2 <title>Other Haskell utility programs</title>
3 <indexterm><primary>utilities, Haskell</primary></indexterm>
5 <para>This section describes other program(s) which we distribute,
6 that help with the Great Haskell Programming Task.</para>
8 <!-- comment: hasktags documentation losely based on that for hstags -->
10 <sect1 id ="hasktags">
11 <title>Ctags and Etags for Haskell: <command>hasktags</command></title>
12 <indexterm><primary><command>hasktags</command></primary></indexterm>
13 <indexterm><primary>CTAGS for Haskell</primary></indexterm>
15 <para><command>hasktags</command> is a very simple Haskell program that produces ctags "tags" and etags "TAGS" files for Haskell programs.</para>
17 <para>When loaded into an editor such an NEdit, Vim, or Emacs, this allows one to easily navigate around a multi-file program, finding definitions of functions, types, and constructors.</para>
19 <para>Invocation Syntax:</para>
25 <para>This will read all the files listed in <option>files</option> and produce a ctags "tags" file and an etags "TAGS" file in the current directory.</para>
27 <para>Example usage</para>
30 find -name \*.\*hs | xargs hasktags
33 <para>This will find all haskell source files in the current directory and below, and create tags files indexing them in the current directory.</para>
35 <para><command>hasktags</command> is a simple program that uses simple
36 parsing rules to find definitions of functions, constructors, and types. It isn't guranteed to find everything, and will sometimes create false index entries, but it usually gets the job done fairly well. In particular, at present, functions are only indexed if a type signature is given for them.</para>
38 <para>Before hasktags, there used to be <command>fptags</command> and <command>hstags</command>, which did essentially the same job, however neither of these seem to be maintained any more.</para>
41 <title>Using tags with your editor</title>
43 <para>With NEdit, load the "tags" file using "File/Load Tags File". Use "Ctrl-D" to search for a tag.</para>
45 <para>With XEmacs, load the "TAGS" file using "visit-tags-table". Use "M-." to search for a tag.</para>
52 <!-- comment: hstags doesn't work anymore
55 <title>Emacs `TAGS' for Haskell: <command>hstags</command></title>
56 <indexterm><primary><command>hstags</command></primary></indexterm>
57 <indexterm><primary>TAGS for Haskell</primary></indexterm>
59 <para>`Tags' is a facility for indexing the definitions of
60 programming-language things in a multi-file program, and then
61 using that index to jump around among these definitions.</para>
63 <para>Rather than scratch your head, saying “Now where did
64 we define `foo'?”, you just do (in Emacs) <Literal>M-. foo
65 RET</Literal>, and You're There! Some people go wild over this
68 <para>GHC comes with a program <command>hstags</command>, which
69 build Emacs-able TAGS files. The invocation syntax is:</para>
72 hstags [GHC-options] file [files...]
75 <para>The best thing is just to feed it your GHC command-line
76 flags. A good Makefile entry might be:</para>
81 hstags $(GHC_FLAGS) *.lhs
84 <para>The only flags of its own are: <Option>-v</Option> to be
85 verbose; <Option>-a</Option> to <Emphasis>APPEND</Emphasis> to the
86 TAGS file, rather than write to it.</para>
88 <para>Shortcomings: (1) Instance declarations don't get into
89 the TAGS file (but the definitions inside them do); as instances
90 aren't named, this is probably just as well.
