File: manual.tex

package info (click to toggle)
curry-tools 1.0.1%2Bdfsg1-1
  • links: PTS, VCS
  • area: main
  • in suites: stretch
  • size: 5,492 kB
  • ctags: 121
  • sloc: makefile: 470; sh: 421
file content (127 lines) | stat: -rw-r--r-- 4,073 bytes parent folder | download
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
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
\section{\texttt{runcurry}: Running Curry Programs}

\code{runcurry}\pindex{runcurry}
is a command usually stored in \code{\cyshome/bin}
(where \cyshome is the installation directory of \CYS;
see Section~\ref{sec-general}).
This command supports the execution of Curry programs
without explicitly invoking the interactive environment.
Hence, it can be useful to write short scripts in Curry
intended for direct execution.
The Curry program must always contain the definition of an operation
\code{main} of type \code{IO ()}.
The execution of the program consists of the evaluation of this operation.

Basically, the command \code{runcurry} supports three modes of operation:
\begin{itemize}
\item
One can execute a Curry program whose file name
is provided as an argument when \code{runcurry} is called.
In this case, the suffix (\ccode{.curry} or \ccode{.lcurry})
must be present and cannot be dropped.
One can write additional commands for the interactive environment,
typically settings of some options, before the Curry program name.
All arguments after the Curry program name are passed as run-time
arguments. For instance, consider the following program stored
in the file \code{ShowArgs.curry}:
\begin{curry}
import System(getArgs)

main = getArgs >>= print
\end{curry}
This program can be executed by the shell command
\begin{curry}
> runcurry ShowArgs.curry Hello World!
\end{curry}
which produces the output
\begin{curry}
["Hello","World!"]
\end{curry}

\item
One can also execute a Curry program whose program text
comes from the standard input. Thus, one can either ``pipe''
the program text into this command or type the program text on
the keyboard. For instance, if we type
\begin{currynomath}
> runcurry
main = putStr . unlines . map show . take 8 $ [1..]
\end{currynomath} % $
(followed by the end-of-file marker Ctrl-D), the output
\begin{curry}
1
2
3
4
5
6
7
8
\end{curry}
is produced.

\item
One can also write the program text in a script file to be executed
like a shell script. In this case, the script must start with the line
\begin{curry}
#!/usr/bin/env runcurry
\end{curry}
followed by the source text of the Curry program.
For instance, we can write a simple Curry script to count the
number of code lines in a Curry program by removing all blank
and comment lines and counting the remaining lines:
\begin{currynomath}
#!/usr/bin/env runcurry

import Char(isSpace)
import System(getArgs)

-- count number of program lines in a file:
countCLines :: String -> IO Int
countCLines f =
  readFile f >>=
  return . length . filter (not . isEmptyLine) . map stripSpaces . lines
 where
  stripSpaces = reverse . dropWhile isSpace . reverse . dropWhile isSpace

  isEmptyLine []        = True
  isEmptyLine [_]       = False
  isEmptyLine (c1:c2:_) = c1=='-' && c2=='-'

-- The main program reads Curry file names from arguments:
main = do
  args <- getArgs
  mapIO_ (\f -> do ls <- countCLines f
                   putStrLn $ "Stripped lines of file "++f++": " ++ show ls)
         args
\end{currynomath} % $
%
If this script is stored in the (executable) file \ccode{codelines.sh},
we can count the code lines of the file \code{Prog.curry} by
the shell command
\begin{curry}
> ./codelines.sh Prog.curry
\end{curry}
%
When this command is executed, the command \code{runcurry}
compiles the program and evaluates the expression \code{main}.
Since the compilation might take some time in more complex scripts,
one can also save the result of the compilation in a binary file.
To obtain this behavior, one has to insert the line
\begin{curry}
#jit
\end{curry}
in the script file, e.g., in the second line.
With this option, a binary of the compiled program is saved
(in the same directory as the script).
Now, when the same script is executed the next time,
the stored binary file is executed (provided that it is still newer
than the script file itself, otherwise it will be recompiled).
This feature combines easy scripting with Curry together with
fast execution.
\end{itemize}

%%% Local Variables:
%%% mode: pdflatex
%%% TeX-master: "manual"
%%% End: