes synchronously.  To execute a command asynchronously (in the 
  background), use [1m/quote[22;0m without the -S option.  

  See: [1m/sh[22;0m, [1m/quote[22;0m 

&/telnet

/telnet

  Usage: 

  [1m/TELNET[22;0m <[4mhost[24m> [<[4mport[24m>]
  ____________________________________________________________________________

  Connect to a line-based telnet host.  The telnet login port is used if 
  <[4mport[24m> is omitted.  

  Note that TF operates strictly in line-by-line mode, but telnetd (the server 
  running on the telnet login port) expects character-by- character mode.  So, 
  simple shell operations and anything else which is basically line-by-line 
  should work without much difficulty, but anything that tries to control the 
  screen or expects single keystroke input will [4mnot[24m work.  [1m/Telnet[22;0m is somewhat 
  useful, but not useful enough to alter the fundamental line-by-line nature 
  of TF.  If you want a general telnet client, you know where to find it.  

  TF supports most of the TELNET protocol (even if a command other than 
  [1m/telnet[22;0m was used to connect).  TF implements the TELNET options ECHO (lets 
  server control echoing of input), SGA (suppress GOAHEAD), EOR (allows use of 
  END-OF-RECORD in [1mprompts[22;0m), NAWS (allows TF to send window size information 
  to the server), TTYPE (allows server to ask about the terminal type), and 
  BINARY (allows transmission of 8-bit characters).  For TTYPE queries, TF 
  responds "TINYFUGUE", "ANSI-ATTR", "ANSI", and "UNKNOWN", in that order.  
  For information on TELNET protocol, see RFC 854 and 1123.  See also: 
  [1mprompts[22;0m.  

  See: [1m/addtelnet[22;0m, [1m/connect[22;0m, [1m%telopt[22;0m, [1m%binary_eol[22;0m, [1mprotocols[22;0m 

&/test

/test

  Usage: 

  [1m/TEST[22;0m <[4m[1mexpression[22;0m[24m>
  ____________________________________________________________________________

  [1m/test[22;0m evaluates the <[4m[1mexpression[22;0m[24m> and returns its value, also setting the 
  special [1mvariable[22;0m [1m%?[22;0m.  The return value may be any type (before version 4.0, 
  only integer values were allowed).  A new [1mvariable[22;0m scope is NOT created.  

  [1m/Test[22;0m can be useful for evaluating an [1mexpression[22;0m for its side effects, 
  ignoring the return value.  For example, the command "[1m/test[22;0m [1mkbdel[22;0m([1mkbpoint()[22;0m 
  - 1)" will perform a backspace, and "[1m/test[22;0m [1mregmatch[22;0m('foo(.*)', 'foobar')" 
  will assign "bar" to [1m%P1[22;0m.  

  Before version 3.5, [1m/test[22;0m was frequently used as the condition of an [1m/IF[22;0m or 
  [1m/WHILE[22;0m statement.  This is no longer needed, since [1m/IF[22;0m and [1m/WHILE[22;0m can now 
  take an [1mexpression[22;0m as a condition.  

  Before version 4.0, [1m/test[22;0m was sometimes used to set the return value of a 
  [1mmacro[22;0m, since a [1mmacro[22;0m's return value is that of the last command executed.  
  The preferred way to do this now is with [1m/return[22;0m or [1m/result[22;0m.  

  See: [1m/return[22;0m, [1m/if[22;0m, [1m/while[22;0m, [1mexpressions[22;0m, [1mevaluation[22;0m, [1mvariables[22;0m 

&/textencode

textencode()

  [1m/require[22;0m textencode.tf

  Function usage: 

  [1mtextencode[22;0m(<[4mstring[24m>)
  [1mtextdecode[22;0m(<[4mencodedstring[24m>)
  ____________________________________________________________________________

  [1mtextencode[22;0m converts <[4mstring[24m> to a form that contains only letters, digits, 
  and underscores.  [1mtextdecode[22;0m converts <[4mencodedstring[24m> (returned by a 
  previous call to [1mtextencode[22;0m) back to the original string.  

  These two functions can be useful for converting arbitrary text, such as a 
  world name or the name of a player on a mud, into a form that is safe to use 
  as part of a tf [1mvariable[22;0m or [1mmacro[22;0m name, or a filename.  

  The following example records the time a player connects to the mud, and is 
  safe even if the player name contains characters that are not legal in tf 
  [1mvariable[22;0m names:
  [1m/def[22;0m [1m-m[22;0mglob [1m-t[22;0m'{*} has connected.' record_connect_time = \
       [1m/set[22;0m connect_time_[1m$[[22;0m[1mtextencode[22;0m([1m{1}[22;0m)]=[1m$[[22;0m[1mtime[22;0m()] 

  See: [1mfunctions[22;0m 

&/fgrep
&/grep
&/egrep
&/readfile
&/writefile
&/head
&/wc
&/tee
&/copyio
&/fmt
&/uniq
&/randline
&textutil
&textutil.tf

Text Utilities

  [1m/REQUIRE[22;0m textutil.tf
  ____________________________________________________________________________

  The library file [1mtextutil.tf[22;0m defines several unix-like commands that are 
  particularly convenient when used with the [1m%|[22;0m pipe to redirect their input 
  or output.  

  In the descriptions below, <[4mfilename[24m> is the name of a file, and <[4min[24m> and 
  <[4mout[24m> are handles of [1mtfio streams[22;0m.  When <[4min[24m> is optional, its default is 
  [1mtfin[22;0m.  

  [1m/fgrep[22;0m [-cvi] <[4mpattern[24m> 
  [1m/grep[22;0m [-cv] <[4mpattern[24m> 
  [1m/egrep[22;0m [-cvi] <[4mpattern[24m> 
          These commands search [1mtfin[22;0m for lines that match the given pattern, 
          and by default prints those lines.  For [1m/fgrep[22;0m, a line must contain 
          <[4mpattern[24m> to match; for [1m/grep[22;0m, the [4mentire[24m line must match the [1mglob[22;0m 
          pattern <[4mpattern[24m>; for [1m/egrep[22;0m, it must match the [1mregexp[22;0m pattern 
          <[4mpattern[24m>.  
          [1mOptions[22;0m: 
          -c      print only the count of matching lines.  
          -v      select only non-matching lines.  
          -i      ignore case (for /fgrep and /egrep only; /grep always 
                  ignores case).  
          Note: these commands are not compatible with those defined in the 
          old library file [1mgrep.tf[22;0m.  

  [1m/readfile[22;0m <[4mfilename[24m> 
          Reads lines from <[4mfilename[24m> and writes them to [1mtfout[22;0m.  

  [1m/writefile[22;0m [-a] <[4mfilename[24m> 
          Reads lines from [1mtfin[22;0m and writes them to file <[4mfilename[24m>.  
          [1mOptions[22;0m: 
          -a      append to file instead of overwriting.  

  [1m/head[22;0m [-n<[4mcount[24m>] [<[4min[24m>] 
          Reads the first <[4mcount[24m> (default 10) lines from <[4min[24m> or [1mtfin[22;0m and 
          writes them to [1mtfout[22;0m.  

  [1m/wc[22;0m [-lwc] [<[4min[24m>] 
          Reads lines from <[4min[24m> or [1mtfin[22;0m and prints the count of lines, 
          space-separated words, and characters that were read.  
          [1mOptions[22;0m: 
          -l      Print the count of lines only.  
          -w      Print the count of words only.  
          -c      Print the count of characters only.  

  [1m/tee[22;0m <[4mout[24m> 
          Reads lines from [1mtfin[22;0m and echoes them to <[4mout[24m> and [1mtfout[22;0m.  

  [1m/copyio[22;0m <[4min[24m> <[4mout[24m> 
          Reads lines from <[4min[24m> and writes them to <[4mout[24m>.  This can be useful, 
          for example, when you want to send text from a [1mtfio[22;0m stream to a 
          command that reads only [1mtfin[22;0m: 

              /copyio <[4min[24m> o %| /<[4mcommand[24m>
              

  [1m/fmt[22;0m    Copies [1mtfin[22;0m to [1mtfout[22;0m, with adjacent non-blank lines joined.  

  [1m/uniq[22;0m   Copies [1mtfin[22;0m to [1mtfout[22;0m, with adjacent duplicate lines removed.  

  [1m/randline[22;0m [<[4min[24m>] 
          Copies one randomly selected line from <[4min[24m> or [1mtfin[22;0m to [1mtfout[22;0m.  

  ____________________________________________________________________________

  See: [1mtfio[22;0m, [1mevaluation[22;0m, [1msubstitution[22;0m, [1moldgrep[22;0m 

&/tick
&/tickon
&/tickoff
&/tickset
&/ticksize

/tick

  Usage: 

  [1m/REQUIRE[22;0m tick.tf

  [1m/tick[22;0m
  [1m/tickoff[22;0m
  [1m/tickon[22;0m
  [1m/tickset[22;0m
  [1m/ticksize[22;0m <[4mn[24m>
  ____________________________________________________________________________

  The [1m/tick*[22;0m commands implement dikumud tick counting, similar to tintin.  
  When the ticker is started with [1m/tickon[22;0m, it will warn you 10 seconds before 
  each tick, and print "TICK" on the tick.  

  The messages can be changed by redefining the /tick_warn (10-second warning) 
  and /tick_action ("TICK") macros.  You can make them perform any tf command, 
  not just printing.  

  It is up to you to start the ticker in synch with the mud.  If the mud 
  prints something on a tick, you can define a [1mtrigger[22;0m on that which calls 
  [1m/tickon[22;0m.  

  [1m/Tick[22;0m displays the time remaining until the next tick.  

  [1m/Tickoff[22;0m stops the ticker.  

  [1m/Tickon[22;0m and [1m/tickset[22;0m reset and start the ticker.  

  [1m/Ticksize[22;0m sets the tick period to <[4mn[24m> seconds (the default is 75).  

  See: [1m/require[22;0m, [1mtiming[22;0m, [1mprompts[22;0m 

&/time

/time

  Usage: 

  [1m/TIME[22;0m [<[4mformat[24m>]
  ____________________________________________________________________________

  Displays the current time.  <[4mFormat[24m> is described under "[1mftime()[22;0m".  If 
  <[4mformat[24m> is omitted, [1m%{time_format}[22;0m is used.  

  See: [1mtime()[22;0m, [1mftime()[22;0m, [1mmktime()[22;0m, [1m%TZ[22;0m, [1m%time_format[22;0m, [1m%clock[22;0m, [1midle()[22;0m 

&/toggle

/toggle

  Usage: 

  [1m/TOGGLE[22;0m <[4mvariable[24m>
  ____________________________________________________________________________

  If <[4mvariable[24m> has a value of 0, its value will be set to "1".  If <[4mvariable[24m> 
  has a non-zero value, its value will be set to "0".  

  See: [1mvariables[22;0m 

&/tr

/tr

  Usage: 

  [1m/REQUIRE[22;0m tr.tf

  [1m/TR[22;0m <[4mdomain[24m> <[4mrange[24m> <[4mstring[24m>
  ____________________________________________________________________________

  <[4mDomain[24m> and <[4mrange[24m> are lists of characters of equal length.  Each 
  character in <[4mstring[24m> that appears in <[4mdomain[24m> is translated to the 
  corresponding character in <[4mrange[24m>, and the resulting string is printed.  

  Example:
  command: [1m/def[22;0m biff = [1m/tr[22;0m OIS.  01Z! $[[1mtoupper[22;0m({*})]
  command: /biff TinyFugue is cool wares, dude. 
  output: T1NYFUGUE 1Z C00L WAREZ, DUDE!

  See: [1m/replace[22;0m, [1mexpressions[22;0m, [1mfunctions[22;0m 

&/act
&/trigpc
&/trigp
&/trigc
&/trig

/trig

  Usage: 

  [1m/TRIG[22;0m <[4mpattern[24m> = <[4mbody[24m>
  [1m/TRIGP[22;0m <[4mpriority[24m> <[4mpattern[24m> = <[4mbody[24m>
  [1m/TRIGC[22;0m <[4mchance[24m> <[4mpattern[24m> = <[4mbody[24m>
  [1m/TRIGPC[22;0m <[4mpriority[24m> <[4mchance[24m> <[4mpattern[24m> = <[4mbody[24m>
  ____________________________________________________________________________

  Creates an unnamed [1mmacro[22;0m that will [1mtrigger[22;0m on <[4m[1mpattern[22;0m[24m> and execute <[4mbody[24m>.  
  If <[4mchance[24m> is given with [1m/trigc[22;0m or [1m/trigpc[22;0m, it will be the percentage 
  probability of the [1mtrigger[22;0m going off; default is 100%.  If <[4mpriority[24m> is 
  given with [1m/trigp[22;0m or [1m/trigpc[22;0m, it will be the [1mpriority[22;0m of the [1mtrigger[22;0m; 
  default is 0.  The matching style of the [1mtrigger[22;0m is determined by the global 
  [1mvariable[22;0m [1m%{matching}[22;0m.  

  If the command fails it returns 0.  Otherwise, it creates a new [1mmacro[22;0m and 
  returns its (positive) number (useful in [1m/undefn[22;0m and [1m/edit[22;0m).  

  [1m/trig[22;0m is equivalent to: [1m/def[22;0m [1m-t[22;0m<[4mpattern[24m> = <[4mbody[24m>. 
  [1m/trigp[22;0m is equivalent to: [1m/def[22;0m [1m-p[22;0m<[4mpriority[24m> [1m-t[22;0m<[4mpattern[24m> = <[4mbody[24m>. 
  [1m/trigc[22;0m is equivalent to: [1m/def[22;0m [1m-c[22;0m<[4mchance[24m> [1m-t[22;0m<[4mpattern[24m> = <[4mbody[24m>. 
  [1m/trigpc[22;0m is equivalent to: [1m/def[22;0m [1m-p[22;0m<[4mpriority[24m> [1m-c[22;0m<[4mchance[24m> [1m-t[22;0m<[4mpattern[24m> = <[4mbody[24m>.

  Note: the [1m/trig[22;0m commands create [1mmacros[22;0m without names.  Thus each [1m/trig[22;0m 
  command will create a new [1mmacro[22;0m macro instead of replacing an old [1mmacro[22;0m.  
  For this reason, it is usually better to use [1m/def[22;0m and give your [1mmacros[22;0m 
  names.  

  See: [1mtriggers[22;0m, [1mevaluation[22;0m, [1mpatterns[22;0m, [1m/def[22;0m, [1m/untrig[22;0m 

&/trigger

/trigger

  Usage: 

  [1m/TRIGGER[22;0m [-ln] [-g] [-w[<[4mworld[24m>]] [-h[<[4mevent[24m>]] <[4mtext[24m>
  ____________________________________________________________________________

  Executes [1mmacros[22;0m with [1mtriggers[22;0m or [1mhook[22;0m arguments that match <[4mtext[24m>, just as 
  if <[4mtext[24m> had come from a [1msocket[22;0m or a hook event had occurred with <[4mtext[24m> as 
  its arguments.  The return value of [1m/trigger[22;0m is the number of (non-[1mquiet[22;0m) 
  [1mmacros[22;0m that were executed.  [1m/Trigger[22;0m is useful for debugging [1mtriggers[22;0m and 
  [1mhooks[22;0m.  

  [1mOptions:[22;0m 
  -g      Match "global" [1mtriggers[22;0m or [1mhooks[22;0m that were not defined with [1m/def[22;0m [1m-w[22;0m 
  -w<[4mworld[24m> 
          Match [1mtriggers[22;0m or [1mhooks[22;0m for <[4mworld[24m>, or the [1mcurrent[22;0m [1mworld[22;0m if <[4mworld[24m> 
          is omitted.  
  -h<[4mevent[24m> 
          Match [1mhooks[22;0m where <[4mevent[24m> matches the hook event and <[4mtext[24m> matches 
          the hook argument pattern.  Without -h, [1m/trigger[22;0m matches [1mtriggers[22;0m, 
          not [1mhooks[22;0m.  
  -n      Do not execute any of the matched macros; instead, display a list of 
          each macro that would have matched, including its [1mfallthru flag[22;0m, 
          [1mpriority[22;0m, and name.  (Note that if any macro in the list would have 
          executed [1msubstitute()[22;0m or [1m/substitute[22;0m, the macros listed after it may 
          not be correct.) 
  -l      Like -n, but list each macro in full, as if by [1m/list[22;0m.  
  If neither -g nor -w options are given, both are assumed.  That is, <[4mtext[24m> 
  is matched against global [1mtriggers[22;0m or [1mhooks[22;0m, as well as [1mtriggers[22;0m or [1mhooks[22;0m 
  for the [1mcurrent[22;0m [1mworld[22;0m.  

  See: [1mtriggers[22;0m, [1mhooks[22;0m, [1mdebugging[22;0m, [1m/def[22;0m 

&/false
&/:
&/true

/true

  Usage: 

  [1m/TRUE[22;0m
  [1m/FALSE[22;0m
  ____________________________________________________________________________

  [1m/True[22;0m does nothing, and returns nonzero.  

  [1m/False[22;0m does nothing, and returns zero.  

  /: is the same as [1m/true[22;0m.  

&/unbind

/unbind

  Usage: 

  [1m/UNBIND[22;0m <[4msequence[24m>
  ____________________________________________________________________________

  Removes a [1mmacro[22;0m with the keybinding <[4msequence[24m>.  

  See: [1mgeneral[22;0m, [1m/bind[22;0m, [1m/purge[22;0m 

&/undef

/undef

  Usage: 

  [1m/UNDEF[22;0m <[4mname[24m>... 
  ____________________________________________________________________________

  For each <[4mname[24m> given, [1m/undef[22;0m removes the definition of the [1mmacro[22;0m with that 
  name.  

  The return value of [1m/undef[22;0m is the number of macros that were removed.  

  See: [1mmacros[22;0m, [1m/def[22;0m, [1m/purge[22;0m, [1m/undefn[22;0m, [1m/undeft[22;0m, [1m/untrig[22;0m, [1m/unhook[22;0m 

&/undefn

/undefn

  Usage: 

  [1m/UNDEFN[22;0m <[4mnumber[24m> ... 
  ____________________________________________________________________________

  Removes [1mmacros[22;0m with the numbers specified in the arguments.  [1mMacro[22;0m numbers 
  can be determined with [1m/list[22;0m, or from the return value of the command used 
  to create the [1mmacro[22;0m.  

  See: [1mmacros[22;0m, [1m/def[22;0m, [1m/list[22;0m, [1m/purge[22;0m, [1m/undef[22;0m 

&/undeft

/undeft

  Usage: 

  [1m/UNDEFT[22;0m <[4mtrigger[24m>
  ____________________________________________________________________________

  Removes a [1mmacro[22;0m with a [1mtrigger[22;0m associated with it that is [1mtrigger[22;0med by the 
  pattern <[4mtrigger[24m>.  <[4mTrigger[24m> is matched against existing [1mtriggers[22;0m using 
  simple comparison.  

  See: [1mmacros[22;0m, [1mtriggers[22;0m, [1m/def[22;0m, [1m/purge[22;0m, [1m/undef[22;0m 

&/unhook

/unhook

  Usage: 

  [1m/UNHOOK[22;0m <[4mevent[24m> [<[4mpattern[24m>]
  ____________________________________________________________________________

  Removes a [1mmacro[22;0m with an associated [1mhook[22;0m on <[4mevent[24m> <[4mpattern[24m>.  

  See: [1mhooks[22;0m, [1m/hook[22;0m, [1m/purge[22;0m, [1m/undef[22;0m 

&/unset

/unset

  Usage: 

  [1m/UNSET[22;0m <[4mname[24m>
  ____________________________________________________________________________

  [1m/Unset[22;0m removes the value of [1mvariable[22;0m <[4mname[24m>.  

  [1m/Unset[22;0m returns 0 if an error occurred, nonzero otherwise.  

  See: [1mvariables[22;0m, [1m/set[22;0m, [1m/setenv[22;0m, [1m/let[22;0m 

&/untrig

/untrig

  Usage: 

  [1m/UNTRIG[22;0m [-a<[4mattrs[24m>] <[4mtrigger[24m>
  ____________________________________________________________________________

  Removes a [1mmacro[22;0m with an associated [1mtrigger[22;0m that is [1mtrigger[22;0med by the pattern 
  <[4mtrigger[24m> and has [1mattributes[22;0m <[4mattrs[24m>.  If -a<[4mattrs[24m> is omitted, -an is 
  assumed.  <[4mTrigger[24m> is matched against existing [1mtriggers[22;0m using simple 
  comparison.  

  See: [1mtriggers[22;0m, [1m/trig[22;0m, [1m/purge[22;0m, [1m/undef[22;0m 

&/unworld

/unworld

  Usage: 

  [1m/UNWORLD[22;0m <[4mname[24m>... 
  ____________________________________________________________________________

  For each <[4mname[24m> given, [1m/unworld[22;0m removes the definition of the world with 
  that name.  The [1mhistory[22;0m for removed worlds will be deleted, but some or all 
  of the lines may still exist in the global [1mhistory[22;0m.  

  The return value of [1m/unworld[22;0m is the number of worlds that were removed.  

  See: [1mworlds[22;0m, [1m/addworld[22;0m 

&/ver
&/version

/version

  Usage: 

  [1m/VERSION[22;0m
  [1m/VER[22;0m
  ____________________________________________________________________________

  [1m/Version[22;0m displays the TinyFugue version you're running and the operating 
  system for which it was compiled (if known).  

  [1m/Ver[22;0m displays an abbreviated version number.  

  The latest version of TF can be found at [1mhttp://tinyfugue.sourceforge.net/[22;0m.  

  See: [1m/changes[22;0m 

&/watchdog

/watchdog

  Usage: 

  [1m/WATCHDOG[22;0m [OFF|ON]
  [1m/WATCHDOG[22;0m <[4mn1[24m> [<[4mn2[24m>]
  ____________________________________________________________________________

  Sets the flag [1m%{watchdog}[22;0m.  This flag determines whether Fugue will watch 
  for identical lines and suppress them.  Fugue looks for lines which have 
  occurred <[4mn1[24m> times out of <[4mn2[24m> (<[4mn1[24m> defaults to 2 and <[4mn2[24m> to 5) and 
  suppress them, so with the default settings Fugue will suppress any lines 
  that have occurred 2 times out of the last 5.  

  The <[4mn1[24m> and <[4mn2[24m> settings for [1m/watchdog[22;0m are distinct from the <[4mn1[24m> and <[4mn2[24m> 
  settings for [1m/watchname[22;0m.  

  The flag [1m%{watchdog}[22;0m defaults to 0 (off).  

  See: [1m%watchdog[22;0m, [1m/watchname[22;0m 

&/watchname

/watchname

  Usage: 

  [1m/WATCHNAME[22;0m [OFF|ON]
  [1m/WATCHNAME[22;0m <[4mn1[24m> [<[4mn2[24m>]
  ____________________________________________________________________________

  Sets the flag [1m%{watchname}[22;0m.  This flag determines whether Fugue will watch 
  for players displaying lots of output.  Fugue looks for names which have 
  begun the line <[4mn1[24m> times out of <[4mn2[24m> (<[4mn1[24m> defaults to 4 and <[4mn2[24m> to 5) and 
  [1mgag[22;0m that person (with a message), so with the default settings Fugue will 
  [1mgag[22;0m any person whose name has begun 4 of the last 5 lines.  

  The <[4mn1[24m> and <[4mn2[24m> settings for [1m/watchname[22;0m are distinct from the <[4mn1[24m> and 
  <[4mn2[24m> settings for [1m/watchdog[22;0m.  

  The flag [1m%{watchname}[22;0m defaults to 0 (off).  

  See: [1m%watchname[22;0m, [1m/watchdog[22;0m 

&/while
&/do
&/done
&/while

/while

  Usage: 

  [1m/WHILE[22;0m ([4mexpr[24m) [4mlist[24m [1m/DONE[22;0m
  [1m/WHILE[22;0m [4mlist[24m [1m/DO[22;0m [4mlist[24m [1m/DONE[22;0m
  ____________________________________________________________________________

  The <[4mlist[24m>s may be any list of commands.  The return value of a <[4mlist[24m> is 
  the return value of the last command executed in the <[4mlist[24m>.  Each <[4mlist[24m> 
  must be terminated by "[1m%;[22;0m".  

  The <[4mlist[24m> or <[4mexpr[24m> following the [1m/WHILE[22;0m is called the condition.  The 
  condition is executed or evaluated, and if its result is non-zero, the next 
  <[4mlist[24m> is executed.  This sequence is repeated until the condition returns 
  zero.  

  The [1m/BREAK[22;0m command can be used within the loop to terminate the loop early.  
  The loop can also be terminated early by catching a SIGINT (usually 
  generated by typing ^C).  If the [1mvariable[22;0m [1m%{max_iter}[22;0m is non-zero, the loop 
  will terminate automatically if the number of iterations reaches that 
  number.  

  When [1m/WHILE[22;0m is used on the command line, "[1m%;[22;0m" command separation will be 
  done even if [1m%sub[22;0m=off.  Of course, full substitution will be done if 
  [1m%sub[22;0m=full.  

  Example: 


    [1m/def[22;0m count = \
        [1m/let[22;0m i=1%; \
        [1m/while[22;0m (i <= {1}) \
            say %{i}%; \
            [1m/let[22;0m i=$[i + 1]%; \
        [1m/done[22;0m

  The command "/count 10" will execute the commands "say 1", "say 2", ...  
  "say 10".  

  See: [1mevaluation[22;0m, [1m/test[22;0m, [1m/break[22;0m, [1m/for[22;0m 

&/world

/world

  Usage: 

  [1m/WORLD[22;0m [-lqnxfb] [<[4mworld[24m>]
  [1m/WORLD[22;0m <[4mhost[24m> <[4mport[24m>
  ____________________________________________________________________________

  If <[4mworld[24m> is already connected, "[1m/world[22;0m <[4mworld[24m>" is equivalent to "[1m/fg[22;0m 
  <[4mworld[24m>", and brings <[4mworld[24m> into the [1mforeground[22;0m.  If <[4mworld[24m> is not 
  connected, "[1m/world[22;0m <[4mworld[24m>" is equivalent to "[1m/connect[22;0m <[4mworld[24m>", and 
  attempts to open a connection to that world.  

  The second form is equivalent to "[1m/connect[22;0m <[4mhost[24m> <[4mport[24m>".  

  The -lqnxfb options are the same as those for [1m/fg[22;0m and [1m/connect[22;0m.  

  See: [1m/connect[22;0m, [1m/fg[22;0m 

&
&hilites
&gags
&underline
&reverse
&flash
&dim
&bell
&bold
&attrs
&attributes
&display attributes
&attribute

display attributes

  Many TF commands take an argument to specify an [1mattribute[22;0m list, containing 
  one or more of: "n" (none), "x" (exclusive), "g" ([1mgag[22;0m), "G" (nohistory), "L" 
  (nolog), "A" (noactivity), "u" (underline), "r" (reverse), "B" (bold), "b" 
  (bell), "h" ([1mhilite[22;0m), "E" (error), "W" (warning), or "C<[4mcolor[24m>" ([1mcolor[22;0m).  
  These [1mattributes[22;0m are used to display text associated with the command.  Use 
  commas to separate attributes within an attribute list; commas may be 
  omitted between single-letter attributes.  For example, "BuCred,Cbgyellow" 
  means bold underlined red text on a yellow background.  

  "None" ("n") is useful for finding macros without attributes (e.g.  "[1m/list[22;0m 
  -an") or for turning off attributes in the middle of a line (e.g.  "[1m/echo[22;0m -p 
  foo @{u}bar@{n} baz").  

  Normally, new attributes are combined with the pre-existing attributes.  But 
  if the new attributes include "x" (exclusive), the pre-existing display 
  attributes are turned off first.  So, for example, if one trigger with [1m-a[22;0mu 
  and another trigger with [1m-P[22;0mr match the same line, the whole line will be 
  underlined and part of it will also be reversed; but if the second trigger 
  had [1m-P[22;0mxr instead, then most of the line would be underlined, and part would 
  be reversed but not underlined.  

  The "G" (nohistory) [1mattribute[22;0m prevents the line from being recorded in 
  [1mhistory[22;0m.  The "L" (nolog) [1mattribute[22;0m prevents the line from being recorded in 
  a [1mlog[22;0m file.  

  The "A" (noactivity) [1mattribute[22;0m prevents the line from causing an [1mACTIVITY[22;0m 
  [1mhook[22;0m or a nonzero [1mmoresize[22;0m().  For example, the following command prevents 
  people connecting and disconnecting from counting as activity: 

        [1m/def[22;0m [1m-a[22;0mA [1m-q[22;0m [1m-t[22;0m"{*} has {*connected.}" noact_connect
    

  The "C<[4mname[24m>" ([1mColor[22;0m) [1mattribute[22;0m allows you to name a color.  The "C" must be 
  followed by the <[4mname[24m> of the color; a comma after the <[4mname[24m> can be used to 
  separate it from attributes that follow it.  Depending on your terminal and 
  how tf was compiled, there may be 8, 16, or 256 colors available.  See: 
  [1mcolor[22;0m.  

  The "h" ([1mhilite[22;0m), "E" (error), and "W" (warning) [1mattributes[22;0m are special.  
  When "h", "E", or "W" is specified, it is replaced with the [1mattributes[22;0m 
  listed in the [1m%{hiliteattr}[22;0m, [1m%{error_attr}[22;0m, or [1m%{warning_attr}[22;0m [1mvariable[22;0m, 
  respectively.  Additionally, error and warning messages generated by tf 
  automatically have the "E" and "W" [1mattributes[22;0m, so you can alter their 
  appearance by setting the corresponding variable.  For example, the commands 

        [1m/set[22;0m [1mhiliteattr[22;0m=r
        [1m/echo[22;0m -ahu foobar

  will display the word "foobar" with reverse and underline [1mattributes[22;0m.  
  [1m%{hiliteattr}[22;0m makes it easy to change the meaning of all your hilite macros 
  at once, without editing each one individually.  

  The "f" (flash) and "d" (dim) [1mattributes[22;0m are accepted for backward 
  compatiblity, but ignored.  

  All [1mattributes[22;0m except 'n' may be combined usefully.  (Even [1mgag[22;0ms can be 
  combined with other [1mattributes[22;0m: combining 'g' and 'B', for example, will [1mgag[22;0m 
  the text initially, but will display it as bold if it is recalled with 
  [1m/recall[22;0m -ag.) 

  It is possible to apply [1mattributes[22;0m to a part of a line, using [1m/partial[22;0m or 
  the [1m-P[22;0m option of [1m/def[22;0m.  If two or more partial [1mattributes[22;0m overlap, their 
  effects will be combined (unless the "x" attribute is used).  For example, 
  overlapping bold and reverse will appear bold and reverse; overlapping blue 
  and red will appear magenta.  

  Ansi [1mattribute[22;0m codes sent by the server will be interpreted by tf if 
  [1m%{emulation}[22;0m is set to "ansi_attr".  See: [1m%emulation[22;0m.  

  As of version 5.0, [1mattributes[22;0m in string values are preserved by just about 
  every string operation, including [1mcommands[22;0m, [1mvariables[22;0m, [1mexpression operators[22;0m, 
  [1mfunctions[22;0m, [1mregexp substitutions[22;0m, [1m$() command substitution[22;0m, and [1mstatus bar 
  field expressions[22;0m.  The [1minline_attr()[22;0m function can be used to convert 
  attribute codes within a string to actual attributes.  

  [1mAttributes[22;0m not supported by your terminal type will be stored, but not 
  displayed.  

&%catch_ctrls

%catch_ctrls

  See: [1m%emulation[22;0m 
&/color_off
&color
&colour
&colours
&256colors
&colors

colors

  Color is enabled by default.  To disable it, use "/color_off"; to re-enable 
  color using ANSI codes, use "/color_on".  

  The color [1mattribute[22;0m allows you to specify a foreground color with "C<[4mname[24m>" 
  or a background color with "Cbg<[4mname[24m>".  Any terminal that supports color 
  should support the 8 basic colors: [30mblack[0m (black), [31mred[0m, [32mgreen[0m, [33myellow[0m, [34mblue[0m, 
  [35mmagenta[0m, [36mcyan[0m, [37mwhite[0m (white).  (If you are reading this in tf, and the 
  previous sentence did not contain colored words, you do not have working 
  color support.  If it contained strange codes, you should do "/color_off" or 
  redefine the codes as described below.) The standard library defines these 8 
  basic colors with ANSI control codes, which will work on most terminals that 
  support color.  

  Many terminals also support brighter versions of the 8 basic colors, but may 
  need to be configured to do so.  On xterm, you may want to disable the 
  "boldColors" resource so that bold plus a normal color does not produce one 
  of these bright colors.  The bright color names are: gray, brightred, 
  brightgreen, brightyellow, brightblue, brightmagenta, brightcyan, or 
  brightwhite.  The standard library defines these 8 bright colors with ISO 
  6429 extension control codes, which will work on most terminals that support 
  16 colors.  

  Some newer terminals can display 256 colors.  If tf was built with the 
  "256colors" [1mfeature[22;0m, tf will recognize the following additional color names. 
  Names names of the form "rgb<[4mR[24m><[4mG[24m><[4mB[24m>" describe a color within a 6x6x6 color 
  cube: <[4mR[24m>, <[4mG[24m> and <[4mB[24m> are each a single digit between 0 and 5 that 
  specifies the brightness of the red, green, or blue component of the color.  
  For example, "rgb020" is a dark green, and "rgb520" is reddish orange.  
  Names of the form "gray<[4mN[24m>" describe a point on a grayscale, where <[4mN[24m> is 
  between 0 (dark) and 23 (light).  The standard library defines the "rgb*" 
  and "gray*" colors with xterm 256 color extension control codes.  

  To test the functionality and appearance of colors in tf, you can "[1m/load[22;0m 
  testcolor.tf".  This will also show the <[4mR[24m>, <[4mG[24m> and <[4mB[24m> values of each 
  color.  

  You can use a defined color in any [1mattribute[22;0m string.  For example, to make 
  [1m/hilite[22;0m'd text appear blue, you can [1m/set[22;0m [1mhiliteattr[22;0m=Cblue.  

  To define your own control codes for terminals that don't accept the 
  predefined codes, you will need to edit the color [1mvariables[22;0m.  The code to 
  enable foreground or background color <[4mname[24m> is stored in a [1mvariable[22;0m called 
  [1m%{start_color_<[4mname[24m>}[22;0m or [1m%{start_color_bg<[4mname[24m>}[22;0m.  The code to turn off 
  colors is stored in [1m%{end_color}[22;0m.  These [1mvariables[22;0m may contain carat 
  notation and backslashed ascii codes in decimal, octal, or hexadecimal 
  (e.g., ESC is ^[, \27, \033, or \0x1B).  

  The default definition of [1m%end_color[22;0m is "\033[39;49;0m", which should work 
  on most ANSI-like terminals.  If this does not work on your terminal, then 
  try "[1m/set[22;0m [1mend_color[22;0m \033[30;47;0m" (for black on white) or "[1m/set[22;0m [1mend_color[22;0m 
  \033[37;40;0m" (for white on black).  

  If [1m%{emulation}[22;0m is set to "ansi_attr" (the default), then ANSI, ISO 6429, 
  and xterm 256 color extension codes sent by the server will be interpreted 
  by tf.  As a result, if the [1m%{start_color_<[4mname[24m>}[22;0m [1mvariables[22;0m are set 
  correctly for your terminal, tf will translate color codes from the server 
  into codes for your terminal, displaying them correctly even if your 
  terminal does not use the same codes the server sends.  See: [1m%emulation[22;0m.  

  Note for "screen(1)" users: to make 8-16 colors work under Screen, you need 
  the following screenrc settings: 

      termcap  xterm AF=\E[3%dm
      terminfo xterm AF=\E[3%p1%dm
      termcap  xterm AB=\E[4%dm
      terminfo xterm AB=\E[4%p1%dm
    

  To make 256 colors work under Screen, it must have been compiled with 
  "--enable-colors256", and you need the following screenrc settings: 

      terminfo xterm Co=256
      termcap  xterm Co=256
      termcap  xterm AF=\E[38;5;%dm
      terminfo xterm AF=\E[38;5;%p1%dm
      termcap  xterm AB=\E[48;5;%dm
      terminfo xterm AB=\E[48;5;%p1%dm
    

  Colors are numbered 0 through 255 in the order in which they are described 
  above, but refering to colors by their enumeration number is generally not 
  recommended, as the numbering is subject to change.  In particular, the 
  numbering and interpretation of background colors changed in version 5.0 
  beta 7.  

  See: [1mattributes[22;0m 

&copy
&warranty
&copying
&copyright

copyright

  TinyFugue - programmable mud client
  Copyright (C) 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2002, 2003, 2004, 
  2005, 2006-2007 [1mKen Keys[22;0m 

  PCRE regexp package is Copyright (C) 1997-1999 University of Cambridge 

  For bug reports, questions, suggestions, etc., see "[1mproblems[22;0m".  

  This program is free software; you can redistribute it and/or modify it 
  under the terms of the [1mGNU General Public License[22;0m as published by the Free 
  Software Foundation; either version 2 of the [1mLicense[22;0m, or (at your option) 
  any later version.  

  This program is distributed in the hope that it will be useful, but WITHOUT 
  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or 
  FITNESS FOR A PARTICULAR PURPOSE.  See the [1mGNU General Public License[22;0m for 
  more details.  

  You should have received a copy of the [1mGNU General Public License[22;0m along with 
  this program; if not, write to the Free Software Foundation, Inc., 675 Mass 
  Ave, Cambridge, MA 02139, USA.  

#sites
#find
#where
#www
#ftp
  The latest information and latest version of TinyFugue can be found at 
  [1mhttp://tinyfugue.sourceforge.net/[22;0m.  Other sites may or may not have the 
  latest version.  

&debug
&debugger
&debugging

Debugging

  Debugging topics: 

    * [1m%kecho[22;0m - echo keyboard input 
    * [1m%mecho[22;0m - echo [1mmacros[22;0m as they execute 
    * [1m%qecho[22;0m - echo generated [1m/quote[22;0m text 
    * [1m%secho[22;0m - echo text sent to server 
    * [1m%pedantic[22;0m - enable extra warnings 
    * [1m%defcompile[22;0m - display syntax errors when macros are defined instead 
      of the first time they are used 
    * [1m%emulation[22;0m=debug - display nonprintable characters 
    * [1m%telopt[22;0m - echo telnet negotiation 
    * [1m/trigger[22;0m -n - see what [1mmacros[22;0m would be triggered 
    * [1m/addworld[22;0m -e - simulated "loopback" server 
    * [1m/runtime[22;0m - measure running time of commands 

  See also: [1mhints[22;0m 

&syntax
&body
&macro body
&reentrance
&execution
&expansion
&evaluation

evaluation

  A Builtin Command is any of the commands listed under "[1mcommands[22;0m".  All 
  builtin commands start with "/".  All builtins have a return value, usually 
  nonzero for success and 0 for failure.  

  A [1mMacro[22;0m Command is a user-defined command.  [1mMacro[22;0m commands also start with 
  '/'.  The return value of a [1mmacro[22;0m is the return value of its body when 
  executed.  

#/!
#/@
#/#
#/
  A command starting with a single "/" is either a [1mMacro[22;0m Command or a [1mBuiltin[22;0m 
  Command.  If the "/" is followed by "!", the return value of the command 
  will be negated.  If the "/" or "/!" is followed by "@", the rest of the 
  word is interpreted as the name of a [1mBuiltin[22;0m Command.  If the "/" or "/!" is 
  followed by "#", the rest of the word is interpreted as the number of a 
  [1mmacro[22;0m.  If neither "@" nor "#" is used (the normal case), the rest of the 
  word is interpreted as a [1mmacro[22;0m if there is one with that name, otherwise it 
  is interpreted as the name of a [1mBuiltin Command[22;0m.  If the name does not match 
  any [1mmacro[22;0m or [1mBuiltin Command[22;0m, the [1mNOMACRO[22;0m [1mhook[22;0m is called.  
#

  A Simple Command is any command that does not start with "/".  The text of 
  such a command is sent directly to the [1mcurrent[22;0m world, if there is one.  The 
  return value of a simple command is 1 if the text is successfully sent to 
  the [1mcurrent[22;0m world, otherwise 0.  To send a line that starts with "/" without 
  having it interpreted as a [1mMacro[22;0m Command or [1mBuiltin[22;0m Command, use a leading 
  "//"; the first "/" will be stripped.  