91 (2) Data-constructor definitions don't get in. Go for the
92 corresponding type constructor instead.</para>
94 <para>Actually, GHC also comes with <command>etags</command>
95 [for C], and <Command>perltags</Command> [for You
96 Know What]. And—I cannot tell a lie—there is
97 Denis Howe's <Command>fptags</Command> [for Haskell,
98 etc.] in the <Filename>ghc/CONTRIB</Filename>
99 section…)</para>
105 <title>“Yacc for Haskell”: <command>happy</command></title>
107 <indexterm><primary>Happy</primary></indexterm>
108 <indexterm><primary>Yacc for Haskell</primary></indexterm>
109 <indexterm><primary>parser generator for Haskell</primary></indexterm>
111 <para>Andy Gill and Simon Marlow have written a parser-generator
113 <Command>happy</Command>.<IndexTerm><Primary>happy parser
114 generator</Primary></IndexTerm> <command>Happy</command> is to
115 Haskell what <command>Yacc</command> is to C.</para>
117 <para>You can get <Command>happy</Command> from <ulink
118 url="http://www.haskell.org/happy/">the Happy
119 Homepage</ulink>.</para>
121 <para><command>Happy</command> is at its shining best when
122 compiled by GHC.</para>
126 <!-- we don't distribute this anymore
128 <title>Pretty-printing Haskell: <Command>pphs</Command></title>
129 <indexterm><primary>pphs</primary></indexterm>
130 <indexterm><primary>pretty-printing Haskell code</primary></indexterm>
132 <para>Andrew Preece has written
133 <Command>pphs</Command>,<IndexTerm><Primary>pphs</Primary></IndexTerm><IndexTerm><Primary>pretty-printing
134 Haskell</Primary></IndexTerm> a utility to pretty-print Haskell
135 code in LaTeX documents. Keywords in bolds, variables in
136 italics—that sort of thing. It is good at lining up program
137 clauses and equals signs, things that are very tiresome to do by
140 <para>The code is distributed with GHC in
141 <Filename>ghc/CONTRIB/pphs</Filename>.</para>
146 <title>Writing Haskell interfaces to C code:
147 <command>hsc2hs</command></title>
148 <indexterm><primary><command>hsc2hs</command></primary>
151 <para>The <command>hsc2hs</command> command can be used to automate
152 some parts of the process of writing Haskell bindings to C code.
153 It reads an almost-Haskell source with embedded special
154 constructs, and outputs a real Haskell file with these constructs
155 processed, based on information taken from some C headers. The
156 extra constructs deal with accessing C data from Haskell.</para>
158 <para>It may also output a C file which contains additional C
159 functions to be linked into the program, together with a C header
160 that gets included into the C code to which the Haskell module
161 will be compiled (when compiled via C) and into the C file. These
162 two files are created when the <literal>#def</literal> construct
163 is used (see below).</para>
165 <para>Actually <command>hsc2hs</command> does not output the Haskell
166 file directly. It creates a C program that includes the headers,
167 gets automatically compiled and run. That program outputs the
170 <para>In the following, “Haskell file” is the main
171 output (usually a <literal>.hs</literal> file), “compiled
172 Haskell file” is the Haskell file after
173 <command>ghc</command> has compiled it to C (i.e. a
174 <literal>.hc</literal> file), “C program” is the
175 program that outputs the Haskell file, “C file” is the
176 optionally generated C file, and “C header” is its
180 <title>Command line syntax</title>
182 <para><command>hsc2hs</command> takes input files as arguments,
183 and flags that modify its behavior:</para>
187 <term><literal>-o FILE</literal> or
188 <literal>––output=FILE</literal></term>
190 <para>Name of the Haskell file.</para>
195 <term><literal>-t FILE</literal> or
196 <literal>––template=FILE</literal></term>
198 <para>The template file (see below).</para>
203 <term><literal>-c PROG</literal> or
204 <literal>––cc=PROG</literal></term>
206 <para>The C compiler to use (default:
207 <command>ghc</command>)</para>
212 <term><literal>-l PROG</literal> or
213 <literal>––ld=PROG</literal></term>
215 <para>The linker to use (default:
216 <command>gcc</command>).</para>
221 <term><literal>-C FLAG</literal> or
222 <literal>––cflag=FLAG</literal></term>
224 <para>An extra flag to pass to the C compiler.</para>
229 <term><literal>-I DIR</literal></term>
231 <para>Passed to the C compiler.</para>
236 <term><literal>-L FLAG</literal> or
237 <literal>––lflag=FLAG</literal></term>
239 <para>An extra flag to pass to the linker.</para>
244 <term><literal>-i FILE</literal> or
245 <literal>––include=FILE</literal></term>
247 <para>As if the appropriate <literal>#include</literal>
248 directive was placed in the source.</para>
253 <term><literal>-D NAME[=VALUE]</literal> or
254 <literal>––define=NAME[=VALUE]</literal></term>
256 <para>As if the appropriate <literal>#define</literal>
257 directive was placed in the source.</para>
262 <term><literal>––no-compile</literal></term>
264 <para>Stop after writing out the intermediate C program to disk.
