OpenCores
URL https://opencores.org/ocsvn/or1k/or1k/trunk

Subversion Repositories or1k

[/] [or1k/] [trunk/] [gdb-5.3/] [etc/] [standards.texi] - Blame information for rev 1765

Details | Compare with Previous | View Log

Line No. Rev Author Line
1 1181 sfurman
\input texinfo @c -*-texinfo-*-
2
@c %**start of header
3
@setfilename standards.info
4
@settitle GNU Coding Standards
5
@c This date is automagically updated when you save this file:
6
@set lastupdate February 14, 2002
7
@c %**end of header
8
 
9
@ifinfo
10
@format
11
START-INFO-DIR-ENTRY
12
* Standards: (standards).        GNU coding standards.
13
END-INFO-DIR-ENTRY
14
@end format
15
@end ifinfo
16
 
17
@c @setchapternewpage odd
18
@setchapternewpage off
19
 
20
@c Put everything in one index (arbitrarily chosen to be the concept index).
21
@syncodeindex fn cp
22
@syncodeindex ky cp
23
@syncodeindex pg cp
24
@syncodeindex vr cp
25
 
26
@c This is used by a cross ref in make-stds.texi
27
@set CODESTD  1
28
@iftex
29
@set CHAPTER chapter
30
@end iftex
31
@ifinfo
32
@set CHAPTER node
33
@end ifinfo
34
 
35
@ifinfo
36
GNU Coding Standards
37
Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001 Free Software Foundation, Inc.
38
 
39
Permission is granted to copy, distribute and/or modify this document
40
under the terms of the GNU Free Documentation License, Version 1.1
41
or any later version published by the Free Software Foundation;
42
with no Invariant Sections, with no
43
Front-Cover Texts, and with no Back-Cover Texts.
44
A copy of the license is included in the section entitled ``GNU
45
Free Documentation License''.
46
@end ifinfo
47
 
48
@titlepage
49
@title GNU Coding Standards
50
@author Richard Stallman, et al.
51
@author last updated @value{lastupdate}
52
@page
53
 
54
@vskip 0pt plus 1filll
55
Copyright @copyright{} 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001 Free Software Foundation, Inc.
56
 
57
Permission is granted to copy, distribute and/or modify this document
58
under the terms of the GNU Free Documentation License, Version 1.1
59
or any later version published by the Free Software Foundation;
60
with no Invariant Sections, with no
61
Front-Cover Texts, and with no Back-Cover Texts.
62
A copy of the license is included in the section entitled ``GNU
63
Free Documentation License''.
64
@end titlepage
65
 
66
@ifinfo
67
@node Top, Preface, (dir), (dir)
68
@top Version
69
 
70
Last updated @value{lastupdate}.
71
@end ifinfo
72
 
73
@menu
74
* Preface::                     About the GNU Coding Standards
75
* Legal Issues::                Keeping Free Software Free
76
* Design Advice::               General Program Design
77
* Program Behavior::            Program Behavior for All Programs
78
* Writing C::                   Making The Best Use of C
79
* Documentation::               Documenting Programs
80
* Managing Releases::           The Release Process
81
* References::                  References to Non-Free Software or Documentation
82
* Copying This Manual::         How to Make Copies of This Manual
83
* Index::
84
 
85
@end menu
86
 
87
@node Preface
88
@chapter About the GNU Coding Standards
89
 
90
The GNU Coding Standards were written by Richard Stallman and other GNU
91
Project volunteers.  Their purpose is to make the GNU system clean,
92
consistent, and easy to install.  This document can also be read as a
93
guide to writing portable, robust and reliable programs.  It focuses on
94
programs written in C, but many of the rules and principles are useful
95
even if you write in another programming language.  The rules often
96
state reasons for writing in a certain way.
97
 
98
This release of the GNU Coding Standards was last updated
99
@value{lastupdate}.
100
 
101
@cindex where to obtain @code{standards.texi}
102
@cindex downloading this manual
103
If you did not obtain this file directly from the GNU project and
104
recently, please check for a newer version.  You can ftp the GNU
105
Coding Standards from any GNU FTP host in the directory
106
@file{/pub/gnu/standards/}.  The GNU Coding Standards are available
107
there in several different formats: @file{standards.text},
108
@file{standards.info}, and @file{standards.dvi}, as well as the
109
Texinfo ``source'' which is divided in two files:
110
@file{standards.texi} and @file{make-stds.texi}.  The GNU Coding
111
Standards are also available on the GNU World Wide Web server:
112
@uref{http://www.gnu.org/prep/standards_toc.html}.
113
 
114
Corrections or suggestions for this document should be sent to
115
@email{bug-standards@@gnu.org}.  If you make a suggestion, please include a
116
suggested new wording for it; our time is limited.  We prefer a context
117
diff to the @file{standards.texi} or @file{make-stds.texi} files, but if
118
you don't have those files, please mail your suggestion anyway.
119
 
120
These standards cover the minimum of what is important when writing a
121
GNU package.  Likely, the needs for additional standards will come up.
122
Sometimes, you might suggest that such standards be added to this
123
document.  If you think your standards would be generally useful, please
124
do suggest them.
125
 
126
You should also set standards for your package on many questions not
127
addressed or not firmly specified here.  The most important point is to
128
be self-consistent---try to stick to the conventions you pick, and try
129
to document them as much as possible.  That way, your program will be
130
more maintainable by others.
131
 
132
@node Legal Issues
133
@chapter Keeping Free Software Free
134
@cindex legal aspects
135
 
136
This @value{CHAPTER} discusses how you can make sure that GNU software
137
avoids legal difficulties, and other related issues.
138
 
139
@menu
140
* Reading Non-Free Code::       Referring to Proprietary Programs
141
* Contributions::               Accepting Contributions
142
* Trademarks::                  How We Deal with Trademark Issues
143
@end menu
144
 
145
@node Reading Non-Free Code
146
@section Referring to Proprietary Programs
147
@cindex proprietary programs
148
@cindex avoiding proprietary code
149
 
150
Don't in any circumstances refer to Unix source code for or during
151
your work on GNU!  (Or to any other proprietary programs.)
152
 
153
If you have a vague recollection of the internals of a Unix program,
154
this does not absolutely mean you can't write an imitation of it, but
155
do try to organize the imitation internally along different lines,
156
because this is likely to make the details of the Unix version
157
irrelevant and dissimilar to your results.
158
 
159
For example, Unix utilities were generally optimized to minimize
160
memory use; if you go for speed instead, your program will be very
161
different.  You could keep the entire input file in core and scan it
162
there instead of using stdio.  Use a smarter algorithm discovered more
163
recently than the Unix program.  Eliminate use of temporary files.  Do
164
it in one pass instead of two (we did this in the assembler).
165
 
166
Or, on the contrary, emphasize simplicity instead of speed.  For some
167
applications, the speed of today's computers makes simpler algorithms
168
adequate.
169
 
170
Or go for generality.  For example, Unix programs often have static
171
tables or fixed-size strings, which make for arbitrary limits; use
172
dynamic allocation instead.  Make sure your program handles NULs and
173
other funny characters in the input files.  Add a programming language
174
for extensibility and write part of the program in that language.
175
 
176
Or turn some parts of the program into independently usable libraries.
177
Or use a simple garbage collector instead of tracking precisely when
178
to free memory, or use a new GNU facility such as obstacks.
179
 
180
@node Contributions
181
@section Accepting Contributions
182
@cindex legal papers
183
@cindex accepting contributions
184
 
185
If the program you are working on is copyrighted by the Free Software
186
Foundation, then when someone else sends you a piece of code to add to
187
the program, we need legal papers to use it---just as we asked you to
188
sign papers initially.  @emph{Each} person who makes a nontrivial
189
contribution to a program must sign some sort of legal papers in order
190
for us to have clear title to the program; the main author alone is not
191
enough.
192
 
193
So, before adding in any contributions from other people, please tell
194
us, so we can arrange to get the papers.  Then wait until we tell you
195
that we have received the signed papers, before you actually use the
196
contribution.
197
 
198
This applies both before you release the program and afterward.  If
199
you receive diffs to fix a bug, and they make significant changes, we
200
need legal papers for that change.
201
 
202
This also applies to comments and documentation files.  For copyright
203
law, comments and code are just text.  Copyright applies to all kinds of
204
text, so we need legal papers for all kinds.
205
 
206
We know it is frustrating to ask for legal papers; it's frustrating for
207
us as well.  But if you don't wait, you are going out on a limb---for
208
example, what if the contributor's employer won't sign a disclaimer?
209
You might have to take that code out again!
210
 
211
You don't need papers for changes of a few lines here or there, since
212
they are not significant for copyright purposes.  Also, you don't need
213
papers if all you get from the suggestion is some ideas, not actual code
214
which you use.  For example, if someone send you one implementation, but
215
you write a different implementation of the same idea, you don't need to
216
get papers.
217
 
218
The very worst thing is if you forget to tell us about the other
219
contributor.  We could be very embarrassed in court some day as a
220
result.
221
 
222
We have more detailed advice for maintainers of programs; if you have
223
reached the stage of actually maintaining a program for GNU (whether
224
released or not), please ask us for a copy.
225
 
226
@node Trademarks
227
@section Trademarks
228
@cindex trademarks
229
 
230
Please do not include any trademark acknowledgements in GNU software
231
packages or documentation.
232
 
233
Trademark acknowledgements are the statements that such-and-such is a
234
trademark of so-and-so.  The GNU Project has no objection to the basic
235
idea of trademarks, but these acknowledgements feel like kowtowing, so
236
we don't use them.  There is no legal requirement for them.
237
 
238
What is legally required, as regards other people's trademarks, is to
239
avoid using them in ways which a reader might read as naming or labeling
240
our own programs or activities.  For example, since ``Objective C'' is
241
(or at least was) a trademark, we made sure to say that we provide a
242
``compiler for the Objective C language'' rather than an ``Objective C
243
compiler''.  The latter is meant to be short for the former, but it does
244
not explicitly state the relationship, so it could be misinterpreted as
245
using ``Objective C'' as a label for the compiler rather than for the
246
language.
247
 
248
@node Design Advice
249
@chapter General Program Design
250
@cindex program design
251
 
252
This @value{CHAPTER} discusses some of the issues you should take into
253
account when designing your program.
254
 
255
@c                         Standard or ANSI C
256
@c
257
@c In 1989 the American National Standards Institute (ANSI) standardized
258
@c C   as  standard  X3.159-1989.    In  December   of  that   year  the
259
@c International Standards Organization ISO  adopted the ANSI C standard
260
@c making  minor changes.   In 1990  ANSI then  re-adopted  ISO standard
261
@c C. This version of C is known as either ANSI C or Standard C.
262
 
263
@c A major revision of the C Standard appeared in 1999.
264
 
265
@menu
266
* Source Language::             Which languges to use.
267
* Compatibility::               Compatibility with other implementations
268
* Using Extensions::            Using non-standard features
269
* Standard C::                  Using Standard C features
270
* Conditional Compilation::     Compiling Code Only If A Conditional is True
271
@end menu
272
 
273
@node Source Language
274
@section Which Languages to Use
275
@cindex programming languges
276
 
277
When you want to use a language that gets compiled and runs at high
278
speed, the best language to use is C.  Using another language is like
279
using a non-standard feature: it will cause trouble for users.  Even if
280
GCC supports the other language, users may find it inconvenient to have
281
to install the compiler for that other language in order to build your
282
program.  For example, if you write your program in C++, people will
283
have to install the GNU C++ compiler in order to compile your program.
284
 
285
C has one other advantage over C++ and other compiled languages: more
286
people know C, so more people will find it easy to read and modify the
287
program if it is written in C.
288
 
289
So in general it is much better to use C, rather than the
290
comparable alternatives.
291
 
292
But there are two exceptions to that conclusion:
293
 
294
@itemize @bullet
295
@item
296
It is no problem to use another language to write a tool specifically
297
intended for use with that language.  That is because the only people
298
who want to build the tool will be those who have installed the other
299
language anyway.
300
 
301
@item
302
If an application is of interest only to a narrow part of the community,
303
then the question of which language it is written in has less effect on
304
other people, so you may as well please yourself.
305
@end itemize
306
 
307
Many programs are designed to be extensible: they include an interpreter
308
for a language that is higher level than C.  Often much of the program
309
is written in that language, too.  The Emacs editor pioneered this
310
technique.
311
 
312
@cindex GUILE
313
The standard extensibility interpreter for GNU software is GUILE, which
314
implements the language Scheme (an especially clean and simple dialect
315
of Lisp).  @uref{http://www.gnu.org/software/guile/}.  We don't reject
316
programs written in other ``scripting languages'' such as Perl and
317
Python, but using GUILE is very important for the overall consistency of
318
the GNU system.
319
 
320
@node Compatibility
321
@section Compatibility with Other Implementations
322
@cindex compatibility with C and @sc{posix} standards
323
@cindex @sc{posix} compatibility
324
 
325
With occasional exceptions, utility programs and libraries for GNU
326
should be upward compatible with those in Berkeley Unix, and upward
327
compatible with Standard C if Standard C specifies their
328
behavior, and upward compatible with @sc{posix} if @sc{posix} specifies
329
their behavior.
330
 
331
When these standards conflict, it is useful to offer compatibility
332
modes for each of them.
333
 
334
@cindex options for compatibility
335
Standard C and @sc{posix} prohibit many kinds of extensions.  Feel
336
free to make the extensions anyway, and include a @samp{--ansi},
337
@samp{--posix}, or @samp{--compatible} option to turn them off.
338
However, if the extension has a significant chance of breaking any real
339
programs or scripts, then it is not really upward compatible.  So you
340
should try to redesign its interface to make it upward compatible.
341
 
342
@cindex @code{POSIXLY_CORRECT}, environment variable
343
Many GNU programs suppress extensions that conflict with @sc{posix} if the
344
environment variable @code{POSIXLY_CORRECT} is defined (even if it is
345
defined with a null value).  Please make your program recognize this
346
variable if appropriate.
347
 
348
When a feature is used only by users (not by programs or command
349
files), and it is done poorly in Unix, feel free to replace it
350
completely with something totally different and better.  (For example,
351
@code{vi} is replaced with Emacs.)  But it is nice to offer a compatible
352
feature as well.  (There is a free @code{vi} clone, so we offer it.)
353
 
354
Additional useful features are welcome regardless of whether
355
there is any precedent for them.
356
 
357
@node Using Extensions
358
@section Using Non-standard Features
359
@cindex non-standard extensions
360
 
361
Many GNU facilities that already exist support a number of convenient
362
extensions over the comparable Unix facilities.  Whether to use these
363
extensions in implementing your program is a difficult question.
364
 
365
On the one hand, using the extensions can make a cleaner program.
366
On the other hand, people will not be able to build the program
367
unless the other GNU tools are available.  This might cause the
368
program to work on fewer kinds of machines.
369
 
370
With some extensions, it might be easy to provide both alternatives.
371
For example, you can define functions with a ``keyword'' @code{INLINE}
372
and define that as a macro to expand into either @code{inline} or
373
nothing, depending on the compiler.
374
 
375
In general, perhaps it is best not to use the extensions if you can
376
straightforwardly do without them, but to use the extensions if they
377
are a big improvement.
378
 
379
An exception to this rule are the large, established programs (such as
380
Emacs) which run on a great variety of systems.  Using GNU extensions in
381
such programs would make many users unhappy, so we don't do that.
382
 
383
Another exception is for programs that are used as part of compilation:
384
anything that must be compiled with other compilers in order to
385
bootstrap the GNU compilation facilities.  If these require the GNU
386
compiler, then no one can compile them without having them installed
387
already.  That would be extremely troublesome in certain cases.
388
 
389
@node Standard C
390
@section Standard C and Pre-Standard C
391
@cindex @sc{ansi} C standard
392
 
393
1989 Standard C is widespread enough now that it is ok to use its
394
features in new programs.  There is one exception: do not ever use the
395
``trigraph'' feature of Standard C.
396
 
397
1999 Standard C is not widespread yet, so please do not require its
398
features in programs.  It is ok to use its features if they are present.
399
 
400
However, it is easy to support pre-standard compilers in most programs,
401
so if you know how to do that, feel free.  If a program you are
402
maintaining has such support, you should try to keep it working.
403
 
404
@cindex function prototypes
405
To support pre-standard C, instead of writing function definitions in
406
standard prototype form,
407
 
408
@example
409
int
410
foo (int x, int y)
411
@dots{}
412
@end example
413
 
414
@noindent
415
write the definition in pre-standard style like this,
416
 
417
@example
418
int
419
foo (x, y)
420
     int x, y;
421
@dots{}
422
@end example
423
 
424
@noindent
425
and use a separate declaration to specify the argument prototype:
426
 
427
@example
428
int foo (int, int);
429
@end example
430
 
431
You need such a declaration anyway, in a header file, to get the benefit
432
of prototypes in all the files where the function is called.  And once
433
you have the declaration, you normally lose nothing by writing the
434
function definition in the pre-standard style.
435
 
436
This technique does not work for integer types narrower than @code{int}.
437
If you think of an argument as being of a type narrower than @code{int},
438
declare it as @code{int} instead.
439
 
440
There are a few special cases where this technique is hard to use.  For
441
example, if a function argument needs to hold the system type
442
@code{dev_t}, you run into trouble, because @code{dev_t} is shorter than
443
@code{int} on some machines; but you cannot use @code{int} instead,
444
because @code{dev_t} is wider than @code{int} on some machines.  There
445
is no type you can safely use on all machines in a non-standard
446
definition.  The only way to support non-standard C and pass such an
447
argument is to check the width of @code{dev_t} using Autoconf and choose
448
the argument type accordingly.  This may not be worth the trouble.
449
 
450
In order to support pre-standard compilers that do not recognize
451
prototypes, you may want to use a preprocessor macro like this:
452
 
453
@example
454
/* Declare the prototype for a general external function.  */
455
#if defined (__STDC__) || defined (WINDOWSNT)
456
#define P_(proto) proto
457
#else
458
#define P_(proto) ()
459
#endif
460
@end example
461
 
462
@node Conditional Compilation
463
@section Conditional Compilation
464
 
465
When supporting configuration options already known when building your
466
program we prefer using @code{if (... )} over conditional compilation,
467
as in the former case the compiler is able to perform more extensive
468
checking of all possible code paths.
469
 
470
For example, please write
471
 
472
@smallexample
473
  if (HAS_FOO)
474
    ...
475
  else
476
    ...
477
@end smallexample
478
 
479
instead of:
480
 
481
@smallexample
482
  #ifdef HAS_FOO
483
    ...
484
  #else
485
    ...
486
  #endif
487
@end smallexample
488
 
489
A modern compiler such as GCC will generate exactly the same code in
490
both cases, and we have been using similar techniques with good success
491
in several projects.
492
 
493
While this is not a silver bullet solving all portability problems,
494
following this policy would have saved the GCC project alone many person
495
hours if not days per year.
496
 
497
In the case of function-like macros like @code{REVERSIBLE_CC_MODE} in
498
GCC which cannot be simply used in @code{if( ...)} statements, there is
499
an easy workaround.  Simply introduce another macro
500
@code{HAS_REVERSIBLE_CC_MODE} as in the following example:
501
 
502
@smallexample
503
  #ifdef REVERSIBLE_CC_MODE
504
  #define HAS_REVERSIBLE_CC_MODE 1
505
  #else
506
  #define HAS_REVERSIBLE_CC_MODE 0
507
  #endif
508
@end smallexample
509
 
510
@node Program Behavior
511
@chapter Program Behavior for All Programs
512
 
513
This @value{CHAPTER} describes conventions for writing robust
514
software.  It also describes general standards for error messages, the
515
command line interface, and how libraries should behave.
516
 
517
@menu
518
* Semantics::                   Writing robust programs
519
* Libraries::                   Library behavior
520
* Errors::                      Formatting error messages
521
* User Interfaces::             Standards about interfaces generally
522
* Graphical Interfaces::        Standards for graphical interfaces
523
* Command-Line Interfaces::     Standards for command line interfaces
524
* Option Table::                Table of long options
525
* Memory Usage::                When and how to care about memory needs
526
* File Usage::                  Which files to use, and where
527
@end menu
528
 