  A Compound Command is one of [1m/IF[22;0m...[1m/ENDIF[22;0m or [1m/WHILE[22;0m...[1m/DONE[22;0m.  These are 
  described under separate help sections.  Their return value is that of the 
  last command executed.  

  A List is a sequence of commands separated by "[1m%;[22;0m" (separator) or "[1m%|[22;0m" 
  (pipe) tokens.  The commands are executed in sequence, but may be aborted 
  early with the [1m/RETURN[22;0m or [1m/BREAK[22;0m commands.  and the return value of the List 
  is the return value of the last command executed in the sequence.  An empty 
  List has a return value of 1.  

  Two commands separated by "[1m%|[22;0m" pipe token mentioned above will have the 
  output stream ([1mtfout[22;0m) of the first connected to the input stream ([1mtfin[22;0m) of 
  the second.  The first command runs to completion before the second command 
  begins; the second command should stop reading [1mtfin[22;0m when it becomes empty.  
  Simple Commands have no [1mtfin[22;0m or [1mtfout[22;0m, so they may not be piped.  The [1mtfout[22;0m 
  of a Compound Command may not be piped directly, but the output of a macro 
  that contains a Compound Command may be piped.  

  Some characters within a [1mmacro[22;0m body undergo substitution.  These special 
  characters are not interpreted as themselves, but cause some evaluation to 
  be performed, and the result substituted in place of these characters.  This 
  is described under "[1msubstitution[22;0m".  

#scope
#dynamic scope
  When an [1mexpansion[22;0m begins, a new scope is created.  Any local [1mvariables[22;0m 
  created during the [1mexpansion[22;0m will be placed in this new scope.  The scope 
  and all [1mvariables[22;0m in it are destroyed when the [1mexpansion[22;0m exits.  

  Any [1mvariable[22;0m reference will refer to the [1mvariable[22;0m with that name in the 
  nearest enclosing (i.e., most recently created) still existing scope.  This 
  is called "dynamic scope".  

  Lexical scope can be simulated to some extent by using [1mvariable[22;0m 
  substitutions with the correct number of "%"s instead of [1mvariable[22;0m 
  references.  (Remember, a "reference" uses the name by itself in an 
  [1mexpression[22;0m, like "[1m/test[22;0m foo"; a "substitution" uses "%" characters, like 
  "[1m/test[22;0m %foo").  

#
  See: [1mcommands[22;0m, [1mmacros[22;0m, [1msubstitution[22;0m, [1m/if[22;0m, [1m/while[22;0m 

&expnonvis
&expnonvisusal
&experimental non-visual mode

experimental non-visual mode

  TF 5.0 beta 5 has a new experimental non-visual mode ("expnonvis") that 
  fixes design flaws in traditional [1mnon-visual[22;0m mode.  I may get rid of 
  traditional non-visual mode in the future, so if you use it, I suggest you 
  try expnonvis mode now and let me know if you don't like it.  To enable 
  expnonvis mode, "[1m/set[22;0m expnonvis=on" and "[1m/set[22;0m visual=off".  You may also 
  want to "[1m/set[22;0m [1mkecho[22;0m=on" (see below).  

  In the new expnonvis mode, input is only ever visible on the bottom line.  
  It scrolls your input buffer left and right as needed to display the part of 
  the input buffer in the neighborhood of the cursor.  The part of the line 
  that is "off the left edge" of the screen is simply not visible.  In 
  traditional non-visual mode, that part of the line would scroll up, 
  polluting the output region with partial input lines.  

  The "only on bottom line" rule applies even when you hit return to execute 
  the input line.  Your input is erased, and the command is executed; it does 
  not scroll up.  If you want to see the input text scroll up, you can "[1m/set[22;0m 
  [1mkecho[22;0m=on"; this will print the entire input, not just the last segment of it 
  that fit within the screen width.  You may also want to set [1m%kecho_attr[22;0m so 
  that the echoed input is easily distinguishable from regular output.  

  The minimum amount of scrolling is determined by the %sidescroll variable, 
  which defaults to 1.  For slow terminals, you may wish to increase this.  
  Any movement that would exceed half the screen width does not use the 
  terminal's scrolling, but instead redraws the line.  

  The current implementation probably has a few bugs; if the screen display 
  ever appears incorrect, use ^R or ^L to redraw it.  I don't think there are 
  any fatal bugs, but it is possible that some remain, so don't try expnonvis 
  unless you don't mind crashing tf.  Terminals without the delete character 
  capability are not yet supported, but will be in the future.  

&logic
&math
&strings
&arithmetic
&expression
&expressions

expressions

  [1mExpressions[22;0m apply operators to numeric and string operands, and return a 
  result.  They can be used in [1m$[...] expression subs[22;0m, the condition of [1m/if[22;0m 
  and [1m/while[22;0m statements, the condition of [1m/def -E[22;0m, and as arguments to 
  [1m/return[22;0m, [1m/result[22;0m, and [1m/test[22;0m commands.  

#float
#real
#integer
#string
#dtime
#atime
#hours:minutes:seconds
#hours:minutes
#hh:mm
#hh:mm:ss
#types
#scalar
#scalars
#operands

Operands

  Operands can be any of: 

    * Integer constants (e.g., 42).  
    * Real decimal point constants ("reals", for short) containing a 
      decimal point (e.g., 12.3456789) or exponent (e.g., 1e-2) or both (e.g., 
      1.23e4).  
    * Time duration ("dtime") values of the form <[4mhours[24m>:<[4mminutes[24m>, 
      <[4mhours[24m>:<[4mminutes[24m>:<[4mseconds[24m>, or <[4mseconds[24m> (where <[4mseconds[24m> may contain a 
      decimal point followed by up to 6 digits), will be interpreted as real 
      seconds (e.g., 0:01:02.3 == 62.3), and can be used anywhere a number is 
      expected.  
    * Absolute time ("atime") values, in the form of a number with up to 6 
      decimal places.  On most systems, this represents the number of seconds 
      since 1970-01-01 00:00:00 UTC.  
    * Strings of characters, surrounded with quotes (", ', or `, with the 
      same kind of quote on each end), like "hello world".  
    * [1mVariable[22;0m references (see below) like visual.  
    * [1mVariable substitutions[22;0m (see below) like {visual} and {1}.  
    * [1mMacro substitutions[22;0m like ${COMPRESS_SUFFIX}.  
    * [1mCommand substitutions[22;0m like $([1m/listworlds[22;0m -s).  

  Named [1mvariables[22;0m may be accessed by simply using their name (with no leading 
  '%').  This is called a [1mvariable[22;0m reference, and is the preferred way of 
  using a [1mvariable[22;0m in an expression.  The special substitutions ([1m*[22;0m, [1m?[22;0m, [1m#[22;0m, [1m<[4mn[24m>[22;0m, 
  [1mL<[4mn[24m>[22;0m, [1mP<[4mn[24m>[22;0m, [1mR[22;0m) may not be used this way.  

  [1mVariable substitutions[22;0m of the form "[1m{selector}[22;0m" and "[1m{selector-default}[22;0m" may 
  be used.  They follow the same rules as [1mvariable substitution[22;0m in macros, 
  except that there is no leading '%', and the '{' and '}' are required.  The 
  special substitutions ([1m*[22;0m, [1m?[22;0m, [1m#[22;0m, [1m<[4mn[24m>[22;0m, [1mL<[4mn[24m>[22;0m, [1mP<[4mn[24m>[22;0m, [1mR[22;0m) are allowed.  

  Macro-style [1mvariable substitutions[22;0m beginning with '%' may also be used, but 
  are not recommended, since the multiple '%'s required in nested [1mmacros[22;0m can 
  quickly get confusing.  It always easier to use one of the above methods.  

#operators

Operators

  In the following list, operators are listed in groups, from highest to 
  lowest precedence.  Operators listed together have equal precedence.  The 
  letters in the table below correspond to the type of objects acted on by the 
  operators: [4mn[24m for numeric (integer or real); [4ms[24m for string; [4me[24m for any 
  expression.  All operators group left-to-right except assignment, which 
  groups right-to-left.  If any binary numeric operator is applied to two 
  integers, the result will be an integer, unless the result would overflow, 
  in which case it is converted to real.  If either operand is a real, the 
  other will be converted to real if it is not already a real, and the result 
  will be a real.  

  ([4me[24m)         Parentheses, for grouping.  

  [4mfunc[24m([4margs[24m)  Perform [1mfunction[22;0m <[4mfunc[24m> on arguments <[4margs[24m>.  (see: [1mfunctions[22;0m).  

  ![4mn[24m          Boolean NOT (1 if [4mn[24m==0, otherwise 0).  
  +[4mn[24m          Unary positive (useful for converting a string to a number).  
  -[4mn[24m          Unary negative.  
  ++[4mv[24m         Equivalent to ([4mv[24m := [4mv[24m + 1).  
  --[4mv[24m         Equivalent to ([4mv[24m := [4mv[24m - 1).  

  [4mn1[24m * [4mn2[24m     Numeric multiplication.  
  [4mn1[24m / [4mn2[24m     Numeric division.  Remember, if both operands are type integer, 
              the result will be truncated to integer.  

  [4mn1[24m + [4mn2[24m     Numeric addition.  
  [4mn1[24m - [4mn2[24m     Numeric subtraction.  

  [4mn1[24m = [4mn2[24m     Numeric equality (but easily confused with assignment; you are 
              advised to use == instead).  
  [4mn1[24m == [4mn2[24m    Numeric equality.  
  [4mn1[24m != [4mn2[24m    Numeric inequality.  
  [4ms1[24m =~ [4ms2[24m    String equality (case sensitive, [1mattribute[22;0m insensitive).  
  [4ms1[24m !~ [4ms2[24m    String inequality (case sensitive, [1mattribute[22;0m insensitive).  
  [4ms1[24m =/ [4ms2[24m    String [4ms1[24m matches [1mglob[22;0m pattern [4ms2[24m.  
  [4ms1[24m !/ [4ms2[24m    String [4ms1[24m does not match [1mglob[22;0m pattern [4ms2[24m.  
  [4mn1[24m < [4mn2[24m     Numeric less than.  
  [4mn1[24m <= [4mn2[24m    Numeric less than or equal.  
  [4mn1[24m > [4mn2[24m     Numeric greater than.  
  [4mn1[24m >= [4mn2[24m    Numeric greater than or equal.  

  [4mn1[24m & [4mn2[24m     Boolean AND.  [4mn2[24m will be evaluated if and only if [4mn1[24m is nonzero. 

  [4mn1[24m | [4mn2[24m     Boolean OR.  [4mn2[24m will be evaluated if and only if [4mn1[24m is zero.  

  [4mn[24m ? [4me1[24m : [4me2[24m 
  [4mn[24m ? : [4me2[24m    Conditional.  If [4mn[24m is nonzero, the result is the value of 
              [1mexpression[22;0m [4me1[24m; otherwise it is the value of [1mexpression[22;0m [4me2[24m.  If 
              [4me1[24m is omitted, the value of [4mn[24m is used in its place.  Note that 
              digits followed by a colon is interpreted as a dtime value, so 
              if the [4me2[24m operand of the ?: operator is an integer, you must 
              separate it from the colon (with a space or parenthesis, for 
              example).  

  [4mv[24m := [4me[24m      Assignment.  The identifier "[4mv[24m" refers to the [1mvariable[22;0m in the 
              nearest scope.  If not found, a new [1mvariable[22;0m is created at the 
              global level, as if by [1m/set[22;0m.  If [4mv[24m is a [1mspecial variable[22;0m, the 
              value of [4me[24m may need to be converted to the type of [4mv[24m, or the 
              assignment may fail altogther if the value is not legal for [4mv[24m.  
              The value of the assignment expression is the new value of [4mv[24m.  
  [4mv[24m += [4mn[24m      Equivalent to [4mv[24m := [4mv[24m + ([4mn[24m).  
  [4mv[24m -= [4mn[24m      Equivalent to [4mv[24m := [4mv[24m - ([4mn[24m).  
  [4mv[24m *= [4mn[24m      Equivalent to [4mv[24m := [4mv[24m * ([4mn[24m).  
  [4mv[24m /= [4mn[24m      Equivalent to [4mv[24m := [4mv[24m / ([4mn[24m).  

  [4me1[24m , [4me2[24m     Comma.  [1mExpressions[22;0m [4me1[24m and [4me2[24m are evaluated; the result is the 
              value of [4me2[24m.  Only useful if [4me1[24m has some side effect.  

  The comparison operators return 0 for false, nonzero for true.  The boolean 
  operators (& and |) stop evaluating as soon as the value of the [1mexpression[22;0m 
  is known ("short-circuit"), and return the value of the last operand 
  evaluated.  This does not affect the value of the [1mexpression[22;0m, but is 
  important when the second operand performs side effects.  

  Normal (non-[1menumerated[22;0m) [1mVariables[22;0m set with any of the assignment operators 
  keep the type of the [1mexpression[22;0m assigned to them.  This is different than 
  [1m/set[22;0m and [1m/let[22;0m, which always assign a string value to the [1mvariables[22;0m.  This 
  distinction is important for real numeric values, which lose precision if 
  converted to a string and back.  
#conversion

  All operands will be automatically converted to the type expected by the 
  operator.  

    * String to numeric: leading signs, digits, colons, and exponents are 
      interpreted as an integer, decimal (real), or dtime (real) value; e.g., 
      "12abc" becomes 12, "12.3junk" becomes 12.3, "0:01:02.3" becomes 
      0:01:02.3, and "xyz" becomes 0.  
    * Integer to real: straightforward.  
    * Real to integer: the fractional part is truncated.  
    * [1mEnumerated variable[22;0m to string: straightforward string value.  
    * [1mEnumerated variable[22;0m to numeric: one integer stands for each of the 
      allowed values.  "Off" is always 0, "on" is always 1, etc.  This makes 
      (![1mvisual[22;0m) and ([1mvisual[22;0m == 0) the same as ([1mvisual[22;0m =~ 'off').  
    * Integer to string: straightforward.  
    * Real to string: decimal notation if the exponent is greater than -5 
      and less than [1m%sigfigs[22;0m, otherwise exponential notation.  
    * Normal (non-[1menumerated[22;0m) [1mvariables[22;0m are treated as whatever type their 
      value has.  

#

Examples

  Given the [1mvariables[22;0m 

      [1m/set[22;0m X=5
      [1m/set[22;0m name=Hawkeye
      [1m/set[22;0m [1mvisual[22;0m=1

  here are some [1mexpressions[22;0m and their values: 

      [1mExpression[22;0m             Value   Comments
      ----                   -----   --------
      3 + X * 2                 13   3 + (5 * 2) = 13.
      "foo" =~ "bar"             0   "foo" is not identical to "bar".
      name =/ 'hawk*'            1   "Hawkeye" matches the [1mglob[22;0m "hawk*".
      X =~ "+5"                  0   X is interpreted as string "5".
      X == "+5"                  1   string "+5" is converted to integer 5.
      visual & (X > 0)           1   visual is nonzero, AND %X is positive.

  See: [1mfunctions[22;0m, [1m/test[22;0m, [1mevaluation[22;0m, [1mpatterns[22;0m 

&file
&files
&filenames
&filename expansion

filename expansion

  Certain strings are treated as filenames in tf ([1m%{TFHELP}[22;0m; [1m%{TFLIBDIR}[22;0m; 
  [1m%{TFLIBRARY}[22;0m; arguments to [1m/load[22;0m, [1mfwrite()[22;0m; etc.).  Those strings undergo 
  filename expansion as described below.  

  If <[4mfile[24m> begins with '~', all characters after the '~' up to the first '/' 
  or end of string are treated as a user name, and the '~' and user name are 
  replaced with the name of the home directory of that user.  If the user name 
  is empty, [1m%{HOME}[22;0m is substituted.  

  For example, if bob's home directory is /users/bob, then the command "[1m/load[22;0m 
  ~bob/macros.tf" will attempt to load the file /users/bob/macros.tf.  

  "~user" expansion is not supported on systems that do not have the 
  getpwnam() function.  

&function
&functions

functions

#macro
#function syntax

  In an [1mexpression[22;0m, a function operates on 0 or more arguments and returns a 
  result.  A function call is made with a function name, followed by a 
  parenthesized list of comma-separated arguments: "[4mname[24m([4marg1[24m, [4marg2[24m, ...  
  [4margN[24m)".  

  There are three kinds of objects that can be called as functions: [1mbuiltin 
  functions[22;0m, [1mmacros[22;0m, and builtin commands.  They are searched in that order, 
  so if a builtin function and a macro have the same name, using that name in 
  a function call will invoke the builtin function.  

  A macro called as a function can be called with any number of arguments; 
  each argument corresponds to a [1mpositional parameter[22;0m ([1m%1[22;0m, [1m%2[22;0m, etc.).  For 
  example, if "spam" is a macro, the function call 
  spam("foo", "bar", "baz")

  will set the parameters the same as in the command invocation 
  /spam foo bar baz

  The function call syntax allows [1mpositional parameters[22;0m to contain spaces, 
  which is not possible in the command syntax.  (Note: prior to version 4.0, a 
  macro called as a function could only take 0 or 1 arguments, and a single 
  argument was broken into positional parameters at whitespace.) A macro can 
  set its return value using [1m/return[22;0m or [1m/result[22;0m.  

  A builtin command called as a function can have 0 or 1 arguments; the 
  argument is treated as a command line.  For example, the function call 
  def("-t'{*} has arrived.' greet = :waves.")

  is the same as the command invocation 
  /def -t'{*} has arrived.' greet = :waves. 

  To evaluate a function for its "side effect" only, you can call it from 
  [1m/test[22;0m and ignore the return value (e.g., "[1m/test[22;0m [1mkbdel[22;0m(0)").  

#builtin
Builtin functions

  In the following list of builtin functions, the first letter of each 
  argument indicates its type: <[4ms[24m> for string, <[4mi[24m> for integer, <[4mr[24m> for real, 
  <[4mn[24m> for any numeric type, or <[4mf[24m> for flag (0 or "off"; or, 1 or "on").  

Mathematical functions

  Angles are in radians.  
#abs
#abs()
  [1mabs[22m([4mn[24m)  Absolute value of <[4mn[24m>.  Result has the same numeric type as <[4mn[24m>.  
#sin
#sin()
  [1msin[22m([4mr[24m)  (real) Sine of <[4mr[24m>.  
#cos
#cos()
  [1mcos[22m([4mr[24m)  (real) Cosine of <[4mr[24m>.  
#tan
#tan()
  [1mtan[22m([4mr[24m)  (real) Tangent of <[4mr[24m>.  
#asin
#asin()
  [1masin[22m([4mr[24m) 
          (real) Arcsine of <[4mr[24m>, in the range [-pi/2, pi/2].  <[4mr[24m> must be in 
          the domain [-1, 1].  
#acos
#acos()
  [1macos[22m([4mr[24m) 
          (real) Arccosine of <[4mr[24m>, in the range [0, pi].  <[4mr[24m> must be in the 
          domain [-1, 1].  
#atan
#atan()
  [1matan[22m([4mr[24m) 
          (real) Arctangent of <[4mr[24m>, in the range [-pi/2, pi/2].  
#exp
#exp()
  [1mexp[22m([4mr[24m)  (real) [4me[24m raised to the power <[4mr[24m>.  
#pow
#pow()
  [1mpow[22m([4mn1[24m, [4mn2[24m) 
          (real) <[4mn1[24m> raised to the power <[4mn2[24m>.  If <[4mn1[24m> is negative, <[4mn2[24m> 
          must be an integer.  
#sqrt
#sqrt()
  [1msqrt[22m([4mn[24m) 
          (real) Square root of <[4mn[24m> (same as [1mpow[22;0m(<[4mn[24m>, 0.5)).  
#log
#log()
#ln
#ln()
#log10
#log10()
  [1mln[22m([4mn[24m)   (real) Natural logarithm of <[4mn[24m>.  <[4mn[24m> must be positive.  The base B 
          logarithm of any number N can be found with the expression [1mln[22;0m(N) / 
          [1mln[22;0m(B).  
  [1mlog10[22m([4mn[24m) 
          (real) Base 10 logarithm of <[4mn[24m>.  <[4mn[24m> must be positive.  
#mod
#mod()
  [1mmod[22m([4mi1[24m,[4mi2[24m) 
          (int) Remainder of <[4mi1[24m> divided by <[4mi2[24m>.  
#trunc
#trunc()
  [1mtrunc[22m([4mr[24m) 
          (int) Integer part of <[4mr[24m>.  
#random
#rand
#rand()
  [1mrand[22m()  (int) Random integer in the range [0, system maximum].  
  [1mrand[22m([4mi[24m) 
          (int) Random integer in the range [0, <[4mi[24m> - 1].  
  [1mrand[22m([4mi1[24m,[4mi2[24m) 
          (int) Random integer in the range [<[4mi1[24m>, <[4mi2[24m>].  
#

Input/output functions

#
  [1mecho[22m([4ms1[24m [,[4mattrs[24m [,[4minline[24m [,[4mdest[24m]]]) 
          (int) Echoes <[4ms1[24m> to the screen or <[4mdest[24m> with [1mattributes[22;0m <[4mattrs[24m>, 
          interpreting inline [1mattribute[22;0m codes if the flag <[4minline[24m> is 1 or 
          "on".  See: "[1mecho()[22;0m".  
#
  [1msend[22m([4ms1[24m[, [4mworld[24m[, [4mflags[24m]]) 
          (int) Sends string <[4ms1[24m> to <[4mworld [24m>.  See [1msend()[22;0m.  
#
  [1mprompt[22m([4ms1[24m) 
          (int) Sets the prompt of the [1mcurrent socket[22;0m to <[4ms1[24m>.  See [1m/prompt[22;0m.  
#fwrite
#fwrite()
  [1mfwrite[22m([4ms1[24m,[4ms2[24m) 
          Writes string <[4ms2[24m> to the end of file <[4ms1[24m>.  [1mfwrite()[22;0m is good for 
          writing a single line, but when writing multiple lines it is more 
          efficient to use [1mtfopen()[22;0m, a series of [1mtfwrite()[22;0m, and a [1mtfclose()[22;0m.  
          [1mDisplay attributes[22;0m in <[4ms2[24m> are not written.  
#tfopen
#tfopen()
  [1mtfopen[22m([4ms1[24m, [4ms2[24m) 
  [1mtfopen[22m() 
          (int) Open a [1mtfio stream[22;0m using file <[4ms1[24m> and mode <[4ms2[24m>.  See [1mtfio[22;0m.  
#tfclose
#tfclose()
  [1mtfclose[22m([4mi[24m) 
          (int) Close the [1mstream[22;0m indicated by handle <[4mi[24m>.  See [1mtfio[22;0m.  
#tfread
#tfread()
  [1mtfread[22m([4mi[24m, [4mv[24m) 
  [1mtfread[22m([4mv[24m) 
          (int) Read into variable <[4mv[24m> from the [1mstream[22;0m indicated by handle 
          <[4mi[24m>.  See [1mtfio[22;0m.  
#tfwrite
#tfwrite()
  [1mtfwrite[22m([4mi[24m, [4ms[24m) 
  [1mtfwrite[22m([4ms[24m) 
          (int) Write <[4ms[24m> to the [1mstream[22;0m indicated by handle <[4mi[24m>.  See [1mtfio[22;0m.  
#tfflush
#tfflush()
  [1mtfflush[22m([4mi[24m) 
          Flushes the [1mstream[22;0m indicated by handle <[4mi[24m>.  
  [1mtfflush[22m([4mi[24m, [4mf[24m) 
          Disables (if <[4mf[24m> is 0 or "off") or enables (if <[4mf[24m> is 1 or "on") 
          automatic flushing for the [1mstream[22;0m indicated by handle <[4mi[24m>.  See 
          [1mtfio[22;0m.  
#read
#read()
  [1mread[22m()  Obsolete.  Use [1mtfread()[22;0m instead.  
#

String functions

  String positions are always counted from 0.  Therefore the first character 
  of a string <[4ms[24m> is [1msubstr[22;0m(s, 0, 1), and the last character is [1msubstr[22;0m(s, 
  [1mstrlen[22;0m(s)-1).  

  Range checking is done on string positions.  Any position given outside the 
  allowed range will be silently forced to the closest value that is in the 
  range.  
#ascii
#ascii()
  [1mascii[22m([4ms[24m) 
          (int) Integer code of the first character of <[4ms[24m>, The character does 
          not have to be ASCII, but may be any character allowed by your 
          [1mlocale[22;0m.  
#char
#char()
  [1mchar[22m([4mi[24m) 
          (str) character with integer code <[4mi[24m>.  If <[4mi[24m> is outside the range 
          allowed by your [1mlocale[22;0m, it will be silently forced into the allowed 
          range.  
#tolower
#tolower()
  [1mtolower[22m([4ms[24m) 
  [1mtolower[22m([4ms[24m, [4mi[24m) 
          (str) Convert the first <[4mi[24m> (default all) characters in <[4ms[24m> to lower 
          case.  
#toupper
#toupper()
  [1mtoupper[22m([4ms[24m) 
  [1mtoupper[22m([4ms[24m, [4mi[24m) 
          (str) Convert the first <[4mi[24m> (default all) characters in <[4ms[24m> to upper 
          case.  
#pad
#pad()
  [1mpad[22m([[4ms[24m, [4mi[24m]...) 
          (str) There may be any number of (<[4ms[24m>, <[4mi[24m>) pairs.  For each pair, 
          <[4ms[24m> is padded with spaces to a length equal to the absolute value of 
          <[4mi[24m>.  If <[4mi[24m> is positive, <[4ms[24m> is right-justified (left-padded); If 
          <[4mi[24m> is negative, <[4ms[24m> is left-justified (right-padded).  The result 
          is the concatenation of all the padded strings.  
#regmatch
#regmatch()
  [1mregmatch[22m([4ms1[24m, [4ms2[24m) 
          (int) If string <[4ms2[24m> matches [1mregexp[22;0m <[4ms1[24m>, regmatch() returns a 
          positive integer indicating the number of captured substrings 
          (including [1m%P0[22;0m).  regmatch() returns 0 if string <[4ms2[24m> does not match 
          [1mregexp[22;0m <[4ms1[24m>.  After a successful match, captured substrings can 
          later be extracted using the P[4mn[24m [1mvariables[22;0m or [1m%P[4mn[24m[22;0m substitutions.  
          (See also: [1mregexp[22;0m) 
#
#replace()
  [1mreplace[22m([4ms1[24m, [4ms2[24m, [4ms3[24m) 
          (int) Returns <[4ms3[24m> with every occurance of <[4ms1[24m> replaced with <[4ms2[24m>.  
          See: "[1m/replace[22;0m".  
#strcat
#strcat()
  [1mstrcat[22m([4ms[24m...) 
          (str) Returns the concatenation of all string arguments.  
#strchr
#strchr()
  [1mstrchr[22m([4ms1[24m, [4ms2[24m) 
  [1mstrchr[22m([4ms1[24m, [4ms2[24m, [4mi[24m) 
          (int) Searches for any character of <[4ms2[24m> in <[4ms1[24m> starting at 
          position <[4mi[24m> (default 0), and returns the position if found, or -1 
          if not found.  If <[4mi[24m> is negative, it is counted as an absolute 
          value from the end of <[4ms[24m>.  
#strcmp
#strcmp()
  [1mstrcmp[22m([4ms1[24m, [4ms2[24m) 
          (int) Returns an integer less than, equal to, or greater than 0 if 
          <[4ms1[24m> is lexicographically less than, equal to, or greater than <[4ms2[24m>, 
          respectively.  
#strcmpattr
#strcmpattr()
  [1mstrcmpattr[22m([4ms1[24m, [4ms2[24m) 
          (int) Like [1mstrcmp()[22;0m, except that in order for the strings to be 
          considered equal, both their text and their attributes must be 
          equal.  In other words, [1mstrcmp[22;0m([1mencode_attr[22;0m(<[4ms1[24m>), [1mencode_attr[22;0m(<[4ms2[24m>)) 
          The ordering of attributes is not documented, and may change between 
          versions of tf.  
#strlen
#strlen()
  [1mstrlen[22m([4ms[24m) 
          (int) Length of string <[4ms[24m>.  
#strncmp
#strncmp()
  [1mstrncmp[22m([4ms1[24m, [4ms2[24m, [4mi[24m) 
          (int) Like [1mstrcmp()[22;0m, but compares only the first <[4mi[24m> characters of 
          <[4ms1[24m> and <[4ms2[24m>.  
#strrchr
#strrchr()
  [1mstrrchr[22m([4ms1[24m, [4ms2[24m) 
  [1mstrrchr[22m([4ms1[24m, [4ms2[24m, [4mi[24m) 
          (int) Searches backward in <[4ms1[24m> starting at position <[4mi[24m> (default: 
          end of <[4ms1[24m>) for any character of <[4ms2[24m>, and returns the position if 
          found, or -1 if not found.  If <[4mi[24m> is negative, it is counted as an 
          absolute value from the end of <[4ms[24m>.  
#strrep
#strrep()
  [1mstrrep[22m([4ms[24m, [4mi[24m) 
          (str) Returns a string containing <[4mi[24m> repetitions of <[4ms[24m>.  
#strstr
#strstr()
  [1mstrstr[22m([4ms1[24m, [4ms2[24m) 
  [1mstrstr[22m([4ms1[24m, [4ms2[24m, [4mi[24m) 
          (int) Searches for <[4ms2[24m> in <[4ms1[24m> starting at position <[4mi[24m> (default 
          0), and returns the position if found, or -1 if not found.  
#substr
#substr()
  [1msubstr[22m([4ms[24m, [4mi1[24m) 
  [1msubstr[22m([4ms[24m, [4mi1[24m, [4mi2[24m) 
          (str) Substring of <[4ms[24m>, starting at position <[4mi1[24m>, with length <[4mi2[24m>. 
          If <[4mi2[24m> is omitted, it defaults to the remaining length of <[4ms[24m>.  If 
          <[4mi1[24m> or <[4mi2[24m> is negative, they are counted as absolute values from 
          the end of <[4ms[24m>.  
#strip_attr
#strip_attr()
  [1mstrip_attr[22m([4ms[24m) 
          (str) Returns <[4ms[24m> with all display [1mattributes[22;0m removed.  
#inline_attr
#inline_attr()
#decode_attr
#decode_attr()
  [1mdecode_attr[22m([4ms1[24m [, [4ms2[24m [, f]]) 
          (str) Returns <[4ms1[24m> with "@{<[4mattr[24m>}" codes interpeted as display 
          [1mattributes[22;0m, as in [1m/echo[22;0m -p.  If present, <[4ms2[24m> is a string of 
          [1mattributes[22;0m that will be applied to the entire string (as in [1m/echo[22;0m 
          -a<[4ms2[24m>).  If <[4mf[24m> is present and equal to 0 or "off", then 
          "@{<[4mattr[24m>}" codes are [4mnot[24m interpeted; this is useful for applying 
          <[4ms2[24m> attributes with no other effects.  
#encode_attr
#encode_attr()
  [1mencode_attr[22m([4ms[24m) 
          (str) Returns <[4ms[24m> with display [1mattributes[22;0m encoded in "@{<[4mattr[24m>}" 
          form.  
#decode_ansi
#decode_ansi()
  [1mdecode_ansi[22m([4ms[24m) 
          (str) Returns <[4ms[24m> with attribute control codes interpeted as display 
          [1mattributes[22;0m, and, if [1m%expand_tabs[22;0m is on, tabs are expanded to spaces 
          according to [1m%tabsize[22;0m.  Any attributes originally on <[4ms[24m> are [4mnot[24m 
          copied to the result.  The attribute control codes recognzied 
          include ANSI codes, ISO 6429 16-color extension codes, and xterm 
          256-color extension codes.  
#encode_ansi
#encode_ansi()
  [1mencode_ansi[22m([4ms[24m) 
          (str) Returns <[4ms[24m> with display [1mattributes[22;0m encoded in terminal 
          control code form.  The control codes generated include ANSI codes, 
          ISO 6429 16-color extension codes, and xterm 256-color extension 
          codes.  
#
  [1mtextencode[22m([4ms[24m) 
          (str) Returns <[4ms[24m> converted to a form containing only letters, 
          digits, and underscores.  See [1mtextencode()[22;0m.  
#
  [1mtextdecode[22m([4ms[24m) 
          (str) Converts <[4ms[24m>, the result of [1mtextencode()[22;0m, back to its original 
          form.  See [1mtextencode()[22;0m.  
#

Keyboard buffer functions

#kbdel
#kbdel()
  [1mkbdel[22m([4mi[24m) 
          (int) Delete from the cursor to position <[4mi[24m> in the input buffer.  
          Returns the new position.  
#kbgoto
#kbgoto()
  [1mkbgoto[22m([4mi[24m) 
          (int) Move the cursor to position <[4mi[24m> in the input buffer.  Returns 
          the new position (which may be different than <[4mi[24m> if <[4mi[24m> would put 
          the cursor outside the buffer).  
#kbhead
#kbhead()
  [1mkbhead[22m() 
          (str) Return the current input up to the cursor.  
#kblen
#kblen()
  [1mkblen[22m() 
          (int) Length of current input line.  
#kbmatch
#kbmatch()
  [1mkbmatch[22m() 
  [1mkbmatch[22m([4mi[24m) 
          (int) Finds one of "()[]{}" under or to the right of the position 
          <[4mi[24m> (default: cursor position), and returns the position of its 
          match, or -1 if not found.  (See also: [1mkeybindings[22;0m) 
#kbpoint
#kbpoint()
  [1mkbpoint[22m() 
          (int) Return the current position of the cursor in input.  
#kbtail
#kbtail()
  [1mkbtail[22m() 
          (str) Return the current input after the cursor.  
#kbwordleft
#kbwordleft()
  [1mkbwordleft[22m() 
  [1mkbwordleft[22m([4mi[24m) 
          (int) Position of the beginning of the word left of <[4mi[24m> within the 
          input buffer.  <[4mi[24m> defaults to the current cursor position.  (See 
          also: [1m%wordpunct[22;0m) 
#kbwordright
#kbwordright()
  [1mkbwordright[22m() 
  [1mkbwordright[22m([4mi[24m) 
          (int) Position just past the end of the word right of <[4mi[24m> within the 
          input buffer.  <[4mi[24m> defaults to the current cursor position.  (See 
          also: [1m%wordpunct[22;0m) 
#keycode
#keycode()
  [1mkeycode[22m([4ms[24m) 
          (str) String generated by typing the key labeled <[4ms[24m>, as defined in 
          the termcap entry corresponding to the value of [1m%TERM[22;0m.  See also: 
          [1mkeybindings[22;0m.  
#

Information functions

#time
#time()
  [1mtime[22m()  (atime) Absolute system time in seconds, to the nearest microsecond 
          (typically measured since 1970-01-01 00:00:00 UTC).  See also: 
          [1mcputime()[22;0m, [1mmktime()[22;0m, [1midle()[22;0m, [1msidle()[22;0m, [1m/time[22;0m, [1mftime()[22;0m.  
#cputime
#cputime()
  [1mcputime[22m() 
          (real) CPU time used by tf, or -1 if not available.  The resolution 
          depends on the operating system.  See also: [1m/runtime[22;0m, [1mtime()[22;0m, [1m/time[22;0m. 
#columns
#columns()
  [1mcolumns[22m() 
          (int) Number of columns on the screen.  See also: [1mhooks (RESIZE)[22;0m, 
          [1mlines()[22;0m, [1mwinlines()[22;0m, [1m%COLUMNS[22;0m.  
#lines
#lines()
  [1mlines[22m() 
          (int) Number of lines on the screen.  To get the number of lines in 
          the output window, use [1mwinlines()[22;0m.  See also: [1mhooks (RESIZE)[22;0m, 
          [1mwinlines()[22;0m, [1mcolumns()[22;0m, [1m%LINES[22;0m.  
#winlines
#winlines()
  [1mwinlines[22m() 
          (int) Number of lines in the output window.  See also: [1mhooks 
          (RESIZE)[22;0m, [1mlines()[22;0m, [1mcolumns()[22;0m.  
#morepaused
#morepaused()
  [1mmorepaused[22m([[4ms1[24m]) 
          (int) Returns 1 if output of world <[4ms1[24m> is paused (by [1mmore[22;0m or ([1mdokey[22;0m 
          pause).  If omitted, <[4ms1[24m> defaults to the current world.  See also: 
          [1mmoresize()[22;0m.  
#morescroll
#morescroll()
  [1mmorescroll[22m([4mi[24m) 
          (int) If <[4mi[24m> is positive, this function scrolls <[4mi[24m> lines of text 
          from the window buffer into the window from the bottom.  If <[4mi[24m> is 
          negative, it reverse-scrolls abs(<[4mi[24m>) lines of text from the window 
          buffer into the window from the top.  If abs(<[4mi[24m>) is larger than one 
          screenful, the actual scrolling is skipped, and only the end result 
          is displayed.  Returns the number of lines actually scrolled.  
#moresize
#moresize()
  [1mmoresize[22m([[4ms1[24m [, [4ms2[24m]]) 
          (int) Returns a line count for world <[4ms2[24m>, or the current world if 
          <[4ms2[24m> is omitted.  If <[4ms1[24m> is omitted or blank, the count is the 
          number of lines below the bottom of the output window (i.e., queued 
          at a [1mmore[22;0m prompt).  If <[4ms1[24m> contains "n", it counts only new lines 
          that have never been seen, not lines that had been displayed and 
          then reverse scrolled off.  If <[4ms1[24m> contains "l", it counts only 
          lines that match the current [1m/limit[22;0m.  "n" and "l" may be combined.  
          If all lines that would be counted have the "A" (noactivity) 
          [1mattribute[22;0m, the result will normally be 0.  But if <[4ms1[24m> contains "a", 
          lines with "A" [1mattributes[22;0m are counted anyway.  In all cases, the 
          count is the number of physical (after wrapping) lines.  Note that a 
          return value of 0 does not necessarily indicate that output is not 
          paused; it may be the case that output is paused and there are just 
          0 lines below the bottom of the window, or that all the lines have 
          the "A" [1mattribute[22;0m.  Use [1mmorepaused()[22;0m, to tell if output is paused.  
          See also: [1mmorepaused()[22;0m, [1mnactive()[22;0m.  
#nactive
#nactive()
  [1mnactive[22m() 
          (int) Number of active worlds (ie, worlds with unseen text).  
  [1mnactive[22m([4ms[24m) 
          (int) Number of unseen lines in world <[4ms[24m>.  
          Note: when [1mnactive()[22;0m (with or without arguments) is called from a 
          [1mtrigger[22;0m, the line that caused the [1mtrigger[22;0m is not counted by 
          [1mnactive()[22;0m because it has not yet been fully processed (for example, 
          a lower [1mpriority[22;0m [1mtrigger[22;0m might [1mgag[22;0m the line).  [1mnactive[22;0m(<[4ms[24m>) is 
          equivalent to [1mmoresize[22;0m("n", <[4ms[24m>).  See also: [1mmoresize()[22;0m.  
#world_info
#world_info()
  [1mworld_info[22m([4ms1[24m, [4ms2[24m) 
          (str) Return the value of field <[4ms2[24m> of world <[4ms1[24m>, 
  [1mworld_info[22m([4ms2[24m) 
          (str) Return the value of field <[4ms2[24m> of the [1mcurrent world[22;0m.  
  [1mworld_info[22m() 
          (str) Return the name of the [1mcurrent world[22;0m.  See [1mworlds[22;0m.  
#fg_world
#fg_world()
  [1mfg_world[22m() 
          (str) Returns the name of the [1mworld[22;0m associated with the [1mforeground[22;0m 
          [1msocket[22;0m.  
#is_connected
#is_connected()
  [1mis_connected[22m() 
          (int) Returns 1 if the [1mcurrent[22;0m [1msocket[22;0m is connected, 0 otherwise.  
  [1mis_connected[22m([4ms[24m) 
          (int) Returns 1 if [1mworld[22;0m <[4ms[24m> is connected, 0 otherwise.  See also 
          [1mis_open()[22;0m.  
#is_open
#is_open()
  [1mis_open[22m() 
          (int) Returns 1 if the [1mcurrent[22;0m [1msocket[22;0m is open, 0 otherwise.  
  [1mis_open[22m([4ms[24m) 
          (int) Returns 1 if [1mworld[22;0m <[4ms[24m> is open, 0 otherwise.  
#idle
#idle()
  [1midle[22m()  (dtime) Number of seconds (to the nearest microsecond) since the 
          last keypress.  
  [1midle[22m([4ms[24m) 
          (dtime) Number of seconds (to the nearest microsecond) since the 
          last text was received on the [1msocket[22;0m connected to [1mworld[22;0m <[4ms[24m>, or -1 
          on error.  
#sidle
#sidle()
  [1msidle[22m() 
  [1msidle[22m([4ms[24m) 
          (dtime) Number of seconds (to the nearest microsecond) since the 
          last text was sent on the [1mcurrent socket[22;0m or the [1msocket[22;0m connected to 
          [1mworld[22;0m <[4ms[24m>, or -1 on error.  
#nlog
#nlog()
  [1mnlog[22m()  (int) Number of open log files.  
#nmail
#nmail()
  [1mnmail[22m() 
          (int) Number of monitored mail files containing unread mail.  See 
          [1mmail[22;0m.  
#nread
#nread()
  [1mnread[22m() 
          (int) Returns a positive number if a [1mread[22;0m from the keyboard is in 
          progress, 0 otherwise.  
#getpid
#getpid()
  [1mgetpid[22m() 
          (int) The operating system's process id for tf.  
#gethostname
#gethostname()
  [1mgethostname[22m() 
          (str) Returns the host's name, or an empty string if the host name 
          is not available.  
#systype
#systype()
  [1msystype[22m() 
          (str) System type: "unix" (includes MacOS X), "os/2", or "cygwin32". 
#

Other functions

#
  [1maddworld[22m([4mname[24m, [4mtype[24m, [4mhost[24m, [4mport[24m, [4mchar[24m, [4mpass[24m, [4mfile[24m, [4muse_proxy[24m) 
          Defines or redefines a [1mworld[22;0m.  See "[1maddworld()[22;0m".  
#
  [1meval[22m([4ms1[24m [, [4ms2[24m]) 
          (str) Evaluates <[4ms1[24m> as a [1mmacro body[22;0m.  See: [1m/eval[22;0m.  
#filename
#filename()
  [1mfilename[22m([4ms[24m) 
          (str) Performs filename expansion on <[4ms[24m> as described under 
          "[1mfilenames[22;0m".  
#
  [1mftime[22m([4ms[24m,[4mn[24m) 
  [1mftime[22m([4ms[24m) 
  [1mftime[22m() 
          (str) Formats a system time <[4mn[24m> (obtained from [1mtime()[22;0m) according to 
          format <[4ms[24m>, or prints an error message and returns an empty string 
          if <[4mn[24m> is out of range.  See: [1mftime()[22;0m.  
#mktime
#mktime()
  [1mmktime[22m([4myear[24m [, [4mmonth[24m [, [4mday[24m [, [4mhour[24m [, [4mminute[24m [, [4msecond[24m [, 
  [4mmicrosecond[24m]]]]]]) 
          (atime) Returns the system time in seconds of the date in the local 
          [1mtime zone[22;0m represented by the arguments.  Returns -1 if the arguments 
          do not represent a valid date.  Omitted [4mmonth[24m or day arguments 
          default to 1; other omitted arguments default to 0.  See: [1m%TZ[22;0m, 
          [1mftime()[22;0m, [1m/time[22;0m, 
#
  [1mgetopts[22m(s1, s2) 
          (int) Parse macro options according to format <s1>.  See 
          "[1mgetopts()[22;0m".  
#test()
  [1mtest[22m(s) 
          Interprets the contents of the string s as an expression and returns 
          the result.  See also: [1m/test[22;0m, [1m/expr[22;0m.  
#status_fields()
  [1mstatus_fields[22m([i]) 
          Returns the list of fields of status row i, or row 0 if i is 
          omitted.  [1mstatus area[22;0m.  
#
  [1msubstitute[22m(s [,attrs [,inline]]) 
          (int) Replaces trigger text.  See "[1m/substitute[22;0m".  
#

  Examples: 

  Capitalize first letter of string <s>: 

        [1mstrcat[22;0m([1mtoupper[22;0m([1msubstr[22;0m(s, 0, 1)), [1msubstr[22;0m(s, 1))

  Extract the number from a string <dbref> of the form "(#123PML)": 

        0 + [1msubstr[22;0m(dbref, [1mstrchr[22;0m(dbref, "#") + 1)

  See: [1mexpressions[22;0m 

&getopts
&getopts()

getopts()

  Usage: 

  getopts(<[4moptions[24m> [, <[4minit[24m>])
  ____________________________________________________________________________

  [1mgetopts()[22;0m is a [1mfunction[22;0m that parses and validates [1mmacro[22;0m options according to 
  the format described by <[4moptions[24m>.  <[4mOptions[24m> is a list of letters that 
  [1mgetopts()[22;0m will accept.  If a letter is followed by ":", the option will be 
  expected to have a string argument; if a letter is followed by "#", the 
  option will be expected to have a [1mexpression[22;0m argument that evaluates to a 
  (possibly signed) integer; if a letter is followed by "@", the option will 
  be expected to have a time argument.  The option syntax accepted by 
  [1mgetopts()[22;0m is a subset of that accepted by builtin tf commands, as described 
  under "[1moptions[22;0m".  

  When an option is found, [1mgetopts()[22;0m creates a new local [1mvariable[22;0m named 
  "opt_X", where "X" is the letter of the option.  If an argument is expected, 
  the [1mvariable[22;0m will get that argument as its value; otherwise, the [1mvariable[22;0m 
  will have a value of "1".  

  If <[4minit[24m> is given, the [1mvariables[22;0m corresponding to each letter of <[4moptions[24m> 
  are initialized to <[4minit[24m> before processing the command line options.  If 
  <[4minit[24m> is omitted, the [1mvariables[22;0m are not initialized, so if [1mvariables[22;0m with 
  the same names already exist and are not set by [1mgetopts()[22;0m, they will be 
  unchanged.  You can use this to set the [1mvariables[22;0m to some default value 
  before calling [1mgetopts()[22;0m.  

  The argument list will be shifted to discard the options that have been 
  parsed, so [1m%{*}[22;0m will contain the remainder of the arguments, without the 
  options.  

  If [1mgetopts()[22;0m encounters an error, it will print an error message and return 
  0; otherwise, it returns nonzero.  

  Using [1mgetopts()[22;0m, [1m/escape[22;0m, and [1m/split[22;0m, it is possible to write [1mmacros[22;0m that 
  behave just like builtin tf commands.  

  Here's a contrived example to illustrate how [1mgetopts()[22;0m works: 


    [1m/def[22;0m foo = \
        [1m/if[22;0m (!getopts("abn#s:", "")) [1m/return[22;0m 0%; [1m/endif[22;0m%; \
        [1m/echo[22;0m option a:  %{opt_a}%;\
        [1m/echo[22;0m option b:  %{opt_b}%;\
        [1m/echo[22;0m option n:  %{opt_n}%;\
        [1m/echo[22;0m option s:  %{opt_s}%;\
        [1m/echo[22;0m args: [1m%{*}[22;0m%;\
        [1m/split[22;0m [1m%{*}[22;0m%;\
        [1m/echo[22;0m name: [1m%{P1}[22;0m%;\
        [1m/echo[22;0m body: [1m%{P2}[22;0m

  Now, all of these commands are equivalent: 

        /foo -a -b -n5 -s"can't stop" -- whiz = bang biff
        /foo -a -b -n5 -s'can\'t stop' whiz = bang biff
        /foo -n5 -ba -s`can't stop` whiz = bang biff
        /foo -as"can't stop" -bn5 whiz = bang biff

  and produce this output: 

        option a:  1
        option b:  1
        option n:  5
        option s:  can't stop
        args: whiz = bang biff
        name: whiz
        body: bang biff

  But the command: 

        /foo -a -x whiz = bang biff

  produces the error: 

        % foo: invalid option 'x'
        % foo: options: -ab -n<integer> -s<string>

  See: [1mexpressions[22;0m, [1mfunctions[22;0m, [1moptions[22;0m, [1m/escape[22;0m, [1m/split[22;0m 

&style
&tips
&hints

hints

  Some hints and style tips: 

    * Use a high-[1mpriority[22;0m [1mtrigger[22;0m on yourself to prevent loops.  Say I 
      want to throw a tomato at anyone who says the word "tomato", and I write 
      the following [1mtrigger[22;0m: 

        [1m/def[22;0m -t"*tomato*" tomato = :throws a tomato at [1m%1[22;0m.

      If Ben uses the word tomato, I will [1mtrigger[22;0m, and then see the text 
      "Hawkeye throws a tomato at Ben." That text contains the word tomato, 
      which will [1mtrigger[22;0m me again, creating an infinite loop.  One way to 
      prevent this is by creating a high-[1mpriority[22;0m [1mtrigger[22;0m on myself which does 
      nothing: 

        [1m/def[22;0m -p99999 -t"{Hawkeye|You}*" anti_loop

      Now, when I see "Hawkeye throws a tomato at Ben", the /anti_loop [1mtrigger[22;0m 
      will catch it before /tomato does, so I won't loop.  

    * Use multiple lines, spacing, and indentation in [1m/load[22;0m files.  
      Normally, commands must be on one line.  But in files read with [1m/load[22;0m, 
      if a line ends in '\', the following line will have leading whitespace 
      stripped and the two lines will be joined.  This makes it much easier 
      (for humans) to read complex [1mmacros[22;0m.  Compare the two identical [1mmacros[22;0m 
      below, and see which is easier to read.  


        [1m/def[22;0m count=[1m/let[22;0m i=1%;[1m/while[22;0m (i<=[1m%1[22;0m) say %i%;[1m/let[22;0m i=$[i+1]%;[1m/done[22;0m


        [1m/def[22;0m count = \
            [1m/let[22;0m i=1%; \
            [1m/while[22;0m ( i <= [1m%1[22;0m ) \
                say %i%; \
                [1m/let[22;0m i=$[i + 1]%; \
            [1m/done[22;0m

    * Use comments in [1m/load[22;0m files.  Complicated [1mmacros[22;0m are much easier to 
      read if you include a short comment describing the arguments to the 
      [1mmacro[22;0m and what it does.  Lines beginning with ';' or '#' are comments, 
      and are ignored during [1m/load[22;0m.  

    * Name all [1mtriggers[22;0m and [1mhooks[22;0m.  If you ever need to [1m/load[22;0m a file a 
      second time, [1mtriggers[22;0m, [1mhilite[22;0ms, [1mhooks[22;0m, and [1mgag[22;0ms without names may be 
      duplicated.  But if they are named, old copies of [1mmacros[22;0m will be 
      replaced with new copies of [1mmacros[22;0m with the same name.  Naming [1mmacros[22;0m 
      also makes them easier to manipulate with commands like [1m/list[22;0m and 
      [1m/undef[22;0m.  

    * Don't use "weird" characters in [1mmacro[22;0m names.  Although any [1mmacro[22;0m 
      name is legal, some characters can have unwanted [1mexpansion[22;0m effects.  
      Weird characters are also harder to read.  You should stick to letters, 
      numbers, and '_' characters.  In particular, avoid '~' characters, since 
      they are used in library [1mmacros[22;0m.  

    * Use local [1mvariables[22;0m instead of global [1mvariables[22;0m if possible.  This 
      avoids conflicts when two [1mmacros[22;0m use a [1mvariable[22;0m with the same name.  If 
      you're using a [1mvariable[22;0m in an [1mexpression[22;0m, use [1m/let[22;0m first to initialize 
      the [1mvariable[22;0m in the local scope.  But remember, when you use a [1mvariable[22;0m 
      reference (by name, as opposed to a [1mvariable[22;0m substitution using "%"), TF 
      uses dynamic scoping (see: [1mscope[22;0m).  

    * Use [1mvariable[22;0m references instead of %-substitutions in [1mexpressions[22;0m.  
      Because [1mmacro[22;0m bodies are [1mexpanded[22;0m, something like "[1m/test[22;0m [1m%1[22;0m" is prone to 
      problems if [1m%1[22;0m contains any special characters.  But by using a [1mvariable[22;0m 
      reference you can avoid this problem; for example, "[1m/test[22;0m {1}".  

    * "[1m/set[22;0m [1mpedantic[22;0m=on" to make tf generate warnings about some potential 
      problems.  

    * "[1m/set[22;0m [1mdefcompile[22;0m=on" to see syntax errors in a [1mmacro[22;0m when you define 
      it, instead of waiting until you first run it.  

    * "[1m/set[22;0m [1mmecho[22;0m=on" to see what commands are being executed, or [1m/connect[22;0m 
      to a normal or [1mconnectionless[22;0m [1msocket[22;0m defined with "[1m/addworld[22;0m -e" to see 
      what you're sending to the [1msocket[22;0m.  

    * "[1m/set[22;0m [1memulation[22;0m=debug" and "[1mtelopt[22;0m=on" to see exactly what the 
      socket is sending to tf.  

    * Use the -n or -l option of [1m/trigger[22;0m to see a list of [1mtrigger[22;0m [1mmacros[22;0m 
      that would match a given line.  

  See also [1mdebugging[22;0m.  

&history

history

  Associated topics: 

  [1mscrollback[22;0m
  [1m/recall[22;0m
  [1m/quote[22;0m
  [1m/histsize[22;0m
  [1m/recordline[22;0m
  ^<[4mstring1[24m>^<[4mstring2[24m>
  Recall previous/next keys ([1mRECALLB[22;0m/[1mRECALLF[22;0m, default ^P and ^N)
  Recall beginning/end keys ([1mRECALLBEG[22;0m/[1mRECALLEND[22;0m, default ^[< and ^[>)
  Search backward/forward keys ([1mSEARCHB[22;0m/[1mSEARCHF[22;0m, default ^[p and ^[n)

  TinyFugue stores lines in 4 different types of [1mhistory[22;0m lists.  Input [1mhistory[22;0m 
  records the last 100 non-repeated commands from the keyboard, including the 
  current line.  Each world has a world [1mhistory[22;0m, which stores 1000 lines of 
  output from that world.  Local history stores 100 lines of output generated 
  by TF, i.e.  anything that didn't come from a world.  Global [1mhistory[22;0m is an 
  integrated list of 1000 lines from TF and every world.  The [1mhistory[22;0m sizes 
  can be changed with the [1m/histsize[22;0m command and the [1m%{histsize}[22;0m [1mvariable[22;0m.  

  [1m/recall[22;0m is used to display text from any of the [1mhistory[22;0m lists.  The [1m/quote[22;0m 
  command may be used to quote out of any [1mhistory[22;0m list using the [1m/quote[22;0m # 
  feature.  

#^^
#^
  Typing ^<[4mstring1[24m>^<[4mstring2[24m> finds the last command in the input [1mhistory[22;0m 
  containing <[4mstring1[24m>, replaces <[4mstring1[24m> with <[4mstring2[24m>, and executes the 
  modified line.  

#
  The recall keys replace the current input with a line from the input [1mhistory[22;0m 
  list.  See [1m/dokey[22;0m for details.  

  See also [1m/log[22;0m.  
&hook
&hooks

hooks

  Associated topics: 
  [1m/def[22;0m    define a [1mmacro[22;0m with any fields 
  [1m/hook[22;0m   define a [1mhook[22;0m [1mmacro[22;0m 
  [1m/unhook[22;0m 
          undefine a [1mhook[22;0m [1mmacro[22;0m 
  [1m/trigger -h[22;0m 
          call a [1mhook[22;0m [1mmacro[22;0m 
  [1m%hook[22;0m   enable [1mhooks[22;0m 
  [1m%max_hook[22;0m 
          maximum [1mhook[22;0m rate 

  [1mHooks[22;0m are a method of calling a [1mmacro[22;0m based on special events within TF, in 
  much the same way as [1mtriggers[22;0m call [1mmacros[22;0m based on text received from a 
  [1msocket[22;0m.  [1mHooks[22;0m allow the user to customize the behavior of TinyFugue and 
  automate special functions.  

  A [1mhook[22;0m definition has two parts: an <[4mevent[24m> and a <[4mpattern[24m>.  When the event 
  occurs, the [1mmacro[22;0m will be executed if the arguments supplied by the event 
  match the [1mmacro[22;0m's <[4mpattern[24m> (see the section on "[1mpatterns[22;0m").  

  If multiple [1mhooks[22;0m match the same event and pattern, one or more are selected 
  as described under "[1mpriority[22;0m".  

  Most [1mhooks[22;0m have a default message associated with them, which will be 
  displayed with the [1mattributes[22;0m of the [1mhook[22;0m if one is defined.  Thus a [1mhook[22;0m 
  with a [1mgag[22;0m [1mattribute[22;0m will suppress the display of the message.  

  [1mHook[22;0m may have [1mmulti-shots[22;0m, in which case it and the [1mmacro[22;0m it is associated 
  with is removed after executing a specified number of times.  

  In the table below, 'A' or 'W' in the message column indicates the location 
  of the message display: 
  A       the message is printed to the the [1malert stream[22;0m (i.e., the status 
          line).  
  W       the message is printed to the appropriate [1mworld's stream[22;0m; if that 
          world is not the foreground world, the message is also printed to 
          the [1malert stream[22;0m.  
  Otherwise, the message is sent to the the [1mtferr stream[22;0m (i.e., the screen).  

    Event Name  Arguments       Default Message or Action
    ----------  ---------       -------------------------
#ACTIVITY
    ACTIVITY    world           A '% Activity in world <[4mworld[24m>'
                                  (called only the first time activity
                                  occurs on a given [1msocket[22;0m.)
#BAMF
    BAMF        world           W '% [1mBamfing[22;0m to <[4mworld[24m>'
#BGTEXT
    BGTEXT      world             Text was printed in background world <[4mworld[24m>
#BACKGROUND
#BGTRIG
    BGTRIG      world           A '% [1mTrigger[22;0m in world <[4mworld[24m>'
#CONFAIL
    CONFAIL     world, reason   W '% [1mConnection[22;0m to <[4mworld[24m> failed: <[4mreason[24m>'
#CONFAIL
    CONFAIL     world, reason   W '% Unable to [1mconnect[22;0m to <[4mworld[24m>: <[4mreason[24m>'
#CONFLICT
    CONFLICT    [1mmacro[22;0m             '% <[4mmacro[24m> conflicts with builtin command.'
#CONNECT
    CONNECT     world, cipher   W '% [1mConnected[22;0m to <[4mworld[24m>[ using <[4mcipher[24m>].'
#ICONFAIL
    ICONFAIL    world, reason   W '% [1mIntermediate connection[22;0m to <[4mworld[24m>
                                  failed: <[4mreason[24m>'
#DISCONNECT
    DISCONNECT  world, reason   W '% [1mConnection[22;0m to <[4mworld[24m> closed: <[4mreason[24m>.'
                                  (Called if you send the server's disconnect
                                  command (e.g., "QUIT") or [1msocket[22;0m closes, but
                                  not if you use [1m/dc[22;0m.)
#KILL
    KILL        pid               ([1mprocess[22;0m ends)
#LOAD
    LOAD        file              '% [1mLoading[22;0m commands from file <[4mfile[24m>'
#LOADFAIL
    LOADFAIL    file, reason      '% <[4mfile[24m>: <[4mreason[24m>'
#LOG
    LOG         file              '% [1mLogging[22;0m to file <[4mfile[24m>'
#LOGIN
    LOGIN       world             ([1mautomatic login[22;0m)
#MAIL
    MAIL        file            A '% You have new mail in <[4mfile[24m>.'
                                  (See: [1mmail[22;0m).
#MORE
    MORE                          '[1m--More--[22;0m' (reverse bold)
#NOMACRO
    NOMACRO     name              '% <[4mname[24m>: No such command or macro'
#PENDING
    PENDING     world           W '% Hostname lookup for <[4mworld[24m> in progress'
    PENDING     world, address  A '% Trying to [1mconnect[22;0m to <[4mworld[24m>: <[4maddress[24m>'
#PREACTIVITY
    PREACTIVITY world             (Activity in world <[4mworld[24m>)
                                  (called only the first time activity
                                  occurs on a given [1msocket[22;0m.)
#PROCESS
    PROCESS     pid               [1mprocess[22;0m starts
#PROMPT
    PROMPT      text              <[4mtext[24m> is a partial (unterminated) line
                                  from the server.  See "[1mprompts[22;0m"
#PROXY
    PROXY       world             ([1mproxy[22;0m connection to <[4mworld[24m> has completed)
#REDEF
    REDEF       obj_type, name    '% Redefined <[4mobj_type[24m> <[4mname[24m>'
#RESIZE
    RESIZE      columns, lines    (window was resized)
                                  (see also: [1mcolumns()[22;0m, [1mlines()[22;0m)
#SEND
    SEND        text              (text sent to [1mcurrent[22;0m [1msocket[22;0m)
                                  (see note below ("[1mhooks[22;0m"))
#SHADOW
    SHADOW      var_name          '% Variable <[4mvar_name[24m> overshadows global'
#SHELL
    SHELL       type, command     '% Executing <[4mtype[24m>: <[4mcommand[24m>'
#SIGHUP
    SIGHUP                        (SIGHUP [1msignal[22;0m caught; tf may terminate)
#SIGTERM
    SIGTERM                       (SIGTERM [1msignal[22;0m caught; tf terminates)
#SIGUSR1
    SIGUSR1                       (SIGUSR1 [1msignal[22;0m caught; no effect)
#SIGUSR2
    SIGUSR2                       (SIGUSR2 [1msignal[22;0m caught; no effect)
#WORLD
    WORLD       world           W ([1mforeground socket[22;0m changes)

#
  Notes: 

  The -w and -T options to [1m/def[22;0m can be used to restrict [1mhooks[22;0m to matching only 
  when the [1mcurrent[22;0m world matches the world or world type.  

  When a [1mmacro[22;0m is defined with the same name as an existing [1mmacro[22;0m, the REDEF 
  [1mhook[22;0m will be called, [4munless[24m the new [1mmacro[22;0m is identical to the original.  

  BGTRIG used to be called BACKGROUND, and the old name still works.  Its "% 
  Trigger in world " message can be quieted for individual triggers by 
  defining them with [1m/def[22;0m [1m-q[22;0m, or for all triggers with "/def -ag -hBGTRIG".  

  The SEND [1mhook[22;0m is called whenever text would be sent to the [1mcurrent[22;0m [1msocket[22;0m.  
  If a SEND [1mhook[22;0m matches the text that would be sent, the text is not sent 
  (unless the hook was defined with [1m/def -q[22;0m), and the [1mhook[22;0m is executed 
  instead.  By default, SEND [1mhooks[22;0m are not invoked from [1msend()[22;0m or [1m/send[22;0m, but 
  there is an option to do so; SEND [1mhooks[22;0m are invoked from any [1mmacro[22;0m or 
  command line that sends plain text.  

  When successfully connected to a new [1msocket[22;0m, these events occur: 1) If this 
  is a [1mproxy[22;0m connection, the PROXY [1mhook[22;0m is called; 2) If there is a file 
  associated with the world, the file will be loaded (and the LOAD [1mhook[22;0m will 
  be called).  3) If this is not a [1mproxy[22;0m connection, the CONNECT [1mhook[22;0m is 
  called; 4) If [1m%{login}[22;0m is on, a character and password are defined, and this 
  is not a [1mproxy[22;0m connection, the [1mLOGIN[22;0m [1mhook[22;0m is called.  

  When a (non-[1mgag[22;0mged) line is displayed in a background world, the PREACTIVITY 
  hook is called immediately before the line is displayed, and the ACTIVITY 
  hook is called immediately after.  Thus, functions like [1mmoresize()[22;0m and 
  [1mnactive()[22;0m will give different results in the two hooks.  Any activity 
  generated by a PREACTIVITY hook will not recursively cause another 
  PREACTIVITY or ACTIVITY event.  

  The SIGHUP, SIGTERM, SIGUSR1, and SIGUSR2 [1mhooks[22;0m are called when the 
  corresponding [1msignal[22;0m is received.  If SIGHUP is received and SIGHUP was not 
  ignored when tf was started, or SIGTERM was received, TF will terminate 
  immediately after executing the [1mhook[22;0m; if the [1mhook[22;0m calls any commands with 
  delayed effects (a [1m/repeat[22;0m or [1m/quote[22;0m without -S, a nonblocking [1m/connect[22;0m, 
  etc), those effects will not occur before termination.  

  A hook's message, if any, is displayed (with its [1mattributes[22;0m) before any of 
  the hooked [1mmacros[22;0m are executed.  Prior to version 5.0, the message was 
  displayed after executing hooked [1mmacros[22;0m, which may have generated their own 
  output, which was sometimes confusing.  

  Examples: 


      [1m/hook[22;0m ACTIVITY|DISCONNECT {TT|SM}* = [1m/world[22;0m [1m%1[22;0m

  will cause TF to automatically switch to TT or SM if either becomes active 
  or disconnected.  

      [1m/def[22;0m -T'tiny.mush' -hSEND mush_escape = [1m/send[22;0m - $([1m/escape[22;0m \%[ [1m%*[22;0m)

  will catch any line sent to a world of type 'tiny.mush', escape all 
  occurrences of '%', '[' and '\' within that line, and send the new line 
  instead of the original.  This is useful for avoiding unwanted 
  interpretation of '%', '[', and '\' on TinyMUSH servers.  

      [1m/hook[22;0m SIGHUP = [1m/log[22;0m on%; [1m/recall[22;0m /10

  will [1mlog[22;0m the last 10 lines of output if you are unexpectedly disconnected 
  from your tf session.  

#CONNETFAIL
  The CONNETFAIL hook, which existed in versions 5.0 alpha 13 through 5.0 beta 
  6, has been replaced with the [1mICONFAIL[22;0m hook.  
#

  See also: [1mmacros[22;0m, [1mtriggers[22;0m, [1mpatterns[22;0m, [1mpriority[22;0m, [1msignals[22;0m.  

&topics

topics

  Topics marked with + are new; those marked with * have changed since the 
  last version.  Many topics also have subtopics that are not listed here 
  (e.g., individual [1mvariables[22;0m, [1mhooks[22;0m, and functions).  


   *[1mcopying[22;0m       copyright; no warranty
    [1mintro[22;0m         introduction to tf
    [1mstartup[22;0m       how to start tf
    [1minterface[22;0m     how input works
    [1mtfrc[22;0m          personal config file
   *[1mvisual[22;0m        split-screen mode
   *[1mcommands[22;0m      list of commands
   *[1mworlds[22;0m        defining worlds
   *[1mpatterns[22;0m      glob and regexp pattern matching
   *[1mvariables[22;0m     state and environment
   *[1mglobals[22;0m       special tf variables
    [1mattributes[22;0m    special text display
    [1mprompts[22;0m       using LP/Diku prompts
    [1mproblems[22;0m      bugs, core dumps, etc.
   *[1mevaluation[22;0m    macro body execution
    [1mmacros[22;0m        user-defined commands
   *[1msockets[22;0m       world connections
    [1mhistory[22;0m       recall and logging
    [1mpriority[22;0m      trigger/hook selection
   *[1mkeybindings[22;0m   keyboard operations
    [1mcolor[22;0m         terminal color codes
   *[1mprotocols[22;0m     protocols supported by TF
    [1mexpressions[22;0m   math and string operations
    [1mtriggers[22;0m      automatic command execution based on incoming text
   *[1mhooks[22;0m         automatic command execution based on tf events
   +[1mmail[22;0m          mail checking
    [1mlibrary[22;0m       macros and variables in stdlib.tf
   *[1mtools[22;0m         extra commands in tools.tf
    [1mutilities[22;0m     useful extra command files
   *[1mprocesses[22;0m     timed commands and command quoting
   *[1msubs[22;0m          arithmetic, command, [1mmacro[22;0m, and variable substitutions
   *[1mfunctions[22;0m     special [1mexpression[22;0m operations
   *[1mhints[22;0m         some hints and style tips for [1mmacro[22;0m programming
   +[1mdebugging[22;0m     debugging your [1mmacros[22;0m
   *[1mtfio[22;0m          output, error, and world streams
   *[1mproxy[22;0m         connecting to outside hosts via a proxy server (firewall)
   +[1mlocale[22;0m        multi-language support
    

&typing
&user
&interface

interface

  Any input line that does not begin with '/' will be sent directly to the 
  [1mforeground[22;0m world, if there is one.  A line starting with more than one '/' 
  will be sent to the forground [1msocket[22;0m after having the first '/' removed.  
  (Exception: lines may be caught with a SEND [1mhook[22;0m before being sent; see 
  "[1mhooks[22;0m").  

  Any input line beginning with a single '/' is a TF command, which will be 
  interpreted as described in "[1mevaluation[22;0m".  

  Input lines of the form "^old^new" will cause TF to search backward in the 
  input [1mhistory[22;0m for a line containing "old", replace that text with "new", and 
  execute the modified command.  See: [1mhistory[22;0m.  

  Many special functions, such as backspace, can be performed by special keys 
  or sequences of keys.  See "[1mdokey[22;0m" for a complete list.  You can also define 
  your own commands and bind them to key sequences.  See [1mbind[22;0m.  

  Normally, user input does not undergo the [1mexpansion[22;0m that [1mmacro[22;0m bodies 
  undergo.  The [1m/eval[22;0m command can be used to [1mexpand[22;0m text before executing it.  
  If the [1m%{sub}[22;0m flag is on (it is off by default), user input undergoes [1mmacro[22;0m 
  body [1mexpansion[22;0m without the [1m%{sub}[22;0m flag.  The [1m%{sub}[22;0m flag also applies to 
  text generated by "^old^new" history commands.  See: [1mhistory[22;0m, [1m/sub[22;0m, 
  [1mvariables[22;0m 

  Control characters may be input literally.  A literal control character will 
  be displayed in the input window in printable form in bold reverse.  Note 
  that since most control keys are also parts of the default keybindings, it 
  will usually be necessary to type ^V ([1m/dokey[22;0m LNEXT) to avoid invoking the 
  keybinding.  

  International characters may be input if your [1mlocale[22;0m is set to a locale that 
  supports them and your system supports locales.  Any input character that is 
  not valid in your locale and has the high bit set (normally generated by 
  holding the "meta" key) will be translated to ESC plus that character with 
  the high bit stripped (assuming [1m%meta_esc[22;0m is on).  This allows M-x and ^[x 
  to invoke the same ^[x keybinding.  See [1mlocale[22;0m, [1m%meta_esc[22;0m, [1m%istrip[22;0m.  

  If standard input is not a terminal, [1mvisual[22;0m mode will not be allowed, and tf 
  will continue to operate even after EOF is read, until /quit or something 
  else terminates it.  

  See also: [1mvisual[22;0m, [1moptions[22;0m 

&intro
&me
&newbie
&tinyfugue
&introduction

introduction

  TinyFugue is a MUD client.  It helps you connect to a MUD, in a much more 
  convenient manner than telnet.  You can connect to a mud world using the 
  same syntax as you would with telnet: "[1mtf[22;0m <[4mhost[24m> <[4mport[24m>".  Or, while running 
  tf, you can use "[1m/connect[22;0m <[4mhost[24m> <[4mport[24m>".  To make things easier, you can 
  give names to worlds, using [1m/addworld[22;0m, and then use "[1mtf[22;0m <[4mname[24m>" and 
  "[1m/connect[22;0m <[4mname[24m>".  If you store a set of [1m/addworld[22;0m commands in a file, TF 
  can read them automatically when it starts.  You can even connect to more 
  than one world at the same time, and switch between them.  See: [1m/connect[22;0m, 
  [1m/fg[22;0m, [1m/addworld[22;0m, [1mworlds[22;0m, [1mtfrc[22;0m.  

  Normally, TF will split the screen into two windows: one for input, and one 
  for output.  TF will display useful information on the line separating the 
  two windows, such as the name of the [1mforeground[22;0m world.  See: [1mwindows[22;0m.  

  Any line you type that starts with a single '/' is a tf command.  Anything 
  else you type will be sent to the mud.  See: [1minterface[22;0m, [1mcommands[22;0m.  

  You can define your own tf commands, called [1mmacros[22;0m.  The simplest type of 
  [1mmacro[22;0m is just an abbreviation or alias for a longer command or commands.  
  But [1mmacros[22;0m can also perform much more powerful tasks.  See: [1mmacros[22;0m, [1m/def[22;0m.  

  You can tell tf to watch for certain patterns in the text from the mud, and 
  then do special things when it sees that pattern: display the text in a 
  special way ([1mhilite[22;0m); not display the text at all ([1mgag[22;0m); execute a [1mmacro[22;0m 
  command ([1mtrigger[22;0m); or do any combination of these.  See: [1mattributes[22;0m, 
  [1mtriggers[22;0m, [1m/hilite[22;0m, [1m/gag[22;0m, [1m/trig[22;0m, [1m/def[22;0m.  

  TF keeps a [1mhistory[22;0m of every line it prints, every line sent by the mud, and 
  every command you enter.  You can see those histories using [1m/recall[22;0m.  You 
  can also have this text saved in a file using [1m/log[22;0m.  See: [1mhistory[22;0m, [1m/recall[22;0m, 
  [1m/log[22;0m.  

  See also: [1mtopics[22;0m 

&keys
&key
&kbbind
&kbfunc
&kbfunc.tf
&kbbind.tf
&keybindings

keybindings

Default keybindings

  TF's default command line editing keys are similar to those in emacs and 
  bash.  In addition, several features may be invoked by more than one 
  keybinding, and TF has keybindings for unique features like switching the 
  [1mforeground[22;0m [1msocket[22;0m.  

  Here, and throughout the TF documentation, the notation "^X" means the 
  character generated by typing the X key while holding the CTRL key.  Also, 
  "^[" can be more easily typed just by pressing the ESC key.  [1m/Def -b[22;0m and 
  [1m/bind[22;0m accept the ^X notation as well as "\<[4mnumber[24m>" notation, where <[4mnumber[24m> 
  is the octal, hexadecimal, or decimal number of the character's ascii value. 
  For example, the escape character can be given in any of these forms: ^[, 
  \033, \0x1B, or \27.  

  In the tables below, keys with "*" in the "Meaning" column make use of [1mkbnum[22;0m 
  (see [1mbelow[22;0m).  

#named keys
Named keys
  To redefine the named keys, see the section titled "[1mMapping Named Keys to 
  functions[22;0m".  

  Key         Command                 Meaning
  ---         -------                 -------
  Up          /kb_up_or_recallb       *cursor up or recall bkwd input [1mhistory[22;0m
  Down        /kb_down_or_recallf     *cursor down or recall fwd input [1mhistory[22;0m
  Right       [1m/dokey[22;0m RIGHT            *cursor right
  Left        [1m/dokey[22;0m LEFT             *cursor left
  Center      (none)

  Esc_Left    [1m/fg[22;0m -<                  *[1mforeground[22;0m previous [1msocket[22;0m
  Esc_Right   [1m/fg[22;0m ->                  *[1mforeground[22;0m next [1msocket[22;0m

  Ctrl_Up     [1m/dokey_recallb[22;0m          *recall backward input
  Ctrl_Down   [1m/dokey_recallf[22;0m          *recall forward input
  Ctrl_Right  [1m/dokey_wright[22;0m           *word right
  Ctrl_Left   [1m/dokey_wleft[22;0m            *word left

  Insert      [1m/test[22;0m [1minsert[22;0m:=![1minsert[22;0m    toggle [1minsert[22;0m mode
  Delete      [1m/dokey[22;0m dch              *delete character
  Home        [1m/dokey_home[22;0m              cursor to beginning of line
  End         [1m/dokey_end[22;0m               cursor to end of line
  PgDn        [1m/dokey_pgdn[22;0m             *scroll forward a screenful
  PgUp        [1m/dokey_pgup[22;0m             *scroll back a screenful
  Tab         [1m/dokey page[22;0m             *scroll forward a screenful

  Ctrl_Home   [1m/dokey_recallbeg[22;0m         recall first line of input
  Ctrl_End    [1m/dokey_recallend[22;0m         recall last line of input
  Ctrl_PgDn   [1m/dokey_flush[22;0m             scroll forward to last screenful
  Ctrl_PgUp   (reserved for future use)

  F1          [1m/help[22;0m                    help
  F2          (none)                   (function key F1)
  ...
  F20         (none)                   (function key F20)

  nkpTab      (none)                   (see "[1mkeypad[22;0m" section below)
  nkpEnt      (none)                   (see "[1mkeypad[22;0m" section below)
  nkp*        (none)                   (see "[1mkeypad[22;0m" section below)
  nkp+        (none)                   (see "[1mkeypad[22;0m" section below)
  nkp,        (none)                   (see "[1mkeypad[22;0m" section below)
  nkp-        (none)                   (see "[1mkeypad[22;0m" section below)
  nkp.        (none)                   (see "[1mkeypad[22;0m" section below)
  nkp/        (none)                   (see "[1mkeypad[22;0m" section below)
  nkp0        (none)                   (see "[1mkeypad[22;0m" section below)
  ...
  nkp9        (none)                   (see "[1mkeypad[22;0m" section below)
  nkp=        (none)                   (see "[1mkeypad[22;0m" section below)

#unnamed keys
Unnamed key sequences

  String  Command                 Meaning
  ------  -------                 -------
  "^A"    [1m/dokey_home[22;0m              cursor to beginning of line
  "^B"    [1m/dokey[22;0m LEFT             *cursor left
  "^D"    [1m/dokey_dch[22;0m              *delete character to the right
  "^E"    [1m/dokey_end[22;0m               cursor to end of line
  "^F"    [1m/dokey[22;0m RIGHT            *cursor right
  "^G"    [1m/beep[22;0m 1                  beep
  "^H"    (internal)              *backspace
  "^I"    [1m/key_tab[22;0m                 perform the function assigned to the TAB key
  "^J"    (internal)               execute current line
  "^K"    [1m/dokey_deol[22;0m              delete to end of line
  "^L"    [1m/dokey[22;0m redraw            redraw (not clear) screen
  "^M"    (internal)               execute current line
  "^N"    [1m/dokey[22;0m recallf          *recall forward input [1mhistory[22;0m
  "^P"    [1m/dokey[22;0m recallb          *recall backward input [1mhistory[22;0m
  "^Q"    [1m/dokey[22;0m LNEXT             input next key literally (may be overridden
                                   by [1mterminal[22;0m)
  "^R"    [1m/dokey[22;0m REFRESH           refresh line
  "^S"    [1m/dokey[22;0m PAUSE             pause screen
  "^T"    [1m/kb_transpose_chars[22;0m     *transpose characters
  "^U"    [1m/kb_backward_kill_line[22;0m   delete to beginning of line
  "^V"    [1m/dokey[22;0m LNEXT             input next key literally
  "^W"    [1m/dokey[22;0m BWORD            *delete backward word (space-delimited)
  "^?"    (internal)              *backspace
  "^X^R"  [1m/load[22;0m [1m~/.tfrc[22;0m            reload personal config file
  "^X^V"  [1m/version[22;0m                 display version information
  "^X^?"  [1m/kb_backward_kill_word[22;0m  *delete backward word (punctuation-delimited)
  "^X["   [1m/dokey[22;0m_hpageback        *scroll back a half screenful
  "^X]"   [1m/dokey[22;0m_hpage            *scroll forward a half screenful
  "^X{"   [1m/dokey[22;0m_pageback         *scroll back a screenful
  "^X}"   [1m/dokey[22;0m_page             *scroll forward a screenful
  "^[^E"  [1m/kb_expand_line[22;0m          [1mexpand[22;0m current input line in place
  "^[^H"  [1m/kb_backward_kill_word[22;0m  *delete backward word (punctuation-delimited)
  "^[^I"  [1m/complete[22;0m                complete current word, depending on context
  "^[^L"  [1m/dokey[22;0m clear             clear screen (can be refilled with [1mscrollback[22;0m)
  "^[^N"  [1m/dokey[22;0m line             *scroll forward one line
  "^[^P"  [1m/dokey[22;0m lineback         *scroll back one line
  "^[^W"  [1m/complete worldname[22;0m      complete TF world name
  "^[$"   [1m/complete macroname[22;0m      complete TF macro name
  "^[%"   [1m/complete variable[22;0m       complete TF variable name
  "^[/"   [1m/complete filename[22;0m       complete file name (unix only)
  "^[ "   [1m/kb_collapse_space[22;0m       change multiple spaces to a single space
  "^[-"   [1m/set[22;0m [1mkbnum[22;0m=-             start [1mkbnum[22;0m entry with -
  "^[0"   [1m/set[22;0m [1mkbnum[22;0m=+0            start [1mkbnum[22;0m entry with 0
  ...
  "^[9"   [1m/set[22;0m [1mkbnum[22;0m=+9            start [1mkbnum[22;0m entry with 9
  "^[;"   [1m/complete user_defined[22;0m   complete from [1m%{completion_list}[22;0m
  "^[="   [1m/kb_goto_match[22;0m           move cursor to matching parenthesis/bracket
  "^[."   [1m/kb_last_argument[22;0m        input last word of previous line
  "^[<"   [1m/dokey[22;0m recallbeg         go to beginning of input history
  "^[>"   [1m/dokey[22;0m recallend         go to end of input history
  "^[J"   [1m/dokey[22;0m selflush          selective flush (similar to "[1m/dokey[22;0m flush"
                                   followed by "[1m/limit[22;0m -a")
  "^[L"   [1m/kb_toggle_limit[22;0m         toggle between [1m/unlimit[22;0m and [1m/relimit[22;0m
  "^[_"   [1m/kb_last_argument[22;0m        input last word of previous line
  "^[b"   [1m/dokey_wleft[22;0m            *cursor to beginning of word
  "^[c"   [1m/kb_capitalize_word[22;0m     *capitalize word
  "^[d"   [1m/kb_kill_word[22;0m           *delete forward word
  "^[f"   [1m/dokey_wright[22;0m           *cursor to end of word
  "^[h"   [1m/dokey[22;0m_hpage            *scroll forward a half screenful
  "^[i"   [1m/complete[22;0m input_history  complete from previously typed words
  "^[j"   [1m/dokey[22;0m flush             jump to last screenful of text
  "^[l"   [1m/kb_downcase_word[22;0m       *convert word to lower case
  "^[n"   [1m/dokey[22;0m searchf          *search forward input [1mhistory[22;0m
  "^[p"   [1m/dokey[22;0m searchb          *search backward input [1mhistory[22;0m
  "^[u"   [1m/kb_upcase_word[22;0m         *convert word to upper case
  "^[v"   [1m/@test[22;0m insert:=!insert   toggle insert mode
  "^[w"   /to_active_or_prev_world [1m/fg[22;0m next active world, or previous world
  "^[{"   [1m/fg[22;0m -<                  *[1mforeground[22;0m previous [1msocket[22;0m
  "^[}"   [1m/fg[22;0m ->                  *[1mforeground[22;0m next [1msocket[22;0m
  "^[^?"  [1m/kb_backward_kill_word[22;0m  *delete backward word (punctuation-delimited)
  "^]"    [1m/bg[22;0m                      put all [1msockets[22;0m in [1mbackground[22;0m

#

Other useful commands not bound by default

  Command                 Meaning
  -------                 -------
  [1m/dokey[22;0m_bspc             *delete character
  [1m/dokey[22;0m UP               *cursor up
  [1m/dokey[22;0m DOWN             *cursor down
  [1m/dokey[22;0m RECALLB          *recall input backward
  [1m/dokey[22;0m RECALLF          *recall input forward
  [1m/dokey[22;0m NEWLINE           execute input line

#terminal
#tty
#stty
Terminal keys
  Some keys are interpeted by the terminal, not TF, so if you want to change 
  them, you must do so outside of TF (e.g.  with stty in unix).  Typical unix 
  terminal keys include: 

      Key   Name   Meaning
      ---   ----   -------
      ^C    int    generates a SIGINT [1msignal[22;0m.
      ^\    quit   generates a SIGQUIT [1msignal[22;0m.
      ^Z    susp   [1msuspends[22;0m the TF process

  When TF starts, it disables the terminal driver's "stop" and "start" keys 
  (typically ^S and ^Q), so they are available for binding within TF.  
#

Using keys

  Keys F1...F12 are the function keys, located across the top of most 
  keyboards.  

  Keys with names of the form "esc_<[4mname[24m>" correspond to the ESC key followed 
  by the <[4mname[24m> key.  There is an "esc_<[4mname[24m>" for every single key in the 
  Named Key table above, but only the ones with default meanings are listed in 
  the table; the rest are available for custom definitions.  

  On recent versions of xterm with the modifyCursorKeys resource, tf can 
  recognize when the CTRL, SHIFT, or META modifier is held down while pressing 
  the editor keys (insert, delete, home, end, pgdn, pgup), arrow keys, or 
  numbered function keys, and calls /key_ctrl_<[4mname[24m>, /key_shift_<[4mname[24m>, or 
  /key_meta_<[4mname[24m>, respectively.  Additionally, by default, each 
  /key_meta_<[4mname[24m> calls the corresponding /key_esc_<[4mname[24m>, so, for example, 
  pressing META-Left has the same effect as ESC Left.  Note that some xterms 
  capture shift_insert, shift_pgup, and shift_pgdn by default for their own 
  use, so tf will not receive these sequences.  If you use another terminal 
  emulator that generates unique character sequences for ctrl-, shift-, and 
  meta-modified keys, you can bind those sequences to call the corresponding 
  /key_<[4mmod[24m>_<[4mname[24m> (and send them to the [1mtf author[22;0m for inclusion in a future 
  release of tf).  

#keypad
#numeric keypad
Numeric keypad
  [1mTF[22m tries to put the keypad in "application mode", which on many terminals 
  will make the keypad keys generate unique character sequences.  Application 
  mode can be disabled by setting [1m%keypad[22;0m to "off".  The meaning of your 
  numeric keypad keys depends on your terminal emulator and its settings, the 
  setting of [1m%keypad[22;0m in tf, and the state of your NumLock key.  Two common 
  configurations of the keypad are shown below.  A <[4mname[24m> on a key in the 
  diagram indicates that it is bound in tf to "/key_<[4mname[24m>".  

              configuration A                   configuration B

       +------+------+------+------+     +------+------+------+------+
       |      |      |      |      |     |      |nkp/  |nkp*  |nkp-  |
       +------+------+------+------+     +------+------+------+------+
       |Home  |Up    |PgUp  |      |     |nkp7  |nkp8  |nkp9  |nkp+  |
       +------+------+------+      |     +------+------+------+      |
       |Left  |Center|Right |      |     |nkp4  |nkp5  |nkp6  |      |
       +------+------+------+------+     +------+------+------+------+
       |End   |Down  |PgDn  |      |     |nkp1  |nkp2  |nkp3  |nkpEnt|
       +------+------+------+      |     +------+------+------+      |
       |   Insert    |Delete|      |     |    nkp0     |nkp.  |      |
       +-------------+------+------+     +-------------+------+------+

  How this works for some specific terminals: 
  X Consortium xterm 
          [1m%keypad[22;0m=on and NumLock on gives configuration B above; [1m%keypad[22;0m=off 
          and NumLock on gives normal digit/punctuation keys; and NumLock off 
          gives configuration A.  
  XFree86/X.Org xterm 
          Identical to X Consortium xterm [4mif[24m you disable the "Alt/numlock 
          modifiers" option (under the ctrl-leftclick menu); if you do not, 
          then [1m%keypad[22;0m=on and NumLock on gives normal digit/punctuation keys, 
          and there is no way to get configuration B.  There is also a "VT220 
          keyboard" option; if that is enabled, [1m%keypad[22;0m=on and NumLock off 
          gives configuration B, and all other combinations of [1m%keypad[22;0m and 
          NumLock give normal digit/punctuation keys.  
  linux (Linux console) 
          [1m%keypad[22;0m=on gives configuration B, with these changes: "NumLock" 
          calls /key_f1, "/" calls /key_f2, "*" calls /key_f3, and "-" calls 
          /key_f4.  With [1m%keypad[22;0m=off, NumLock chooses between configuration A 
          and normal digit/punctuation keys.  (Prior to TF 5.0 beta 7, it was 
          often impossible to set [1m%keypad[22;0m=on because many (if not all) "linux" 
          termcap entries were missing a necessary code; TF now supplies that 
          code automatically if it is missing and [1m%TERM[22;0m is "linux".) 
  konsole and gnome-terminal 
          As far as I can tell, [1m%keypad[22;0m has no effect, NumLock chooses between 
          configuration A and normal digit/punctuation keys, and there is no 
          way to get configuration B.  
  PuTTY   [1m%keypad[22;0m=on and NumLock on gives configuration B above; [1m%keypad[22;0m=off 
          and NumLock on gives normal digit/punctuation keys; and NumLock off 
          gives a configuration similar to configuration A.  
  Mac OSX Terminal 
          By default, Terminal's keypad always acts like normal 
          digit/punctuation keys.  But if you turn on "strict vt100 keypad 
          behavior" under Terminal | Window Settings | Emulation, then 
          [1m%keypad[22;0m=on will give a configuration similar to configuration B.  
#

  In some environments, unnamed key sequences consisting of "^[" (ESC) 
  followed by one other character may also be typed by holding the META key 
  while typing the other character instead of typing ESC before the other 
  character.  See [1m%meta_esc[22;0m.  

  The one-time warning about certain new keybindings in 5.0 can be disabled by 
  setting the variable [1mwarn_5keys[22;0m=off.  
#mapping_named_keys

Mapping Named Keys to functions

  Named keys have two levels of mapping: first the character sequence 
  generated by the key is bound (with [1m/def -b[22;0m) to call a macro named 
  key_<[4mname[24m>; then the macro key_<[4mname[24m> is defined to execute a command.  If 
  you wish to change the functionality of any named key, you should do so by 
  redefining key_<[4mname[24m>.  For example, if you want Insert to invoke your own 
  macro /foo, you should redefine "/def key_insert = /foo".  You should only 
  make a direct keybinding if a key on your terminal generates a character 
  sequence not covered by TF's default bindings; and then you should only bind 
  the character sequence to call key_<[4mname[24m> (but first, see the "[1mkeypad[22;0m" 
  section above).  For example, if your Insert key generates "^[Q", you can 
  bind it with "[1m/def -b[22;0m'^[Q' = /key_insert".  You should never redefine any of 
  the predefined /dokey_* or /kb_* commands.  

  There are several advantages to this two-level mapping: redefining a key's 
  function is independent of the terminal; and adding keybindings for new 
  terminals is independent of the functions invoked by a named key.  

  Examples of popular alternatives to the standard key definitions: 

  Make PgUp and PgDn to scroll a half screen instead of a full screen: 

      /def key_pgdn = /dokey_hpage
      /def key_pgup = /dokey_hpageback

  Make up and down arrow keys perform movement only: 

      /def key_up = /dokey_up
      /def key_down = /dokey_down

  Make up and down arrow keys perform input recall only: 

      /def key_up = /dokey_recallb
      /def key_down = /dokey_recallf

  Before version 5.0, [1m/def -B[22;0m was the only way to bind a named key to a [1mmacro[22;0m. 
  This, however, has been superceded by the use of "key_<[4mname[24m>" macros.  
  Whereas [1m/def -B[22;0m depends strictly on termcap entries, the bindings to 
  "key_<[4mname[24m>" macros are automatically generated from TF's own list of 
  standard keybindings in addition to termcap entries.  Termcap entries are 
  often incomplete or not well matched to your terminal emulator; TF's 
  additional keybindings fill in the gaps.  So, to redefine the meaning of a 
  named key, you should redefine "[1m/def[22;0m key_<[4mname[24m> = ...", not "[1m/def[22;0m [1m-B[22;0m<[4mname[24m> = 
  ...".  The names recognized by [1m/def -B[22;0m are different than the names in the 
  Named Key table.  For reference, they are: the function keys "F0", "F1",...  
  "F19"; the keypad keys "KP1" (upper left), "KP2" (center), "KP3" (upper 
  right), "KP4" (lower left), "KP5" (lower right); the arrow keys "Up", 
  "Down", "Right", "Left"; and the other keys, "Backspace", "Clear EOL", 
  "Clear EOS", "Clear Screen", "Delete", "Delete Line", "Home", "Home Down", 
  "Insert", "Insert Line", "PgDn", "PgUp", "Scroll Down", "Scroll Up".  They 
  must be spelled as shown, but capitalization is ignored.  The function 
  [1mkeycode()[22;0m can be used to find the string generated by a key (as defined in 
  the termcap entry for [1m%TERM[22;0m).  
#mapping_char_seqs

Mapping character sequences to functions

  [1m/Def -b[22;0m (or [1m/bind[22;0m) allows you to bind a character sequence to a [1mmacro body[22;0m.  
  Typing that sequence at the keyboard (which may mean pressing a single key 
  that generates the sequence) will then execute the [1mmacro body[22;0m.  

  TF's input handler recognizes ^H and ^? as backspace and ^J and ^M as 
  newline, even when they are not bound to anything.  However, if a keybinding 
  is defined for any of these keys, it will override the internal handling of 
  that key.  

  At [1mstartup[22;0m, TF also examines the terminal driver settings for character 
  sequences corresponding to the [1m/dokey[22;0m functions BWORD, DLINE, REFRESH, and 
  LNEXT, and binds them accordingly in addition to the default bindings listed 
  above.  

Mapping character sequences to Named Keys

  Because [1mTF[22m runs in a terminal and not in a windowing system, it does not see 
  actual keystrokes, but only the characters generated by a keystroke.  For 
  example, the up arrow key on many terminals generates "^[[A", and that is 
  what [1mTF[22m receives.  Thus, [1mTF[22m uses a set of definitions like "[1m/def 
  -b[22;0m'<[4mcharsequence[24m>' = /key_<[4mname[24m>" to map chracter sequences to the keys that 
  generate them.  If two different keys generate the same sequence of 
  characters, there is no way for [1mTF[22m to tell them apart.  

  At startup, [1mTF[22m automatically binds character sequences to the named key 
  macros according to vt100, vt220, ANSI, and xterm definitions, plus OS/2 
  definitions if running on OS/2, as well as the termcap entry corresponding 
  to your [1m%TERM[22;0m variable.  If the named keys on your terminal generate 
  character sequences that are not recognized by TF, you will need to bind 
  them yourself with "[1m/def -b[22;0m'<[4mcharsequence[24m>' = /key_<[4mname[24m>".  For example, if 
  your terminal's PgUp key generates "^[[3~", TF will think you pressed 
  Delete, since that is the character sequence generated by Delete on most 
  terminals.  To tell TF about PgUp on your terminal, you should do "[1m/def 
  -b[22;0m'^[[3~' = /key_pgup".  

#Terminal
#terminal.app
#osx
#os x
#OS X Terminal
  Note for Mac OS X Terminal.app users: by default, Terminal.app traps PageUp 
  and PageDown keys itself and does not send them to the application (tf).  It 
  does however send Shift-PageUp and Shift-PageDown to the application, so you 
  can use these to scroll in tf running inside Terminal.  You can also tell 
  Terminal to send the unshifted keys to tf by redefining them in Terminal | 
  Window Settings | Keyboard.  

#teraterm
#niftytelnet
#broken emulators
  Note: some broken terminal emulators (TeraTerm, NiftyTelnet) send incorrect 
  character sequences for the editor keypad (insert, delete, home, end, pgup, 
  pgdn).  For TeraTerm users, the preferred fix is to copy 
  [1m%TFLIBDIR[22;0m/teraterm.keyboard.cnf to KEYBOARD.CNF in their TeraTerm directory; 
  this will help all applications you run within TeraTerm, not just TF.  Users 
  of either terminal emulator may work around the problem with "[1m/load[22;0m 
  kb_badterm.tf".  
#

  Note that before version 3.5 alpha 21 or beta 1, it was usually harmless to 
  "[1m/set[22;0m [1mTERM[22;0m=vt100" on terminals that accepted a superset of vt100 display 
  codes.  However, the termcap key definitions are often different for 
  terminals that are otherwise similar (e.g., vt100 and xterm share many 
  display codes, but the key definitions are different), so setting [1m%TERM[22;0m 
  incorrectly may interfere with the operation of named keys.  Xterm users 
  should also note that since 5.0, TF has its own [1mscrollback[22;0m, and xterm's 
  scrollback will not work properly even if you try to trick TF with 
  [1mTERM[22;0m=vt100.  

#kbnum

"Kbnum" argument

  With the default keybindings, ESC followed by an optional "-" and any number 
  of digits sets the global variable [1m%kbnum[22;0m.  By default, the current [1m%kbnum[22;0m 
  value is displayed near the right end of the [1mstatus line[22;0m.  Then, when any 
  other keybinding is typed, that keybinding may use the value of [1m%kbnum[22;0m.  
  Whether the keybinding uses the value or not, [1m%kbnum[22;0m is cleared after the 
  keybinding has run.  Most keybindings that use [1m%kbnum[22;0m use it as a repeat 
  count.  For example, typing "ESC 1 2 x" is the same as typing "x" 12 times.  
  For keybindings that have a sense of direction, negative values of [1m%kbnum[22;0m 
  reverse that direction: for example, typing "ESC - 4 PgDn" is like typing 
  "PgUp" 4 times.  The "^G" ([1m/beep[22;0m) keybinding does not honor [1m%kbnum[22;0m, so it 
  can be used to cancel [1m%kbnum[22;0m with no effect.  The variable [1m%max_kbnum[22;0m sets 
  an upper limit on the value of [1m%kbnum[22;0m that can be entered by the ESC and 
  digit keys, to prevent typos from sending TF into very long loops.  

  The interpretation of [1m%kbnum[22;0m must be done by the command called from the 
  keybinding; it is not done automatically by TF.  So, for [1m%kbnum[22;0m to be 
  meaningful in a macro you write, you must implement those semantics 
  yourself.  Additionally, most of the standard "/kb_*" and "[1m/dokey[22;0m" commands 
  that use [1m%kbnum[22;0m are optimized to not simply repeat the command a number of 
  times, but instead calculate only the end result.  For example, ESC 300 TAB 
  does not laboriously scroll 300 screenfuls of text onto the screen, but 
  figures out what the 300th screenful looks like and draws that immediately.  
  It does this because /dokey_page calls "[1m/test[22;0m [1mmorescroll[22;0m( [1mwinlines[22;0m() * 
  (kbnum[1m?:[22;0m1))".  

  To set [1m%kbnum[22;0m by means other than the default keybindings above, simply [1m/set[22;0m 
  it as you would any other variable.  Once it is set, all typed digits are 
  appended to it.  When any non-digit key is typed, that key will be executed, 
  and [1m%kbnum[22;0m will be cleared.  
#kbnum

Other key bindings

#kbstack.tf
#kb-bash.tf
#kb-emacs.tf
#kbregion.tf
#cut
#paste
#cut and paste
#bash
#emacs
#extra keybindings

  Some additional keyboard operations can be defined by [1m/load[22;0ming these library 
  files: 
  kb-old.tf 
          keybindings like those in TF 4.0 and earlier 
  kb-emacs.tf 
          additional emacs-like keybindings 
  kbregion.tf 
          cut-and-paste operations 
  kbstack.tf 
          save the current input line with ESC DOWN and restore it later with 
          ESC UP.  
  See the comments at the top of each file for further documentation.  
#
  ____________________________________________________________________________

  See also: [1m/dokey[22;0m, [1m/bind[22;0m, [1m/complete[22;0m, [1m%wordpunct[22;0m, [1msignals[22;0m.  

&stdlib.tf
&local.tf
&lib
&library
&library
&standard library

standard library

  When [1mTF[22;0m is started, commands are loaded from the standard library 
  ([1m%{TFLIBDIR}[22;0m/stdlib.tf).  If the installer has created an optional local 
  library ([1m%{TFLIBDIR}[22;0m/local.tf), that will also be loaded.  [1mMacros[22;0m defined in 
  the standard library are marked with the invisible option ("[1m-i[22;0m") so they 
  will not be processed by [1m/list[22;0m, [1m/save[22;0m and [1m/purge[22;0m unless forced.  Redefining 
  or undefining such a [1mmacro[22;0m will clear the [1m-i[22;0m option, so customized [1mmacros[22;0m 
  with the same names as library [1mmacros[22;0m can be created, listed, saved, and 
  purged.  

  See also: [1mutilities[22;0m 

#filename macros

Filenames:

  These [1mmacros[22;0m may be redefined to any filename.  LOGFILE contains the default 
  filename used by [1m/log[22;0m.  MACROFILE, HILITEFILE, GAGFILE, TRIGFILE, BINDFILE, 
  HOOKFILE, and WORLDFILE contain the default filenames used by the [1m/load[22;0m* and 
  [1m/save[22;0m* families of commands.  
#

#list*

List commands:

  [1m/listdef[22;0m <[4mspec[24m> 
          equivalent to '[1m/list[22;0m <[4mspec[24m>'.  
  [1m/listhilite[22;0m <[4mspec[24m> 
          lists [1mhilite[22;0ms on <[4mspec[24m>.  
  [1m/listgag[22;0m <[4mspec[24m> 
          lists [1mgag[22;0ms on <[4mspec[24m>.  
  [1m/listtrig[22;0m <[4mspec[24m> 
          lists [1mtriggers[22;0m on <[4mspec[24m>.  
  [1m/listbind[22;0m <[4mspec[24m> 
          lists key bindings matching <[4mspec[24m> 
  [1m/listhook[22;0m <[4mspec[24m> 
          lists [1mhooks[22;0m matching <[4mspec[24m>.  

  See: [1m/list[22;0m 

#purge*

Purge commands:

  [1m/purgedef[22;0m <[4mspec[24m> 
          purges [1mmacros[22;0m whose name matches <[4mspec[24m> 
  [1m/purgehilite[22;0m <[4mspec[24m> 
          purges [1mmacros[22;0m with [1mhilite[22;0ms on <[4mspec[24m> 
  [1m/purgegag[22;0m <[4mspec[24m> 
          purges [1mmacros[22;0m with [1mgag[22;0ms on <[4mspec[24m> 
  [1m/purgetrig[22;0m <[4mspec[24m> 
          purges [1mmacros[22;0m with [1mtriggers[22;0m on <[4mspec[24m> 
  [1m/purgedeft[22;0m <[4mspec[24m> 
          purges named [1mmacros[22;0m with [1mtriggers[22;0m on <[4mspec[24m> 
  [1m/purgebind[22;0m <[4mspec[24m> 
          purges key bindings matching <[4mspec[24m>.  
  [1m/purgehook[22;0m <[4mspec[24m> 
          purges [1mhooks[22;0m matching <[4mspec[24m>.  

  See: [1m/purge[22;0m 

#load*

Load commands:

  [1m/loaddef[22;0m, [1m/loadhilite[22;0m, [1m/loadgag[22;0m, [1m/loadtrig[22;0m, [1m/loadbind[22;0m, [1m/loadhook[22;0m, 
  [1m/loadworld[22;0m.  All take a <[4mfile[24m> argument; if the argument is omitted, the 
  appropriate default [1mfilename macro[22;0m is used.  

  See: [1m/load[22;0m 

#save*

Save commands:

  [1m/savedef[22;0m, [1m/savehilite[22;0m, [1m/savegag[22;0m, [1m/savetrig[22;0m, [1m/savebind[22;0m, [1m/savehook[22;0m, 
  [1m/saveworld[22;0m.  All take a <[4mfile[24m> argument.  If <[4mfile[24m> is omitted, the 
  appropriate default [1mfilename macro[22;0m is used.  

  See: [1m/save[22;0m 

#compress
#COMPRESS_READ
#$COMPRESS_READ
#COMPRESS_SUFFIX
#$COMPRESS_SUFFIX
#compression

File compression:

  The helpfile, personal config file, and files read with [1m/load[22;0m may be stored 
  compressed on disk.  If TF can not find a file with the specified name, it 
  will add ${COMPRESS_SUFFIX} to the filename and try to read it by piping it 
  through ${COMPRESS_READ}.  ${COMPRESS_READ} should contain the name of a 
  shell command that takes a filename as an argument, and prints its output on 
  standard output.  The default values for ${COMPRESS_SUFFIX} and 
  ${COMPRESS_READ} defined in the library are ".Z" and "zcat" for unix, ".zip" 
  and "unzip -p" for os/2.  Undefining ${COMPRESS_SUFFIX} will disable this 
  feature.  Note: [1m/save[22;0m, [1m/saveworld[22;0m, and [1m/log[22;0m do not write compressed files.  

#retry
#retry_off

World connection commands:

  [1m/retry[22;0m <[4mworld[24m> [<[4mdelay[24m>] 
          Try to connect to <[4mworld[24m>; repeat every <[4mdelay[24m> seconds until 
          successful.  
  [1m/retry_off[22;0m [<[4mworld[24m>] 
          Cancels "[1m/retry[22;0m <[4mworld[24m>" (default: all worlds) 

#hilite_whisper
#hilite whisper
#hilite_page
#hilite page

Hilite commands:

  [1m/hilite_whisper[22;0m, [1m/hilite_page[22;0m, [1m/nohilite_whisper[22;0m, and [1m/nohilite_page[22;0m turn on 
  or off [1mhiliting[22;0m several different page and whisper formats.  

#

Backward compatible commands:

  [1m/reply[22;0m, [1m/act[22;0m, [1m/nolog[22;0m, [1m/nologin[22;0m, [1m/nologme[22;0m, [1m/noquiet[22;0m, and [1m/nowrap[22;0m are provided 
  for compatibility.  

&8-bit
&8 bit
&characters
&character set
&iso-8859-1
&iso-8859
&iso 8859
&latin1
&international
&i18n
&internationalization
&internationalisation
&locale

locale

  On many systems, "[1m/setenv[22;0m LC_CTYPE=en_US" will allow you to use characters 
  in the 8-bit ISO 8859 character set.  If this does not work on your system, 
  or you want to use a non-English locale, or you just want to learn more, 
  keep reading.  

  A locale defines a set of rules for a language and culture.  If the platform 
  on which TF runs supports locales, TF will support the following categories 
  of locale rules: 
  LC_CTYPE 
          determines what characters are allowed, and whether they should be 
          treated as letters, digits, puctuation, or control characters.  When 
          using a locale with an 8-bit character set, make sure that [1m%istrip[22;0m 
          is off and [1m%meta_esc[22;0m is off or nonprint, so you can type 8-bit 
          characters with the meta key.  
  LC_TIME 
          determines the names and formats used in displaying dates and times 
          with [1m/time[22;0m, [1mftime()[22;0m, etc.  

  The user can set the locale either by having special variables defined in 
  the environment before starting TF (preferred), or by setting them while TF 
  is running (they will automatically be exported to the environment even if 
  [1m/set[22;0m is used).  The exact rules for setting locale depend on the platform, 
  and should be found your system's documentation for setlocale().  The rules 
  are usually something like this: 

    * If the variable [1mLC_ALL[22;0m is set, its value is used as the locale for 
      all supported categories.  
    * Otherwise, if the variable with the name of a category (e.g., 
      LC_CTYPE) is set, its value is used as the locale for that category.  
    * Otherwise, if the variable [1mLANG[22;0m is set, its value is used as the 
      locale for any supported categories that were not covered by the first 
      two rules.  
    * If none of those are set for a category, the default "C" locale is 
      used for that category, which allows the 7-bit ASCII character set and 
      US English date and time formats.  

  The valid values for the locale variables depend on your system.  On a POSIX 
  system, the valid values can be listed with the shell command "locale -a".  

  Bugs: 

    * LC_COLLATE and LC_MESSAGES categories are not supported.  
    * In glob [1mpatterns[22;0m, there is no way to specify a range of all letters 
      that works in all locales.  E.g., "[A-Za-z]" works in the standard "C" 
      locale, but not necessarily in others.  (However, in regexp [1mpatterns[22;0m, 
      locale information [1mis[22m used to define character type operators like "\w" 
      and "\W", case insensitivity, etc.) 
    * TF will convert character 0x80 to the character 0x00.  This is not 
      usually an issue, since character 0x80 is not a printable character in 
      the character sets of most locales (including all ISO character sets).  

  If your system has locale support, but does not have any locales installed, 
  you can get the POSIX 1003.2 WG15-collection locale definitions from 
  [1mftp://dkuug.dk/i18n/[22;0m or [1mftp://i44ftp.info.uni-karlsruhe.de/pub/linux/ctype/[22;0m. 

  Note that even though TF supports locales with non-ASCII character sets, not 
  all MUD servers support non-ASCII character sets.  Many servers simply 
  discard characters that are not printable ASCII.  Among servers that do 
  support non-ASCII characters, the most commonly used set is ISO-8859-1 
  (Latin1).  When choosing a locale for TF, you should choose one that uses 
  the same character set as the servers you use.  

  Note to linux users and other users of GNU libc: at least some versions of 
  GNU localedef generate invalid LC_TIME information from the WG15-collection 
  sources, and the GNU libc causes any program that tries to use the invalid 
  LC_TIME information to crash.  Workarounds: delete the LC_TIME data; or, do 
  not set any of the LC_ALL, LC_TIME, or LANG variables.  

&autologin
&login

login

  If the [1m%{login}[22;0m flag is on when you [1mconnect[22;0m to a world, and that world was 
  [1mdefined[22;0m with a character, password, and optional worldtype, TF will attempt 
  to automatically [1mlogin[22;0m to that world.  

  [1mAutologin[22;0m is done by a [1mhook[22;0m defined in the [1mstandard library[22;0m.  The [1mhook[22;0m for 
  the default worldtype uses TinyMUD [1mlogin[22;0m format; there are also [1mhooks[22;0m for 
  "tiny", "lp", "lpp", and "telnet" worldtypes.  You can also define your own 
  LOGIN [1mhooks[22;0m.  

  See: [1mhooks[22;0m, [1mvariables[22;0m, [1m/addworld[22;0m 

&macros

macros

  A [1mmacro[22;0m is basically a named set of commands.  The simplest kind of [1mmacro[22;0m 
  has a name and a body.  The body is a list of one or more commands, 
  separated by '%;' tokens.  These commands are executed when the [1mmacro[22;0m is 
  called.  For example, if you define a [1mmacro[22;0m like 

      [1m/def[22;0m time_warp = :jumps to the left!%;:steps to the right!

  and call it by typing 

      /time_warp

  you will execute the commands 

      :jumps to the left!
      :steps to the right!

  A [1mmacro[22;0m name is the way of calling it from the command line or from another 
  [1mmacro[22;0m.  You can execute a [1mmacro[22;0m by typing '/' followed by the name of the 
  [1mmacro[22;0m.  If a [1mmacro[22;0m and builtin have the same name, the [1mmacro[22;0m will be called. 
  Typing '/@' followed by the name will always call the builtin command.  

  A [1mmacro[22;0m body, or execution text, is the commands and/or text executed when 
  the [1mmacro[22;0m is called.  This text is evaluated according to the rules 
  described under "[1mevaluation[22;0m".  

  [1mMacros[22;0m actually have many more fields, described below.  All fields 
  (including name and body) are optional.  

  name    The name of the [1mmacro[22;0m.  Names should begin with a letter, and 
          contain letters, numbers, or '_' characters.  

  body    One or more commands to be executed when [1mmacro[22;0m is called.  The body 
          is compiled to an efficient internal format the first time it is 
          needed, so each future call can execute it more quickly.  

  number  All [1mmacros[22;0m are automatically numbered sequentially.  This field can 
          not be changed.  

  trigger 
          when text matches the [1mtrigger[22;0m pattern, the [1mmacro[22;0m may be called.  

  hook    the [1mmacro[22;0m can be called when a TF [1mhook[22;0m event occurs.  

  keybinding 
          the [1mmacro[22;0m will be called when its keybinding is typed.  

  shots   the [1mmacro[22;0m will be deleted after it is [1mtrigger[22;0med or [1mhook[22;0med a certain 
          number of times.  

  [1mpriority[22;0m 
          when multiple [1mtriggers[22;0m match the same text, the one with the highest 
          [1mpriority[22;0m is selected (see "[1mpriority[22;0m").  

  [1mfall-thru[22;0m 
          on a [1mtrigger[22;0m or [1mhook[22;0m, allows additional [1mmacros[22;0m of lower [1mpriority[22;0m to 
          be run (see "[1mpriority[22;0m").  

  world   the [1mmacro[22;0m can only be [1mtrigger[22;0med/[1mhook[22;0med by text/events from a 
          particular world.  

  worldtype 
          the [1mmacro[22;0m can only be [1mtrigger[22;0med/hooked by text/events from a 
          particular type of world.  

  expression 
          the [1mmacro[22;0m can only be [1mtrigger[22;0med/[1mhook[22;0med if [1mexpression[22;0m is non-zero.  

  attributes 
          bold, underline, etc.  for displaying [1mtrigger[22;0m text.  

  probability 
          when [1mtrigger[22;0med, the [1mmacro[22;0m has a certain probability of being 
          executed.  

  invisibility 
          prevents handling of [1mmacro[22;0m by [1m/list[22;0m, [1m/save[22;0m, or [1m/purge[22;0m.  
  [1mMacros[22;0m may be called in several ways: 

    * a command of the form "/[4mname[24m" or "/#[4mnumber[24m" 
    * triggered by text from a [1msocket[22;0m (see "[1mtriggers[22;0m") 
    * hooked by a tinyfugue event (see "[1mhooks[22;0m") 
    * by keybindings 

  Associated commands: 
  [1m/def[22;0m    define a named [1mmacro[22;0m, with any fields 
  [1m/trig[22;0m   define a [1mtrigger[22;0m [1mmacro[22;0m 
  [1m/hilite[22;0m 
          define a [1mhilite[22;0m [1mmacro[22;0m 
  [1m/gag[22;0m    define a [1mgag[22;0m [1mmacro[22;0m 
  [1m/bind[22;0m   define a keybinding [1mmacro[22;0m 
  [1m/hook[22;0m   define a [1mhook[22;0m [1mmacro[22;0m 
  [1m/undef[22;0m  undefine a named [1mmacro[22;0m 
  [1m/unhook[22;0m 
          undefine a [1mhook[22;0m [1mmacro[22;0m 
  [1m/unbind[22;0m 
          undefine a keybinding [1mmacro[22;0m 
  [1m/undefn[22;0m 
          undefine a [1mmacro[22;0m by number 
  [1m/undeft[22;0m 
          undefine a [1mmacro[22;0m by [1mtrigger[22;0m 
  [1m/purge[22;0m  undefine a set of [1mmacros[22;0m 
  [1m/list[22;0m   display a list of [1mmacros[22;0m 
  [1m/load[22;0m   load commands from a file 
  [1m/save[22;0m   save [1mmacro[22;0m definitions to a file 

  See also: [1mtriggers[22;0m, [1mgags[22;0m, [1mhilites[22;0m, [1mhooks[22;0m 

&mail
&mail check
&mail checking

mail checking

  If [1m%{maildelay}[22;0m is nonzero, TF will check for mail every [1m%{maildelay}[22;0m 
  seconds.  TF checks for mail in each file in the space-separated list of 
  files in the [1m%{TFMAILPATH}[22;0m [1mvariable[22;0m (literal spaces in TFMAILPATH may be 
  quoted by preceeding them with a backslash).  If [1m%{TFMAILPATH}[22;0m is not set, 
  TF will check in the single file named by the [1m%{MAIL}[22;0m [1mvariable[22;0m.  

  TF considers a mailfile to have unread mail if the file has been written 
  more recently than it has been read.  When this changes for any of the 
  monitored files, TF updates the mail indicator on the [1mstatus line[22;0m (actually, 
  the "@mail" [1mstatus[22;0m).  When TF determines that a mailfile contains new mail, 
  it calls the [1mMAIL hook[22;0m, which by default prints "You have new mail".  If a 
  mailfile is not empty the [4mfirst[24m time TF checks it, TF just prints "You have 
  mail" without calling the [1mMAIL hook[22;0m.  

  If an error occurs while checking any file, an error message will be 
  displayed only once, until that error clears up (or changes to a different 
  error), but TF will continue to check that file.  To disable checking, even 
  after an error, you must remove the file from [1m%{TFMAILPATH}[22;0m or [1m%{MAIL}[22;0m.  

  The [1mnmail()[22;0m [1mfunction[22;0m returns the number of monitored mail files containing 
  unread mail.  

  MAIL and/or MAILPATH variables are usually set in the environment before [1mTF[22;0m 
  starts.  If [1m%{MAIL}[22;0m is not set when TF starts, TF will try to set it to the 
  name of the system mail directory plus your user name (if the system mail 
  directory was defined when TF was installed).  If MAILPATH (which uses ":" 
  as a delimiter) is set when TF starts, it is transferred to [1m%{TFMAILPATH}[22;0m 
  (which uses space as a delimiter).  

  See: [1mnmail()[22;0m, [1mvariables[22;0m, [1mspecial variables[22;0m, [1m/set[22;0m, [1mmailing list[22;0m.  

&majordomo
&listserv
&mail list
&mailing list

mailing list

  The TinyFugue mailing list is an email forum for discussion of topics 
  related to TinyFugue.  To subscribe, follow the instructions at 
  [1mhttp://tinyfugue.sourceforge.net/[22;0m 

&mccpv1
&mccpv2
&mccp

Mud Client Compression Protocol

  TF supports versions 1 and 2 of the Mud Client Compression Protocol (MCCP) 
  described at [1mhttp://www.randomly.org/projects/MCCP/[22;0m.  MCCP allows a server 
  to compress the data stream it sends to the client (TF), which may improve 
  throughput on a poor connection.  

  MCCP is transparent to the user.  When TF connects to a server that supports 
  MCCP, it will be enabled automatically, unless the [1mmccp[22;0m variable is off.  
  The [1mlistsockets[22;0m command will indicate that MCCP is enabled.  

  MCCP v1 is broken, and may not be supported in the future if it is found to 
  interfere with valid protocols.  If you use a server that has only MCCP v1, 
  you should encourage the owner to upgrade to add support for v2.  

  See also: [1mprotocols[22;0m, [1mtelnet[22;0m 

&visual
&visual mode
&nonvisual
&non-visual
&screen
&mode

mode

  TinyFugue has two main interface modes: Visual and non-visual.  Visual mode 
  will be enabled by default, unless your [1m%{TERM}[22;0m does not support it, or 
  [1m%{visual}[22;0m is explicitly turned off in [1m.tfrc[22;0m, or [1mtf[22;0m is started with the -v 
  option.  Visual mode can be turned off or on with the "[1m/visual[22;0m" command.  

#visual

Visual mode

  The Visual interface has two windows: the bottom window is for input, the 
  top for output.  TF maintains a separate [1mvirtual window[22;0m for each open 
  [1msocket[22;0m; only the [1mforeground[22;0m world's window is displayed.  If your terminal 
  can scroll in a region, output will [1mscroll[22;0m; otherwise if your terminal can 
  delete and insert lines, TF will simulate [1mscrolling[22;0m; otherwise it will wrap 
  from bottom to top, clearing two lines ahead.  The [1m%{scroll}[22;0m [1mvariable[22;0m can be 
  set to explicitly choose [1mscrolling[22;0m or wrapping.  The [1m%{isize}[22;0m, [1m%{cleardone}[22;0m, 
  and [1m%{clearfull}[22;0m [1mvariables[22;0m can be used to customize the visual display.  
  See: [1m%isize[22;0m, [1m%cleardone[22;0m, [1m%clearfull[22;0m.  