265 The file name for the intermediate C program is the input file name
266 with <literal>.hsc</literal> replaced with <literal>_hsc_make.c</literal>.</para>
271 <term><literal>-?</literal> or <literal>––help</literal></term>
273 <para>Display a summary of the available flags and exit successfully.</para>
278 <term><literal>-V</literal> or <literal>––version</literal></term>
280 <para>Output version information and exit successfully.</para>
285 <para>The input file should end with .hsc (it should be plain
286 Haskell source only; literate Haskell is not supported at the
287 moment). Output files by default get names with the
288 <literal>.hsc</literal> suffix replaced:</para>
294 <entry><literal>.hs</literal></entry>
295 <entry>Haskell file</entry>
298 <entry><literal>_hsc.h</literal></entry>
299 <entry>C header</entry>
302 <entry><literal>_hsc.c</literal></entry>
303 <entry>C file</entry>
309 <para>The C program is compiled using the Haskell compiler. This
310 provides the include path to <filename>HsFFI.h</filename> which
311 is automatically included into the C program.</para>
314 <sect2><title>Input syntax</title>
316 <para>All special processing is triggered by
317 the <literal>#</literal> operator. To output
318 a literal <literal>#</literal>, write it twice:
319 <literal>##</literal>. Inside string literals and comments
320 <literal>#</literal> characters are not processed.</para>
322 <para>A <literal>#</literal> is followed by optional
323 spaces and tabs, an alphanumeric keyword that describes
324 the kind of processing, and its arguments. Arguments look
325 like C expressions separated by commas (they are not
326 written inside parens). They extend up to the nearest
327 unmatched <literal>)</literal>, <literal>]</literal> or
328 <literal>}</literal>, or to the end of line if it occurs outside
329 any <literal>() [] {} '' "" /**/</literal> and is not preceded
330 by a backslash. Backslash-newline pairs are stripped.</para>
332 <para>In addition <literal>#{stuff}</literal> is equivalent
333 to <literal>#stuff</literal> except that it's self-delimited
334 and thus needs not to be placed at the end of line or in some
337 <para>Meanings of specific keywords:</para>
342 <term><literal>#include <file.h></literal></term>
343 <term><literal>#include "file.h"</literal></term>
345 <para>The specified file gets included into the C program,
346 the compiled Haskell file, and the C header.
347 <literal><HsFFI.h></literal> is included
348 automatically.</para>
353 <term><literal>#define name</literal></term>
354 <term><literal>#define name value</literal></term>
355 <term><literal>#undef name</literal></term>
357 <para>Similar to <literal>#include</literal>. Note that
358 <literal>#includes</literal> and
359 <literal>#defines</literal> may be put in the same file
360 twice so they should not assume otherwise.</para>
365 <term><literal>#let name parameters = "definition"</literal></term>
367 <para>Defines a macro to be applied to the Haskell
368 source. Parameter names are comma-separated, not
369 inside parens. Such macro is invoked as other
370 <literal>#</literal>-constructs, starting with
371 <literal>#name</literal>. The definition will be
372 put in the C program inside parens as arguments of
373 <literal>printf</literal>. To refer to a parameter,
374 close the quote, put a parameter name and open the
375 quote again, to let C string literals concatenate.