529
@node Semantics
530
@section Writing Robust Programs
531
 
532
@cindex arbitrary limits on data
533
Avoid arbitrary limits on the length or number of @emph{any} data
534
structure, including file names, lines, files, and symbols, by allocating
535
all data structures dynamically.  In most Unix utilities, ``long lines
536
are silently truncated''.  This is not acceptable in a GNU utility.
537
 
538
@cindex @code{NUL} characters
539
Utilities reading files should not drop NUL characters, or any other
540
nonprinting characters @emph{including those with codes above 0177}.
541
The only sensible exceptions would be utilities specifically intended
542
for interface to certain types of terminals or printers
543
that can't handle those characters.
544
Whenever possible, try to make programs work properly with
545
sequences of bytes that represent multibyte characters, using encodings
546
such as UTF-8 and others.
547
 
548
@cindex error messages
549
Check every system call for an error return, unless you know you wish to
550
ignore errors.  Include the system error text (from @code{perror} or
551
equivalent) in @emph{every} error message resulting from a failing
552
system call, as well as the name of the file if any and the name of the
553
utility.  Just ``cannot open foo.c'' or ``stat failed'' is not
554
sufficient.
555
 
556
@cindex @code{malloc} return value
557
@cindex memory allocation failure
558
Check every call to @code{malloc} or @code{realloc} to see if it
559
returned zero.  Check @code{realloc} even if you are making the block
560
smaller; in a system that rounds block sizes to a power of 2,
561
@code{realloc} may get a different block if you ask for less space.
562
 
563
In Unix, @code{realloc} can destroy the storage block if it returns
564
zero.  GNU @code{realloc} does not have this bug: if it fails, the
565
original block is unchanged.  Feel free to assume the bug is fixed.  If
566
you wish to run your program on Unix, and wish to avoid lossage in this
567
case, you can use the GNU @code{malloc}.
568
 
569
You must expect @code{free} to alter the contents of the block that was
570
freed.  Anything you want to fetch from the block, you must fetch before
571
calling @code{free}.
572
 
573
If @code{malloc} fails in a noninteractive program, make that a fatal
574
error.  In an interactive program (one that reads commands from the
575
user), it is better to abort the command and return to the command
576
reader loop.  This allows the user to kill other processes to free up
577
virtual memory, and then try the command again.
578
 
579
@cindex command-line arguments, decoding
580
Use @code{getopt_long} to decode arguments, unless the argument syntax
581
makes this unreasonable.
582
 
583
When static storage is to be written in during program execution, use
584
explicit C code to initialize it.  Reserve C initialized declarations
585
for data that will not be changed.
586
@c ADR: why?
587
 
588
Try to avoid low-level interfaces to obscure Unix data structures (such
589
as file directories, utmp, or the layout of kernel memory), since these
590
are less likely to work compatibly.  If you need to find all the files
591
in a directory, use @code{readdir} or some other high-level interface.
592
These are supported compatibly by GNU.
593
 
594
@cindex signal handling
595
The preferred signal handling facilities are the BSD variant of
596
@code{signal}, and the @sc{posix} @code{sigaction} function; the
597
alternative USG @code{signal} interface is an inferior design.
598
 
599
Nowadays, using the @sc{posix} signal functions may be the easiest way
600
to make a program portable.  If you use @code{signal}, then on GNU/Linux
601
systems running GNU libc version 1, you should include
602
@file{bsd/signal.h} instead of @file{signal.h}, so as to get BSD
603
behavior.  It is up to you whether to support systems where
604
@code{signal} has only the USG behavior, or give up on them.
605
 
606
@cindex impossible conditions
607
In error checks that detect ``impossible'' conditions, just abort.
608
There is usually no point in printing any message.  These checks
609
indicate the existence of bugs.  Whoever wants to fix the bugs will have
610
to read the source code and run a debugger.  So explain the problem with
611
comments in the source.  The relevant data will be in variables, which
612
are easy to examine with the debugger, so there is no point moving them
613
elsewhere.
614
 
615
Do not use a count of errors as the exit status for a program.
616
@emph{That does not work}, because exit status values are limited to 8
617
bits (0 through 255).  A single run of the program might have 256
618
errors; if you try to return 256 as the exit status, the parent process
619
will see 0 as the status, and it will appear that the program succeeded.
620
 
621
@cindex temporary files
622
@cindex @code{TMPDIR} environment variable
623
If you make temporary files, check the @code{TMPDIR} environment
624
variable; if that variable is defined, use the specified directory
625
instead of @file{/tmp}.
626
 
627
In addition, be aware that there is a possible security problem when
628
creating temporary files in world-writable directories.  In C, you can
629
avoid this problem by creating temporary files in this manner:
630
 
631
@example
632
fd = open(filename, O_WRONLY | O_CREAT | O_EXCL, 0600);
633
@end example
634
 
635
@noindent
636
or by using the @code{mkstemps} function from libiberty.
637
 
638
In bash, use @code{set -C} to avoid this problem.
639
 
640
@node Libraries
641
@section Library Behavior
642
@cindex libraries
643
 
644
Try to make library functions reentrant.  If they need to do dynamic
645
storage allocation, at least try to avoid any nonreentrancy aside from
646
that of @code{malloc} itself.
647
 
648
Here are certain name conventions for libraries, to avoid name
649
conflicts.
650
 
651
Choose a name prefix for the library, more than two characters long.
652
All external function and variable names should start with this
653
prefix.  In addition, there should only be one of these in any given
654
library member.  This usually means putting each one in a separate
655
source file.
656
 
657
An exception can be made when two external symbols are always used
658
together, so that no reasonable program could use one without the
659
other; then they can both go in the same file.
660
 
661
External symbols that are not documented entry points for the user
662
should have names beginning with @samp{_}.  The @samp{_} should be
663
followed by the chosen name prefix for the library, to prevent
664
collisions with other libraries.  These can go in the same files with
665
user entry points if you like.
666
 
667
Static functions and variables can be used as you like and need not
668
fit any naming convention.
669
 
670
@node Errors
671
@section Formatting Error Messages
672
@cindex formatting error messages
673
@cindex error messages, formatting
674
 
675
Error messages from compilers should look like this:
676
 
677
@example
678
@var{source-file-name}:@var{lineno}: @var{message}
679
@end example
680
 
681
@noindent
682
If you want to mention the column number, use this format:
683
 
684
@example
685
@var{source-file-name}:@var{lineno}:@var{column}: @var{message}
686
@end example
687
 
688
@noindent
689
Line numbers should start from 1 at the beginning of the file, and
690
column numbers should start from 1 at the beginning of the line.  (Both
691
of these conventions are chosen for compatibility.)  Calculate column
692
numbers assuming that space and all ASCII printing characters have
693
equal width, and assuming tab stops every 8 columns.
694
 
695
Error messages from other noninteractive programs should look like this:
696
 
697
@example
698
@var{program}:@var{source-file-name}:@var{lineno}: @var{message}
699
@end example
700
 
701
@noindent
702
when there is an appropriate source file, or like this:
703
 
704
@example
705
@var{program}: @var{message}
706
@end example
707
 
708
@noindent
709
when there is no relevant source file.
710
 
711
If you want to mention the column number, use this format:
712
 
713
@example
714
@var{program}:@var{source-file-name}:@var{lineno}:@var{column}: @var{message}
715
@end example
716
 
717
In an interactive program (one that is reading commands from a
718
terminal), it is better not to include the program name in an error
719
message.  The place to indicate which program is running is in the
720
prompt or with the screen layout.  (When the same program runs with
721
input from a source other than a terminal, it is not interactive and
722
would do best to print error messages using the noninteractive style.)
723
 
724
The string @var{message} should not begin with a capital letter when
725
it follows a program name and/or file name.  Also, it should not end
726
with a period.
727
 
728
Error messages from interactive programs, and other messages such as
729
usage messages, should start with a capital letter.  But they should not
730
end with a period.
731
 
732
@node User Interfaces
733
@section Standards for Interfaces Generally
734
 
735
@cindex program name and its behavior
736
@cindex behavior, dependent on program's name
737
Please don't make the behavior of a utility depend on the name used
738
to invoke it.  It is useful sometimes to make a link to a utility
739
with a different name, and that should not change what it does.
740
 
741
Instead, use a run time option or a compilation switch or both
742
to select among the alternate behaviors.
743
 
744
@cindex output device and program's behavior
745
Likewise, please don't make the behavior of the program depend on the
746
type of output device it is used with.  Device independence is an
747
important principle of the system's design; do not compromise it merely
748
to save someone from typing an option now and then.  (Variation in error
749
message syntax when using a terminal is ok, because that is a side issue
750
that people do not depend on.)
751
 
752
If you think one behavior is most useful when the output is to a
753
terminal, and another is most useful when the output is a file or a
754
pipe, then it is usually best to make the default behavior the one that
755
is useful with output to a terminal, and have an option for the other
756
behavior.
757
 
758
Compatibility requires certain programs to depend on the type of output
759
device.  It would be disastrous if @code{ls} or @code{sh} did not do so
760
in the way all users expect.  In some of these cases, we supplement the
761
program with a preferred alternate version that does not depend on the
762
output device type.  For example, we provide a @code{dir} program much
763
like @code{ls} except that its default output format is always
764
multi-column format.
765
 
766
@node Graphical Interfaces
767
@section Standards for Graphical Interfaces
768
@cindex graphical user interface
769
 
770
@cindex gtk
771
When you write a program that provides a graphical user interface,
772
please make it work with X Windows and the GTK toolkit unless the
773
functionality specifically requires some alternative (for example,
774
``displaying jpeg images while in console mode'').
775
 
776
In addition, please provide a command-line interface to control the
777
functionality.  (In many cases, the graphical user interface can be a
778
separate program which invokes the command-line program.)  This is
779
so that the same jobs can be done from scripts.
780
 
781
@cindex corba
782
@cindex gnome
783
Please also consider providing a CORBA interface (for use from GNOME), a
784
library interface (for use from C), and perhaps a keyboard-driven
785
console interface (for use by users from console mode).  Once you are
786
doing the work to provide the functionality and the graphical interface,
787
these won't be much extra work.
788
 
789
@node Command-Line Interfaces
790
@section Standards for Command Line Interfaces
791
@cindex command-line interface
792
 
793
@findex getopt
794
It is a good idea to follow the @sc{posix} guidelines for the
795
command-line options of a program.  The easiest way to do this is to use
796
@code{getopt} to parse them.  Note that the GNU version of @code{getopt}
797
will normally permit options anywhere among the arguments unless the
798
special argument @samp{--} is used.  This is not what @sc{posix}
799
specifies; it is a GNU extension.
800
 
801
@cindex long-named options
802
Please define long-named options that are equivalent to the
803
single-letter Unix-style options.  We hope to make GNU more user
804
friendly this way.  This is easy to do with the GNU function
805
@code{getopt_long}.
806
 
807
One of the advantages of long-named options is that they can be
808
consistent from program to program.  For example, users should be able
809
to expect the ``verbose'' option of any GNU program which has one, to be
810
spelled precisely @samp{--verbose}.  To achieve this uniformity, look at
811
the table of common long-option names when you choose the option names
812
for your program (@pxref{Option Table}).
813
 
814
It is usually a good idea for file names given as ordinary arguments to
815
be input files only; any output files would be specified using options
816
(preferably @samp{-o} or @samp{--output}).  Even if you allow an output
817
file name as an ordinary argument for compatibility, try to provide an
818
option as another way to specify it.  This will lead to more consistency
819
among GNU utilities, and fewer idiosyncracies for users to remember.
820
 
821
@cindex standard command-line options
822
All programs should support two standard options: @samp{--version}
823
and @samp{--help}.
824
 
825
@table @code
826
@cindex @samp{--version} option
827
@item --version
828
This option should direct the program to print information about its name,
829
version, origin and legal status, all on standard output, and then exit
830
successfully.  Other options and arguments should be ignored once this
831
is seen, and the program should not perform its normal function.
832
 
833
@cindex canonical name of a program
834
@cindex program's canonical name
835
The first line is meant to be easy for a program to parse; the version
836
number proper starts after the last space.  In addition, it contains
837
the canonical name for this program, in this format:
838
 
839
@example
840
GNU Emacs 19.30
841
@end example
842
 
843
@noindent
844
The program's name should be a constant string; @emph{don't} compute it
845
from @code{argv[0]}.  The idea is to state the standard or canonical
846
name for the program, not its file name.  There are other ways to find
847
out the precise file name where a command is found in @code{PATH}.
848
 
849
If the program is a subsidiary part of a larger package, mention the
850
package name in parentheses, like this:
851
 
852
@example
853
emacsserver (GNU Emacs) 19.30
854
@end example
855
 
856
@noindent
857
If the package has a version number which is different from this
858
program's version number, you can mention the package version number
859
just before the close-parenthesis.
860
 
861
If you @strong{need} to mention the version numbers of libraries which
862
are distributed separately from the package which contains this program,
863
you can do so by printing an additional line of version info for each
864
library you want to mention.  Use the same format for these lines as for
865
the first line.
866
 
867
Please do not mention all of the libraries that the program uses ``just
868
for completeness''---that would produce a lot of unhelpful clutter.
869
Please mention library version numbers only if you find in practice that
870
they are very important to you in debugging.
871
 
872
The following line, after the version number line or lines, should be a
873
copyright notice.  If more than one copyright notice is called for, put
874
each on a separate line.
875
 
876
Next should follow a brief statement that the program is free software,
877
and that users are free to copy and change it on certain conditions.  If
878
the program is covered by the GNU GPL, say so here.  Also mention that
879
there is no warranty, to the extent permitted by law.
880
 
881
It is ok to finish the output with a list of the major authors of the
882
program, as a way of giving credit.
883
 
884
Here's an example of output that follows these rules:
885
 
886
@smallexample
887
GNU Emacs 19.34.5
888
Copyright (C) 1996 Free Software Foundation, Inc.
889
GNU Emacs comes with NO WARRANTY,
890
to the extent permitted by law.
891
You may redistribute copies of GNU Emacs
892
under the terms of the GNU General Public License.
893
For more information about these matters,
894
see the files named COPYING.
895
@end smallexample
896
 
897
You should adapt this to your program, of course, filling in the proper
898
year, copyright holder, name of program, and the references to
899
distribution terms, and changing the rest of the wording as necessary.
900
 
901
This copyright notice only needs to mention the most recent year in
902
which changes were made---there's no need to list the years for previous
903
versions' changes.  You don't have to mention the name of the program in
904
these notices, if that is inconvenient, since it appeared in the first
905
line.
906
 
907
Translations of the above lines must preserve the validity of the
908
copyright notices (@pxref{Internationalization}).  If the translation's
909
character set supports it, the @samp{(C)} should be replaced with the
910
copyright symbol, as follows:
911
 
912
@ifinfo
913
(the official copyright symbol, which is the letter C in a circle);
914
@end ifinfo
915
@ifnotinfo
916
@copyright{}
917
@end ifnotinfo
918
 
919
Write the word ``Copyright'' exactly like that, in English.  Do not
920
translate it into another language.  International treaties recognize
921
the English word ``Copyright''; translations into other languages do not
922
have legal significance.
923
 
924
 
925
@cindex @samp{--help} option
926
@item --help
927
This option should output brief documentation for how to invoke the
928
program, on standard output, then exit successfully.  Other options and
929
arguments should be ignored once this is seen, and the program should
930
not perform its normal function.
931
 
932
@cindex address for bug reports
933
@cindex bug reports
934
Near the end of the @samp{--help} option's output there should be a line
935
that says where to mail bug reports.  It should have this format:
936
 
937
@example
938
Report bugs to @var{mailing-address}.
939
@end example
940
@end table
941
 
942
@node Option Table
943
@section Table of Long Options
944
@cindex long option names
945
@cindex table of long options
946
 
947
Here is a table of long options used by GNU programs.  It is surely
948
incomplete, but we aim to list all the options that a new program might
949
want to be compatible with.  If you use names not already in the table,
950
please send @email{bug-standards@@gnu.org} a list of them, with their
951
meanings, so we can update the table.
952
 
953
@c Please leave newlines between items in this table; it's much easier
954
@c to update when it isn't completely squashed together and unreadable.
955
@c When there is more than one short option for a long option name, put
956
@c a semicolon between the lists of the programs that use them, not a
957
@c period.   --friedman
958
 
959
@table @samp
960
@item after-date
961
@samp{-N} in @code{tar}.
962
 
963
@item all
964
@samp{-a} in @code{du}, @code{ls}, @code{nm}, @code{stty}, @code{uname},
965
and @code{unexpand}.
966
 
967
@item all-text
968
@samp{-a} in @code{diff}.
969
 
970
@item almost-all
971
@samp{-A} in @code{ls}.
972
 
973
@item append
974
@samp{-a} in @code{etags}, @code{tee}, @code{time};
975
@samp{-r} in @code{tar}.
976
 
977
@item archive
978
@samp{-a} in @code{cp}.
979
 
980
@item archive-name
981
@samp{-n} in @code{shar}.
982
 
983
@item arglength
984
@samp{-l} in @code{m4}.
985
 
986
@item ascii
987
@samp{-a} in @code{diff}.
988
 
989
@item assign
990
@samp{-v} in @code{gawk}.
991
 
992
@item assume-new
993
@samp{-W} in Make.
994
 
995
@item assume-old
996
@samp{-o} in Make.
997
 
998
@item auto-check
999
@samp{-a} in @code{recode}.
1000
 
1001
@item auto-pager
1002
@samp{-a} in @code{wdiff}.
1003
 
1004
@item auto-reference
1005
@samp{-A} in @code{ptx}.
1006
 
1007
@item avoid-wraps
1008
@samp{-n} in @code{wdiff}.
1009
 
1010
@item background
1011
For server programs, run in the background.
1012
 
1013
@item backward-search
1014
@samp{-B} in @code{ctags}.
1015
 
1016
@item basename
1017
@samp{-f} in @code{shar}.
1018
 
1019
@item batch
1020
Used in GDB.
1021
 
1022
@item baud
1023
Used in GDB.
1024
 
1025
@item before
1026
@samp{-b} in @code{tac}.
1027
 
1028
@item binary
1029
@samp{-b} in @code{cpio} and @code{diff}.
1030
 
1031
@item bits-per-code
1032
@samp{-b} in @code{shar}.
1033
 
1034
@item block-size
1035
Used in @code{cpio} and @code{tar}.
1036
 
1037
@item blocks
1038
@samp{-b} in @code{head} and @code{tail}.
1039
 
1040
@item break-file
1041
@samp{-b} in @code{ptx}.
1042
 
1043
@item brief
1044
Used in various programs to make output shorter.
1045
 
1046
@item bytes
1047
@samp{-c} in @code{head}, @code{split}, and @code{tail}.
1048
 
1049
@item c@t{++}
1050
@samp{-C} in @code{etags}.
1051
 
1052
@item catenate
1053
@samp{-A} in @code{tar}.
1054
 
1055
@item cd
1056
Used in various programs to specify the directory to use.
1057
 
1058
@item changes
1059
@samp{-c} in @code{chgrp} and @code{chown}.
1060
 
1061
@item classify
1062
@samp{-F} in @code{ls}.
1063
 
1064
@item colons
1065
@samp{-c} in @code{recode}.
1066
 
1067
@item command
1068
@samp{-c} in @code{su};
1069
@samp{-x} in GDB.
1070
 
1071
@item compare
1072
@samp{-d} in @code{tar}.
1073
 
1074
@item compat
1075
Used in @code{gawk}.
1076
 
1077
@item compress
1078
@samp{-Z} in @code{tar} and @code{shar}.
1079
 
1080
@item concatenate
1081
@samp{-A} in @code{tar}.
1082
 
1083
@item confirmation
1084
@samp{-w} in @code{tar}.
1085
 
1086
@item context
1087
Used in @code{diff}.
1088
 
1089
@item copyleft
1090
@samp{-W copyleft} in @code{gawk}.
1091
 
1092
@item copyright
1093
@samp{-C} in @code{ptx}, @code{recode}, and @code{wdiff};
1094
@samp{-W copyright} in @code{gawk}.
1095
 