  The two windows are separated by a [1mstatus line[22;0m, which can be formatted by 
  the user as described under [1mstatus line[22;0m.  

  If you are using a terminal emulator that emulates different terminal types, 
  the recommended type to use is vt220, vt100, or ansi (in that order), with 
  [1m%{TERM}[22;0m set to the same value.  Scrolling may appear jumpy under ansi, but 
  will be smooth under vt220 and vt100.  vt220 also provides some additional 
  features that may make command line editing smoother (especially over a slow 
  modem).  

#nonvisual

Non-visual mode

  In the non-visual interface, input and output are both displayed on the 
  bottom line.  If you are typing and output appears, your input is cleared, 
  the output is displayed and everything above it scrolls, and your input is 
  redisplayed on the last line.  If your input has wrapped around to a second 
  or third line, only the last line will be cleared and redisplayed.  

#
  ____________________________________________________________________________

  In both modes, the output window is redrawn whenever necessary: when its 
  size changes, when the mode changes, when [1m%wrap[22;0m, [1m%wrapsize[22;0m, or [1m%wrapspace[22;0m 
  change, or when TF resumes after [1m/suspend[22;0m or [1m/sh[22;0m.  

  In both modes, output text is wrapped around at a right margin of one less 
  than the number of columns on your screen (typically 79) unless [1mwrapping[22;0m has 
  been turned off.  In addition, when text is wrapped, all wrapped lines after 
  the first will be indented 4 spaces to help distinguish them from the 
  beginning of an original line (configurable by setting [1m%wrapspace[22;0m).  See: 
  [1mcolumns()[22;0m, [1m%wrap[22;0m, [1m%wrapsize[22;0m, [1m%wrappunct[22;0m, [1m%wrapspace[22;0m.  

  If the [1m%{more}[22;0m flag is on, output is suspended when the screen is full, and 
  you can use the TAB key to continue.  See: [1m/more[22;0m, [1m/dokey[22;0m.  

&-
&--
&options

options

  Many commands take options to modify their behavior, following these rules 
  (similar to UNIX conventions, but not identical): 

    * All options must be immediately preceded by '-'.  
    * Options may be grouped after a single '-'.  
    * Some options may take string, numeric, or time arguments.  There 
      must be no space between the option and the argument.  
    * String option-arguments may be delimited by a space, double quotes, 
      single quotes, or backquotes.  
    * A literal delimiter character or '\' within a delimited string must 
      be escaped by preceding it with '\'.  
    * A numeric option-argument may be given as an [1mexpression[22;0m that 
      evaluates to a numeric value.  If the expression contains spaces or 
      quotes, they must be quoted or escaped as in a string option-argument.  
    * All options must precede normal arguments.  
    * A '-' or '--' by itself may be used to mark the end of the options.  
      This is useful when the first regular argument begins with '-'.  
    * A '-?' or invalid option will produce a list of valid options.  

  See also: [1mgetopts()[22;0m.  

&patterns

patterns

  Patterns are used throughout TF, including [1mtriggers[22;0m, [1mhooks[22;0m, [1m/purge[22;0m, [1m/list[22;0m, 
  [1m/limit[22;0m, [1m/recall[22;0m, and [1mexpressions[22;0m.  There are four styles of pattern matching 
  available: 
  simple  target string and pattern string must be identical 
  glob    similar to shell filename patterns 
  regexp  perl-compatible regular expressions 
  substr  target string must contain pattern string 
  The style used by a particular command is determined either by the use of 
  the -m option or the setting of the global [1mvariable[22;0m [1m%{matching}[22;0m.  

#comparison
#simple
#simple matching

Simple matching ("simple")

  The pattern is compared directly to the string.  There are no special 
  characters.  Case is significant.  

#substr
#contain

Substring matching ("substr")

  The string must contain the pattern.  There are no special characters.  Case 
  is significant.  

#smatch
#globbing
#glob

Globbing ("glob")

  Globbing is the default matching style, and was the only style available 
  before version 3.2.  It is similar to filename expansion ("globbing") used 
  by many shells (but unlike shells, tf uses glob only for comparison, not 
  expansion).  

  There are several special sequences that can be used in tf globbing: 

    * The '*' character matches any number of characters.  

    * The '?' character matches any one character.  

    * Square brackets ([...]) can be used to match any one of a sequence 
      of characters.  Ranges can be specified by giving the first and last 
      characters with a '-' between them.  If '^' is the first character, the 
      sequence will match any character NOT specified.  

    * Curly braces ({...}) can be used to match any one of a list of 
      words.  Different words can be matched by listing each within the 
      braces, separated by a '|' (or) character.  Both ends of {...} will only 
      match a space or end of string.  Therefore "{foo}*" and "{foo}p" do not 
      match "foop", and "*{foo}" and "p{foo}" do not match "pfoo".  

      Patterns containing "{...}" can easily be meaningless.  A valid {...} 
      pattern must: (a) contain no spaces, (b) follow a wildcard, space, or 
      beginning of string, (c) be followed by a wildcard, space, or end of 
      string.  

      The pattern "{}" will match the empty string.  

    * Any other character will match itself, ignoring case.  A special 
      character can be made to match itself by preceding it with '\' to remove 
      its special meaning.  

  Examples:
  "d?g" matches "dog", "dig" and "dug" but not "dg" or "drug". 
  "d*g" matches "dg", "dog", "drug", "debug", "dead slug", etc. 
  "{d*g}" matches "dg", "dog", "drug", "debug", but not "dead slug". 
  "M[rs]." matches "Mr." and "Ms."
  "M[a-z]" matches "Ma", "Mb", "Mc", etc. 
  "[^a-z]" matches any character that is not in the English alphabet. 
  "{storm|chup*}*" matches "chupchup fehs" and "Storm jiggles". 
  "{storm|chup*}*" does NOT match "stormette jiggles". 

#re
#regex
#regexp
#regexps
#regular expressions

Regular expressions ("regexp")

  TF implements regular expressions with the package PCRE 2.08, Copyright (c) 
  1997-1999 University of Cambridge.  The PCRE regexp syntax is documented on 
  its own page under the topic "[1mpcre[22;0m".  

  The syntax and semantics of these regular expressions is nearly identical to 
  those in perl 5, and is roughly a superset of those used in versions of tf 
  prior to 5.0.  There is one incompatability with old tf regexps: the "{" 
  character is now special, and must be written "\{" to match a literal "{".  
  To help with the transition to the new syntax, you will be warned if you use 
  a regexp containing "{", unless you turn off the [1mwarn_curly_re[22;0m variable.  

  If all letters in a regexp are lower case, the regexp will default to using 
  caseless matching.  If a regexp contains any upper case letters, it will 
  default to case-sensitive matching.  Of course, you can explicitly specify 
  caseless matching by including "(?i)" at the beginning of the regexp, or 
  case-sensitive by including "(?-i)".  

  Regexps will honor the [1mlocale[22;0m that was set when the regexp was defined.  
  [1mLocale[22;0m affects caseless matching, and determines whether characters are 
  letters, digits, or whatever.  So, for example, while the regexp "[A-Za-z]" 
  will match only English letters, "[^\W\d_]" will match any letter defined by 
  the [1mlocale[22;0m.  

  After a regexp match, [1m%Pn[22;0m substitutions can be used to get the value of the 
  string that matched various parts of the regexp.  See [1m%Pn[22;0m.  

  For those of you who care about code details: TF compiles PCRE regexps with 
  the PCRE_DOLLAR_ENDONLY and PCRE_DOTALL options.  

  See also: [1mregmatch()[22;0m, [1msubstitution[22;0m.  

Comparison of glob and regexps. 

  In a glob, '*' and '?' by themselves match text.  In a regexp, '*' and '?' 
  are only meaningful in combination with the pattern they follow.  Regexps 
  are not "anchored"; that is, the match may occur anywhere in the string, 
  unless you explicitly use '^' and/or '$' to anchor it.  Globs are anchored, 
  and must match the entire string.  

      regexp              equivalent glob
      ------              -----------------
      "part of line"      "*part of line*"
      "^entire line$"     "entire line"
      "\bword\b"          "*{word}*"
      "^(you|hawkeye) "   "{you|hawkeye} *"
      "foo.*bar"          "*foo*bar*"
      "f(oo|00)d"         "*{*food*|*f00d*}*"
      "line\d"            "*line[0-9]*"
      "^[^ ]+ whispers,"  "{*} whispers,*"
      "foo(xy)?bar"       "*{*foobar*|*fooxybar*}*"
      "zoo+m"             none
      "foo ?bar"          none
      "(foo bar|frodo)"   none

Notes. 

    * For best performance, make the beginning of your patterns as 
      specific as possible.  
    * Do not use ".*" or "^.*" at the beginning of a regexp.  It is [4mvery[24m 
      inefficient, and not needed.  Use [1m%PL[22;0m instead if you need to retrieve 
      the substring to the left of the match.  
    * If a glob and regexp can do the same job, the glob is usually 
      [4mslightly[24m faster.  But if using a glob instead of a regexp would mean you 
      need some extra code, then that extra code will cost much more than the 
      regexp would have.  So if only a regexp can do what you need, don't 
      hesitate to use it.  

&pcre
&pcre syntax
  This document was extracted from the pcre.3.html documentation, Copyright 
  (c) 1997-1999 University of Cambridge, and minimally adapted for use in 
  TinyFugue.  

    * 
#SEC13
      REGULAR EXPRESSION DETAILS 

      The syntax and semantics of the regular expressions supported by PCRE 
      are described below.  Regular expressions are also described in the Perl 
      documentation and in a number of other books, some of which have copious 
      examples.  Jeffrey Friedl's "Mastering Regular Expressions", published 
      by O'Reilly (ISBN 1-56592-257-3), covers them in great detail.  The 
      description here is intended as reference documentation.  

      A regular expression is a pattern that is matched against a subject 
      string from left to right.  Most characters stand for themselves in a 
      pattern, and match the corresponding characters in the subject.  As a 
      trivial example, the pattern 


        The quick brown fox

      matches a portion of a subject string that is identical to itself.  The 
      power of regular expressions comes from the ability to include 
      alternatives and repetitions in the pattern.  These are encoded in the 
      pattern by the use of <[4mmeta-characters[24m>, which do not stand for 
      themselves but instead are interpreted in some special way.  

      There are two different sets of meta-characters: those that are 
      recognized anywhere in the pattern except within square brackets, and 
      those that are recognized in square brackets.  Outside square brackets, 
      the meta-characters are as follows: 


        \      general escape character with several uses
        ^      assert start of subject (or line, in multiline mode)
        $      assert end of subject (or line, in multiline mode)
        .      match any character except newline (by default)
        [      start character class definition
        |      start of alternative branch
        (      start subpattern
        )      end subpattern
        ?      extends the meaning of (
               also 0 or 1 quantifier
               also quantifier minimizer
        *      0 or more quantifier
        +      1 or more quantifier
        {      start min/max quantifier

      Part of a pattern that is in square brackets is called a "character 
      class".  In a character class the only meta-characters are: 


        \      general escape character
        ^      negate the class, but only if the first character
        -      indicates character range
        ]      terminates the character class

      The following sections describe the use of each of the meta-characters.  
    * 
#SEC14
      BACKSLASH 

      The backslash character has several uses.  Firstly, if it is followed by 
      a non-alphameric character, it takes away any special meaning that 
      character may have.  This use of backslash as an escape character 
      applies both inside and outside character classes.  

      For example, if you want to match a "*" character, you write "\*" in the 
      pattern.  This applies whether or not the following character would 
      otherwise be interpreted as a meta-character, so it is always safe to 
      precede a non-alphameric with "\" to specify that it stands for itself.  
      In particular, if you want to match a backslash, you write "\\".  

      If a pattern is compiled with the "x" (PCRE_EXTRA) option, whitespace in 
      the pattern (other than in a character class) and characters between a 
      "#" outside a character class and the next newline character are 
      ignored.  An escaping backslash can be used to include a whitespace or 
      "#" character as part of the pattern.  

      A second use of backslash provides a way of encoding non-printing 
      characters in patterns in a visible manner.  There is no restriction on 
      the appearance of non-printing characters, apart from the binary zero 
      that terminates a pattern, but when a pattern is being prepared by text 
      editing, it is usually easier to use one of the following escape 
      sequences than the binary character it represents: 


        \a     alarm, that is, the BEL character (hex 07)
        \cx    "control-x", where x is any character
        \e     escape (hex 1B)
        \f     formfeed (hex 0C)
        \n     newline (hex 0A)
        \r     carriage return (hex 0D)
        \t     tab (hex 09)
        \xhh   character with hex code hh
        \ddd   character with octal code ddd, or backreference

      The precise effect of "\cx" is as follows: if "x" is a lower case 
      letter, it is converted to upper case.  Then bit 6 of the character (hex 
      40) is inverted.  Thus "\cz" becomes hex 1A, but "\c{" becomes hex 3B, 
      while "\c;" becomes hex 7B.  

      After "\x", up to two hexadecimal digits are read (letters can be in 
      upper or lower case).  

      After "\0" up to two further octal digits are read.  In both cases, if 
      there are fewer than two digits, just those that are present are used.  
      Thus the sequence "\0\x\07" specifies two binary zeros followed by a BEL 
      character.  Make sure you supply two digits after the initial zero if 
      the character that follows is itself an octal digit.  

      The handling of a backslash followed by a digit other than 0 is 
      complicated.  Outside a character class, PCRE reads it and any following 
      digits as a decimal number.  If the number is less than 10, or if there 
      have been at least that many previous capturing left parentheses in the 
      expression, the entire sequence is taken as a <[4mback reference[24m>.  A 
      description of how this works is given later, following the discussion 
      of parenthesized subpatterns.  

      Inside a character class, or if the decimal number is greater than 9 and 
      there have not been that many capturing subpatterns, PCRE re-reads up to 
      three octal digits following the backslash, and generates a single byte 
      from the least significant 8 bits of the value.  Any subsequent digits 
      stand for themselves.  For example: 


        \040   is another way of writing a space
        \40    is the same, provided there are fewer than 40
                  previous capturing subpatterns
        \7     is always a back reference
        \11    might be a back reference, or another way of
                  writing a tab
        \011   is always a tab
        \0113  is a tab followed by the character "3"
        \113   is the character with octal code 113 (since there
                  can be no more than 99 back references)
        \377   is a byte consisting entirely of 1 bits
        \81    is either a back reference, or a binary zero
                  followed by the two characters "8" and "1"

      Note that octal values of 100 or greater must not be introduced by a 
      leading zero, because no more than three octal digits are ever read.  

      All the sequences that define a single byte value can be used both 
      inside and outside character classes.  In addition, inside a character 
      class, the sequence "\b" is interpreted as the backspace character (hex 
      08).  Outside a character class it has a different meaning (see below).  

      The third use of backslash is for specifying generic character types: 


        \d     any decimal digit
        \D     any character that is not a decimal digit
        \s     any whitespace character
        \S     any character that is not a whitespace character
        \w     any "word" character
        \W     any "non-word" character

      Each pair of escape sequences partitions the complete set of characters 
      into two disjoint sets.  Any given character matches one, and only one, 
      of each pair.  

      A "word" character is any letter or digit or the underscore character, 
      that is, any character which can be part of a Perl "word".  The 
      definition of letters and digits is controlled by PCRE's character 
      tables, and may vary if locale- specific matching is taking place (see 
      "Locale support" above).  For example, in the "fr" (French) locale, some 
      character codes greater than 128 are used for accented letters, and 
      these are matched by \w.  

      These character type sequences can appear both inside and outside 
      character classes.  They each match one character of the appropriate 
      type.  If the current matching point is at the end of the subject 
      string, all of them fail, since there is no character to match.  

      The fourth use of backslash is for certain simple assertions.  An 
      assertion specifies a condition that has to be met at a particular point 
      in a match, without consuming any characters from the subject string.  
      The use of subpatterns for more complicated assertions is described 
      below.  The backslashed assertions are 


        \b     word boundary
        \B     not a word boundary
        \A     start of subject (same as "^" in tf)
        \Z     end of subject (same as "$" in tf)
        \z     end of subject (same as "$" in tf)

      These assertions may not appear in character classes (but note that "\b" 
      has a different meaning, namely the backspace character, inside a 
      character class).  

      A word boundary is a position in the subject string where the current 
      character and the previous character do not both match \w or \W (i.e.  
      one matches \w and the other matches \W), or the start or end of the 
      string if the first or last character matches \w, respectively.  
    * 
#SEC15
      CIRCUMFLEX AND DOLLAR 

      Outside a character class, in the default matching mode, the circumflex 
      character is an assertion which is true only if the current matching 
      point is at the start of the subject string.  Inside a character class, 
      circumflex has an entirely different meaning (see below).  

      Circumflex need not be the first character of the pattern if a number of 
      alternatives are involved, but it should be the first thing in each 
      alternative in which it appears if the pattern is ever to match that 
      branch.  If all possible alternatives start with a circumflex, that is, 
      if the pattern is constrained to match only at the start of the subject, 
      it is said to be an "anchored" pattern.  (There are also other 
      constructs that can cause a pattern to be anchored.) 

      A dollar character is an assertion which is true only if the current 
      matching point is at the end of the subject string.  Dollar need not be 
      the last character of the pattern if a number of alternatives are 
      involved, but it should be the last item in any branch in which it 
      appears.  Dollar has no special meaning in a character class.  
    * 
#SEC17
      SQUARE BRACKETS 

      An opening square bracket introduces a character class, terminated by a 
      closing square bracket.  A closing square bracket on its own is not 
      special.  If a closing square bracket is required as a member of the 
      class, it should be the first data character in the class (after an 
      initial circumflex, if present) or escaped with a backslash.  

      A character class matches a single character in the subject; the 
      character must be in the set of characters defined by the class, unless 
      the first character in the class is a circumflex, in which case the 
      subject character must not be in the set defined by the class.  If a 
      circumflex is actually required as a member of the class, ensure it is 
      not the first character, or escape it with a backslash.  

      For example, the character class [aeiou] matches any lower case vowel, 
      while [^aeiou] matches any character that is not a lower case vowel.  
      Note that a circumflex is just a convenient notation for specifying the 
      characters which are in the class by enumerating those that are not.  It 
      is not an assertion: it still consumes a character from the subject 
      string, and fails if the current pointer is at the end of the string.  

      When caseless matching is set, any letters in a class represent both 
      their upper case and lower case versions, so for example, a caseless 
      [aeiou] matches "A" as well as "a", and a caseless [^aeiou] does not 
      match "A", whereas a caseful version would.  

      The minus (hyphen) character can be used to specify a range of 
      characters in a character class.  For example, [d-m] matches any letter 
      between d and m, inclusive.  If a minus character is required in a 
      class, it must be escaped with a backslash or appear in a position where 
      it cannot be interpreted as indicating a range, typically as the first 
      or last character in the class.  

      It is not possible to have the literal character "]" as the end 
      character of a range.  A pattern such as [W-]46] is interpreted as a 
      class of two characters ("W" and "-") followed by a literal string 
      "46]", so it would match "W46]" or "-46]".  However, if the "]" is 
      escaped with a backslash it is interpreted as the end of range, so 
      [W-\]46] is interpreted as a single class containing a range followed by 
      two separate characters.  The octal or hexadecimal representation of "]" 
      can also be used to end a range.  

      Ranges operate in ASCII collating sequence.  They can also be used for 
      characters specified numerically, for example [\000-\037].  If a range 
      that includes letters is used when caseless matching is set, it matches 
      the letters in either case.  For example, [W-c] is equivalent to 
      [][\^_`wxyzabc], matched caselessly, and if character tables for the 
      "fr" locale are in use, [\xc8-\xcb] matches accented E characters in 
      both cases.  

      The character types \d, \D, \s, \S, \w, and \W may also appear in a 
      character class, and add the characters that they match to the class.  
      For example, [\dABCDEF] matches any hexadecimal digit.  A circumflex can 
      conveniently be used with the upper case character types to specify a 
      more restricted set of characters than the matching lower case type.  
      For example, the class [^\W_] matches any letter or digit, but not 
      underscore.  

      All non-alphameric characters other than \, -, ^ (at the start) and the 
      terminating ] are non-special in character classes, but it does no harm 
      if they are escaped.  
    * 
#SEC18
      VERTICAL BAR 

      Vertical bar characters are used to separate alternative patterns.  For 
      example, the pattern 


        gilbert|sullivan

      matches either "gilbert" or "sullivan".  Any number of alternatives may 
      appear, and an empty alternative is permitted (matching the empty 
      string).  The matching process tries each alternative in turn, from left 
      to right, and the first one that succeeds is used.  If the alternatives 
      are within a subpattern (defined below), "succeeds" means matching the 
      rest of the main pattern as well as the alternative in the subpattern.  
    * 
#SEC19
      INTERNAL OPTION SETTING 

      The settings of PCRE_CASELESS, PCRE_EXTENDED, and PCRE_UNGREEDY can be 
      changed from within the pattern by a sequence of Perl option letters 
      enclosed between "(?" and ")".  The option letters are 


        i  for PCRE_CASELESS
        x  for PCRE_EXTENDED
        U  for PCRE_UNGREEDY (not in perl)

      For example, (?x) sets extended matching.  It is also possible to unset 
      these options by preceding the letter with a hyphen, and a combined 
      setting and unsetting such as (?x-i), which sets extended and while 
      unsetting caseless, is also permitted.  If a letter appears both before 
      and after the hyphen, the option is unset.  

      The scope of these option changes depends on where in the pattern the 
      setting occurs.  For settings that are outside any subpattern (defined 
      below), the effect is the same as if the options were set or unset at 
      the start of matching.  The following patterns all behave in exactly the 
      same way: 


        (?i)ABC
        A(?i)BC
        AB(?i)C
        ABC(?i)

      Such "top level" settings apply to the whole pattern (unless there are 
      other changes inside subpatterns).  If there is more than one setting of 
      the same option at top level, the rightmost setting is used.  

      If an option change occurs inside a subpattern, the effect is different. 
      This is a change of behaviour in Perl 5.005.  An option change inside a 
      subpattern affects only that part of the subpattern that follows it, so 


        (a(?-i)b)c

      matches abc, Abc, abC and AbC, and no other strings (remember, in tf, 
      regexps are caseless by default if they do not contain any capital 
      letters).  By this means, options can be made to have different settings 
      in different parts of the pattern.  Any changes made in one alternative 
      do carry on into subsequent branches within the same subpattern.  For 
      example, 


        X(a(?i)b|c)

      matches "Xab", "XaB", "Xc", and "XC", even though when matching "C" the 
      first branch is abandoned before the option setting.  This is because 
      the effects of option settings happen at compile time.  There would be 
      some very weird behaviour otherwise.  
    * 
#SEC20
      SUBPATTERNS 

      Subpatterns are delimited by parentheses (round brackets), which can be 
      nested.  Marking part of a pattern as a subpattern does two things: 

      1.  It localizes a set of alternatives.  For example, the pattern 


        cat(aract|erpillar|)

      matches one of the words "cat", "cataract", or "caterpillar".  Without 
      the parentheses, it would match "cataract", "erpillar" or the empty 
      string.  

      2.  It sets up the subpattern as a capturing subpattern (as defined 
      above).  When the whole pattern matches, that portion of the subject 
      string that matched the subpattern is remembered for the TinyFugue %Pn 
      substitutions.  Opening parentheses are counted from left to right 
      (starting from 1) to obtain the numbers of the capturing subpatterns.  

      For example, if the string "the red king" is matched against the pattern 


        the ((red|white) (king|queen))

      the captured substrings are "red king", "red", and "king", and are 
      numbered 1, 2, and 3.  

      The fact that plain parentheses fulfil two functions is not always 
      helpful.  There are often times when a grouping subpattern is required 
      without a capturing requirement.  If an opening parenthesis is followed 
      by "?:", the subpattern does not do any capturing, and is not counted 
      when computing the number of any subsequent capturing subpatterns.  For 
      example, if the string "the white queen" is matched against the pattern 


        the ((?:red|white) (king|queen))

      the captured substrings are "white queen" and "queen", and are numbered 
      1 and 2.  The maximum number of captured substrings is 99, and the 
      maximum number of all subpatterns, both capturing and non-capturing, is 
      200.  

      As a convenient shorthand, if any option settings are required at the 
      start of a non-capturing subpattern, the option letters may appear 
      between the "?" and the ":".  Thus the two patterns 


        (?i:saturday|sunday)
        (?:(?i)saturday|sunday)

      match exactly the same set of strings.  Because alternative branches are 
      tried from left to right, and options are not reset until the end of the 
      subpattern is reached, an option setting in one branch does affect 
      subsequent branches, so the above patterns match "SUNDAY" as well as 
      "Saturday".  
    * 
#SEC21
      REPETITION 

      Repetition is specified by quantifiers, which can follow any of the 
      following items: 


        a single character, possibly escaped
        the . metacharacter
        a character class
        a back reference (see next section)
        a parenthesized subpattern (unless it is an assertion - see below)

      The general repetition quantifier specifies a minimum and maximum number 
      of permitted matches, by giving the two numbers in curly brackets 
      (braces), separated by a comma.  The numbers must be less than 65536, 
      and the first must be less than or equal to the second.  For example: 


        z{2,4}

      matches "zz", "zzz", or "zzzz".  A closing brace on its own is not a 
      special character.  If the second number is omitted, but the comma is 
      present, there is no upper limit; if the second number and the comma are 
      both omitted, the quantifier specifies an exact number of required 
      matches.  Thus 


        [aeiou]{3,}

      matches at least 3 successive vowels, but may match many more, while 


        \d{8}

      matches exactly 8 digits.  An opening curly bracket that appears in a 
      position where a quantifier is not allowed, or one that does not match 
      the syntax of a quantifier, is taken as a literal character.  For 
      example, {,6} is not a quantifier, but a literal string of four 
      characters.  

      The quantifier {0} is permitted, causing the expression to behave as if 
      the previous item and the quantifier were not present.  

      For convenience (and historical compatibility) the three most common 
      quantifiers have single-character abbreviations: 


        *    is equivalent to {0,}
        +    is equivalent to {1,}
        ?    is equivalent to {0,1}

      It is possible to construct infinite loops by following a subpattern 
      that can match no characters with a quantifier that has no upper limit, 
      for example: 


        (a?)*

      Earlier versions of Perl and PCRE used to give an error at compile time 
      for such patterns.  However, because there are cases where this can be 
      useful, such patterns are now accepted, but if any repetition of the 
      subpattern does in fact match no characters, the loop is forcibly 
      broken.  

      By default, the quantifiers are "greedy", that is, they match as much as 
      possible (up to the maximum number of permitted times), without causing 
      the rest of the pattern to fail.  The classic example of where this 
      gives problems is in trying to match comments in C programs.  These 
      appear between the sequences /* and */ and within the sequence, 
      individual * and / characters may appear.  An attempt to match C 
      comments by applying the pattern 


        /\*.*\*/

      to the string 


        /* first command */  not comment  /* second comment */

      fails, because it matches the entire string due to the greediness of the 
      .* item.  

      However, if a quantifier is followed by a question mark, then it ceases 
      to be greedy, and instead matches the minimum number of times possible, 
      so the pattern 


        /\*.*?\*/

      does the right thing with the C comments.  The meaning of the various 
      quantifiers is not otherwise changed, just the preferred number of 
      matches.  Do not confuse this use of question mark with its use as a 
      quantifier in its own right.  Because it has two uses, it can sometimes 
      appear doubled, as in 


        \d??\d

      which matches one digit by preference, but can match two if that is the 
      only way the rest of the pattern matches.  

      If the "U" (PCRE_UNGREEDY) option is set (an option which is not 
      available in Perl) then the quantifiers are not greedy by default, but 
      individual ones can be made greedy by following them with a question 
      mark.  In other words, it inverts the default behaviour.  

      When a parenthesized subpattern is quantified with a minimum repeat 
      count that is greater than 1 or with a limited maximum, more store is 
      required for the compiled pattern, in proportion to the size of the 
      minimum or maximum.  

      If a pattern starts with .* or .{0,}, then the pattern is implicitly 
      anchored, because whatever follows will be tried against every character 
      position in the subject string, so there is no point in retrying the 
      overall match at any position after the first.  PCRE treats such a 
      pattern as though it were preceded by \A.  

      When a capturing subpattern is repeated, the value captured is the 
      substring that matched the final iteration.  For example, after 


        (tweedle[dume]{3}\s*)+

      has matched "tweedledum tweedledee" the value of the captured substring 
      is "tweedledee".  However, if there are nested capturing subpatterns, 
      the corresponding captured values may have been set in previous 
      iterations.  For example, after 


        /(a|(b))+/

      matches "aba" the value of the second captured substring is "b".  
    * 
#SEC22
      BACK REFERENCES 

      Outside a character class, a backslash followed by a digit greater than 
      0 (and possibly further digits) is a back reference to a capturing 
      subpattern earlier (i.e.  to its left) in the pattern, provided there 
      have been that many previous capturing left parentheses.  

      However, if the decimal number following the backslash is less than 10, 
      it is always taken as a back reference, and causes an error only if 
      there are not that many capturing left parentheses in the entire 
      pattern.  In other words, the parentheses that are referenced need not 
      be to the left of the reference for numbers less than 10.  See the 
      section entitled "Backslash" above for further details of the handling 
      of digits following a backslash.  

      A back reference matches whatever actually matched the capturing 
      subpattern in the current subject string, rather than anything matching 
      the subpattern itself.  So the pattern 


        (sens|respons)e and \1ibility

      matches "sense and sensibility" and "response and responsibility", but 
      not "sense and responsibility".  If caseful matching is in force at the 
      time of the back reference, then the case of letters is relevant.  For 
      example, 


        ((?i)rah)\s+\1

      matches "rah rah" and "RAH RAH", but not "RAH rah", even though the 
      original capturing subpattern is matched caselessly.  

      There may be more than one back reference to the same subpattern.  If a 
      subpattern has not actually been used in a particular match, then any 
      back references to it always fail.  For example, the pattern 


        (a|(bc))\2

      always fails if it starts to match "a" rather than "bc".  Because there 
      may be up to 99 back references, all digits following the backslash are 
      taken as part of a potential back reference number.  If the pattern 
      continues with a digit character, then some delimiter must be used to 
      terminate the back reference.  If the "x" (PCRE_EXTENDED) option is set, 
      this can be whitespace.  Otherwise an empty comment can be used.  

      A back reference that occurs inside the parentheses to which it refers 
      fails when the subpattern is first used, so, for example, (a\1) never 
      matches.  However, such references can be useful inside repeated 
      subpatterns.  For example, the pattern 


        (a|b\1)+

      matches any number of "a"s and also "aba", "ababaa" etc.  At each 
      iteration of the subpattern, the back reference matches the character 
      string corresponding to the previous iteration.  In order for this to 
      work, the pattern must be such that the first iteration does not need to 
      match the back reference.  This can be done using alternation, as in the 
      example above, or by a quantifier with a minimum of zero.  
    * 
#SEC23
      ASSERTIONS 

      An assertion is a test on the characters following or preceding the 
      current matching point that does not actually consume any characters.  
      The simple assertions coded as \b, \B, \A, \Z, \z, ^ and $ are described 
      above.  More complicated assertions are coded as subpatterns.  There are 
      two kinds: those that look ahead of the current position in the subject 
      string, and those that look behind it.  

      An assertion subpattern is matched in the normal way, except that it 
      does not cause the current matching position to be changed.  Lookahead 
      assertions start with (?= for positive assertions and (?! for negative 
      assertions.  For example, 


        \w+(?=;)

      matches a word followed by a semicolon, but does not include the 
      semicolon in the match, and 


        foo(?!bar)

      matches any occurrence of "foo" that is not followed by "bar".  Note 
      that the apparently similar pattern 


        (?!foo)bar

      does not find an occurrence of "bar" that is preceded by something other 
      than "foo"; it finds any occurrence of "bar" whatsoever, because the 
      assertion (?!foo) is always true when the next three characters are 
      "bar".  A lookbehind assertion is needed to achieve this effect.  

      Lookbehind assertions start with (?<= for positive assertions and (?<! 
      for negative assertions.  For example, 


        (?<!foo)bar

      does find an occurrence of "bar" that is not preceded by "foo".  The 
      contents of a lookbehind assertion are restricted such that all the 
      strings it matches must have a fixed length.  However, if there are 
      several alternatives, they do not all have to have the same fixed 
      length.  Thus 


        (?<=bullock|donkey)

      is permitted, but 


        (?<!dogs?|cats?)

      causes an error at compile time.  Branches that match different length 
      strings are permitted only at the top level of a lookbehind assertion.  
      This is an extension compared with Perl 5.005, which requires all 
      branches to match the same length of string.  An assertion such as 


        (?<=ab(c|de))

      is not permitted, because its single top-level branch can match two 
      different lengths, but it is acceptable if rewritten to use two 
      top-level branches: 


        (?<=abc|abde)

      The implementation of lookbehind assertions is, for each alternative, to 
      temporarily move the current position back by the fixed width and then 
      try to match.  If there are insufficient characters before the current 
      position, the match is deemed to fail.  Lookbehinds in conjunction with 
      once-only subpatterns can be particularly useful for matching at the 
      ends of strings; an example is given at the end of the section on 
      once-only subpatterns.  

      Several assertions (of any sort) may occur in succession.  For example, 


        (?<=\d{3})(?<!999)foo

      matches "foo" preceded by three digits that are not "999".  Notice that 
      each of the assertions is applied independently at the same point in the 
      subject string.  First there is a check that the previous three 
      characters are all digits, then there is a check that the same three 
      characters are not "999".  This pattern does <[4mnot[24m> match "foo" preceded 
      by six characters, the first of which are digits and the last three of 
      which are not "999".  For example, it doesn't match "123abcfoo".  A 
      pattern to do that is 


        (?<=\d{3}...)(?<!999)foo

      This time the first assertion looks at the preceding six characters, 
      checking that the first three are digits, and then the second assertion 
      checks that the preceding three characters are not "999".  

      Assertions can be nested in any combination.  For example, 


        (?<=(?<!foo)bar)baz

      matches an occurrence of "baz" that is preceded by "bar" which in turn 
      is not preceded by "foo", while 


        (?<=\d{3}(?!999)...)foo

      is another pattern which matches "foo" preceded by three digits and any 
      three characters that are not "999".  

      Assertion subpatterns are not capturing subpatterns, and may not be 
      repeated, because it makes no sense to assert the same thing several 
      times.  If any kind of assertion contains capturing subpatterns within 
      it, these are counted for the purposes of numbering the capturing 
      subpatterns in the whole pattern.  However, substring capturing is 
      carried out only for positive assertions, because it does not make sense 
      for negative assertions.  

      Assertions count towards the maximum of 200 parenthesized subpatterns.  
    * 
#SEC24
      ONCE-ONLY SUBPATTERNS 

      With both maximizing and minimizing repetition, failure of what follows 
      normally causes the repeated item to be re-evaluated to see if a 
      different number of repeats allows the rest of the pattern to match.  
      Sometimes it is useful to prevent this, either to change the nature of 
      the match, or to cause it fail earlier than it otherwise might, when the 
      author of the pattern knows there is no point in carrying on.  

      Consider, for example, the pattern \d+foo when applied to the subject 
      line 


        123456bar

      After matching all 6 digits and then failing to match "foo", the normal 
      action of the matcher is to try again with only 5 digits matching the 
      \d+ item, and then with 4, and so on, before ultimately failing.  
      Once-only subpatterns provide the means for specifying that once a 
      portion of the pattern has matched, it is not to be re-evaluated in this 
      way, so the matcher would give up immediately on failing to match "foo" 
      the first time.  The notation is another kind of special parenthesis, 
      starting with (?> as in this example: 


        (?>\d+)bar

      This kind of parenthesis "locks up" the part of the pattern it contains 
      once it has matched, and a failure further into the pattern is prevented 
      from backtracking into it.  Backtracking past it to previous items, 
      however, works as normal.  

      An alternative description is that a subpattern of this type matches the 
      string of characters that an identical standalone pattern would match, 
      if anchored at the current point in the subject string.  

      Once-only subpatterns are not capturing subpatterns.  Simple cases such 
      as the above example can be thought of as a maximizing repeat that must 
      swallow everything it can.  So, while both \d+ and \d+? are prepared to 
      adjust the number of digits they match in order to make the rest of the 
      pattern match, (?>\d+) can only match an entire sequence of digits.  

      This construction can of course contain arbitrarily complicated 
      subpatterns, and it can be nested.  

      Once-only subpatterns can be used in conjunction with lookbehind 
      assertions to specify efficient matching at the end of the subject 
      string.  Consider a simple pattern such as 


        abcd$

      when applied to a long string which does not match it.  Because matching 
      proceeds from left to right, PCRE will look for each "a" in the subject 
      and then see if what follows matches the rest of the pattern.  If the 
      pattern is specified as 


        ^.*abcd$

      then the initial .* matches the entire string at first, but when this 
      fails, it backtracks to match all but the last character, then all but 
      the last two characters, and so on.  Once again the search for "a" 
      covers the entire string, from right to left, so we are no better off.  
      However, if the pattern is written as 


        ^(?>.*)(?<=abcd)

      then there can be no backtracking for the .* item; it can match only the 
      entire string.  The subsequent lookbehind assertion does a single test 
      on the last four characters.  If it fails, the match fails immediately.  
      For long strings, this approach makes a significant difference to the 
      processing time.  
    * 
#SEC25
      CONDITIONAL SUBPATTERNS 

      It is possible to cause the matching process to obey a subpattern 
      conditionally or to choose between two alternative subpatterns, 
      depending on the result of an assertion, or whether a previous capturing 
      subpattern matched or not.  The two possible forms of conditional 
      subpattern are 


        (?(condition)yes-pattern)
        (?(condition)yes-pattern|no-pattern)

      If the condition is satisfied, the yes-pattern is used; otherwise the 
      no-pattern (if present) is used.  If there are more than two 
      alternatives in the subpattern, a compile-time error occurs.  

      There are two kinds of condition.  If the text between the parentheses 
      consists of a sequence of digits, then the condition is satisfied if the 
      capturing subpattern of that number has previously matched.  Consider 
      the following pattern, which contains non-significant white space to 
      make it more readable (assume the "x" PCRE_EXTENDED option) and to 
      divide it into three parts for ease of discussion: 


        ( \( )?    [^()]+    (?(1) \) )

      The first part matches an optional opening parenthesis, and if that 
      character is present, sets it as the first captured substring.  The 
      second part matches one or more characters that are not parentheses.  
      The third part is a conditional subpattern that tests whether the first 
      set of parentheses matched or not.  If they did, that is, if subject 
      started with an opening parenthesis, the condition is true, and so the 
      yes-pattern is executed and a closing parenthesis is required.  
      Otherwise, since no-pattern is not present, the subpattern matches 
      nothing.  In other words, this pattern matches a sequence of 
      non-parentheses, optionally enclosed in parentheses.  

      If the condition is not a sequence of digits, it must be an assertion.  
      This may be a positive or negative lookahead or lookbehind assertion.  
      Consider this pattern, again containing non-significant white space, and 
      with the two alternatives on the second line: 


        (?(?=[^a-z]*[a-z])
        \d{2}[a-z]{3}-\d{2}  |  \d{2}-\d{2}-\d{2} )

      The condition is a positive lookahead assertion that matches an optional 
      sequence of non-letters followed by a letter.  In other words, it tests 
      for the presence of at least one letter in the subject.  If a letter is 
      found, the subject is matched against the first alternative; otherwise 
      it is matched against the second.  This pattern matches strings in one 
      of the two forms dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are 
      digits.  
    * 
#SEC26
      COMMENTS 

      The sequence (?# marks the start of a comment which continues up to the 
      next closing parenthesis.  Nested parentheses are not permitted.  The 
      characters that make up a comment play no part in the pattern matching 
      at all.  

      If the "x" PCRE_EXTENDED option is set, an unescaped # character outside 
      a character class introduces a comment that continues up to the next 
      newline character in the pattern.  
    * 
#SEC27
      PERFORMANCE 

      Certain items that may appear in patterns are more efficient than 
      others.  It is more efficient to use a character class like [aeiou] than 
      a set of alternatives such as (a|e|i|o|u).  In general, the simplest 
      construction that provides the required behaviour is usually the most 
      efficient.  Jeffrey Friedl's book contains a lot of discussion about 
      optimizing regular expressions for efficient performance.  

      Beware of patterns that contain nested indefinite repeats.  These can 
      take a long time to run when applied to a string that does not match.  
      Consider the pattern fragment 


        (a+)*

      This can match "aaaa" in 33 different ways, and this number increases 
      very rapidly as the string gets longer.  (The * repeat can match 0, 1, 
      2, 3, or 4 times, and for each of those cases other than 0, the + 
      repeats can match different numbers of times.) When the remainder of the 
      pattern is such that the entire match is going to fail, PCRE has in 
      principle to try every possible variation, and this can take an 
      extremely long time.  

      An optimization catches some of the more simple cases such as 


        (a+)*b

      where a literal character follows.  Before embarking on the standard 
      matching procedure, PCRE checks that there is a "b" later in the subject 
      string, and if there is not, it fails the match immediately.  However, 
      when there is no following literal this optimization cannot be used.  
      You can see the difference by comparing the behaviour of 


        (a+)*\d

      with the pattern above.  The former gives a failure almost instantly 
      when applied to a whole line of "a" characters, whereas the latter takes 
      an appreciable time with strings longer than about 20 characters.  
    * 
#SEC28
      AUTHOR 

      Philip Hazel <ph10@cam.ac.uk> 
      University Computing Service, 
      New Museums Site, 
      Cambridge CB2 3QG, England.  
      Phone: +44 1223 334714 

      Last updated: 29 July 1999 
      Copyright (c) 1997-1999 University of Cambridge.  

&priorities
&fallthru
&fall-thru
&selection
&priority
&priority rules

priority

  When more than one [1mmacro[22;0m is matched by a [1mtrigger[22;0m or [1mhook[22;0med event, the 
  following rules are used to select which of the [1mmacros[22;0m will be applied 
  (i.e., have its attributes applied to the text, and its body executed): 

    * [1mMacros[22;0m are compared in order of decreasing [1mpriority[22;0m.  
    * [1mFall-thrus[22;0m of a given [1mpriority[22;0m are compared before [1mnon-fall-thrus[22;0m of 
      the same [1mpriority[22;0m.  
    * Each matching [1mfall-thru[22;0m [1mmacro[22;0m is applied immediately when it is 
      found.  
    * When the first matching [1mnon-fall-thru[22;0m [1mmacro[22;0m is found, all the 
      [1mnon-fall-thrus[22;0m of equal [1mpriority[22;0m are collected, and the search ends.  
      One of the [1mnon-fall-thrus[22;0m is chosen at random and applied.  

  So, in the simple case when there are no [1mfall-thrus[22;0m, the highest [1mpriority[22;0m 
  match is chosen.  If there is more than one of the highest [1mpriority[22;0m, one of 
  those is chosen at random.  

  These [1mpriority[22;0m rules apply even to [1mmacros[22;0m defined or undefined by a [1mmacro[22;0m 
  found during the search.  For example, if a mud line triggers a [1mfall-thru[22;0m 
  [1mmacro[22;0m /foo, and /foo defines a new [1mtrigger[22;0m [1mmacro[22;0m /bar which also matches the 
  line, then /bar may be triggered if it has lower [1mpriority[22;0m than /foo.  

  A [1mmacro's[22;0m [1mpriority[22;0m is set with [1m/def[22;0m [1m-p[22;0m; its [1mfall-thru[22;0m option is set with 
  [1m/def[22;0m [1m-F[22;0m.  

  Use the [1m/trigger[22;0m -n command to display a list of the triggers [1mtriggers[22;0m or 
  [1mhooks[22;0m will match a given string.  

  See: [1mtriggers[22;0m, [1mhooks[22;0m, [1mmacros[22;0m, [1m/def[22;0m 

&bug
&bugs
&core
&crash
&error
&report
&hawkeye
&kkeys
&kenkeys
&author
&support
&problems

problems

  If you have an old version of TF, chances are your bug has already been 
  fixed.  Current information and the latest version of TF can be found at 
  [1mhttp://tinyfugue.sourceforge.net/[22;0m.  

  For general bug reports, questions, etc, visit the website above 
  (preferred), or email kenkeys@users.sourceforge.net.  For problems specific 
  to the OS/2 version, contact Andreas Sahlbach at asa@stardiv.de.  When 
  reporting a problem or bug, please provide this information: 

    * The full version number of TF (type "[1m/version[22;0m" in tf).  Please give 
      the [4mfull[24m number, don't just say something like "beta 4" or "the latest 
      version".  
    * The operating system name and version.  (On unix systems, type 
      "uname -a" in the shell to get the exact version information.) 
    * If tf won't install, send the output of the installation process (on 
      UNIX, that's the output of configure and make).  Don't leave out parts 
      just because you don't know what they mean or think they're irrelevant.  
    * If you have a bug or core: do NOT send the core file, but do send 
      the debugging dump file (tf.<[4mNNNNN[24m>.dump) if tf generated one.  If not, 
      give me ALL messages from tf (not just the last line).  In either case, 
      tell me what you did or what happened before the problem, and whether 
      the problem is repeatable.  
    * Optional: If you have a core, you know how to use a debugger, tf was 
      compiled with core dumps enabled, and tf did not generate a debugging 
      dump file, a manual stack trace would be useful (use the 'bt full' 
      command in gdb or 'where' in dbx).  If you don't know how, at least 
      provide the other information described above.  

#
  The following bugs are known.  Don't bother reporting them.  

    * The [1m%{lp}[22;0m and [1m%{emulation}[22;0m [1mvariables[22;0m should work on a per-[1msocket[22;0m 
      basis (This is partially overcome with WORLD [1mhooks[22;0m).  
    * If a shell quote ([1m/quote[22;0m !) reads a partial line from the child 
      process, tf will hang until the line is completed.  
    * [1m/recall[22;0m by timestamp doesn't work when switching to/from daylight 
      savings time (but [1m/recall[22;0m by age always works).  

&tinyprocesses
&process
&proc
&processes

processes

  Associated topics: 

  [1m/quote[22;0m
  [1m/repeat[22;0m
  [1m/ps[22;0m
  [1m/kill[22;0m
  [1m%ptime[22;0m
  [1m%lpquote[22;0m

  The [1m/quote[22;0m and [1m/repeat[22;0m commands in Fugue are done by setting up internal 
  [1mprocesses[22;0m that run concurrently with normal input and output.  

  [1m/ps[22;0m can be used to get a listing of the currently running [1mprocesses[22;0m and 
  their process ID's (for use with [1m/kill[22;0m).  

  [1m/kill[22;0m can be used to terminate a [1mprocess[22;0m.  

  [1mProcesses[22;0m can be either synchronous or asynchronous.  Synchronous [1mprocesses[22;0m 
  (started with the -S option) run immediately when they are started, and run 
  to completion (unless TF is [1minterrupted[22;0m) before any other commands are 
  executed.  Synchronous [1mprocesses[22;0m are new in version 3.3 beta 10.  

  Asynchronous [1mprocesses[22;0m are merely scheduled to be run by a [1m/quote[22;0m or [1m/repeat[22;0m 
  command; the actual execution occurs at some later time.  They can be run 
  based on two different criteria: 

  1.  Normally, [1mprocesses[22;0m run whenever a specific period of time has elapsed.  
  The delay can be specified when the [1mprocess[22;0m is started, or will default to 
  the value of [1m%{ptime}[22;0m.  

  2.  If the [1m%{lpquote}[22;0m flag is on or the [1mprocess[22;0m was started with the -P 
  option, a [1mprocess[22;0m run whenever a [1mprompt[22;0m is received from the server, 
  indicating that the previous command has completed.  If the process was 
  started with a -w option, only prompts from the specified world will trigger 
  its execution.  Example: 

          [1m/quote[22;0m -P [1m/send[22;0m `/_echo n%; /_echo w%; /_echo w%; /_echo s

  will send the commands "n", "w", "w", and "s", waiting between each one 
  until the [1mprompt[22;0m following the previous command is seen.  

  If an asynchronous [1m/quote[22;0m or [1m/repeat[22;0m is followed immediately by another 
  command, the other command will run first, because the asynchronous [1mprocess[22;0m 
  was only scheduled, not actually executed (even with -n or -0 options).  Use 
  a synchronous [1m/quote[22;0m or [1m/repeat[22;0m to force the [1mprocess[22;0m to run before any other 
  commands.  

  Bodies of [1m/repeat[22;0m undergo [1mmacro[22;0m body [1mexpansion[22;0m when they are executed; text 
  generated by [1m/quote[22;0m does not.  

  See also: [1mutilities[22;0m ([1m/at[22;0m, [1m/tick[22;0m) 

&goahead
&eor
&end-of-record
&prompt protocol

prompt protocol

  TF will recognize the TELNET protocol commands GOAHEAD or END-OF-RECORD as 
  the end of a [1mprompt[22;0m.  If you are responsible for a server that has [1mprompts[22;0m, 
  and wish to make it more friendly to TF users, choose one of these options: 

  GOAHEAD: Send IAC GA (\377 \371) after each [1mprompt[22;0m.  This is the easier of 
  the two options.  In many servers, this can be done at the beginning of the 
  routine that reads user input.  Disadvantage: could possibly cause problems 
  in clients that don't understand TELNET protocol (but usually, they will 
  just pass it through to the terminal, which will usually ignore it).  

  END-OF-RECORD: Send IAC WILL EOR (\377 \373 \031) when the user connects.  
  If the client responds with IAC DO EOR, then you can send IAC END-OF-RECORD 
  (\377 \357) after each [1mprompt[22;0m; otherwise, do nothing special in [1mprompts[22;0m.  
  Disadvantage: requires extra state per descriptor and more understanding of 
  telnet protocol.  Advantage: minimizes potential problems for clients that 
  do not recognize telnet protocol.  

  To debug telnet option negotiations, you may find it useful to "[1m/set[22;0m [1mtelopt[22;0m 
  on" in TF.  

  For more information on TELNET protocol, see RFCs 854, 855, 885, and 1123.  

  See also: [1m/telnet[22;0m, [1mtelopt[22;0m, [1mprompts[22;0m, [1mprotocols[22;0m 

&lp
&diku
&prompt
&prompts

prompts

  Most LP muds, Diku muds, telnetd, and some other types of servers send 
  unterminated [1mprompts[22;0m, that is, [1mprompts[22;0m that do not end with newline or any 
  other special character.  Normally, TF will not display text until a newline 
  is received, so you may not see the [1mprompt[22;0m until after you press return.  
  But if the [1m%{lp}[22;0m flag is on, TF will attempt to separate these [1mprompts[22;0m from 
  normal text and display them correctly.  

  The recommended way to use the [1m%{lp}[22;0m flag is to define your worlds with one 
  of the [1m/addlp[22;0m, [1m/adddiku[22;0m, or [1m/addtelnet[22;0m commands.  The [1m%{lp}[22;0m flag will be 
  turned on automatically when you switch to such a world, and turned off for 
  the other predefined world types.  See: [1m/addworld[22;0m.  

  TF also provides a PROMPT [1mhook[22;0m, which allows you to tell it what to look for 
  in a [1mprompt[22;0m.  When an unterminated line is received, the PROMPT [1mhook[22;0m is 
  called immediately.  If there is no match, TF will use the timeout method 
  described below (if [1m%{lp}[22;0m is on).  But if there is a matching PROMPT [1mhook[22;0m, 
  TF will forget about the line (unless the hook was defined with [1m/def[22;0m [1m-q[22;0m) and 
  let the [1mhook[22;0m deal with it.  By combining the PROMPT [1mhook[22;0m with the [1m/prompt[22;0m 
  command, you can recognize most [1mprompts[22;0m immediately without having to use 
  the [1m%{lp}[22;0m timing mechanism.  The typical way of doing this is: 

      [1m/def[22;0m -h"PROMPT *> " catch_prompt = [1m/test[22;0m [1mprompt[22;0m([1m{*}[22;0m)

  So, whenever TF receives an unterminated line that ends in "> ", 
  catch_prompt will see it, and use [1m/prompt[22;0m to copy it to the current [1mprompt[22;0m.  

  If an unterminated line is not matched by any PROMPT [1mhook[22;0m, and it is not 
  followed by more text within a short period of time, TF will assume it is a 
  [1mprompt[22;0m.  This method is not foolproof.  If the delay is too short, broken 
  lines will look like [1mprompts[22;0m, and will briefly appear in the input window 
  until the rest of the line arrives, at which time both parts of the line 
  will be printed as normal output.  If the delay is too long, there will be 
  an annoying delay before displaying real [1mprompts[22;0m.  

  The delay can be varied by setting the [1mvariable[22;0m [1mprompt_wait[22;0m.  Its default 
  value is 0.25 seconds.  

  All of this hackery can be avoided if the server sends unambiguous [1mprompts[22;0m.  
  TF will recognize "*\b" (that is, "*" followed by backspace) and anything 
  ending with [1mGOAHEAD[22;0m or [1mEND-OF-RECORD[22;0m telnet characters.  When TF sees such 
  text, it does not wait for a delay, but calls the PROMPT hook immediately; 
  if there is no match, TF displays the prompt immediately.  To avoid some 
  minor glitches, you should leave the [1m%{lp}[22;0m flag off when connected to such a 
  server.  If you are responsible for a server and wish to make it more 
  TF-friendly, see "[1mprompt protocol[22;0m".  

  See also: [1m%login[22;0m, [1mprompt protocol[22;0m, [1m/addworld[22;0m 

&protocol
&ip
&ipv4
&ipv6
&ssl
&rfc
&rfcs
&protocols

Protocols

  TF supports the following protocols: 

    * TCP over IPv4 (RFC 791) 
    * TCP over IPv6 (RFC 2460, 3493), if supported by the host 
    * TELNET Protocol (RFC 854, 855) (See: [1mtelnet[22;0m) 
    * Generic proxy servers (See: [1mproxy[22;0m) 
    * ANSI display attributes (See: [1m%emulation[22;0m) 
    * EOR and GOAHEAD prompt protocol (See: [1mprompt protocol[22;0m) 
    * Mud Client Compression Protocol version 2, if TF was compiled with 
      zlib (See: [1mmccp[22;0m) 
    * Secure Socket Layer (SSL), if TF was compiled with libssl.  (See: 
      [1maddworld[22;0m, [1mconnect[22;0m) 

  RFCs can be obtained from 

    * [1mhttp://www.rfc-editor.org/rfc.html[22;0m 
    * [1mhttp://info.internet.isi.edu/1/in-notes/rfc[22;0m 
    * [1mhttp://www.garlic.com/~lynn/rfcietf.html[22;0m 
    * [1mhttp://www.cis.ohio-state.edu/hypertext/information/rfc.html[22;0m 
    * [1mftp://wuarchive.wustl.edu/doc/rfc/[22;0m 
    * [1mftp://nis.nsf.net/documents/rfc/[22;0m 
    * [1mftp://src.doc.ic.ac.uk/rfc/[22;0m 

  and other sites.  

&firewall
&proxy
&/proxy_connect
&/proxy_command
&proxy server

proxy server

  If [1m%{proxy_host}[22;0m is set, all connections will go through a proxy server 
  (firewall) defined by [1m%proxy_host[22;0m and [1m%proxy_port[22;0m.  Note that [1m%{proxy_host}[22;0m 
  should usually [4mnot[24m be set if TF has been compiled to use SOCKS.  

  When the connection to [1m%proxy_host[22;0m [1m%proxy_port[22;0m is made, only the PROXY [1mhook[22;0m 
  is called; the CONNECT and LOGIN [1mhooks[22;0m which are normally called after 
  making a connection are not called when a proxy is used.  A PROXY [1mhook[22;0m 
  defined in the standard library calls [1m/proxy_command[22;0m, which by default sends 
  "telnet [1m${world_host}[22;0m [1m${world_port}[22;0m", and then invoke the CONNECT and LOGIN 
  [1mhooks[22;0m (which, by default, bring the [1mworld[22;0m into the [1mforeground[22;0m and perform an 
  [1mautomatic login[22;0m).  

  Before the connection to the proxy server is made, [1m${world_host}[22;0m, 
  [1m${world_port}[22;0m, error messages, and [1m/listsockets[22;0m all refer to the proxy 
  server; after the connection is made, they refer to the target server 
  defined in [1m/addworld[22;0m.  

  The proxy connection command is done with this standard [1mmacro[22;0m: 

  [1m/def[22;0m -i proxy_connect = telnet [1m${world_host}[22;0m [1m${world_port}[22;0m

  If your proxy server requires a different command, you should redefine 
  proxy_connect.  That will be sufficient for most proxy servers.  (Before 
  version 5.0, a custom connect command required you to redefine 
  proxy_command.  This should be avoided now if possible.) 

  If your proxy server has more complex requirements, or you want better error 
  detection, you will need to redefine the proxy_command [1mmacro[22;0m.  By default, 
  proxy_command immediately calls /proxy_connect, enables [1mlocalecho[22;0m, and 
  invokes the CONNECT and LOGIN [1mhooks[22;0m.  There are several reasons you might 
  want to redefine proxy_command: 

    * The default proxy_command can not detect when proxy_connect fails, 
      so it will always send your login command even if the proxy server did 
      not connect to the target server.  
    * Your proxy server may not accept commands immediately, so 
      proxy_command should wait for some indication that the proxy server is 
      ready before sending commands.  

  For example, say you use a Gauntlet telnet proxy that leaves [1mlocalecho[22;0m off; 
  prints a "tn-gw->" [1mprompt[22;0m; requires you to send "telnet <[4mhostname[24m> <[4mport[24m>" 
  to connect; after a successful connection, prints "Connected to <[4mhostname[24m>"; 
  and after a failed connection prints an error message and prints another 
  [1mprompt[22;0m.  So, you could use this definition: 

  [1m/def[22;0m proxy_command =\
      [1m/def[22;0m -p10000 -w -1 -h'PROMPT tn-gw->' =\
          /proxy_connect%%; \
          [1m/localecho[22;0m on%%; \
          [1m/def[22;0m -p10002 -w -1 -h'PROMPT tn-gw->' proxy_error_[1m$${world_name}[22;0m =\
              [1m/undef[22;0m proxy_success_[1m$$${world_name}[22;0m%%%;\
              [1m/dc[22;0m%%;\
          [1m/def[22;0m -p10002 -w -1 -t'Connected to *' proxy_success_[1m$${world_name}[22;0m =\
              [1m/undef[22;0m proxy_error_[1m$$${world_name}[22;0m%%%;\
              [1m/trigger[22;0m -hCONNECT [1m$$${world_name}[22;0m%%%;\
              [1m/if[22;0m ([1m$$${world_character}[22;0m !~ "" & [1m$$${world_login}[22;0m) \
                  [1m/trigger[22;0m -hLOGIN [1m${world_name}[22;0m%%%;\
              [1m/endif[22;0m

  The first [1m/def[22;0m waits for the first [1mprompt[22;0m before doing anything.  It then 
  sends the connection command, turns [1mlocalecho[22;0m back on, and sets up [1mmacros[22;0m to 
  catch the results of the connection command.  The success [1mtrigger[22;0m undefines 
  the error [1mhook[22;0m, and invokes the CONNECT and LOGIN [1mhooks[22;0m.  The error [1mhook[22;0m 
  undefines the success [1mtrigger[22;0m and disconnects from the proxy.  

  See: [1m/addworld[22;0m, [1m%proxy_host[22;0m, [1m%proxy_port[22;0m 

&redirection

redirection

  If TF is started with input or output redirected, [1m%more[22;0m will be ignored and 
  SIGINT (^C) will kill TF without prompting.  TF will not exit when EOF is 
  reached; the [1m/quit[22;0m command must be given explicitly.  

  On UNIX systems, it is possible to write a tf script starting with the 
  lines: 

      #!/bin/sh
      exec tf -n $* <$0

  and following with any tf commands.  The file can then be executed directly 
  like a shell script.  

&scrolling
&scrollback
&windows
&window
&virtual window
&virtual windows

virtual windows

  Starting in version 5.0, TF maintains a separate virtual window for each 
  open [1msocket[22;0m, including the "(no world)" pseudo-socket.  Normally, a window 
  scrolls when text is written to it.  If the [1mmore[22;0m flag is set, automatic 
  scrolling will stop when the window becomes full.  You can manually scroll 
  forwards and backwards in each [1msocket[22;0m's window using the keys in the table 
  below.  

  Per-socket windows make it unnecessary to finish reading the text on one 
  [1msocket[22;0m before switching to another.  When you bring a new [1msocket[22;0m into the 
  [1mforeground[22;0m, the old [1msocket[22;0m's window is hidden, but remembers all of its text 
  and current position; when you return that old [1msocket[22;0m to the [1mforeground[22;0m, the 
  text is redrawn at the remembered position, and you can resume reading where 
  you left off.  A [1mdividing line[22;0m makes it easy to find the point where the old 
  text ends and the new text begins.  The text of a window is also refilled 
  after resuming from [1m/suspend[22;0m or [1m/sh[22;0m, and even when the terminal's size 
  changes.  

  In the table below, the "[1m/dokey[22;0m" columns indicate the argument to the [1m/dokey[22;0m 
  command that performs the scrolling, and the "keys" column indicates the 
  default keystrokes that perform the scrolling.  

      scroll       ....forward....   ...backward....
      amount       [1m/dokey[22;0m  keys      [1m/dokey[22;0m     keys
      -----------  ------- -------   ---------- ----
      normal       [1mPgDn[22;0m    PgDn      [1mPgUp[22;0m       PgUp
      1/2 screen   [1mhpage[22;0m   ^[h ^X]   [1mhpageback[22;0m  ^X[
      1 screen     [1mpage[22;0m    TAB ^X}   [1mpageback[22;0m   ^X{
      1 line       [1mline[22;0m    ^[^N      [1mlineback[22;0m   ^[^P

  Note that the line-scrolling keys may be typable as meta-ctrl-n and 
  meta-ctrl-p (depending on your [1m%meta_esc[22;0m and [1mlocale[22;0m).  "Normal" scrolling is 
  a full screenful by default.  If you prefer PgUp and PgDn to scroll a half 
  screen instead, you should redefine 

      [1m/def[22;0m key_pgdn = /dokey_hpage
      [1m/def[22;0m key_pgup = /dokey_hpageback

  Some terminal emulators do not send PgUp and PgDn keys to tf.  If you have 
  such a terminal, you may wish to 

      [1m/bind[22;0m ^F = /dokey_page
      [1m/bind[22;0m ^B = /dokey_pageback

  If you're an emacs user, you may want to bind 

      [1m/bind[22;0m ^V = /dokey_page
      [1m/bind[22;0m ^[v = /dokey_pageback

  (or, "/load kb-emacs.tf").  

  A virtual screen can be redrawn with ^L, or cleared with ^[^L (ESC ctrl-L).  
  Once lines are cleared from a screen, they can be redrawn by scrolling back 
  to them.  They are not automatically redrawn when you hide the screen and 
  then unhide it again.  

  Some [1mhooks[22;0m need to print messages that do not make sense at the bottom of 
  the [1mforeground[22;0m window (as they did before version 5.0).  For example, if you 
  have world Foo in the [1mforeground[22;0m, and get activity in world Bar, it would 
  not make sense for the [1mACTIVITY hook[22;0m to print "% Activity in world Bar" to 
  Foo's window.  Firstly, you might want to know about the activity even if 
  you are not at the end of Foo's window buffer.  Secondly, after you read the 
  text in Bar and returned to Foo, the message would still be at the bottom of 
  Foo's window buffer, misleadingly.  Many messages of this type are now 
  delivered as "[1malerts[22;0m".  An [1malert[22;0m appears temporarily on the [1mstatus line[22;0m, 
  where you can see it immediately and it will not outlive its usefulness.  
  Also, because text from different worlds is not mixed in 5.0, the [1mWORLD hook[22;0m 
  no longer prints "--- World <[4mname[24m> ---".  

  The [1m/limit[22;0m command will filter the text displayed in a window.  The counters 
  in the [1mmore[22;0m prompt will count only the lines that match the [1mlimit[22;0m.  

  If your terminal emulator has its own scrollback, it probably will not work 
  very well with tf.  To avoid confusion and avoid polluting your terminal's 
  scrollback with garbage, tf tries to switch to the terminal's "alternate 
  buffer", which does not keep scrollback.  But not all terminals and 
  configurations allow this (for example, xterm does, but only if the termcap 
  or terminfo entry contains the correct codes, and it has not been disabled 
  with xterm's titeInhibit resource).  If the terminal can not switch to an 
  alternate buffer, the terminal's scrollback may appear to work for a while, 
  but will become jumbled as soon as you switch worlds in tf or use tf's 
  scrollback.  You are advised to not attempt to use your terminal's 
  scrollback at all while running tf.  

  See also: [1minterface[22;0m, [1mvisual[22;0m, [1m/limit[22;0m, [1mkeybindings[22;0m.  
&interrupt
&hangup
&sigwinch
&signals

signals

  TF catches several signals from the operating system and handles them 
  specially: 
  SIGINT (normally generated by typing ^C) 
          Aborts any running [1mmacro[22;0m or blocking [1mhostname resolution[22;0m or [1mconnect[22;0m, 
          and, if [1minteractive[22;0m is on, offers a menu of choices:
          C) continue tf; X) exit; T) disable triggers; P) kill processes.  If 
          [1minteractive[22;0m is off, tf exits without prompting.  
  SIGQUIT (normally generated by typing ^\) 
          If [1minteractive[22;0m is on, TF prompts the user to quit.  If the answer is 
          'y', or [1minteractive[22;0m is off, TF will dump a core file if configured 
          to do so, and exit.  
  SIGTERM 
          Calls the SIGTERM [1mhook[22;0m, and then exits TF.  
  SIGHUP (normally generated when the terminal disconnects) 
          Calls the SIGHUP [1mhook[22;0m, and then exits TF if SIGHUP was not ignored 
          when tf was started.  
  SIGUSR1 
          Calls the SIGUSR1 [1mhook[22;0m.  TF does not exit.  
  SIGUSR2 
          Calls the SIGUSR2 [1mhook[22;0m.  TF does not exit.  
  SIGTSTP (normally generated by typing ^Z) 
          Suspends the TF process, like [1m/suspend[22;0m.  
  SIGWINCH (normally generated by resizing the terminal) 
          Redraws the screen, and calls the [1mRESIZE[22;0m [1mhook[22;0m.  

  See also: [1mhooks[22;0m, [1m/signal[22;0m 

&
&sockets

sockets

  Associated topics: 
  [1m/connect[22;0m 
          open a [1msocket[22;0m connection to a world 
  [1m/dc[22;0m     close (disconnect) a [1msocket[22;0m 
  [1m/fg[22;0m     bring a [1msocket[22;0m into the [1mforeground[22;0m 
  [1m%login[22;0m  enable [1mautomatic login[22;0m 
  [1m/listsockets[22;0m 
          display a list of open [1msockets[22;0m 
  [1mfg_world()[22;0m 
          name of foreground world 
  [1midle()[22;0m  idle time 
  [1mnactive()[22;0m 
          number of active sockets, or number of undisplayed lines 
  [1mis_connected()[22;0m 
          tests whether a [1msocket[22;0m is connected 
  [1mis_open()[22;0m 
          tests whether a [1msocket[22;0m is open 
  [1m%background[22;0m 
          determines when to process text from [1mbackground[22;0m [1msockets[22;0m 
  [1m%bg_output[22;0m 
          determines how to display text from [1mbackground[22;0m [1msockets[22;0m 

#current
#foreground
#background
#foreground/background/current
  A [1msocket[22;0m is a world-in-use, including a network connection (usually) and a 
  virtual window for displaying text.  TF can have multiple [1msockets[22;0m open 
  simultaneously.  Only one of these can be displayed at a time; this is 
  called the [1mforeground[22;0m [1msocket[22;0m.  In [1mvisual mode[22;0m, the name of the world on the 
  [1mforeground[22;0m [1msocket[22;0m is displayed on the [1mstatus line[22;0m.  Other [1msockets[22;0m are in the 
  [1mbackground[22;0m.  Text from any [1msocket[22;0m is [1mtrigger[22;0med and stored in [1mhistory[22;0m 
  immediately, but is not displayed until that [1msocket[22;0m is brought into the 
  [1mforeground[22;0m.  Handling of events in [1mbackground[22;0m [1msockets[22;0m can be customized with 
  the [1m%{bg_output}[22;0m and [1m%{background}[22;0m flags.  

  The [1mcurrent[22;0m [1msocket[22;0m is the [1msocket[22;0m to which commands are sent.  The [1mcurrent[22;0m 
  [1msocket[22;0m is almost always the same as the [1mforeground[22;0m [1msocket[22;0m, except: 1) when a 
  [1mmacro[22;0m is [1mtriggered[22;0m from any [1msocket[22;0m, that [1msocket[22;0m becomes the [1mcurrent[22;0m [1msocket[22;0m 
  for the duration of that [1mmacro[22;0m execution; 2) when a [1m/repeat[22;0m or [1m/quote[22;0m with 
  world redirection runs (-w option), that world's [1msocket[22;0m becomes the [1mcurrent[22;0m 
  [1msocket[22;0m for the duration of the [1mprocess[22;0m execution.  

#
  Text from a [1msocket[22;0m goes through a number of checks before being displayed.  
  If the text matches any [1mtrigger[22;0m patterns, a [1mmacro[22;0m may be executed, or the 
  text may be [1mgag[22;0mged or [1mhilite[22;0md.  If the text was not [1mgag[22;0mged, TF also checks 
  to see if it should be suppressed because of [1m%quiet[22;0m, [1m/watchdog[22;0m or 
  [1m/watchname[22;0m.  Finally, the text is added to the world's [1mhistory[22;0m and the 
  global [1mhistory[22;0m, and is queued for display.  

  You can open a new [1msocket[22;0m in several ways: 

    * By giving the world name or address on the command line when 
      [1mstarting tf[22;0m.  
    * By using a [1m/connect[22;0m or [1m/world[22;0m command.  
    * By "[1mbamfing[22;0m" through a portal between muds (see "[1mbamf[22;0m").  

  You can switch between [1mforeground[22;0m [1msockets[22;0m with the [1m/fg[22;0m command; the [1m/dokey[22;0m 
  socketb and [1m/dokey[22;0m socketf commands, which by default are [1mbound[22;0m to ESC-left 
  and ESC-right; and with the ESC-w [1mkeybinding[22;0m, which switches to the next 
  world with activity, or if there is none, to the last world you were on.  

  If the [1m%{quitdone}[22;0m flag is on, and you disconnect from all worlds (either 
  with [1m/dc[22;0m or because the other end of the [1msocket[22;0m's network connection 
  closes), TF will exit.  

  If the [1m%{sockmload}[22;0m flag is on, a world's [1mmacro[22;0m file will be loaded when you 
  switch to the [1msocket[22;0m for that world (either with the next and previous 
  [1msocket[22;0m keys or with the [1m/world[22;0m command).  

  TF supports several TELNET options; see [1mtelnet[22;0m.  

  If [1m%{proxy_host}[22;0m is defined, all connections will go through a [1mproxy[22;0m server. 
  See: [1mproxy[22;0m.  

  Normally, certain types of disconnection can only be detected when you try 
  to send something on a connection.  TF uses the socket option SO_KEEPALIVE 
  to detect such disconnections even when idle, but it usually takes at least 
  2 hours to detect.  The time limit is usually a property of the operating 
  system, and can not be set by TF or an unprivileged user.  
#loopback
#connectionless
#connectionless socket

  A "connectionless" socket is created when you [1m/connect[22;0m to a [1mworld[22;0m that does 
  not have a host or port defined.  If the world also has the echo flag set, 
  any text you "send" to the socket is immediately "received", as if you were 
  connected to an echo server.  

  See also: [1mworlds[22;0m 

&flags
&globals
&global variables
&environment
&enumerated variable
&enumerated variables
&special
&special variable
&special variables

special variables

  Many options in TF can be controlled by setting special global [1mvariables[22;0m.  
  Many [1mvariables[22;0m have a limited number of permitted values, with corresponding 
  integer values; these are called enumerated [1mvariables[22;0m.  All flags are 
  enumerated [1mvariables[22;0m which can have the values "off" (0) or "on" (1).  
  Numeric [1mvariables[22;0m can have any integer value (within the range allowed by 
  your system).  Attempting to unset numeric [1mvariable[22;0m or give it a string 
  value will force its value to 0.  Dtime [1mvariables[22;0m represent a time duration 
  or period; their values can be written as a number of seconds or in 
  [4mhours[24m:[4mminutes[24m[:[4mseconds[24m] format, with up to 6 decimal places (microseconds).  
  A variable's type (enumerated, numeric, dtime, or string) affects its 
  behavior in [1mexpressions[22;0m.  

Special substitute-only variables

  The following special [1mvariables[22;0m may be used only in [1msubstitutions[22;0m, never as 
  a variable [1mreference[22;0m in an [1mexpression[22;0m.  

##
#%#
  [1m#[22m       The number of words in a [1mmacro[22;0m's argument text.  

#?
#%?
  [1m?[22m       The string return value of the most recently executed [1mcommand[22;0m 
          (builtin or [1mmacro[22;0m).  ([1mMacros[22;0m) called as functions return their value 
          and do [4mnot[24m set %?.) 

#
  [1m1,2...[22m 
  [1mL1,L2...[22m 
  [1m*[22m 
  [1mR[22m       Positional parameters.  See "[1msubstitution[22;0m".  (As of 5.0 beta 7, 
          these are case sensitive.) 

#
  [1mP<[4mn[24m>[22m 
  [1mPL[22m 
  [1mPR[22m      The text matched by the <[4mn[24m>th parenthesized subexpression, or the 
          text to the left or right of the matched text, in the last 
          successful [1mregexp[22;0m comparison.  See [1m%Pn[22;0m for more details.  (As of 5.0 
          beta 7, these are case sensitive.) 

#

Special global variables

  The following special global [1mvariables[22;0m can be examined and set.  In the 
  following list, a '=' following a [1mvariable[22;0m name indicates its default value. 
  For [1mvariables[22;0m that do not have defaults listed, the default is dependent on 
  your system or configuration.  

#COLUMNS
#%COLUMNS
  [1mCOLUMNS[22m 
          If this variable is set in the environment when TF starts, TF will 
          use its value instead of the value from the terminal driver.  See 
          [1m%LINES[22;0m, [1mcolumns()[22;0m.  

#HOME
#%HOME
  [1mHOME[22m    Your home directory, used by [1m/cd[22;0m and [1mfilename expansion[22;0m.  This is 
          usually inherited from the environment when TF starts.  

#LANG
#%LANG
  [1mLANG[22m    The current locale.  See [1mlocale[22;0m.  Automatically exported to the 
          environment when set.  

#LC_ALL
#%LC_ALL
  [1mLC_ALL[22m  The current locale.  See [1mlocale[22;0m.  Automatically exported to the 
          environment when set.  

#LC_CTYPE
#%LC_CTYPE
  [1mLC_CTYPE[22m 
          The current locale for character classification.  See [1mlocale[22;0m.  
          Automatically exported to the environment when set.  

#LC_TIME
#%LC_TIME
  [1mLC_TIME[22m 
          The current locale for time formatting.  See [1mlocale[22;0m.  Automatically 
          exported to the environment when set.  

#LINES
#%LINES
  [1mLINES[22m   If this variable is set in the environment when TF starts, TF will 
          use its value instead of the value from the terminal driver.  See 
          [1m%COLUMNS[22;0m, [1mlines()[22;0m.  

#%MAIL
  [1mMAIL[22m    The name of a file which TF may check for mail.  See: [1mmail[22;0m.  

#SHELL
#%SHELL
  [1mSHELL[22m   Shell used by [1m/sh[22;0m and [1m/quote[22;0m !.  This is usually inherited from the 
          environment when TF starts.  

#terminal
#term
#TERM
#%TERM
  [1mTERM[22m    Terminal type.  Changing the value of [1m%TERM[22;0m at any time will cause 
          TF to re-initialize its display functions to use the new value.  The 
          value of [1m%TERM[22;0m should agree with your actual terminal or emulator.  
          If your emulator supports multiple terminal types, the recommended 
          type to use is vt220, vt100, or ansi (in that order).  [1m%TERM[22;0m is 
          usually inherited from the environment when TF starts.  See also: 
          [1mmode[22;0m.  

#TFHELP
#%TFHELP
  [1mTFHELP[22m=[1m%{TFLIBDIR}[22;0m/tf-help 
          The name of the file used by [1m/help[22;0m.  

#TFLIBDIR
#%TFLIBDIR
  [1mTFLIBDIR[22m 
          The name of the TF library directory, which should contain the help 
          file (tf-help), the standard library (stdlib.tf), the local library 
          (local.tf), and many useful utility files.  The default value of 
          [1mTFLIBDIR[22;0m is set when TF is installed, but can be overridden by 
          setting it in the environment before starting TF.  This directory 
          will be searched by [1m/load[22;0m if [1mTFPATH[22;0m is blank or not set.  See also: 
          [1m/load[22;0m.  

#TFLIBRARY
#%TFLIBRARY
  [1mTFLIBRARY[22m=[1m%{TFLIBDIR}[22;0m/stdlib.tf 
          The name of the library file loaded at [1mstartup[22;0m.  This can be set in 
          the environment before starting TF, to load from an alternate 
          library file.  

#MAILPATH
#TFMAILPATH
#%TFMAILPATH
  [1mTFMAILPATH[22m 
          A space-separated list of files which TF may check for mail.  
          Literal spaces in a filename must be preceded by "\".  See: [1mmail[22;0m.  

#TFPATH
#%TFPATH
  [1mTFPATH[22m= 
          A space-separated list of directories that will be searched by 
          [1m/load[22;0m.  Literal spaces in a directory name must be preceded by "\".  
          If this is set, [1m%{TFLIBDIR}[22;0m will be ignored by [1m/load[22;0m, so be sure to 
          include the value of [1m%{TFLIBDIR}[22;0m in [1m%{TFPATH}[22;0m if you want to be able 
          to [1m/load[22;0m files with relative names from the standard library 
          directory.  See also: [1m/load[22;0m.  

#timezone
#time zone
#TZ
#%TZ
  [1mTZ[22m      On most systems, the timezone used to display formatted times.  In 
          the United States, the value is usually the local timezone name, 
          followed by the difference in hours from GMT, followed by an 
          optional daylight saving timezone name; for example, "PST8PDT".  For 
          details, see your system documentation for tzset(3) or environ(5).  
          This is usually inherited from the environment when TF starts, and 
          is automatically exported to the environment when set.  

#alert_attr
#%alert_attr
  [1malert_attr[22m=Br 
          The [1mattributes[22;0m used to display [1malert[22;0m text on the [1mstatus line[22;0m.  

#alert_time
#%alert_time
  [1malert_time[22m=5.0 
          (dtime) The number of seconds that [1malert[22;0m text is displayed on the 
          [1mstatus line[22;0m.  See [1mtfio[22;0m.  

#background
#%background
  [1mbackground[22m=on 
          (flag) If on, text from [1mbackground[22;0m worlds is processed and recorded 
          immediately upon receipt.  Otherwise, the text is ignored until the 
          [1msocket[22;0m is brought into the [1mforeground[22;0m.  In either case, the text is 
          not displayed until the [1msocket[22;0m is brought into the [1mforeground[22;0m (but 
          see [1m%{bg_output}[22;0m).  

#backslash
#%backslash
  [1mbackslash[22m=on 
          (flag) Enables use of '\' to quote the following character literally 
          during [1mmacro[22;0m [1mexpansion[22;0m.  Generally, this should only be turned off 
          if you are having problems with '\' in [1mmacros[22;0m written before version 
          3.0.  