376 Or use <literal>printf</literal>'s format directives.
377 Values of arguments must be given as strings, unless the
378 macro stringifies them itself using the C preprocessor's
379 <literal>#parameter</literal> syntax.</para>
384 <term><literal>#def C_definition</literal></term>
386 <para>The definition (of a function, variable, struct or
387 typedef) is written to the C file, and its prototype or
388 extern declaration to the C header. Inline functions are
389 handled correctly. struct definitions and typedefs are
390 written to the C program too. The
391 <literal>inline</literal>, <literal>struct</literal> or
392 <literal>typedef</literal> keyword must come just after
393 <literal>def</literal>.</para>
398 <term><literal>#if condition</literal></term>
399 <term><literal>#ifdef name</literal></term>
400 <term><literal>#ifndef name</literal></term>
401 <term><literal>#elif condition</literal></term>
402 <term><literal>#else</literal></term>
403 <term><literal>#endif</literal></term>
404 <term><literal>#error message</literal></term>
405 <term><literal>#warning message</literal></term>
407 <para>Conditional compilation directives are passed
408 unmodified to the C program, C file, and C header. Putting
409 them in the C program means that appropriate parts of the
410 Haskell file will be skipped.</para>
415 <term><literal>#const C_expression</literal></term>
417 <para>The expression must be convertible to
418 <literal>long</literal> or <literal>unsigned
419 long</literal>. Its value (literal or negated literal)
420 will be output.</para>
425 <term><literal>#const_str C_expression</literal></term>
427 <para>The expression must be convertible to const char
428 pointer. Its value (string literal) will be output.</para>
433 <term><literal>#type C_type</literal></term>
435 <para>A Haskell equivalent of the C numeric type will be
436 output. It will be one of
437 <literal>{Int,Word}{8,16,32,64}</literal>,
438 <literal>Float</literal>, <literal>Double</literal>,
439 <literal>LDouble</literal>.</para>
444 <term><literal>#peek struct_type, field</literal></term>
446 <para>A function that peeks a field of a C struct will be
447 output. It will have the type
448 <literal>Storable b => Ptr a -> IO b</literal>.
450 The intention is that <literal>#peek</literal> and
451 <literal>#poke</literal> can be used for implementing the
452 operations of class <literal>Storable</literal> for a
453 given C struct (see the
454 <literal>Foreign.Storable</literal> module in the library
455 documentation).</para>
460 <term><literal>#poke struct_type, field</literal></term>
462 <para>Similarly for poke. It will have the type
463 <literal>Storable b => Ptr a -> b -> IO ()</literal>.</para>
468 <term><literal>#ptr struct_type, field</literal></term>
470 <para>Makes a pointer to a field struct. It will have the type
471 <literal>Ptr a -> Ptr b</literal>.</para>
476 <term><literal>#offset struct_type, field</literal></term>
478 <para>Computes the offset, in bytes, of
479 <literal>field</literal> in
480 <literal>struct_type</literal>. It will have type
481 <literal>Int</literal>.</para>
486 <term><literal>#size struct_type</literal></term>
488 <para>Computes the size, in bytes, of
489 <literal>struct_type</literal>. It will have type
490 <literal>Int</literal>.</para>
495 <term><literal>#enum type, constructor, value, value, ...</literal></term>
497 <para>A shortcut for multiple definitions which use
498 <literal>#const</literal>. Each <literal>value</literal>
499 is a name of a C integer constant, e.g. enumeration value.
500 The name will be translated to Haskell by making each
501 letter following an underscore uppercase, making all the rest
502 lowercase, and removing underscores. You can supply a different
503 translation by writing <literal>hs_name = c_value</literal>
504 instead of a <literal>value</literal>, in which case
505 <literal>c_value</literal> may be an arbitrary expression.
506 The <literal>hs_name</literal> will be defined as having the
507 specified <literal>type</literal>. Its definition is the specified
508 <literal>constructor</literal> (which in fact may be an expression
509 or be empty) applied to the appropriate integer value. You can
510 have multiple <literal>#enum</literal> definitions with the same
511 <literal>type</literal>; this construct does not emit the type
512 definition itself.</para>
520 <title>Custom constructs</title>
522 <para><literal>#const</literal>, <literal>#type</literal>,
523 <literal>#peek</literal>, <literal>#poke</literal> and
524 <literal>#ptr</literal> are not hardwired into the
525 <command>hsc2hs</command>, but are defined in a C template that is
526 included in the C program: <filename>template-hsc.h</filename>.
527 Custom constructs and templates can be used too. Any
528 <literal>#</literal>-construct with unknown key is expected to
529 be handled by a C template.</para>
531 <para>A C template should define a macro or function with name
532 prefixed by <literal>hsc_</literal> that handles the construct
533 by emitting the expansion to stdout. See
534 <filename>template-hsc.h</filename> for examples.</para>
536 <para>Such macros can also be defined directly in the
537 source. They are useful for making a <literal>#let</literal>-like
538 macro whose expansion uses other <literal>#let</literal> macros.
539 Plain <literal>#let</literal> prepends <literal>hsc_</literal>
540 to the macro name and wraps the defininition in a
541 <literal>printf</literal> call.</para>
550 ;;; Local Variables: ***
552 ;;; sgml-parent-document: ("users_guide.sgml" "book" "chapter") ***