1096
@item core
1097
Used in GDB.
1098
 
1099
@item count
1100
@samp{-q} in @code{who}.
1101
 
1102
@item count-links
1103
@samp{-l} in @code{du}.
1104
 
1105
@item create
1106
Used in @code{tar} and @code{cpio}.
1107
 
1108
@item cut-mark
1109
@samp{-c} in @code{shar}.
1110
 
1111
@item cxref
1112
@samp{-x} in @code{ctags}.
1113
 
1114
@item date
1115
@samp{-d} in @code{touch}.
1116
 
1117
@item debug
1118
@samp{-d} in Make and @code{m4};
1119
@samp{-t} in Bison.
1120
 
1121
@item define
1122
@samp{-D} in @code{m4}.
1123
 
1124
@item defines
1125
@samp{-d} in Bison and @code{ctags}.
1126
 
1127
@item delete
1128
@samp{-D} in @code{tar}.
1129
 
1130
@item dereference
1131
@samp{-L} in @code{chgrp}, @code{chown}, @code{cpio}, @code{du},
1132
@code{ls}, and @code{tar}.
1133
 
1134
@item dereference-args
1135
@samp{-D} in @code{du}.
1136
 
1137
@item device
1138
Specify an I/O device (special file name).
1139
 
1140
@item diacritics
1141
@samp{-d} in @code{recode}.
1142
 
1143
@item dictionary-order
1144
@samp{-d} in @code{look}.
1145
 
1146
@item diff
1147
@samp{-d} in @code{tar}.
1148
 
1149
@item digits
1150
@samp{-n} in @code{csplit}.
1151
 
1152
@item directory
1153
Specify the directory to use, in various programs.  In @code{ls}, it
1154
means to show directories themselves rather than their contents.  In
1155
@code{rm} and @code{ln}, it means to not treat links to directories
1156
specially.
1157
 
1158
@item discard-all
1159
@samp{-x} in @code{strip}.
1160
 
1161
@item discard-locals
1162
@samp{-X} in @code{strip}.
1163
 
1164
@item dry-run
1165
@samp{-n} in Make.
1166
 
1167
@item ed
1168
@samp{-e} in @code{diff}.
1169
 
1170
@item elide-empty-files
1171
@samp{-z} in @code{csplit}.
1172
 
1173
@item end-delete
1174
@samp{-x} in @code{wdiff}.
1175
 
1176
@item end-insert
1177
@samp{-z} in @code{wdiff}.
1178
 
1179
@item entire-new-file
1180
@samp{-N} in @code{diff}.
1181
 
1182
@item environment-overrides
1183
@samp{-e} in Make.
1184
 
1185
@item eof
1186
@samp{-e} in @code{xargs}.
1187
 
1188
@item epoch
1189
Used in GDB.
1190
 
1191
@item error-limit
1192
Used in @code{makeinfo}.
1193
 
1194
@item error-output
1195
@samp{-o} in @code{m4}.
1196
 
1197
@item escape
1198
@samp{-b} in @code{ls}.
1199
 
1200
@item exclude-from
1201
@samp{-X} in @code{tar}.
1202
 
1203
@item exec
1204
Used in GDB.
1205
 
1206
@item exit
1207
@samp{-x} in @code{xargs}.
1208
 
1209
@item exit-0
1210
@samp{-e} in @code{unshar}.
1211
 
1212
@item expand-tabs
1213
@samp{-t} in @code{diff}.
1214
 
1215
@item expression
1216
@samp{-e} in @code{sed}.
1217
 
1218
@item extern-only
1219
@samp{-g} in @code{nm}.
1220
 
1221
@item extract
1222
@samp{-i} in @code{cpio};
1223
@samp{-x} in @code{tar}.
1224
 
1225
@item faces
1226
@samp{-f} in @code{finger}.
1227
 
1228
@item fast
1229
@samp{-f} in @code{su}.
1230
 
1231
@item fatal-warnings
1232
@samp{-E} in @code{m4}.
1233
 
1234
@item file
1235
@samp{-f} in @code{info}, @code{gawk}, Make, @code{mt}, and @code{tar};
1236
@samp{-n} in @code{sed};
1237
@samp{-r} in @code{touch}.
1238
 
1239
@item field-separator
1240
@samp{-F} in @code{gawk}.
1241
 
1242
@item file-prefix
1243
@samp{-b} in Bison.
1244
 
1245
@item file-type
1246
@samp{-F} in @code{ls}.
1247
 
1248
@item files-from
1249
@samp{-T} in @code{tar}.
1250
 
1251
@item fill-column
1252
Used in @code{makeinfo}.
1253
 
1254
@item flag-truncation
1255
@samp{-F} in @code{ptx}.
1256
 
1257
@item fixed-output-files
1258
@samp{-y} in Bison.
1259
 
1260
@item follow
1261
@samp{-f} in @code{tail}.
1262
 
1263
@item footnote-style
1264
Used in @code{makeinfo}.
1265
 
1266
@item force
1267
@samp{-f} in @code{cp}, @code{ln}, @code{mv}, and @code{rm}.
1268
 
1269
@item force-prefix
1270
@samp{-F} in @code{shar}.
1271
 
1272
@item foreground
1273
For server programs, run in the foreground;
1274
in other words, don't do anything special to run the server
1275
in the background.
1276
 
1277
@item format
1278
Used in @code{ls}, @code{time}, and @code{ptx}.
1279
 
1280
@item freeze-state
1281
@samp{-F} in @code{m4}.
1282
 
1283
@item fullname
1284
Used in GDB.
1285
 
1286
@item gap-size
1287
@samp{-g} in @code{ptx}.
1288
 
1289
@item get
1290
@samp{-x} in @code{tar}.
1291
 
1292
@item graphic
1293
@samp{-i} in @code{ul}.
1294
 
1295
@item graphics
1296
@samp{-g} in @code{recode}.
1297
 
1298
@item group
1299
@samp{-g} in @code{install}.
1300
 
1301
@item gzip
1302
@samp{-z} in @code{tar} and @code{shar}.
1303
 
1304
@item hashsize
1305
@samp{-H} in @code{m4}.
1306
 
1307
@item header
1308
@samp{-h} in @code{objdump} and @code{recode}
1309
 
1310
@item heading
1311
@samp{-H} in @code{who}.
1312
 
1313
@item help
1314
Used to ask for brief usage information.
1315
 
1316
@item here-delimiter
1317
@samp{-d} in @code{shar}.
1318
 
1319
@item hide-control-chars
1320
@samp{-q} in @code{ls}.
1321
 
1322
@item html
1323
In @code{makeinfo}, output HTML.
1324
 
1325
@item idle
1326
@samp{-u} in @code{who}.
1327
 
1328
@item ifdef
1329
@samp{-D} in @code{diff}.
1330
 
1331
@item ignore
1332
@samp{-I} in @code{ls};
1333
@samp{-x} in @code{recode}.
1334
 
1335
@item ignore-all-space
1336
@samp{-w} in @code{diff}.
1337
 
1338
@item ignore-backups
1339
@samp{-B} in @code{ls}.
1340
 
1341
@item ignore-blank-lines
1342
@samp{-B} in @code{diff}.
1343
 
1344
@item ignore-case
1345
@samp{-f} in @code{look} and @code{ptx};
1346
@samp{-i} in @code{diff} and @code{wdiff}.
1347
 
1348
@item ignore-errors
1349
@samp{-i} in Make.
1350
 
1351
@item ignore-file
1352
@samp{-i} in @code{ptx}.
1353
 
1354
@item ignore-indentation
1355
@samp{-I} in @code{etags}.
1356
 
1357
@item ignore-init-file
1358
@samp{-f} in Oleo.
1359
 
1360
@item ignore-interrupts
1361
@samp{-i} in @code{tee}.
1362
 
1363
@item ignore-matching-lines
1364
@samp{-I} in @code{diff}.
1365
 
1366
@item ignore-space-change
1367
@samp{-b} in @code{diff}.
1368
 
1369
@item ignore-zeros
1370
@samp{-i} in @code{tar}.
1371
 
1372
@item include
1373
@samp{-i} in @code{etags};
1374
@samp{-I} in @code{m4}.
1375
 
1376
@item include-dir
1377
@samp{-I} in Make.
1378
 
1379
@item incremental
1380
@samp{-G} in @code{tar}.
1381
 
1382
@item info
1383
@samp{-i}, @samp{-l}, and @samp{-m} in Finger.
1384
 
1385
@item init-file
1386
In some programs, specify the name of the file to read as the user's
1387
init file.
1388
 
1389
@item initial
1390
@samp{-i} in @code{expand}.
1391
 
1392
@item initial-tab
1393
@samp{-T} in @code{diff}.
1394
 
1395
@item inode
1396
@samp{-i} in @code{ls}.
1397
 
1398
@item interactive
1399
@samp{-i} in @code{cp}, @code{ln}, @code{mv}, @code{rm};
1400
@samp{-e} in @code{m4};
1401
@samp{-p} in @code{xargs};
1402
@samp{-w} in @code{tar}.
1403
 
1404
@item intermix-type
1405
@samp{-p} in @code{shar}.
1406
 
1407
@item iso-8601
1408
Used in @code{date}
1409
 
1410
@item jobs
1411
@samp{-j} in Make.
1412
 
1413
@item just-print
1414
@samp{-n} in Make.
1415
 
1416
@item keep-going
1417
@samp{-k} in Make.
1418
 
1419
@item keep-files
1420
@samp{-k} in @code{csplit}.
1421
 
1422
@item kilobytes
1423
@samp{-k} in @code{du} and @code{ls}.
1424
 
1425
@item language
1426
@samp{-l} in @code{etags}.
1427
 
1428
@item less-mode
1429
@samp{-l} in @code{wdiff}.
1430
 
1431
@item level-for-gzip
1432
@samp{-g} in @code{shar}.
1433
 
1434
@item line-bytes
1435
@samp{-C} in @code{split}.
1436
 
1437
@item lines
1438
Used in @code{split}, @code{head}, and @code{tail}.
1439
 
1440
@item link
1441
@samp{-l} in @code{cpio}.
1442
 
1443
@item lint
1444
@itemx lint-old
1445
Used in @code{gawk}.
1446
 
1447
@item list
1448
@samp{-t} in @code{cpio};
1449
@samp{-l} in @code{recode}.
1450
 
1451
@item list
1452
@samp{-t} in @code{tar}.
1453
 
1454
@item literal
1455
@samp{-N} in @code{ls}.
1456
 
1457
@item load-average
1458
@samp{-l} in Make.
1459
 
1460
@item login
1461
Used in @code{su}.
1462
 
1463
@item machine
1464
No listing of which programs already use this;
1465
someone should check to
1466
see if any actually do, and tell @email{gnu@@gnu.org}.
1467
 
1468
@item macro-name
1469
@samp{-M} in @code{ptx}.
1470
 
1471
@item mail
1472
@samp{-m} in @code{hello} and @code{uname}.
1473
 
1474
@item make-directories
1475
@samp{-d} in @code{cpio}.
1476
 
1477
@item makefile
1478
@samp{-f} in Make.
1479
 
1480
@item mapped
1481
Used in GDB.
1482
 
1483
@item max-args
1484
@samp{-n} in @code{xargs}.
1485
 
1486
@item max-chars
1487
@samp{-n} in @code{xargs}.
1488
 
1489
@item max-lines
1490
@samp{-l} in @code{xargs}.
1491
 
1492
@item max-load
1493
@samp{-l} in Make.
1494
 
1495
@item max-procs
1496
@samp{-P} in @code{xargs}.
1497
 
1498
@item mesg
1499
@samp{-T} in @code{who}.
1500
 
1501
@item message
1502
@samp{-T} in @code{who}.
1503
 
1504
@item minimal
1505
@samp{-d} in @code{diff}.
1506
 
1507
@item mixed-uuencode
1508
@samp{-M} in @code{shar}.
1509
 
1510
@item mode
1511
@samp{-m} in @code{install}, @code{mkdir}, and @code{mkfifo}.
1512
 
1513
@item modification-time
1514
@samp{-m} in @code{tar}.
1515
 
1516
@item multi-volume
1517
@samp{-M} in @code{tar}.
1518
 
1519
@item name-prefix
1520
@samp{-a} in Bison.
1521
 
1522
@item nesting-limit
1523
@samp{-L} in @code{m4}.
1524
 
1525
@item net-headers
1526
@samp{-a} in @code{shar}.
1527
 
1528
@item new-file
1529
@samp{-W} in Make.
1530
 
1531
@item no-builtin-rules
1532
@samp{-r} in Make.
1533
 
1534
@item no-character-count
1535
@samp{-w} in @code{shar}.
1536
 
1537
@item no-check-existing
1538
@samp{-x} in @code{shar}.
1539
 
1540
@item no-common
1541
@samp{-3} in @code{wdiff}.
1542
 
1543
@item no-create
1544
@samp{-c} in @code{touch}.
1545
 
1546
@item no-defines
1547
@samp{-D} in @code{etags}.
1548
 
1549
@item no-deleted
1550
@samp{-1} in @code{wdiff}.
1551
 
1552
@item no-dereference
1553
@samp{-d} in @code{cp}.
1554
 
1555
@item no-inserted
1556
@samp{-2} in @code{wdiff}.
1557
 
1558
@item no-keep-going
1559
@samp{-S} in Make.
1560
 
1561
@item no-lines
1562
@samp{-l} in Bison.
1563
 
1564
@item no-piping
1565
@samp{-P} in @code{shar}.
1566
 
1567
@item no-prof
1568
@samp{-e} in @code{gprof}.
1569
 
1570
@item no-regex
1571
@samp{-R} in @code{etags}.
1572
 
1573
@item no-sort
1574
@samp{-p} in @code{nm}.
1575
 
1576
@item no-split
1577
Used in @code{makeinfo}.
1578
 
1579
@item no-static
1580
@samp{-a} in @code{gprof}.
1581
 
1582
@item no-time
1583
@samp{-E} in @code{gprof}.
1584
 
1585
@item no-timestamp
1586
@samp{-m} in @code{shar}.
1587
 
1588
@item no-validate
1589
Used in @code{makeinfo}.
1590
 
1591
@item no-wait
1592
Used in @code{emacsclient}.
1593
 
1594
@item no-warn
1595
Used in various programs to inhibit warnings.
1596
 
1597
@item node
1598
@samp{-n} in @code{info}.
1599
 
1600
@item nodename
1601
@samp{-n} in @code{uname}.
1602
 
1603
@item nonmatching
1604
@samp{-f} in @code{cpio}.
1605
 
1606
@item nstuff
1607
@samp{-n} in @code{objdump}.
1608
 
1609
@item null
1610
@samp{-0} in @code{xargs}.
1611
 
1612
@item number
1613
@samp{-n} in @code{cat}.
1614
 
1615
@item number-nonblank
1616
@samp{-b} in @code{cat}.
1617
 
1618
@item numeric-sort
1619
@samp{-n} in @code{nm}.
1620
 
1621
@item numeric-uid-gid
1622
@samp{-n} in @code{cpio} and @code{ls}.
1623
 
1624
@item nx
1625
Used in GDB.
1626
 
1627
@item old-archive
1628
@samp{-o} in @code{tar}.
1629
 
1630
@item old-file
1631
@samp{-o} in Make.
1632
 
1633
@item one-file-system
1634
@samp{-l} in @code{tar}, @code{cp}, and @code{du}.
1635
 
1636
@item only-file
1637
@samp{-o} in @code{ptx}.
1638
 
1639
@item only-prof
1640
@samp{-f} in @code{gprof}.
1641
 
1642
@item only-time
1643
@samp{-F} in @code{gprof}.
1644
 
1645
@item options
1646
@samp{-o} in @code{getopt}, @code{fdlist}, @code{fdmount},
1647
@code{fdmountd}, and @code{fdumount}.
1648
 
1649
@item output
1650
In various programs, specify the output file name.
1651
 
1652
@item output-prefix
1653
@samp{-o} in @code{shar}.
1654
 
1655
@item override
1656
@samp{-o} in @code{rm}.
1657
 
1658
@item overwrite
1659
@samp{-c} in @code{unshar}.
1660
 
1661
@item owner
1662
@samp{-o} in @code{install}.
1663
 
1664
@item paginate
1665
@samp{-l} in @code{diff}.
1666
 
1667
@item paragraph-indent
1668
Used in @code{makeinfo}.
1669
 
1670
@item parents
1671
@samp{-p} in @code{mkdir} and @code{rmdir}.
1672
 
1673
@item pass-all
1674
@samp{-p} in @code{ul}.
1675
 
1676
@item pass-through
1677
@samp{-p} in @code{cpio}.
1678
 
1679
@item port
1680
@samp{-P} in @code{finger}.
1681
 
1682
@item portability
1683
@samp{-c} in @code{cpio} and @code{tar}.
1684
 
1685
@item posix
1686
Used in @code{gawk}.
1687
 
1688
@item prefix-builtins
1689
@samp{-P} in @code{m4}.
1690
 
1691
@item prefix
1692
@samp{-f} in @code{csplit}.
1693
 
1694
@item preserve
1695
Used in @code{tar} and @code{cp}.
1696
 
1697
@item preserve-environment
1698
@samp{-p} in @code{su}.
1699
 
1700
@item preserve-modification-time
1701
@samp{-m} in @code{cpio}.
1702
 
1703
@item preserve-order
1704
@samp{-s} in @code{tar}.
1705
 
1706
@item preserve-permissions
1707
@samp{-p} in @code{tar}.
1708
 
1709
@item print
1710
@samp{-l} in @code{diff}.
1711
 
1712
@item print-chars
1713
@samp{-L} in @code{cmp}.
1714
 
1715
@item print-data-base
1716
@samp{-p} in Make.
1717
 
1718
@item print-directory
1719
@samp{-w} in Make.
1720
 
1721
@item print-file-name
1722
@samp{-o} in @code{nm}.
1723
 
1724
@item print-symdefs
1725
@samp{-s} in @code{nm}.
1726
 
1727
@item printer
1728
@samp{-p} in @code{wdiff}.
1729
 
1730
@item prompt
1731
@samp{-p} in @code{ed}.
1732
 
1733
@item proxy
1734
Specify an HTTP proxy.
1735
 
1736
@item query-user
1737
@samp{-X} in @code{shar}.
1738
 
1739
@item question
1740
@samp{-q} in Make.
1741
 
1742
@item quiet
1743
Used in many programs to inhibit the usual output.  @strong{Note:} every
1744
program accepting @samp{--quiet} should accept @samp{--silent} as a
1745
synonym.
1746
 
1747
@item quiet-unshar
1748
@samp{-Q} in @code{shar}
1749
 
1750
@item quote-name
1751
@samp{-Q} in @code{ls}.
1752
 
1753
@item rcs
1754
@samp{-n} in @code{diff}.
1755
 
1756
@item re-interval
1757
Used in @code{gawk}.
1758
 
1759
@item read-full-blocks
1760
@samp{-B} in @code{tar}.
1761
 
1762
@item readnow
1763
Used in GDB.
1764
 
1765
@item recon
1766
@samp{-n} in Make.
1767
 
1768
@item record-number
1769
@samp{-R} in @code{tar}.
1770
 
1771
@item recursive
1772
Used in @code{chgrp}, @code{chown}, @code{cp}, @code{ls}, @code{diff},
1773
and @code{rm}.
1774
 
1775
@item reference-limit
1776
Used in @code{makeinfo}.
1777
 
1778
@item references
1779
@samp{-r} in @code{ptx}.
1780
 
1781
@item regex
1782
@samp{-r} in @code{tac} and @code{etags}.
1783
 
1784
@item release
1785
@samp{-r} in @code{uname}.
1786
 
1787
@item reload-state
1788
@samp{-R} in @code{m4}.
1789
 
1790
@item relocation
1791
@samp{-r} in @code{objdump}.
1792
 
1793
@item rename
1794
@samp{-r} in @code{cpio}.
1795
 
1796
@item replace
1797
@samp{-i} in @code{xargs}.
1798
 
1799
@item report-identical-files
1800
@samp{-s} in @code{diff}.
1801
 
1802
@item reset-access-time
1803
@samp{-a} in @code{cpio}.
1804
 
1805
@item reverse
1806
@samp{-r} in @code{ls} and @code{nm}.
1807
 
1808
@item reversed-ed
1809
@samp{-f} in @code{diff}.
1810
 
1811
@item right-side-defs
1812
@samp{-R} in @code{ptx}.
1813
 
1814
@item same-order
1815
@samp{-s} in @code{tar}.
1816
 
1817
@item same-permissions
1818
@samp{-p} in @code{tar}.
1819
 
1820
@item save
1821
@samp{-g} in @code{stty}.
1822
 
1823
@item se
1824
Used in GDB.
1825
 
1826
@item sentence-regexp
1827
@samp{-S} in @code{ptx}.
1828
 
1829
@item separate-dirs
1830
@samp{-S} in @code{du}.
1831
 
1832
@item separator
1833
@samp{-s} in @code{tac}.
1834
 
1835
@item sequence
1836
Used by @code{recode} to chose files or pipes for sequencing passes.
1837
 