#bamf
#%bamf
  [1mbamf[22m=off 
          (enumerated) 
          off     (0): server "[1mportals[22;0m" are ignored.  
          on      (1): Unter-style [1mbamfing[22;0m is enabled (disconnect).  
          old     (2): Old-style [1mbamfing[22;0m is enabled (no disconnect).  
          See [1m/bamf[22;0m.  

#bg_output
#%bg_output
  [1mbg_output[22m=on 
          (flag) When a [1mworld[22;0m is brought into the [1mforeground[22;0m, [1m%bg_output[22;0m 
          determines how to display output that was produced while the [1mworld[22;0m 
          was in the [1mbackground[22;0m: If on, the window display resumes where it 
          left off; if off, the window display jumps to the end, showing only 
          the last screenful.  Turning [1m%bg_output[22;0m off is equivalent to always 
          using the -q option with [1m/fg[22;0m.  The [1m%bg_output[22;0m flag has no effect on 
          other processing, including [1mtriggers[22;0m and [1mhistory[22;0m.  This flag is 
          ignored if the [1m%{background}[22;0m flag is off.  [1m%{background}[22;0m is tested 
          when the [1mworld[22;0m is [1mforeground[22;0med (in versions before 5.0, it was 
          tested when the text was received).  (See also: [1m/fg[22;0m -q) 

#binary_eol
#%binary_eol
  [1mbinary_eol[22m=LF 
          Determines what to send as end-of-line marker in [1mTELNET[22;0m BINARY mode. 
          Valid values are "LF", "CR", and "CRLF".  (See: [1m/telnet[22;0m) 

#borg
#%borg
  [1mborg[22m=on 
          (flag) Enables [1mtrigger[22;0m bodies ([1mattributes[22;0m are unaffected).  (See: 
          [1mtriggers[22;0m, [1m%max_trig[22;0m) 

#clearfull
#%clearfull
  [1mclearfull[22m=off 
          (flag) In [1mvisual[22;0m mode, clear input window rather than scroll when 
          full.  Always on if terminal can not scroll.  

#cleardone
#%cleardone
  [1mcleardone[22m=off 
          (flag) In [1mvisual[22;0m mode, enables clearing of input window when return 
          is pressed.  

#%clock
  [1mclock[22m   This variable is no longer supported.  To disable the status bar 
          clock, use "[1m/clock[22;0m off".  To make the clock display in 12-hour 
          format, do "[1m/clock[22;0m %I:%M".  See [1m/clock[22;0m.  

#clock_format
#%clock_format
  [1mclock_format[22m=%H:%M 
          The format of the clock displayed on the [1mstatus line[22;0m.  To make the 
          clock display in 12-hour format, "/set [1mclock_format[22;0m=%I:%M".  See 
          also: [1m/clock[22;0m, [1m%time_format[22;0m.  

#connect
#%connect
  [1mconnect[22m=nonblocking 
          Set to "blocking" or "nonblocking" to determine how [1m/connect[22;0m works.  
          Default is "nonblocking" on platforms that support it.  Nonblocking 
          allows you to continue doing other things while TF tries to 
          establish a new connection.  See also [1m%gethostbyname[22;0m.  

#defcompile
#%defcompile
  [1mdefcompile[22m=off 
          (flag) If off, [1mmacro[22;0m bodies are compiled the first time they are 
          executed; if on, [1mmacro[22;0m bodies are compiled immediately when they are 
          defined.  Since syntax checking is performed during compilation, 
          setting [1mdefcompile[22;0m=on will allow you to see the syntax errors in a 
          macro when you define it instead of waiting until execution.  

#%e
  [1me[22m=2.718281828...  
          Euler's number.  

#expand_tabs
#%expand_tabs
  [1mexpand_tabs[22m=on 
          (flag) If on (and [1m%emulation[22;0m is "print", "ansi_strip", or 
          "ansi_attr"), tabs received from a server are expanded to spaces 
          (according to [1m%tabsize[22;0m) immediately, before any [1mtrigger[22;0m processing.  
          If off, tab characters are left in received lines.  

#raw
#canon
#print
#ansi
#ansi_strip
#ansi_attr
#emulation
#%emulation
  [1memulation[22m=ansi_attr 
          Determines how special codes sent by the server should be 
          interpreted by TF.  The set of printable characters is determined by 
          the current [1mlocale[22;0m.  Valid values for [1m%emulation[22;0m are: 
          raw:    No conversion is done; lines are not wrapped; all 
                  nonprintable characters are displayed, and their effect is 
                  undefined (depending mainly on your terminal).  TF's input 
                  display is not guaranteed correct; use at your own risk.  
                  This mode allows the server to have most of the control over 
                  the screen, but is not guaranteed to give the desired 
                  effect, and will interfere with [1mtrigger[22;0m matching.  For best 
                  results, [1m%visual[22;0m should be "off", and TF [1mattributes[22;0m should 
                  not be used.  "Raw" is not recommended unless you know what 
                  you're doing.  
          print:  Tabs are expanded (if [1m%expand_tabs[22;0m is on); backspaces are 
                  interpreted; lines are wrapped; nonprintable characters 
                  removed.  
          ansi_strip: 
                  Like "print", but ANSI display codes are also removed.  
          ansi_attr: 
                  Like "ansi_strip", but ANSI [1mdisplay attribute[22;0m codes will be 
                  converted to TF's internal format and displayed correctly 
                  (on any terminal).  Other ANSI display codes (e.g., cursor 
                  motion) will be removed.  Recommended, especially for 
                  servers that send vt100/ansi display [1mattribute[22;0m codes.  
          debug:  converts nonprinting characters to a printable form.  See 
                  also: [1m%telopt[22;0m.  
          See also: [1m%istrip[22;0m, [1m%meta_esc[22;0m, [1m%tabsize[22;0m, [1m%expand_tabs[22;0m, [1mlocale[22;0m, 
          [1mattributes[22;0m, [1mdebugging[22;0m.  

#end_color
#%end_color
  [1mend_color[22m 
          The code that should be sent to your terminal to return to normal 
          [1mcolor[22;0m after a [1m%{start_color_*}[22;0m code.  See: [1mcolor[22;0m.  

#error_attr
#%error_attr
  [1merror_attr[22m 
          Defines the [1mattributes[22;0m used by the "E" [1mattribute[22;0m.  Can be any 
          combination of [1mattributes[22;0m, including color names.  See: [1mattributes[22;0m.  

#gag
#%gag
  [1mgag[22m=on  (flag) Enable the [1mgag[22;0m [1mattribute[22;0m.  (See: [1m/gag[22;0m, [1m/nogag[22;0m) 

#gethostbyname
#%gethostbyname
  [1mgethostbyname[22m=nonblocking 
          Set to "blocking" or "nonblocking" to determine how [1m/connect[22;0m does 
          hostname resolution.  See also [1m%connect[22;0m.  

#gpri
#%gpri
  [1mgpri[22m=0  Priority of subsequent [1m/gag[22;0ms.  (See: [1m/gag[22;0m) 

#hook
#%hook
  [1mhook[22m=on 
          (flag) Enable [1mhooks[22;0m.  (See: [1mhooks[22;0m, [1m/hook[22;0m, [1m%max_hook[22;0m.) Note that 
          [1mautologin[22;0m and automatic [1m%{lp}[22;0m setting will not work if [1m%{hook}[22;0m is 0. 

#hilite
#%hilite
  [1mhilite[22m=on 
          (flag) Enable display [1mattributes[22;0m, whether from a [1mtrigger[22;0m, the 
          server, or whatever.  (See: [1m/hilite[22;0m, [1m/nohilite[22;0m) 

#hiliteattr
#%hiliteattr
  [1mhiliteattr[22m=B 
          Defines the [1mattributes[22;0m used by [1mhilite[22;0ms.  Can be any combination of 
          [1mattributes[22;0m, including color names.  (See: [1mattributes[22;0m, [1m/hilite[22;0m) 

#histsize
#%histsize
  [1mhistsize[22m=1000 
          When a new world [1mhistory[22;0m is created, it will have space for 
          [1m%{histsize}[22;0m lines.  A world [1mhistory[22;0m is created the first time text 
          is sent to it.  (See also: [1m/histsize[22;0m) 

#hpri
#%hpri
  [1mhpri[22m=0  Priority of subsequent [1m/hilite[22;0ms.  

#insert
#typeover
#%insert
  [1minsert[22m=on 
          (flag) If on, keyboard input is inserted; if off, input overstrikes 
          existing text.  

#interactive
#%interactive
  [1minteractive[22m 
          (flag) If off, TF will not prompt for [1m/quit[22;0m, returning from [1m/sh[22;0m, 
          [1mSIGINT[22;0m (^C), or [1mSIGQUIT[22;0m (^\).  Defaults to on if standard input and 
          output are attatched to a terminal, off otherwise.  

#isize
#%isize
  [1misize[22m=3 
          Size of input window in [1mvisual[22;0m mode.  The output window will be 
          redrawn when this is changed.  See also: [1mlines()[22;0m, [1mwinlines()[22;0m.  

#istrip
#%istrip
  [1mistrip[22m=off 
          (flag) If on, the meta (high) bit will be stripped from all input 
          characters.  Note that this will prevent [1m%meta_esc[22;0m and [1mlocales[22;0m with 
          8-bit characters from working correctly.  

#%kbnum
  [1mkbnum[22m=  A value that can be set by typing ESC followed by digits, to be used 
          as an argument (repeat count) for a subsequent keybinding.  See: 
          [1mkeybindings[22;0m.  

#kecho
#%kecho
  [1mkecho[22m=off 
          (flag) Enables echoing of keyboard input, preceded by [1m%{kprefix}[22;0m.  
          See also: [1m%{kecho_attr}[22;0m.  [1m%{secho}[22;0m.  [1m/localecho[22;0m, [1m/addworld[22;0m -e.  

#kecho_attr
#%kecho_attr
  [1mkecho_attr[22m 
          Attributes used for lines echoed by [1m%{kecho}[22;0m.  

#keepalive
#%keepalive
  [1mkeepalive[22m=on 
          (flag) Enable periodic "pings" (TCP keepalive) of servers, to 
          prevent network timeouts and detect dropped connections.  Note: the 
          timing of keepalive messages is a system parameter that can not be 
          changed from tf.  

#%keypad
  [1mkeypad[22m=on 
          (flag) Enable application keypad mode, if supported by the terminal. 
          Application keypad mode makes the numeric keypad generate characters 
          different than the usual digit characters, so they may be 
          distinguished from the digit keys across the top of the keyboard.  
          See: [1mkeybindings[22;0m.  

#kprefix
#%kprefix
  [1mkprefix[22m= 
          Prefix prepended to lines echoed by [1m%{kecho}[22;0m.  

#%login
  [1mlogin[22m=on 
          (flag) Enable [1mautomatic login[22;0m [1mhook[22;0m.  (See: [1mautomatic login[22;0m, [1mhooks[22;0m, 
          [1m/world[22;0m) 

#lp
#%lp
  [1mlp[22m=off  (flag) Displays partial lines as [1mprompts[22;0m, after a short timeout.  
          Useful for LP and Diku MUDs.  (See: [1mprompts[22;0m) 

#lpquote
#%lpquote
  [1mlpquote[22m=off 
          (flag) If on, all [1m/quote[22;0m and [1m/repeat[22;0m processes run when an LP [1mprompt[22;0m 
          is received instead of when a timer expires.  The -P option of 
          [1m/quote[22;0m and [1m/repeat[22;0m provides the same feature on a per-process basis. 
          (See: [1mprocesses[22;0m) 

#maildelay
#%maildelay
  [1mmaildelay[22m=0:01:00.0 (60 seconds) 
          (dtime) Delay between mail checks.  Setting this to 0 disables mail 
          checking.  The file to be checked is named by the [1m%{MAIL}[22;0m [1mvariable[22;0m.  

#matching
#%matching
  [1mmatching[22m=glob 
          (enumerated) Determines the default [1mpattern matching[22;0m style.  
          "[1msimple[22;0m": 
                  straightforward string comparison.  
          "[1mglob[22;0m": 
                  shell-like matching (as before version 3.2).  
          "[1mregexp[22;0m": 
                  regular expression.  
          See also: [1mpatterns[22;0m, [1mregmatch()[22;0m, [1m%Pn[22;0m.  

#max_hook
#%max_hook
  [1mmax_hook[22m=1000 
          Maximum number of [1mhooks[22;0m allowed in a 10 second period.  When this 
          value is exceeded, a message is printed and [1m%hook[22;0m is automatically 
          turned off to disable hooks.  This helps prevent infinite hook 
          loops.  A value of 0 will allow unlimited hooks.  

#max_instr
#iteration
#iterations
#instruction
#instructions
#%max_instr
  [1mmax_instr[22m=1000000 
          Maximum number of instructions in a [1mmacro[22;0m execution.  A value of 0 
          will allow unlimited instructions.  An "instruction" is a basic 
          internal tf operation, such as addition, testing an /if or /while 
          condition, a substitution, sending a line of text to a server, or 
          joining two commands with a "%|" pipe.  

#max_kbnum
#%max_kbnum
  [1mmax_kbnum[22m=999 
          The maximum value of [1mkbnum[22;0m that can be set via the keyboard.  See: 
          [1mkeybindings[22;0m.  

#max_recur
#recursion
#recursions
#%max_recur
  [1mmax_recur[22m=100 
          Maximum depth of recursive [1mmacro[22;0m calls or [1mtriggers[22;0m.  This helps 
          prevent infinite macro loops.  A value of 0 will allow unlimited 
          recursion.  

#max_trig
#%max_trig
  [1mmax_trig[22m=1000 
          Maximum number of [1mtriggers[22;0m allowed in a 10 second period.  When this 
          value is exceeded, a message is printed and [1m%borg[22;0m is automatically 
          turned off to disable triggers.  This helps prevent infinite trigger 
          loops.  A value of 0 will allow unlimited triggers.  

#%mccp
  [1mmccp[22m=on (if tf was compiled with MCCP support) 
          (flag) If on, MCCPv2 is allowed on new connections.  See [1mmccp[22;0m.  

#mecho
#%mecho
  [1mmecho[22m=off 
          (enumerated) 
          "off" (0): 
                  do not echo [1mmacro[22;0m [1mexpansions[22;0m.  
          "on" (1): 
                  echo [1mexpansions[22;0m of non-invisible [1mmacros[22;0m.  
          "all" (2): 
                  echo [1mexpansions[22;0m of all [1mmacros[22;0m.  
          [1m%{mprefix}[22;0m will be prepended once for each recursion level when 
          [1mmacro[22;0m [1mexpansion[22;0m echoing is enabled.  See also: [1m%{mecho_attr}[22;0m, 
          [1mdebugging[22;0m.  

#mecho_attr
#%mecho_attr
  [1mmecho_attr[22m 
          Attributes used for lines echoed by [1m%{mecho}[22;0m.  

#meta_esc
#meta
#%meta_esc
  [1mmeta_esc[22m=nonprint 
          (enumerated) If [1m%istrip[22;0m is off, typed characters with their meta 
          (high) bit set may have the meta bit stripped and be prefixed with 
          an ESC character.  This allows META-x and ESC x to invoke the same 
          keybinding.  Possible values of [1m%meta_esc[22;0m: 
          "off" (0): 
                  Never convert a meta bit to ESC.  
          "on" (1): 
                  Always convert a meta bit to ESC.  
          "nonprint" (2): 
                  Convert a meta bit to ESC only if the meta bit makes the 
                  character unprintable in the current [1mlocale[22;0m.  
          Meta bit conversion can be prevented for a single keystroke by 
          preceeding it with the [1mLNEXT[22;0m key (^V), regardless of the state of 
          [1m%meta_esc[22;0m.  

#more
#%more
  [1mmore[22m=off 
          (flag) Displays output one screenfull at a time.  (See: [1m/more[22;0m) 

#mprefix
#%mprefix
  [1mmprefix[22m=+ 
          Prefix prepended to lines echoed by [1m%{mecho}[22;0m.  

#oldslash
#%oldslash
  [1moldslash[22m=on 
          (flag) If on, sequences of more than one '/' in a [1mmacro[22;0m body will be 
          compressed by one during [1mmacro[22;0m [1mexpansion[22;0m.  This allows [1mmacros[22;0m 
          written before version 3.0 to work properly.  With oldslash=off, 
          only slashes at the beginning of a body are handled specially.  You 
          are encouraged to turn this off.  (See: [1mevaluation[22;0m) 

#%pi
  [1mpi[22m=3.141592654...  
          The ratio of a circle's circumference to its diameter.  

#pedantic
#%pedantic
  [1mpedantic[22m=off 
          (flag) If on, TF will generate warnings about some potential 
          problems in your macro code.  Often the warnings indicate code that 
          is technically valid but may not do what you intended.  See also 
          [1mdebugging[22;0m.  

#prompt_sec
#%prompt_sec
#prompt_usec
#%prompt_usec
  [1mprompt_sec[22m 
  [1mprompt_usec[22m 
          Obsolete.  Use [1m%{prompt_wait}[22;0m instead.  
#prompt_wait
#%prompt_wait
  [1mprompt_wait[22m=0.25 
          (dtime) The delay (in seconds) used to recognize unterminated 
          [1mprompts[22;0m.  (See: [1mprompts[22;0m).  

#proxy_host
#%proxy_host
#proxy_port
#%proxy_port
  [1mproxy_host[22m= 
  [1mproxy_port[22m=23 
          These two [1mvariables[22;0m describe the [1mproxy[22;0m server used for opening 
          connections.  (See: [1mproxy[22;0m).  

#ptime
#%ptime
  [1mptime[22m=1.0 
          (dtime) Default delay (in seconds) between [1m/quote[22;0m and [1m/repeat[22;0m 
          [1mprocess[22;0m runs.  

#qecho
#%qecho
  [1mqecho[22m=off 
          (flag) Echoing of [1m/quote[22;0m text.  See also: [1m%{qprefix}[22;0m, [1m%{qecho_attr}[22;0m, 
          [1mdebugging[22;0m.  

#qecho_attr
#%qecho_attr
  [1mqecho_attr[22m 
          Attributes used for lines echoed by [1m%{qecho}[22;0m.  

#qprefix
#%qprefix
  [1mqprefix[22m= 
          Prefix prepended to lines echoed by [1m%{qecho}[22;0m.  

#quiet login
#quiet
#%quiet
  [1mquiet[22m=off 
          (flag) [1mGag[22;0m text after [1mlogin[22;0m until the mud sends "Use the WHO 
          command", "### end of messages ###", or 25 lines.  Note: This will 
          not function correctly on MUDs which don't send those strings or 25 
          lines in the introductory text.  

#quitdone
#%quitdone
  [1mquitdone[22m=off 
          (flag) Quit upon disconnection from last [1msocket[22;0m.  

#redef
#%redef
  [1mredef[22m=on 
          (flag) Allows redefinition of existing worlds, keybindings, and 
          named [1mmacros[22;0m.  

#refreshtime
#%refreshtime
  [1mrefreshtime[22m=100000 
          (int) The delay (in microseconds) for redisplaying your keyboard 
          input after it is overwritten by incoming text in [1mnon-visual[22;0m mode.  
          If you have a slow connection between you and tf, you may wish to 
          increase this delay.  The default is 100000 (1/10 second).  

#scroll
#%scroll
  [1mscroll[22m=on 
          (flag) In [1mvisual[22;0m mode, scroll output instead of wrapping from bottom 
          to top.  

#secho
#%secho
  [1msecho[22m=off 
          (flag) Echoing of text before sending it to the server (above the 
          TELNET layer).  See also: [1m%{sprefix}[22;0m, [1m%{secho_attr}[22;0m, [1m%{kecho}[22;0m.  
          [1m%{telopt}[22;0m, [1mdebugging[22;0m.  

#secho_attr
#%secho_attr
  [1msecho_attr[22m 
          Attributes used for lines echoed by [1m%{secho}[22;0m.  

#shpause
#%shpause
  [1mshpause[22m=on 
          (flag) Wait for a keypress after returning from [1m/sh[22;0m (unless 
          [1m%interactive[22;0m is off).  

#sigfigs
#%sigfigs
  [1msigfigs[22m=15 
          The maximum number of significant digits to display when printing a 
          floating point number.  Note that 16 or more may introduce rounding 
          error.  Also note that some real numbers with up to 6 decimal places 
          are stored with fixed points, not floating points, so are not 
          affected by [1msigfigs[22;0m (or rounding error).  

#snarf
#%snarf
  [1msnarf[22m=off 
          (flag) Don't send empty lines to the server.  

#sockmload
#%sockmload
  [1msockmload[22m=off 
          (flag) Load [1mmacro[22;0m files when [1mforegrounding[22;0m a world ("[1m/dokey[22;0m 
          socketf", "[1m/dokey[22;0m socketb", or "[1m/fg[22;0m").  Normally, a world's [1mmacro[22;0m 
          file is loaded only when TF first connects to it.  (Note: the WORLD 
          [1mhook[22;0m is more useful than sockmload).  

#sprefix
#%sprefix
  [1msprefix[22m= 
          Prefix prepended to lines echoed by [1m%{secho}[22;0m.  

#start_color
#%start_color
#start_color_*
#%start_color_*
#start_color_name
#%start_color_name
#start_color_<name>
#%start_color_<name>
#start_color_bgname
#%start_color_bgname
#start_color_bg<name>
#%start_color_bg<name>
  [1mstart_color_<[4mname[24m>[22m 
  [1mstart_color_bg<[4mname[24m>[22m 
          The control code that should be sent to your terminal to produce 
          foreground or background [1mcolor[22;0m <[4mname[24m>.  See: [1mcolor[22;0m.  

#status_attr
#%status_attr
  [1mstatus_attr[22m 
          The [1mattributes[22;0m used to display the [1mstatus area[22;0m in [1mvisual mode[22;0m.  See: 
          [1mstatus area[22;0m.  

#%status_fields
  [1mstatus_fields[22m 
          [1mDeprecated.[22m The list of fields displayed on row 0 of the [1mstatus area[22;0m 
          in [1mvisual mode[22;0m.  See: [1mstatus area[22;0m.  

#status_height
#%status_height
  [1mstatus_height[22m=1 
          The number of rows in the [1mstatus area[22;0m in [1mvisual mode[22;0m.  See: [1mstatus 
          area[22;0m.  

#status_pad
#%status_pad
  [1mstatus_pad[22m=_ 
          The padding character used in displaying the [1mstatus area[22;0m in [1mvisual 
          mode[22;0m.  See: [1mstatus area[22;0m.  

#tab
#%tab
#tabs
#tabsize
#%tabsize
  [1mtabsize[22m=8 
          Tabs will be replaced with spaces to pad to a multiple of 
          [1m%{tabsize}[22;0m.  See also: [1m%expand_tabs[22;0m, [1m%emulation[22;0m.  

#telopt
#%telopt
  [1mtelopt[22m=off 
          (flag) Display [1mtelnet[22;0m option negotiations (for debugging purposes).  
          See also: [1m%emulation[22;0m=debug, [1mdebugging[22;0m.  

#textdiv
#separator
#separator.tf
#divider
#=====
#dividing line
#%textdiv
  [1mtextdiv[22m=on 
          (enumerated) When you bring a socket into the foreground, TF can 
          help you distinguish old text that has been displayed before from 
          new text that is being displayed for the first time by printing a 
          dividing line between the old and new text or by clearing the old 
          text.  The setting of [1m%textdiv[22;0m controls this behavior: 
          "off" (0): 
                  Never print a divider or clear the screen; just draw old and 
                  new text normally.  
          "on" (1): 
                  Print a [1m%textdiv_str[22;0m divider between old and new text.  The 
                  divider is temporary: when it scrolls off the screen, or the 
                  screen is backgrounded, it disappears forever.  
          "always" (2): 
                  Print a [1m%textdiv_str[22;0m divider after the old text even if 
                  there is no new text.  
          "clear" (3): 
                  Clear (don't redraw) all old text before displaying new 
                  text.  Old text can be manually redisplayed by [1mscrolling 
                  back[22;0m.  
          See also: [1m%textdiv_str[22;0m, [1m/fg[22;0m.  

#textdiv_str
#%textdiv_str
  [1mtextdiv_str[22m====== 
          The dividing line printed between old and new text when bringing a 
          socket to the foreground.  See [1m%textdiv[22;0m.  

#tfhost
#%tfhost
  [1mtfhost[22m= 
          Name or address to use for the client (tf) end of connections.  See 
          also: [1maddworld[22;0m, [1mconnect[22;0m.  

#sub
#%sub
  [1msub[22m=off 
          See: [1m/sub[22;0m.  

#time_format
#%time_format
  [1mtime_format[22m=%H:%M 
          The format used to display times in [1m/recall[22;0m and [1m/time[22;0m.  The default 
          displays hours and minutes.  See [1mftime()[22;0m for a description of the 
          format.  See also: [1m%clock_format[22;0m.  

#visual
#%visual
  [1mvisual[22m=on 
          (flag) Divides the screen into an input window and an output window. 
          The output window will be redrawn when this is changed.  (See: [1mmode[22;0m) 

#warn_5keys
#%warn_5keys
  [1mwarn_5keys[22m=on 
          (flag) If on, TF will warn the first time some of the new 5.0 
          [1mkeybindings[22;0m are used.  

#warn_curly_re
#%warn_curly_re
  [1mwarn_curly_re[22m=on 
          (flag) If on, TF will warn when using a [1mregexp[22;0m containing '{', which 
          has a new meaning in version 5.0.  

#warn_status
#%warn_status
  [1mwarn_status[22m=on 
          (flag) If on, TF will warn when directly setting [1m%status_fields[22;0m, 
          [1m%status_int_more[22;0m, [1m%status_int_world[22;0m, or [1m%status_int_clock[22;0m, which 
          have new default values and new ways to set them in version 5.0.  
          See [1mstatus line[22;0m.  

#warning_attr
#%warning_attr
  [1merror_attr[22m 
          Defines the [1mattributes[22;0m used by the "W" [1mattribute[22;0m.  Can be any 
          combination of [1mattributes[22;0m, including color names.  See: [1mattributes[22;0m.  

#watchdog
#%watchdog
  [1mwatchdog[22m=off 
          (flag) [1mGag[22;0m repeated lines.  (See: [1m/watchdog[22;0m) 

#watchname
#%watchname
  [1mwatchname[22m=off 
          (flag) [1mGag[22;0m overactive players.  (See: [1m/watchname[22;0m) 

#wordpunct
#%wordpunct
  [1mwordpunct[22m=_ 
          List of punctuation that will be considered to be part of a word 
          instead of delimiting the ends of a word, by [1mkbwordleft()[22;0m and 
          [1mkbwordright()[22;0m (and therefore by [1m/dokey[22;0m WLEFT, WRIGHT, etc).  

#wordwrap
#wrap
#%wrap
  [1mwrap[22m=on 
          (flag) Enable wordwrap on the screen.  TF will try to break lines at 
          spaces or other punctuation to fit them within [1m%{wrapsize}[22;0m columns.  
          [1m%{wrap}[22;0m is ignored if [1m%{emulation}[22;0m is "raw".  See also: 
          [1m%{wrappunct}[22;0m, [1m%{wrapsize}[22;0m, [1m%{wrapspace}[22;0m.  

#wraplog
#%wraplog
  [1mwraplog[22m=off 
          (flag) Enable wordwrap in log files.  See also: [1m%wrap[22;0m.  

#wrappunct
#%wrappunct
  [1mwrappunct[22m=10 
          When [1mwrapping[22;0m, allow wrapping at any punctuation if wrapping only at 
          spaces would have caused more than [1m%wrappunct[22;0m characters to wrap.  
          This can make long URLs look nicer, but harder to cut and paste.  
          Setting [1m%wrappunct[22;0m to 0 disables wrapping at punctuation other than 
          spaces.  

#wrapsize
#%wrapsize
  [1mwrapsize[22m=79 
          Lines (input and output) extending past this column will be split.  
          Default value is one less than the number of columns on your 
          terminal (typically 80).  Output is not wrapped if [1m%{emulation}[22;0m is 
          "raw".  See also: [1m%wrap[22;0m, [1m%wrappunct[22;0m, [1m%wrapspace[22;0m, [1mcolumns()[22;0m.  

#wrapspace
#indent
#indenting
#indentation
#%wrapspace
  [1mwrapspace[22m=4 
          Wrapped text is indented by this many spaces.  See also: [1m%wrap[22;0m, 
          [1m%wrapsize[22;0m.  

#
  The following builtin commands set the corresponding [1mvariables[22;0m, and also 
  perform additional functions: [1m/gag[22;0m, [1m/hilite[22;0m, [1m/hook[22;0m, [1m/nogag[22;0m, [1m/nohilite[22;0m, 
  [1m/watchdog[22;0m, and [1m/watchname[22;0m 

  The standard library also defines the following [1mmacros[22;0m to set the values of 
  the corresponding [1mvariables[22;0m: [1m/background[22;0m, [1m/bamf[22;0m, [1m/borg[22;0m, [1m/clearfull[22;0m, 
  [1m/cleardone[22;0m, [1m/gpri[22;0m, [1m/hpri[22;0m, [1m/insert[22;0m, [1m/isize[22;0m, [1m/login[22;0m, [1m/lp[22;0m, [1m/lpquote[22;0m, [1m/kecho[22;0m, 
  [1m/mecho[22;0m, [1m/more[22;0m, [1m/ptime[22;0m, [1m/qecho[22;0m, [1m/quiet[22;0m, [1m/quitdone[22;0m, [1m/redef[22;0m, [1m/shpause[22;0m, 
  [1m/sockmload[22;0m, [1m/sub[22;0m, [1m/visual[22;0m and [1m/wrapspace[22;0m.  

  Note: The [1mvariables[22;0m 'L' and 'R' are reserved (see: [1mvariables[22;0m).  You should 
  not assign values to them.  

  See: [1mvariables[22;0m, [1m/set[22;0m 

&status
&status fields
&%status_fields
&visual bar
&visual line
&status bar
&status_line
&status line
&status area

status line

  In [1mvisual[22;0m mode, the input and output windows are separated by a status line, 
  which by default looks something like this: 

    [1;7mMore 156[22;27m_[4mWorldName[24m____________(Read)_(Active: [4mn[24m)_(Log)_(Mail)_(Over)_12:34

    * "[1mMore[22;0m" indicates how many [1mmore[22;0m lines of text are waiting to be seen. 
    * "<[4mWorldName[24m>" is the name of the [1mforeground[22;0m [1msocket[22;0m's world.  
    * "(Read)" indicates that keyboard input is being read by [1mread()[22;0m.  
    * The "(Active: [4mn[24m)" indicator shows the number of [1msockets[22;0m with unseen 
      text.  
    * "(Log)" indicates that there is one or more [1mlog[22;0m file open.  
    * "(Mail)" or "Mail [4mn[24m" indicates the number of files named by [1m%MAIL[22;0m or 
      [1m%MAILPATH[22;0m that contain unread mail.  
    * "(Over)" indicates that typed characters will overstrike instead of 
      insert (that is, [1m%insert[22;0m is off).  
    * The current time is displayed at the right end of the status line.  

Configuring the status area

  The status area may contain 1 or more rows; the number is determined by 
  [1m%status_height[22;0m.  The rows are numbered from the top starting at 0.  Each row 
  is defined as a list of fields.  A status field is defined as follows: 

    * an optional field name 
    * an optional ":" and number indicating the field width 
    * an optional ":" and [1mattribute[22;0m 

  The current list of status fields for row <[4mN[24m> can be fetched with 
  [1mstatus_fields(<[4mN[24m>)[22;0m.  
#%status_field_defaults
#status_rm
#status_edit
#status_defaults
#status_save
#status_restore
#status_add
#/clock
#/status_rm
#/status_edit
#/status_defaults
#/status_save
#/status_restore
#/status_add
  The following commands modify the fields of the status area: 
  /clock off 
          Remove the clock from the status bar (equivalent to "/status_rm 
          @clock").  

  /clock on 
          Add a clock to the end of status row 0 if there is not already a 
          clock on status row 0.  The width of the @clock field will be set 
          exactly wide enough to hold a time formatted according to 
          [1m%clock_format[22;0m.  

  /clock [<[4mformat[24m>] 
          Add a clock to the end of status row 0 if there is not already a 
          clock on status row 0; in either case, use <[4mformat[24m> to control the 
          format of the clock (see [1mftime()[22;0m for the meaning of <[4mformat[24m>).  If 
          <[4mformat[24m> is omitted, it defaults to "%H:%M".  The width of the 
          @clock field will be set exactly wide enough to hold a time 
          formatted according to <[4mformat[24m>.  

          Example: display a clock in 12-hour format: 
                  /clock %I:%M 

  /status_defaults 
          Restore list of status fields for all rows and their formats 
          (%status_int_* and %status_var_*) to their default values.  
          (Previous versions of tf had a [1m%status_field_defaults[22;0m variable; this 
          is now deprecated.) 

  /status_save <[4mname[24m> 
          Save the current list of fields in row 0 into memory slot with label 
          <[4mname[24m>.  <[4mName[24m> must be a legal variable name.  (Saved fields will 
          be forgotten when tf exits.) 

  /status_restore <[4mname[24m> 
          Restore the list of fields in row 0 that was previously saved with 
          "/status_save <[4mname[24m>".  

  /status_rm [-r<[4mN[24m>] <[4mname[24m> 
          Remove status field <[4mname[24m> from status row <[4mN[24m>.  If -r is not 
          specified, all rows are searched.  Only the first matching field is 
          removed.  If there are unnamed pad fields on both sides of the named 
          field, the one with the smaller width is also removed; if the named 
          field is at the beginning or end of a row, the neighboring pad field 
          (if any) is removed.  

          Example: Remove the @mail field from the status bar: 
                  /status_rm @mail 

  /status_add [<[4moptions[24m>] <[4mname[24m>[:<[4mwidth[24m>[:<[4mattributes[24m>]] ...  
          Add status field <[4mname[24m> to the status bar with optional <[4mwidth[24m> and 
          <[4mattributes[24m>.  Options: 
          -r<[4mN[24m>   add to row <[4mN[24m> (default 0) 
          -A      add after all other fields (i.e., at end) 
          -A<[4mfield[24m> 
                  add after existing field <[4mfield[24m> 
          -B      add before all other fields (i.e., at beginning) 
          -B<[4mfield[24m> 
                  add before existing field <[4mfield[24m> 
          -s<[4mN[24m>   insert padding of <[4mN[24m> spaces between the new field and the 
                  neighbor selected by -A or -B (default 1) 
          -x      don't add the field if one with the same name is already 
                  present 
          -c      clear all existing fields before adding new fields 
          If neither -A nor -B is given, -A is assumed.  

          Example: Add a new field after the world name to display the 
          contents of the variable "hp": 
                  /status_add -A@world hp:4 

          Multiple fields may be specified, but padding is not automatically 
          added between them; you must specify padding explicly.  For example, 
                  /status_add -Aclock foo:4 :1 bar:4 :2 baz:4 
          is equivalent to 
                  /status_add -Aclock foo:4 
                  /status_add -Afoo bar:4 
                  /status_add -Abar -s2 baz:4 

  /status_edit [-r<[4mN[24m>] <[4mname[24m>[:<[4mwidth[24m>[:<[4mattributes[24m>]] 
          If field <[4mname[24m> currently exists in any status row, replace it with 
          <[4mname[24m>[:<[4mwidth[24m>[:<[4mattributes[24m>]].  Neighboring padding is unchanged.  
          If -r is given, only row <[4mN[24m> is searched.  Only the first matching 
          field is edited.  

          Example: Change the @log field to say "L" instead of "(Log)", and 
          change the field's width to match: 
                  /set status_int_log=nlog() ? "L" : "" 
                  /status_edit @log:1 

#

  For backward compatiblity, you can get and set the status fields for row 0 
  via the %status_fields [1mvariable[22;0m, but doing so is deprecated.  

  The default list of status fields is: 

      @more:8:Br :1 @world :1 @read:6 :1 @active:11 :1 @log:5 :1 @mail:6 :1 insert:6 :1 kbnum:4 :1 @clock:5
    

  There are several types of fields: 

    * Unnamed fields create padding between the fields on either side of 
      it.  Each of the ":1" fields in the default [1mstatus_fields[22;0m puts a space 
      of 1 character between the other fields.  
    * Field names beginning with "@" correspond to internal states.  For 
      example, "@more" will be updated whenever the number of unseen lines 
      changes.  
    * Field names containing only letter, digits, and underscores 
      correspond to [1mvariables[22;0m.  Whenever there is a change in the value of the 
      [1mvariable[22;0m with the same name, the field will be updated.  The value an 
      unset variable is considered to be the empty string.  For example, 
      whenever the [1m%insert[22;0m variable changes, the "insert" field is updated.  
      Any [1mvariable[22;0m may be monitored in this manner.  
    * A field whose name is in quotes (", ', or `) has its name (without 
      the quotes) printed literally on the status bar, and is never updated.  
      Use the \ character to escape a quote inside the string.  The default 
      [1mstatus_fields[22;0m does not contain any of these literal fields.  

  Any [1mvariable[22;0m may be monitored, but there is a fixed list of internal 
  statuses.  The internal statuses available are: 
  @more   Updated when there is a change in the number of lines below the 
          bottom of the window.  
  @world  Updated when when the [1mforeground[22;0m [1mworld[22;0m changes.  During the 
          evaluation of the format expression, the [1mcurrent socket[22;0m is the new 
          [1msocket[22;0m.  
  @read   Updated when entering or exiting a [1mread()[22;0m function call.  
  @active 
          Updated when the number of active [1mworlds[22;0m changes.  During the 
          evaluation of the format expression, the [1mcurrent socket[22;0m is the 
          [1msocket[22;0m that became active.  
  @log    Updated when the number of open [1mlog[22;0m files changes.  
  @mail   Updated when mail arrives (See "[1mmail[22;0m").  
  @clock  Updated every minute, on the minute.  

  A field's width determines how many columns it will take up on the screen.  
  If the width of a string literal field field is omitted, it defaults to the 
  length of the string literal.  One other field width may be omitted or set 
  to 0, which means that field will use whatever columns are unused by the 
  other fields.  Normally, fields are left-justified within the width, but a 
  negative field width will right-justify the field within the absolute value 
  of the width.  A width of "-0" can be used to right-justify the 
  variable-width field.  If the formatted text is wider than the field width, 
  it will be truncated to fit within the specified width.  Fields may also be 
  truncated if they would not fit on the screen.  

  The [1mattributes[22;0m explicily given in the field definiton are combined with 
  those in the corresponding %status_attr_int_<[4mfieldname[24m> (for internal state 
  fields) or %status_attr_var_<[4mvarname[24m> (for variable fields).  The combined 
  [1mattributes[22;0m are applied to the field text when it is displayed, but not to 
  the padding used to bring the field to the specified width.  The entire 
  status line, including padding, is displayed with the [1mattributes[22;0m given by 
  [1m%status_attr[22;0m, which is none by default.  

  To bring fields up to their specified width, they are padded with 
  [1m%status_pad[22;0m, which is "_" by default.  By setting [1mstatus_pad[22;0m to " " and 
  [1mstatus_attr[22;0m to "r", you can create a status line that looks more like the 
  one in emacs or the irc client.  

  When a status field is updated, the text displayed for that field is 
  determined by evaluating the [1mexpression[22;0m contained in the [1mvariable[22;0m 
  status_int_<[4mname[24m> (for internal state @<[4mname[24m>) or status_var_<[4mname[24m> (for 
  variable <[4mname[24m>).  Also, for [1mvariable[22;0m fields, if status_var_<[4mname[24m> is not 
  set, the value of the [1mvariable[22;0m will be displayed directly.  Changing a 
  format variable will cause the status line to update.  

