File: README

package info (click to toggle)
horae 066-3
  • links: PTS
  • area: contrib
  • in suites: lenny
  • size: 17,272 kB
  • ctags: 1,656
  • sloc: perl: 64,718; lisp: 744; sh: 82; makefile: 67; ansic: 35
file content (107 lines) | stat: -rw-r--r-- 4,750 bytes parent folder | download | duplicates (7)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
Bruce Ravel
<ravel@phys.washington.edu>
Time-stamp: <2001/01/18 21:43:55 bruce>


This is the README file for atoms.cgi -- the version of Atoms intended
to run on the web.  There is no automatic installation procedure
because there are several things that the web administrator must take
care of which depend on the specific configuration of the server.  It
seems easier to me to just have you directly edit atoms.cgi rather
than use some silly little interactive configuration tool.

1. Make sure that Atoms is installed on the server and that it works.
   In fact, it is not important that atoms.pl or tkatoms.pl actually
   be installed.  atoms.cgi does not call either of those programs.
   What is important is that all of Atoms's library files be installed
   correctly.  If you can run atoms.pl or tkatoms.pl, then all of the
   library files are well installed.  If you have not installed Atoms,
   you should go ahead and do so.  Don't forget to install the
   AtomsBundle (on NT the files in the AtomsBundle are included with
   the Atoms distribution itself).

2. Take a look at the file atoms.cgi.  Near the beginning are some
   variables that need to be set appropriately for your web server.
   They are:

	$extra_INC	if Atoms is installed somewhere outside of the 
			default search path for perl modules, then set 
			this variable to that location.  The variable
			is unshifted to @INC when atoms.cgi starts

	$atoms_help	this takes the URL to the place where you will 
			install the file WebAtoms_help.html.

        $ADB_search     the URL of the Atoms Database search form

	$ADB_directory  this takes the actual location (not the
	                location relative to the server's root) on the
	                server disk where input files may be found.
	                The purpose of this is so that atoms.cgi can
	                be on the back end of a search CGI script.

3. Put atoms.cgi in a place where your web server knows to look for
   CGI programs.  Make sure it is executable.

4. Put WebAtoms_help.html in the place indicated by the $atoms_help
   variable.  You can use the copy on my web page
	http://feff.phys.washington.edu/~ravel/software/atoms/doc/WebAtoms_help.html
   but I would rather that you install a copy on your own server.

5. Run atoms.cgi once by hand from the command line like this:
        atoms.cgi -install
   This sets up the set of hard links to the script which are used 
   to serve up the correct file names when the client clicks the
   "save-as" button.

6. If you encounter problems, there are two things that you could
   include in a bug report that would greatly improve my chances of
   solving the problem.  (1) Any relevant lines from the server log
   and/or error files.  (2) Save the page where you entered input data 
   as a flat text file and mail me the text file.


Calling this CGI script directly will serve up a page with a blank
fill-in form.  You can import data from a file in $ADB_directory with
a URL like this:
        http:/what.ever/path/to/atoms.cgi?file=foo.inp
In that case, the form will be filled in with data from the specified
input file.

This CGI script has only been tested with Apache.  Please let me know
how things work with other servers.

----------------------------------------------------------------------------

Here is how atoms.cgi is intended to work:

1.  First time called it should display a fill-out form for entering
    all the data.  The page has this structure:

      +-----------------------+
      |  header information   |
      +-----------------------+
      |     warning area      |
      +-----------------------+
      |      data form        |
      +-----------------------+
      |  trailer information  |
      +-----------------------+

    The warning area should be blank initially.  The data form should
    be empty, unless data import has happened, in which case many of
    the entry widgets will be filled.  The header contains the html
    headers, and a brief explanation of the page.  The trailer
    contains copyright information, some links, and closing tags.  The 
    form has lots of links to a help file on it.

2.  If any trapable errors are encountered, the form will be
    redisplayed with state completely saved.  The warning area will be 
    filled with red text explaining the problems.

3.  If the calculation can proceed, then the browser will be
    redirected to the appropriate symbolic link with the complete
    query list in the URL.  This will display the chosen output as
    text/plain and the default "save-as" name will be appropriate.  If
    there were any warnings that did not preclude continuing the
    calculation, they are written to the top of the output file.