1838
@item shell
1839
@samp{-s} in @code{su}.
1840
 
1841
@item show-all
1842
@samp{-A} in @code{cat}.
1843
 
1844
@item show-c-function
1845
@samp{-p} in @code{diff}.
1846
 
1847
@item show-ends
1848
@samp{-E} in @code{cat}.
1849
 
1850
@item show-function-line
1851
@samp{-F} in @code{diff}.
1852
 
1853
@item show-tabs
1854
@samp{-T} in @code{cat}.
1855
 
1856
@item silent
1857
Used in many programs to inhibit the usual output.
1858
@strong{Note:} every program accepting
1859
@samp{--silent} should accept @samp{--quiet} as a synonym.
1860
 
1861
@item size
1862
@samp{-s} in @code{ls}.
1863
 
1864
@item socket
1865
Specify a file descriptor for a network server to use for its socket,
1866
instead of opening and binding a new socket.  This provides a way to
1867
run, in a nonpriveledged process, a server that normally needs a
1868
reserved port number.
1869
 
1870
@item sort
1871
Used in @code{ls}.
1872
 
1873
@item source
1874
@samp{-W source} in @code{gawk}.
1875
 
1876
@item sparse
1877
@samp{-S} in @code{tar}.
1878
 
1879
@item speed-large-files
1880
@samp{-H} in @code{diff}.
1881
 
1882
@item split-at
1883
@samp{-E} in @code{unshar}.
1884
 
1885
@item split-size-limit
1886
@samp{-L} in @code{shar}.
1887
 
1888
@item squeeze-blank
1889
@samp{-s} in @code{cat}.
1890
 
1891
@item start-delete
1892
@samp{-w} in @code{wdiff}.
1893
 
1894
@item start-insert
1895
@samp{-y} in @code{wdiff}.
1896
 
1897
@item starting-file
1898
Used in @code{tar} and @code{diff} to specify which file within
1899
a directory to start processing with.
1900
 
1901
@item statistics
1902
@samp{-s} in @code{wdiff}.
1903
 
1904
@item stdin-file-list
1905
@samp{-S} in @code{shar}.
1906
 
1907
@item stop
1908
@samp{-S} in Make.
1909
 
1910
@item strict
1911
@samp{-s} in @code{recode}.
1912
 
1913
@item strip
1914
@samp{-s} in @code{install}.
1915
 
1916
@item strip-all
1917
@samp{-s} in @code{strip}.
1918
 
1919
@item strip-debug
1920
@samp{-S} in @code{strip}.
1921
 
1922
@item submitter
1923
@samp{-s} in @code{shar}.
1924
 
1925
@item suffix
1926
@samp{-S} in @code{cp}, @code{ln}, @code{mv}.
1927
 
1928
@item suffix-format
1929
@samp{-b} in @code{csplit}.
1930
 
1931
@item sum
1932
@samp{-s} in @code{gprof}.
1933
 
1934
@item summarize
1935
@samp{-s} in @code{du}.
1936
 
1937
@item symbolic
1938
@samp{-s} in @code{ln}.
1939
 
1940
@item symbols
1941
Used in GDB and @code{objdump}.
1942
 
1943
@item synclines
1944
@samp{-s} in @code{m4}.
1945
 
1946
@item sysname
1947
@samp{-s} in @code{uname}.
1948
 
1949
@item tabs
1950
@samp{-t} in @code{expand} and @code{unexpand}.
1951
 
1952
@item tabsize
1953
@samp{-T} in @code{ls}.
1954
 
1955
@item terminal
1956
@samp{-T} in @code{tput} and @code{ul}.
1957
@samp{-t} in @code{wdiff}.
1958
 
1959
@item text
1960
@samp{-a} in @code{diff}.
1961
 
1962
@item text-files
1963
@samp{-T} in @code{shar}.
1964
 
1965
@item time
1966
Used in @code{ls} and @code{touch}.
1967
 
1968
@item timeout
1969
Specify how long to wait before giving up on some operation.
1970
 
1971
@item to-stdout
1972
@samp{-O} in @code{tar}.
1973
 
1974
@item total
1975
@samp{-c} in @code{du}.
1976
 
1977
@item touch
1978
@samp{-t} in Make, @code{ranlib}, and @code{recode}.
1979
 
1980
@item trace
1981
@samp{-t} in @code{m4}.
1982
 
1983
@item traditional
1984
@samp{-t} in @code{hello};
1985
@samp{-W traditional} in @code{gawk};
1986
@samp{-G} in @code{ed}, @code{m4}, and @code{ptx}.
1987
 
1988
@item tty
1989
Used in GDB.
1990
 
1991
@item typedefs
1992
@samp{-t} in @code{ctags}.
1993
 
1994
@item typedefs-and-c++
1995
@samp{-T} in @code{ctags}.
1996
 
1997
@item typeset-mode
1998
@samp{-t} in @code{ptx}.
1999
 
2000
@item uncompress
2001
@samp{-z} in @code{tar}.
2002
 
2003
@item unconditional
2004
@samp{-u} in @code{cpio}.
2005
 
2006
@item undefine
2007
@samp{-U} in @code{m4}.
2008
 
2009
@item undefined-only
2010
@samp{-u} in @code{nm}.
2011
 
2012
@item update
2013
@samp{-u} in @code{cp}, @code{ctags}, @code{mv}, @code{tar}.
2014
 
2015
@item usage
2016
Used in @code{gawk}; same as @samp{--help}.
2017
 
2018
@item uuencode
2019
@samp{-B} in @code{shar}.
2020
 
2021
@item vanilla-operation
2022
@samp{-V} in @code{shar}.
2023
 
2024
@item verbose
2025
Print more information about progress.  Many programs support this.
2026
 
2027
@item verify
2028
@samp{-W} in @code{tar}.
2029
 
2030
@item version
2031
Print the version number.
2032
 
2033
@item version-control
2034
@samp{-V} in @code{cp}, @code{ln}, @code{mv}.
2035
 
2036
@item vgrind
2037
@samp{-v} in @code{ctags}.
2038
 
2039
@item volume
2040
@samp{-V} in @code{tar}.
2041
 
2042
@item what-if
2043
@samp{-W} in Make.
2044
 
2045
@item whole-size-limit
2046
@samp{-l} in @code{shar}.
2047
 
2048
@item width
2049
@samp{-w} in @code{ls} and @code{ptx}.
2050
 
2051
@item word-regexp
2052
@samp{-W} in @code{ptx}.
2053
 
2054
@item writable
2055
@samp{-T} in @code{who}.
2056
 
2057
@item zeros
2058
@samp{-z} in @code{gprof}.
2059
@end table
2060
 
2061
@node Memory Usage
2062
@section Memory Usage
2063
@cindex memory usage
2064
 
2065
If a program typically uses just a few meg of memory, don't bother making any
2066
effort to reduce memory usage.  For example, if it is impractical for
2067
other reasons to operate on files more than a few meg long, it is
2068
reasonable to read entire input files into core to operate on them.
2069
 
2070
However, for programs such as @code{cat} or @code{tail}, that can
2071
usefully operate on very large files, it is important to avoid using a
2072
technique that would artificially limit the size of files it can handle.
2073
If a program works by lines and could be applied to arbitrary
2074
user-supplied input files, it should keep only a line in memory, because
2075
this is not very hard and users will want to be able to operate on input
2076
files that are bigger than will fit in core all at once.
2077
 
2078
If your program creates complicated data structures, just make them in
2079
core and give a fatal error if @code{malloc} returns zero.
2080
 
2081
@node File Usage
2082
@section File Usage
2083
@cindex file usage
2084
 
2085
Programs should be prepared to operate when @file{/usr} and @file{/etc}
2086
are read-only file systems.  Thus, if the program manages log files,
2087
lock files, backup files, score files, or any other files which are
2088
modified for internal purposes, these files should not be stored in
2089
@file{/usr} or @file{/etc}.
2090
 
2091
There are two exceptions.  @file{/etc} is used to store system
2092
configuration information; it is reasonable for a program to modify
2093
files in @file{/etc} when its job is to update the system configuration.
2094
Also, if the user explicitly asks to modify one file in a directory, it
2095
is reasonable for the program to store other files in the same
2096
directory.
2097
 
2098
@node Writing C
2099
@chapter Making The Best Use of C
2100
 
2101
This @value{CHAPTER} provides advice on how best to use the C language
2102
when writing GNU software.
2103
 
2104
@menu
2105
* Formatting::                  Formatting Your Source Code
2106
* Comments::                    Commenting Your Work
2107
* Syntactic Conventions::       Clean Use of C Constructs
2108
* Names::                       Naming Variables, Functions, and Files
2109
* System Portability::          Portability between different operating systems
2110
* CPU Portability::             Supporting the range of CPU types
2111
* System Functions::            Portability and ``standard'' library functions
2112
* Internationalization::        Techniques for internationalization
2113
* Mmap::                        How you can safely use @code{mmap}.
2114
@end menu
2115
 
2116
@node Formatting
2117
@section Formatting Your Source Code
2118
@cindex formatting source code
2119
 
2120
@cindex open brace
2121
@cindex braces, in C source
2122
It is important to put the open-brace that starts the body of a C
2123
function in column zero, and avoid putting any other open-brace or
2124
open-parenthesis or open-bracket in column zero.  Several tools look
2125
for open-braces in column zero to find the beginnings of C functions.
2126
These tools will not work on code not formatted that way.
2127
 
2128
It is also important for function definitions to start the name of the
2129
function in column zero.  This helps people to search for function
2130
definitions, and may also help certain tools recognize them.  Thus,
2131
the proper format is this:
2132
 