  All this may sound rather complex, so an example might help.  The default 
  value of [1mstatus_fields[22;0m is: 

    @more:8:Br :1 @world :1 @read:6 :1 @active:11 :1 @log:5 :1 @mail:6 :1 insert:6 :1 kbnum:4 :1 @clock:5
    

  and the corresponding format [1mvariables[22;0m are: 

    [1m/set[22;0m status_int_more \
         [1mmoresize()[22;0m == 0 ? "" : \
         [1mmoresize()[22;0m > 9999 ? "MuchMore" : \
         [1mpad[22;0m("More", 4, [1mmoresize()[22;0m, 4)
    [1m/set[22;0m status_int_world   [1mstrcat[22;0m( \
         [1mfg_world[22;0m() !~ "" & ![1mis_open[22;0m([1mfg_world[22;0m()) ? "!" : "",  [1mfg_world[22;0m())
    [1m/set[22;0m status_int_read    [1mnread()[22;0m ? "(Read)" : ""
    [1m/set[22;0m status_int_active  [1mnactive()[22;0m ? [1mpad[22;0m("(Active:",0,[1mnactive()[22;0m,2,")") : ""
    [1m/set[22;0m status_int_log     [1mnlog()[22;0m ? "(Log)" : ""
    [1m/set[22;0m status_int_mail \
         ![1mnmail[22;0m() ? "" : \
         [1mnmail[22;0m()==1 ? "(Mail)" : \
         [1mpad[22;0m("Mail", 0, [1mnmail[22;0m(), 2)
    [1m/set[22;0m status_var_insert  [1minsert[22;0m ? "" : "(Over)"
    [1m/set[22;0m status_int_clock   [1mftime[22;0m([1mclock_format[22;0m)
    

  The first field is "@more:8:Br".  So, whenever the number of unseen lines 
  changes, TF looks for the [1mvariable[22;0m status_int_more, and evaluates the 
  [1mexpression[22;0m it contains.  The result of the [1mexpression[22;0m is printed in the 
  first 8 columns of the status line, with [1mattributes[22;0m "Br" (bold and reverse). 
  The [1mexpression[22;0m was carefully written so that it will never be more than 8 
  characters, because it would be confusing to generate something like 
  "More:12345" and then have it truncated to "More:123" because of the field 
  width of 8.  

  Since the "@world" field has no explicit width, its width is determined 
  dynamically.  The fields on its left are pushed to the left side of the 
  screen, the fields on its right are pushed to the right side of the screen, 
  and the "@world" field uses whatever space remains in the middle.  
#prompt example

  Another example: Say your mud has a [1mprompt[22;0m like "H:42 M:17> " that shows 
  your hit points and mana, and you want it displayed on the status line like 
  " 42, 17", after the world name.  To do this, call "/status_add -Aworld 
  hp_mana:7", and define a [1mprompt[22;0m [1mhook[22;0m: 

    [1m/def[22;0m [1m-mregexp[22;0m [1m-h[22;0m"PROMPT ^H:([^ ]*) M:([^ ]*)> $" hp_mana_hook = \
        [1m/set[22;0m hp=[1m%P1[22;0m%; \
        [1m/set[22;0m mana=[1m%P2[22;0m%; \
        [1m/set[22;0m hp_mana=[1m$[[22;0m[1mpad[22;0m(hp, 3, ",", 0, mana, 3)]%; \
        [1m/test[22;0m [1mprompt[22;0m([1m{*}[22;0m)

#

  See: [1mvisual[22;0m 

&subs
&substitution

substitution

  Before a [1mmacro[22;0m body or arguments to [1m/eval[22;0m are executed, special character 
  sequences are replaced with new text as described below.  

#%;
#newline
#command separator

Command separation. 
%;

  Separates commands within a [1mmacro[22;0m body.  See [1mevaluation[22;0m.  

#%|

Pipe. 
%|

  Separates commands within a [1mmacro[22;0m body, and connects the output of the first 
  to the input of the second.  See [1mevaluation[22;0m.  

#character substitution
#\n
#\\
#ascii

Character substitution. 
\[4mn[24m
\[4mc[24m

  In the first form, the character whose ASCII code is <[4mn[24m> is substituted.  If 
  <[4mn[24m> starts with "0x", it is interpreted as a hexadecimal number; otherwise, 
  if <[4mn[24m> starts with "0", it is interpreted as octal; otherwise, it is 
  interpreted as decimal.  In the second form, the character <[4mc[24m> is 
  substituted.  This is useful for escaping any special meaning <[4mc[24m> has; in 
  particular, "\\" is substituted with "\".  If the [1mvariable[22;0m [1m%{backslash}[22;0m is 
  off, the \[4mc[24m form does not have this special interpretation.  

#//

Slash compression. 
//... 

  If [1m%{oldslash}[22;0m is on, sequences of slashes are replaced with a sequence of 
  one fewer slashes.  A single slash, however, is left alone.  This feature 
  remains for backward compatibility only; you are encouraged to turn 
  [1m%{oldslash}[22;0m off to disable this.  

#$[
#$[]

[1mExpression[22;0m evaluation. 
$[[4m[1mexpression[22;0m[24m]

  The <[4m[1mexpression[22;0m[24m> is evaluated and its string value is substituted in its 
  place.  See "[1mexpressions[22;0m".  

#$(
#$()
#command subs
#command substitution

Command substitution. 
$([4mcommand[24m)

  <[4mCommand[24m> is [1mevaluated[22;0m as if it were the body of a [1mmacro[22;0m: it goes through 
  [1msubstitution[22;0m, and is executed in a new [1mscope[22;0m.  If <[4mcommand[24m> contains any ')' 
  characters, they must be escaped by preceding them with '\' so they are not 
  interpreted as the end of the substitution.  The echoed output of <[4mcommand[24m> 
  is substituted in place of the $(...) construct (much like `...` in most 
  shells).  If <[4mcommand[24m> produces more than one line of output, they will be 
  concatenated, with a space between each, to form one line.  

  Example: 

          [1m/def[22;0m showver = :is using tf version $([1m/ver[22;0m)

  could be used to tell other mudders what version of tf you're using.  

#$
#${
#${}
#macro subs
#macro substitution

[1mMacro[22;0m substitution. 
${[4mname[24m}
$[4mname[24m$

  The body of the [1mmacro[22;0m <[4mname[24m> is substituted.  The second form is supported 
  only for backward compatibility, and its use is discouraged.  In the first 
  form, the brackets may be omitted if the subsequent text could not be 
  confused as part of the name.  

  Example: The text "${foo}" would be replaced with the body of the [1mmacro[22;0m 
  named "foo".  

#$$

Dollar compression. 
$$... 

  Sequences of '$'s are replaced by a sequence of one fewer '$'s.  A single 
  '$', however, is left alone, unless it introduces one of the substitutions 
  described above.  This is used to put a literal '$' in text that goes 
  through macro substitution.  

#%
#%{
#%{}
#%n
#%0
#%1
#%-1
#%-n
#%R
#%L
#%*
#%#
#%?
#variable subs
#variable substitution
#positional parameters
#arguments
#parameters
#variables and parameters

Variable and Argument substitution. 
%[4mselector[24m
%{[4mselector[24m}
%{[4mselector[24m-[4mdefault[24m}

  The value of a [1mvariable[22;0m or an argument to the [1mmacro[22;0m is substituted, as 
  determined by <[4mselector[24m>.  The brackets are recommended for clarity, but may 
  be omitted if there is no default and the text following it can not be 
  misinterpreted as part of the selector.  The selector can be any of: 

  <[4mname[24m>  The value of the [1mvariable[22;0m <[4mname[24m> is substituted.  Names are case 
          sensitive.  

  0       selects the name of the executing macro.  (Before version 4.0, "0" 
          was equivalent to "*").  

  #       selects the count of positional parameters.  

  *       selects all positional parameters.  

  ?       selects the return value of the most recently executed command 
          (builtin or macro).  

  1, 2, 3, etc.  
          selects the corresponding positional parameter.  There is no maximum 
          parameter number; any number greater than [1m%{#}[22;0m will simply produce 
          an empty substitution.  

  -1, -2, -3, etc.  
          selects all positional parameters except the first, all except the 
          first two, all except the first three, etc.  

  L1, L2, etc.  
          selects the last positional parameter, second-to-last, etc.  "L" is 
          the same as "L1".  (As of 5.0 beta 7, these are case sensitive.) 

  -L1, -L2, etc.  
          selects all positional parameters except the last, all except the 
          last two, etc.  "-L" is the same as "-L1".  (As of 5.0 beta 7, these 
          are case sensitive.) 

  P[4mn[24m      selects the text matching the <[4mn[24m>th parenthesized subexpression from 
          the last [1mregular expression[22;0m match.  See [1m%P[4mn[24m[22;0m.  (As of 5.0 beta 7, 
          these are case sensitive.) 

  R       selects a positional parameter at random.  (see also: [1mrand()[22;0m) (As of 
          5.0 beta 7, this is case sensitive.) 

  [1mVariable[22;0m name and selectors are case sensitive (prior to 5.0 beta 7, "L[4mn[24m", 
  "P[4mn[24m" and "R" selectors were not).  No substitutions are performed on 
  <[4mselector[24m>.  

  If the substitution determined by the <[4mselector[24m> would be empty, and a 
  <[4mdefault[24m> value is given, the default will be substituted instead.  Thus 
  "[1m%{1[22;0m-foofle}" is replaced with the first word if there is one, or "foofle" 
  if not.  The <[4mdefault[24m> value may contain [1mvariable[22;0m, [1mmacro[22;0m, [1mexpression[22;0m, and 
  [1mcommand[22;0m [1msubstitutions[22;0m.  

  The meaning of "positional parameters" depends on how the [1mmacro[22;0m was called.  
  If called with the traditional "/[4mname[24m ..." command syntax, each 
  space-separated word is a positional parameter.  If called with the 
  "[4mname[24m(...)" [1mfunction syntax[22;0m, each function argument is a positional 
  parameter; if more than one is selected, they are concatenated, with a space 
  between each.  If called as a [1mtrigger[22;0m, the positional parameters are the 
  words in the text that [1mtrigger[22;0med the [1mmacro[22;0m.  In a [1mhook[22;0m call, the positional 
  parameters are the hook arguments.  In an [1m/eval[22;0m statement, they are 
  inherited from the caller.  

  Note that in [1mexpressions[22;0m, it is easiest to omit the % and just use the 
  {[4mselector[24m[-[4mdefault[24m]} part.  If the selector is a variable name and no 
  default is desired, the name may be used directly in an [1mexpressions[22;0m without 
  % or {...}.  

#%{PL}
#%PL
#%{PR}
#%PR
#%{Pn}
#%Pn
#%P
#subexpressions
#regexp subexpressions

Regexp subexpressions. 
%{P[4mn[24m}
%{PL}
%{PR}

  This is actually a special case of [1mvariable substitution[22;0m.  The [1m%P[22;0m variables 
  get their values from the last successful regexp match in scope.  [1m%P0[22;0m 
  expands to the text matched by the entire [1mregexp[22;0m.  [1m%P[4mn[24m[22;0m expands to the text 
  matched by the <[4mn[24m>th parenthesised subexpression of the [1mregexp[22;0m.  [1m%PL[22;0m and [1m%PR[22;0m 
  expand to the text to the left and right, respectively, of the text matched 
  by the entire [1mregexp[22;0m.  The "scope" of a [1mregexp[22;0m match is the lifetime of the 
  [1mmacro[22;0m expansion it [1mtrigger[22;0med, [1mhook[22;0med, or in which it occurred (i.e., with 
  [1mregmatch()[22;0m).  

  For example, after the text "Jabba the Hutt goes east." matches the [1mregexp[22;0m 

    " goes ([^ ]*)\.$"

  then the following expansions will be available until the [1mmacro[22;0m exits: PL = 
  "Jabba the Hutt"; P0 = " goes east."; P1 = "east".  

  The number <[4mn[24m> can be any nonnegative number.  If there is no subexpression 
  corresponding to <[4mn[24m>, the substitution will be ignored.  When parentheses 
  are nested, <[4mn[24m> refers to the order of the opening parentheses.  

  The [1m%P[4mn[24m[22;0m subs will always refer to the first [1mregexp[22;0m match on the line, even 
  if a partial [1mhilite[22;0m ([1m/def -P[22;0m) causes the [1mregexp[22;0m to be applied more than 
  once.  

#%%
#percent compression

Percent compression. 
%%... 

  Sequences of '%'s are replaced by a sequence of one fewer '%'s.  A single 
  '%', however, is left alone unless it introduces one of the substitutions 
  described above.  This is used to put a literal '%' in text that goes 
  through macro substitution.  

#

Examples 

  Here are a couple of simple examples.  

  Definition: [1m/def[22;0m advice = whisper [1m%1[22;0m = Let the wookie win. 
  Command: /advice R2D2
  Sends: whisper R2D2 = Let the wookie win. 

  Definition: [1m/set[22;0m ending=meister
  Definition: [1m/def[22;0m greet = :waves to [1m%{1[22;0m-Jack}%{ending}. 
  Command: /greet
  Sends: :waves to Jackmeister. 
  Command: /greet Dave
  Sends: :waves to Davemeister. 

  For some more complex examples, look at the files in TFLIBDIR.  

  See: [1mevaluation[22;0m, [1mexpressions[22;0m 

&summary

summary

  Type "[1m/help[22;0m [1mintro[22;0m" for basic information on using TF. 
  Type "[1m/help[22;0m [1mtopics[22;0m" for a list of other help topics. 
  Type "[1m/help[22;0m [1mcommands[22;0m" for a complete list of TF builtin commands. 
  Type "[1m/help[22;0m [1m/help[22;0m" for instructions on using [1m/help[22;0m. 

  If you are having problems with TF and wish to contact the author, see 
  "[1mproblems[22;0m".  

  If you are having trouble reading the help sections because text is 
  scrolling off the screen, try typing "[1m/more[22;0m on" before [1m/help[22;0m, and then when 
  you get a "[1m--More--[22;0m" prompt, press [1mTAB[22;0m or [1mPageDown[22;0m when you're ready to 
  continue.  

&command line
&commandline
&startup
&initialization
&invocation
&tf

tf

  Syntax: 

  tf [-L<[4mdir[24m>] [-f[<[4mfile[24m>]] [-c<[4mcommand[24m>] [-vlqn] [<[4mworld[24m>]
  tf [-L<[4mdir[24m>] [-f[<[4mfile[24m>]] [-c<[4mcommand[24m>] [-vlq] <[4mhost[24m> <[4mport[24m> 
  ____________________________________________________________________________

  At startup, TF takes the following steps: 

    * Initializes [1mspecial variables[22;0m.  Any [1mvariables[22;0m defined in the 
      environment will override TF's default values for the [1mvariables[22;0m with the 
      same name.  
    * Loads commands from the [1mstandard macro library (stdlib.tf)[22;0m, the 
      optional [1mlocal macro library (local.tf)[22;0m, and your [1mpersonal configuration 
      file[22;0m (see [1mtfrc[22;0m).  
    * Executes <[4mcommand[24m>, if one was given.  
    * Enables [1mvisual mode[22;0m if -v was not given and [1m%visual[22;0m has not been 
      explicitly set to "off".  
    * Tries to connect to <[4mworld[24m>, or <[4mhost[24m> <[4mport[24m>.  If no [1mworld[22;0m is 
      given, and the -n option is not given, TF will try to connect to the 
      first [1mworld[22;0m defined with [1maddworld()[22;0m in the configuration file(s).  If no 
      [1mworlds[22;0m are defined, or TF can not connect to the specified [1mworld[22;0m, TF 
      will start up in unconnected mode.  

  Options: 
  -L<[4mdir[24m> 
          Use <[4mdir[24m> instead of [1m%TFLIBDIR[22;0m as library directory.  
  -f<[4mfile[24m> 
          Load <[4mfile[24m> instead of the normal personal config file.  
  -f      Do not load any personal config file at startup.  
  -c<[4mcommand[24m> 
          Execute <[4mcommand[24m> after loading config file.  <[4mCommand[24m> is treated 
          as if it had been typed on the tf command line (i.e., the value of 
          [1m%sub[22;0m is significant).  
  -n      Do not connect to a [1mworld[22;0m automatically at startup if no <[4mworld[24m> or 
          <[4mhost[24m>/<[4mport[24m> are specified.  
  -l      Disable [1mautomatic login[22;0m.  (see: [1mlogin[22;0m) 
  -q      Enable [1mquiet login[22;0m.  (see: [1m%quiet[22;0m) 
  -v      Disable automatic switch to [1mvisual mode[22;0m.  

  The library directory is determined by the first of the following which has 
  a value: -L option; [1m%TFLIBDIR[22;0m environment [1mvariable[22;0m; or, compiled-in default. 
  The standard library file is determined by the first of the following which 
  has a value: [1mTFLIBRARY[22;0m environment [1mvariable[22;0m; or, appending "/stdlib.tf" to 
  [1m%TFLIBDIR[22;0m.  

  TF honors several [1mlocale[22;0m categories, which can be set to make TF work better 
  with languages other than English.  See [1mlocale[22;0m.  

  See [1mhttp://tinyfugue.sourceforge.net/[22;0m for the latest info on TF.  

  See also: [1mintro[22;0m, [1mtfrc[22;0m, [1mlibrary[22;0m, [1mworlds[22;0m, [1m/addworld[22;0m 

&tfout
&tferr
&alert
&streams
&tfio

tfio

  TF normally does its output through "streams", which are analagous to the 
  streams of C stdio.  

  Output from most tf commands, including [1m/echo[22;0m, are output to the "[1mtfout[22;0m" 
  stream, which is normally attached to the screen.  [1mtfout[22;0m may be redirected 
  with a [1mcommand /quote[22;0m, [1m$() command substitution[22;0m, or [1m%| pipe[22;0m.  

  Many TF error messages, hook messages, and the output of "[1m/echo -e[22;0m" are 
  output to the "[1mtferr[22;0m" stream, which is always attached to the screen, and 
  may not be redirected.  

  Some TF error messages, hook messages, and the output of "[1m/echo -A[22;0m" are 
  output to the "[1malert[22;0m" stream.  In [1mvisual[22;0m mode, text sent to the alert stream 
  is displayed briefly on the status line [1mstatus line[22;0m, where it can be seen 
  immediately even if you're at a [1mmore[22;0m prompt.  The duration of the alert 
  display is determined by [1m%alert_time[22;0m.  In [1mnonvisual[22;0m mode, text sent to the 
  alert stream is redirected to the tferr stream.  

  Text from a world or "[1m/echo -w[22;0m" is sent to a [1mstream[22;0m for that world.  Text 
  sent to a world [1mstream[22;0m will be stored in the [1mhistory[22;0m of that world.  If that 
  world is the [1mforeground[22;0m world, the text is sent to the screen immediately; 
  otherwise, it will not be displayed until world is brought into the 
  [1mforeground[22;0m.  

  Commands that read input (using [1mtfread()[22;0m) read by default from "[1mtfin[22;0m", which 
  is normally attached to the keyboard.  [1mtfin[22;0m may be redirected with a [1m%| 
  pipe[22;0m.  

  All [1mstreams[22;0m have a handle which can be used as an argument to the [1mtfio[22;0m 
  functions.  The handles for [1mtfin[22;0m, [1mtfout[22;0m, and [1mtferr[22;0m are "i", "o", and "e", 
  respectively.  The handles for [1mstreams[22;0m opened with [1mtfopen()[22;0m are integers.  

tfopen()

  The [1mtfopen[22;0m([4mname[24m, [4mmode[24m) function can be used to open arbitrary [1mstreams[22;0m.  If 
  called with no arguments, [1mtfopen()[22;0m opens an unnamed "q" mode [1mstream[22;0m.  The 
  <[4mmode[24m> argument describes the usage of the [1mstream[22;0m: 
  "w"     Open a file "<[4mname[24m>" for writing.  Write operations will overwrite 
          existing file contents, if any.  
  "a"     Open a file "<[4mname[24m>" for appending.  Write operations will occur 
          after existing file contents, if any.  
  "r"     Open a file "<[4mname[24m>" for reading.  (see also: "[1m/quote '[22;0m").  
  "p"     Execute a shell command "<[4mname[24m>" and read its output (see also: 
          "[1m/quote ![22;0m").  
  "q"     Open a queue for reading and writing.  The <[4mname[24m> argument will 
          appear in the output of [1m/liststreams[22;0m, but has no other meaning.  
  A "q" mode [1mstream[22;0m may be thought of as a place to hold lines for passing 
  between two or more commands.  

  If successful, the [1mtfopen()[22;0m function returns a positive number which is the 
  handle of the new [1mstream[22;0m, which should be used in subsequent calls to 
  [1mtfread()[22;0m, [1mtfwrite()[22;0m, and [1mtfclose()[22;0m.  If it fails, the [1mtfopen()[22;0m function 
  returns -1.  

  A call to [1mtfwrite()[22;0m or [1mtfread()[22;0m on a [1mstream[22;0m opened with a mode that does not 
  allow that operation will return -1.  

  The [1m/liststreams[22;0m command will display a list of open [1mstreams[22;0m.  

tfclose()

  When a [1mstream[22;0m opened by [1mtfopen()[22;0m is no longer needed, it should be closed 
  with [1mtfclose[22;0m([4mhandle[24m), which will flush the [1mstream[22;0m and release its resources. 
  [1mtfclose()[22;0m can be used on the [1mtfout stream[22;0m (handle "o") within a [1mmacro[22;0m body 
  to prevent further output from subsequent commands in that [1mmacro[22;0m body; 
  closing the [1mtfin stream[22;0m (handle "i") will prevent further reads; and closing 
  the [1mtferr stream[22;0m (handle "e") is not allowed.  

tfwrite()

  The [1mtfwrite[22;0m([4mhandle[24m, [4mline[24m) function writes a <[4mline[24m> of text to the [1mstream[22;0m 
  designated by <[4mhandle[24m>.  If <[4mhandle[24m> is omitted, the [1mtfout stream[22;0m is used 
  (so [1mtfwrite[22;0m([4mline[24m) is equivalent to [1mecho[22;0m([4mline[24m)).  [1mDisplay attributes[22;0m of [4mline[24m 
  are stripped if it is written outside of tf (i.e., to a file or pipe).  

  If an OS file (mode "w" or "a") is set to autoflush (the default), then each 
  line written is flushed to the file immediately.  If you are writing a large 
  number of lines, it is more efficient to disable autoflushing with 
  [1mtfflush[22;0m([4mhandle[24m, "off"), and manually force a flush with [1mtfflush[22;0m([4mhandle[24m) or 
  [1mtfclose[22;0m([4mhandle[24m) after writing the large block.  [1mtfflush()[22;0m has no meaning on 
  files of mode "p", "q", or "r".  [1mStreams[22;0m are flushed automatically when 
  closed.  

tfread()

  The [1mtfread[22;0m([4mhandle[24m, [4mvariable[24m) function reads a line from the [1mstream[22;0m 
  designated by <[4mhandle[24m>.  If <[4mhandle[24m> is omitted, the [1mtfin stream[22;0m is used.  
  If successful, the line is assigned to <[4mvariable[24m>, and [1mtfread()[22;0m returns the 
  (non-negative) length of the line.  If <[4mvariable[24m> did not already exist, it 
  is created at the global level, as if by [1m/set[22;0m.  If there are no lines 
  available to read, or an error occurs, [1mtfread()[22;0m returns -1.  For "r" and "p" 
  mode [1mstreams[22;0m, a -1 return value indicates end-of-file; the only valid 
  operation on the [1mstream[22;0m after that is [1mtfclose()[22;0m.  But for a "q" mode [1mstream[22;0m, 
  a -1 return value may just mean there are currently no lines in the queue; 
  more lines may be added by [1mtfwrite()[22;0m, and then [1mtfread()[22;0m will be able to read 
  them.  

Keyboard Reading

  [1mtfread()[22;0m from the keyboard is special.  It can only be done from a command 
  line command; trying to do it directly or indirectly from a trigger, hook, 
  keybinding, or process is an error, and will make the [1mtfread()[22;0m return -1.  
  It reads a line of input from the keyboard until the newline key is pressed 
  or "[1m/dokey[22;0m newline" is executed.  During the read, all existing [1mkeybindings[22;0m 
  continue to work normally.  Any text already in the input buffer is not 
  cleared when the read starts.  Text entered after the read starts is 
  appended to the existing text, and when the read ends, its result is the 
  entire input buffer.  Lines entered during a read are not saved in the input 
  [1mhistory[22;0m (but you can use "[1m/recordline[22;0m -i" to save them explicitly).  

  A read from the keyboard (and the [1mmacro[22;0m that called it) can be interrupted 
  with a SIGINT, normally generated by typing CTRL-C.  

  During a keyboard read, if a [1mmacro[22;0m calls [1m/dokey[22;0m newline, the newline will 
  not be executed immediately, but will be held until the rest of the commands 
  in the [1mmacro[22;0m are processed.  For example, consider the keybinding "[1m/def[22;0m 
  [1m-b[22;0m'^[^M' = /dokey newline%; /send go".  Normally, typing ^[^M would execute 
  the current input buffer, then send "go" to the server.  But during a 
  keyboard read, typing ^[^M would send "go" first, and then do the newline 
  that completes the read.  

  The library file [1mtextutil.tf[22;0m defines several commands that are useful with 
  [1mtfio[22;0m.  

  See: [1minterface[22;0m, [1m/liststreams[22;0m, [1m/input[22;0m, [1mexpressions[22;0m, [1mnread()[22;0m, [1mfunctions[22;0m, 
  [1mtextutil.tf[22;0m 

&config
&configuration
&customization
&customizing
&tfrc
&tinytalk
&.tinytalk
&.tfrc

.tfrc

  At [1mstartup[22;0m, TF attempts to load and execute commands from the personal 
  config file named "~/.tfrc", "~/tfrc", "./.tfrc" or "./tfrc".  This file can 
  contain any commands you want executed automatically when TF starts.  

  Some useful commands to include in your personal config file: 

  [1m/addworld[22;0m 
          Define a [1mworld[22;0m.  TF will automatically connect to the first [1mworld[22;0m if 
          not started with the "-n" option.  
  [1m/def[22;0m    Define a [1mmacro[22;0m (including [1mtriggers[22;0m, [1mhilites[22;0m, [1mgags[22;0m, [1mkeybindings[22;0m, and 
          [1mhooks[22;0m).  
  [1m/set[22;0m    Set a [1mvariable[22;0m.  There are many [1mspecial variables[22;0m that change the 
          behavior of tf, listed under "[1mspecial variables[22;0m".  
  [1m/load[22;0m   Load commands from another file.  
  [1m/require[22;0m 
          Load a library file.  

  [1mTFLIBDIR[22;0m contains a sample "tfrc" file that you may want to copy and modify 
  to fit your tastes.  

  For backward compatibility, TF will load ~/.tinytalk if it exists.  The use 
  of ~/.tinytalk is discouraged.  

  See: [1mstartup[22;0m, [1mlibrary[22;0m, [1mspecial variables[22;0m, [1m/load[22;0m 

&timer
&timing

timing

  See: [1mprocesses[22;0m, [1m/repeat[22;0m, [1m/quote[22;0m, utilities ([1m/at[22;0m, [1m/tick[22;0m), [1m%clock[22;0m, [1m/time[22;0m.  
&tools
&/reedit
&/edmac
&/edvar
&/edworld
&/name
&/getline
&/xtitle
&xterm
&tools.tf

tools.tf

  Usage: 

  [1m/REQUIRE[22;0m tools.tf
  ____________________________________________________________________________

  [1m/EDMAC[22;0m <[4mmacroname[24m> 
  [1m/EDVAR[22;0m <[4mvariablename[24m> 
  [1m/EDWORLD[22;0m <[4mworldname[24m> 
          Stick an existing [1mmacro[22;0m, [1mvariable[22;0m, or [1mworld[22;0m definition in the input 
          window for editing.  

  [1m/NAME[22;0m [<[4mname[24m>] 
          Change your character name (on a TinyMUD style mud).  

  [1m/GETLINE[22;0m <[4mn[24m> 
          Grab the <[4mn[24m>th line from [1mhistory[22;0m and stick it in the input buffer.  

  [1m/XTITLE[22;0m <[4mtext[24m> 
          Put <[4mtext[24m> on the titlebar of an xterm.  

  See: [1m/sh[22;0m, [1m/edit[22;0m, [1m/recall[22;0m, [1mtfrc[22;0m 

&triggers

triggers

  Before we get into the gory details, here's a simple example of a trigger: 

    [1m/def[22;0m [1m-t[22;0m"{*} has arrived." greet = :waves to [1m%1[22;0m.

  This command defines a macro called "greet".  Whenever text like "Bob has 
  arrived." is received, /greet will be executed automatically, sending the 
  text ":waves to Bob." to the server.  

  Associated commands: 
  [1m/def[22;0m    define a [1mmacro[22;0m with any fields 
  [1m/trig[22;0m   define a [1mtrigger[22;0m [1mmacro[22;0m 
  [1m/trigp[22;0m  define a [1mtrigger[22;0m [1mmacro[22;0m with [1mpriority[22;0m 
  [1m/trigc[22;0m  define a [1mtrigger[22;0m [1mmacro[22;0m with probability 
  [1m/trigpc[22;0m 
          define a [1mtrigger[22;0m [1mmacro[22;0m with probability and [1mpriority[22;0m 
  [1m/gag[22;0m    define a [1mtrigger[22;0m [1mmacro[22;0m to [1mgag[22;0m text 
  [1m/hilite[22;0m 
          define a [1mtrigger[22;0m [1mmacro[22;0m to [1mhilite[22;0m text 
  [1m/trigger[22;0m 
          call a [1mtrigger[22;0m [1mmacro[22;0m manually 
  [1m/substitute[22;0m 
          modify the text that invoked the [1mtrigger[22;0m 

  [1mTriggers[22;0m are a method of calling a [1mmacro[22;0m based on incoming text.  When a 
  line of text from a [1msocket[22;0m matches the [1mtrigger[22;0m [1mpattern[22;0m of a [1mmacro[22;0m, that 
  [1mmacro[22;0m becomes a candidate for automatic execution.  

  If multiple [1mmacros[22;0m have [1mtriggers[22;0m which match the same text, one or more are 
  chosen for execution as described under "[1mpriority[22;0m".  

  The <[4mtext[24m> which [1mtriggers[22;0m a [1mmacro[22;0m is given to the [1mmacro[22;0m as arguments, as if 
  it had been called with ``/<[4mmacro[24m> <[4mtext[24m>''.  Positional parameters (e.g., 
  [1m%1[22;0m) refer the the corresponding word in the [1mtrigger[22;0ming text.  If the [1mtrigger[22;0m 
  is a [1mregexp[22;0m, subexpression parameters refer to the text matched by the 
  corresponding parenthesised subexpression (see also: [1m%Pn[22;0m).  

  If the selected [1mmacro[22;0m(s) have display [1mattributes[22;0m, the [1mattributes[22;0m are used to 
  display the text which [1mtrigger[22;0med the [1mmacro[22;0m.  

  If a [1mmacro[22;0m has the world field set, it can only be [1mtrigger[22;0med by text from 
  that world.  

  If a [1mmacro[22;0m has a probability less than 100%, it might not be executed even 
  if it is [1mtrigger[22;0med.  

  [1mTriggers[22;0m can be disabled by turning the [1m%{borg}[22;0m flag off.  

  If the [1m%{background}[22;0m flag is turned off, text from [1mbackground[22;0m [1msockets[22;0m will 
  not cause [1mtrigger[22;0ming until that [1msocket[22;0m is brought into the [1mforeground[22;0m.  

  [1mTriggers[22;0m can also be invoked manually with the command [1m/trigger[22;0m.  The 
  command "[1m/trigger[22;0m -n" can be used to test which [1mtriggers[22;0m would match a given 
  line.  

  The [1m/def[22;0m command is the only way to define a multi-shot [1mtrigger[22;0m.  All other 
  commands which define [1mtriggers[22;0m will create permanent [1mtriggers[22;0m.  

  Note that tf may run slowly if there are many [1mtriggers[22;0m defined, since every 
  [1mtrigger[22;0m must be compared against every received line of text.  Choose your 
  [1mtriggers[22;0m carefully.  See also "[1mpatterns[22;0m".  

  [1mTriggers[22;0m are only matched against normal lines.  To have a macro invoked by 
  a [1mprompt[22;0m, use the [1mprompt[22;0m [1mhook[22;0m.  

  By default, TF expands tabs and removes ANSI display codes and other non 
  printable characters from received lines before comparing them against 
  [1mtriggers[22;0m, so your [1mtriggers[22;0m need to match only visible text.  But if you 
  change [1m%expand_tabs[22;0m or [1m%emulation[22;0m, received lines may still contain 
  invisible codes when compared against [1mtriggers[22;0m.  

  Trigger patterns are not expanded for variable substitutions or anything 
  else.  To get the effect of a variable trigger, write a macro that redefines 
  the trigger.  For example, 

      [1m/def[22;0m set_victim = \
          [1m/def[22;0m -t"[1m%{1}[22;0m has arrived." kill_victim = \
              kill [1m%%[22;0m{1}
    

  Then, to change the victim to "Bill", type "/set_victim Bill".  

  See also: [1mpatterns[22;0m, [1mmacros[22;0m, [1mgags[22;0m, [1mhilites[22;0m, [1mhooks[22;0m, [1mpriority[22;0m, [1m%max_trig[22;0m 

&util
&utils
&map
&/psh
&space_page
&/speedwalk
&tintin
&/watch
&utilities

utilities

  The library directory [1m%{TFLIBDIR}[22;0m contains many useful utility files ending 
  in ".tf".  To use any one of them, simply [1m/load[22;0m or [1m/require[22;0m the file.  For 
  example, to enable ESC-TAB completion automatically, just "[1m/require[22;0m 
  completion.tf" from your [1m.tfrc[22;0m file.  Some of the more useful files: 

  alias.tf 
          [1m/alias[22;0m, etc: create commands without '/'.  
  at.tf   [1m/at[22;0m: execute commands at a specified time.  
  filexfer.tf 
          [1m/putfile[22;0m, [1m/getfile[22;0m: transfer files to/from a mud.  
  kb-os2.tf 
          Extra default key bindings for OS/2 keyboards.  
  kbbind.tf 
          Default [1mkeybindings[22;0m.  
  kbfunc.tf 
          Macros used by kbbind.tf.  
  map.tf  Mapping commands (like tintin).  
  psh.tf  [1m/psh[22;0m: like [1m/sh[22;0m, but uses your favorite shell.  
  [1mquoter.tf[22;0m 
          Various quoting [1mmacros[22;0m.  
  rwho.tf 
          Remote WHO from a mudwho server.  
  spc-page.tf 
          Old-style SPACE key scrolling at [1m--More--[22;0m prompt.  
  spedwalk.tf 
          Single character movement (like tintin).  
  spell.tf 
          Spelling checker.  
  tick.tf 
          Diku tick counter (like tintin).  
  tintin.tf 
          tintin-like commands.  
  tr.tf   [1m/tr[22;0m: character translation 
  watch.tf 
          [1m/watch[22;0m: Watch for a particular player.  

  There are also other files, not listed here.  

  For complete instructions on any of these utilities, see the help section 
  for that topic if there is one, or read the comments at the top of each 
  file.  Sorry, I haven't gotten around to documenting them very well.  

  Note to unix users: many library files were renamed in version 3.5, but the 
  old names still work (via soft links).  

&variables
&variable

variables

  Associated commands: 
  [1m/listvar[22;0m 
          list values of [1mvariables[22;0m.  
  [1m/set[22;0m    set the value of a global [1mvariable[22;0m 
  [1m/let[22;0m    set the value of a local [1mvariable[22;0m 
  [1m/setenv[22;0m 
          set the value of an environment [1mvariable[22;0m 
  [1m/unset[22;0m  unset a [1mvariable[22;0m 
  [1m/export[22;0m 
          move an global [1mvariable[22;0m to the environment 
  [1m/edvar[22;0m  edit a [1mvariable[22;0m's value 
  [1m:= operator[22;0m 
          assign a value of any type to a [1mvariable[22;0m.  

  A TinyFugue [1mvariable[22;0m has a name and a value.  Names are case sensitive, and 
  should start with a letter and contain only letters, numbers, and 
  underscores.  A value can be a text string (including [1mdisplay attributes[22;0m), 
  integer, or real number, but some [1mspecial variables[22;0m will automatically 
  convert an assigned value to a particular type.  

  [1mVariables[22;0m may either be local, global, or exported.  Global [1mvariables[22;0m are 
  visible to all tf commands; they are defined with [1m/set[22;0m or [1m/setenv[22;0m, or 
  imported from the environment when tf starts.  Local [1mvariables[22;0m are created 
  with [1m/let[22;0m or assignment [1mexpressions[22;0m, and only exist in the scope in which 
  they were created.  Exported [1mvariables[22;0m are global [1mvariables[22;0m which are also 
  visible to subshells, so they can be used by commands [1m/sh[22;0m, the '!' option of 
  [1m/quote[22;0m, and file uncompression.  [1mVariables[22;0m are exported if they were defined 
  with [1m/setenv[22;0m, explicitly exported with [1m/export[22;0m, or imported from tf's parent 
  environment.  

  The value of a [1mvariable[22;0m can be obtained using a '%' substitution (see 
  "[1msubstitution[22;0m"), or by simply using its name in an [1mexpression[22;0m (see 
  "[1mexpressions[22;0m").  

  See "[1mspecial variables[22;0m" for a list of special variables.  

&worlds

worlds

  Associated commands: 
  [1m/addworld[22;0m 
          define a new world 
  [1m/world[22;0m  connect to a defined world 
  [1m/dc[22;0m     disconnect from a world 
  [1m/unworld[22;0m 
          undefine a world 
  [1m/purgeworld[22;0m 
          undefine a group of worlds 
  [1m/saveworld[22;0m 
          save world definitions to a file 
  [1m/loadworld[22;0m 
          load world definitions from a file 
  [1m/listworlds[22;0m 
          display world definitions 
  [1m/edworld[22;0m 
          edit a world definition 
  [1mworld_info()[22;0m 
          get world information 

#$world_name
#$world_character
#$world_password
#$world_host
#$world_port
#$world_mfile
#$world_type
#fields
  Fugue stores a list of "worlds" that it knows about.  Each world has several 
  fields associated with it: 
  name    a label used to refer to the world 
  type    an optional string for matching [1m/def -T[22;0m 
  character 
          optional login name 
  password 
          optional login password 
  host    server's internet host name, IPv4 address, or (if your platform 
          supports it) IPv6 address 
  port    server's TCP port number or name 
  mfile   optional [1mmacro[22;0m file 
  login   "1" if [1mautomatic login[22;0m is enabled for the world's [1msocket[22;0m, "0" 
          otherwise.  
  proxy   "1" if this world's [1msocket[22;0m is using a [1mproxy[22;0m, "0" otherwise 
  src     optional name or address used for client (tf) end of connection.  
  cipher  current cipher used by SSL connection to world.  

  The character name, password, and type are used by [1mautomatic login[22;0m, if the 
  [1m%{login}[22;0m flag is on.  

  The [1mmacro[22;0m file is [1mloaded[22;0m when a [1msocket[22;0m is opened to the world.  It can 
  contain any commands you want executed automatically when you connect to 
  that world.  If the flag [1m%{sockmload}[22;0m is on, this file will also be [1mloaded[22;0m 
  whenever you switch to a world with the SOCKETB and SOCKETF keys (see 
  [1msockets[22;0m, [1m/dokey[22;0m, [1mhooks[22;0m (CONNECT)).  

  World information can be accessed with the macro expansion 
  ${world_[4mfieldname[24m} or the [1mfunction[22;0m [1mworld_info[22;0m([4mworldname[24m, [4mfieldname[24m), where 
  <[4mfieldname[24m> is one of the fields described above.  

  For example: 
  [1m/eval[22;0m say I am [1m${world_character}[22;0m on [1m${world_name}[22;0m.  
  This would tell the rest of the world some stuff they probably don't care 
  about, namely the label your Fugue has assigned to the [1mcurrent[22;0m world and the 
  character name under which it logged on.  
#

  Fugue also keeps track of a world named "default", which is just a dummy 
  world with a character name and password, and optionally a [1mmacro[22;0m file.  If a 
  default world is defined, worlds without character, password, or file fields 
  will use the values from the default world.  

  See also: [1msockets[22;0m 

&