2133
@example
2134
static char *
2135
concat (s1, s2)        /* Name starts in column zero here */
2136
     char *s1, *s2;
2137
@{                     /* Open brace in column zero here */
2138
  @dots{}
2139
@}
2140
@end example
2141
 
2142
@noindent
2143
or, if you want to use Standard C syntax, format the definition like
2144
this:
2145
 
2146
@example
2147
static char *
2148
concat (char *s1, char *s2)
2149
@{
2150
  @dots{}
2151
@}
2152
@end example
2153
 
2154
In Standard C, if the arguments don't fit nicely on one line,
2155
split it like this:
2156
 
2157
@example
2158
int
2159
lots_of_args (int an_integer, long a_long, short a_short,
2160
              double a_double, float a_float)
2161
@dots{}
2162
@end example
2163
 
2164
The rest of this section gives our recommendations for other aspects of
2165
C formatting style, which is also the default style of the @code{indent}
2166
program in version 1.2 and newer.  It corresponds to the options
2167
 
2168
@smallexample
2169
-nbad -bap -nbc -bbo -bl -bli2 -bls -ncdb -nce -cp1 -cs -di2
2170
-ndj -nfc1 -nfca -hnl -i2 -ip5 -lp -pcs -psl -nsc -nsob
2171
@end smallexample
2172
 
2173
We don't think of these recommendations as requirements, because it
2174
causes no problems for users if two different programs have different
2175
formatting styles.
2176
 
2177
But whatever style you use, please use it consistently, since a mixture
2178
of styles within one program tends to look ugly.  If you are
2179
contributing changes to an existing program, please follow the style of
2180
that program.
2181
 
2182
For the body of the function, our recommended style looks like this:
2183
 
2184
@example
2185
if (x < foo (y, z))
2186
  haha = bar[4] + 5;
2187
else
2188
  @{
2189
    while (z)
2190
      @{
2191
        haha += foo (z, z);
2192
        z--;
2193
      @}
2194
    return ++x + bar ();
2195
  @}
2196
@end example
2197
 
2198
@cindex spaces before open-paren
2199
We find it easier to read a program when it has spaces before the
2200
open-parentheses and after the commas.  Especially after the commas.
2201
 
2202
When you split an expression into multiple lines, split it
2203
before an operator, not after one.  Here is the right way:
2204
 
2205
@cindex expressions, splitting
2206
@example
2207
if (foo_this_is_long && bar > win (x, y, z)
2208
    && remaining_condition)
2209
@end example
2210
 
2211
Try to avoid having two operators of different precedence at the same
2212
level of indentation.  For example, don't write this:
2213
 
2214
@example
2215
mode = (inmode[j] == VOIDmode
2216
        || GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j])
2217
        ? outmode[j] : inmode[j]);
2218
@end example
2219
 
2220
Instead, use extra parentheses so that the indentation shows the nesting:
2221
 
2222
@example
2223
mode = ((inmode[j] == VOIDmode
2224
         || (GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j])))
2225
        ? outmode[j] : inmode[j]);
2226
@end example
2227
 
2228
Insert extra parentheses so that Emacs will indent the code properly.
2229
For example, the following indentation looks nice if you do it by hand,
2230
 
2231
@example
2232
v = rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000
2233
    + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000;
2234
@end example
2235
 
2236
@noindent
2237
but Emacs would alter it.  Adding a set of parentheses produces
2238
something that looks equally nice, and which Emacs will preserve:
2239
 
2240
@example
2241
v = (rup->ru_utime.tv_sec*1000 + rup->ru_utime.tv_usec/1000
2242
     + rup->ru_stime.tv_sec*1000 + rup->ru_stime.tv_usec/1000);
2243
@end example
2244
 
2245
Format do-while statements like this:
2246
 
2247
@example
2248
do
2249
  @{
2250
    a = foo (a);
2251
  @}
2252
while (a > 0);
2253
@end example
2254
 
2255
@cindex formfeed
2256
@cindex control-L
2257
Please use formfeed characters (control-L) to divide the program into
2258
pages at logical places (but not within a function).  It does not matter
2259
just how long the pages are, since they do not have to fit on a printed
2260
page.  The formfeeds should appear alone on lines by themselves.
2261
 
2262
@node Comments
2263
@section Commenting Your Work
2264
@cindex commenting
2265
 
2266
Every program should start with a comment saying briefly what it is for.
2267
Example: @samp{fmt - filter for simple filling of text}.
2268
 
2269
Please write the comments in a GNU program in English, because English
2270
is the one language that nearly all programmers in all countries can
2271
read.  If you do not write English well, please write comments in
2272
English as well as you can, then ask other people to help rewrite them.
2273
If you can't write comments in English, please find someone to work with
2274
you and translate your comments into English.
2275
 
2276
Please put a comment on each function saying what the function does,
2277
what sorts of arguments it gets, and what the possible values of
2278
arguments mean and are used for.  It is not necessary to duplicate in
2279
words the meaning of the C argument declarations, if a C type is being
2280
used in its customary fashion.  If there is anything nonstandard about
2281
its use (such as an argument of type @code{char *} which is really the
2282
address of the second character of a string, not the first), or any
2283
possible values that would not work the way one would expect (such as,
2284
that strings containing newlines are not guaranteed to work), be sure
2285
to say so.
2286
 
2287
Also explain the significance of the return value, if there is one.
2288
 
2289
Please put two spaces after the end of a sentence in your comments, so
2290
that the Emacs sentence commands will work.  Also, please write
2291
complete sentences and capitalize the first word.  If a lower-case
2292
identifier comes at the beginning of a sentence, don't capitalize it!
2293
Changing the spelling makes it a different identifier.  If you don't
2294
like starting a sentence with a lower case letter, write the sentence
2295
differently (e.g., ``The identifier lower-case is @dots{}'').
2296
 
2297
The comment on a function is much clearer if you use the argument
2298
names to speak about the argument values.  The variable name itself
2299
should be lower case, but write it in upper case when you are speaking
2300
about the value rather than the variable itself.  Thus, ``the inode
2301
number NODE_NUM'' rather than ``an inode''.
2302
 
2303
There is usually no purpose in restating the name of the function in
2304
the comment before it, because the reader can see that for himself.
2305
There might be an exception when the comment is so long that the function
2306
itself would be off the bottom of the screen.
2307
 
2308
There should be a comment on each static variable as well, like this:
2309
 
2310
@example
2311
/* Nonzero means truncate lines in the display;
2312
   zero means continue them.  */
2313
int truncate_lines;
2314
@end example
2315
 
2316
@cindex conditionals, comments for
2317
@cindex @code{#endif}, commenting
2318
Every @samp{#endif} should have a comment, except in the case of short
2319
conditionals (just a few lines) that are not nested.  The comment should
2320
state the condition of the conditional that is ending, @emph{including
2321
its sense}.  @samp{#else} should have a comment describing the condition
2322
@emph{and sense} of the code that follows.  For example:
2323
 
2324
@example
2325
@group
2326
#ifdef foo
2327
  @dots{}
2328
#else /* not foo */
2329
  @dots{}
2330
#endif /* not foo */
2331
@end group
2332
@group
2333
#ifdef foo
2334
  @dots{}
2335
#endif /* foo */
2336
@end group
2337
@end example
2338
 
2339
@noindent
2340
but, by contrast, write the comments this way for a @samp{#ifndef}:
2341
 
2342
@example
2343
@group
2344
#ifndef foo
2345
  @dots{}
2346
#else /* foo */
2347
  @dots{}
2348
#endif /* foo */
2349
@end group
2350
@group
2351
#ifndef foo
2352
  @dots{}
2353
#endif /* not foo */
2354
@end group
2355
@end example
2356
 
2357
@node Syntactic Conventions
2358
@section Clean Use of C Constructs
2359
@cindex syntactic conventions
2360
 
2361
@cindex implicit @code{int}
2362
@cindex function argument, declaring
2363
Please explicitly declare the types of all objects.  For example, you
2364
should explicitly declare all arguments to functions, and you should
2365
declare functions to return @code{int} rather than omitting the
2366
@code{int}.
2367
 
2368
@cindex compiler warnings
2369
@cindex @samp{-Wall} compiler option
2370
Some programmers like to use the GCC @samp{-Wall} option, and change the
2371
code whenever it issues a warning.  If you want to do this, then do.
2372
Other programmers prefer not to use @samp{-Wall}, because it gives
2373
warnings for valid and legitimate code which they do not want to change.
2374
If you want to do this, then do.  The compiler should be your servant,
2375
not your master.
2376
 
2377
Declarations of external functions and functions to appear later in the
2378
source file should all go in one place near the beginning of the file
2379
(somewhere before the first function definition in the file), or else
2380
should go in a header file.  Don't put @code{extern} declarations inside
2381
functions.
2382
 
2383
@cindex temporary variables
2384
It used to be common practice to use the same local variables (with
2385
names like @code{tem}) over and over for different values within one
2386
function.  Instead of doing this, it is better declare a separate local
2387
variable for each distinct purpose, and give it a name which is
2388
meaningful.  This not only makes programs easier to understand, it also
2389
facilitates optimization by good compilers.  You can also move the
2390
declaration of each local variable into the smallest scope that includes
2391
all its uses.  This makes the program even cleaner.
2392
 
2393
Don't use local variables or parameters that shadow global identifiers.
2394
 
2395
@cindex multiple variables in a line
2396
Don't declare multiple variables in one declaration that spans lines.
2397
Start a new declaration on each line, instead.  For example, instead
2398
of this:
2399
 
2400
@example
2401
@group
2402
int    foo,
2403
       bar;
2404
@end group
2405
@end example
2406
 
2407
@noindent
2408
write either this:
2409
 
2410
@example
2411
int foo, bar;
2412
@end example
2413
 
2414
@noindent
2415
or this:
2416
 
2417
@example
2418
int foo;
2419
int bar;
2420
@end example
2421
 
2422
@noindent
2423
(If they are global variables, each should have a comment preceding it
2424
anyway.)
2425
 
2426
When you have an @code{if}-@code{else} statement nested in another
2427
@code{if} statement, always put braces around the @code{if}-@code{else}.
2428
Thus, never write like this:
2429
 
2430
@example
2431
if (foo)
2432
  if (bar)
2433
    win ();
2434
  else
2435
    lose ();
2436
@end example
2437
 
2438
@noindent
2439
always like this:
2440
 
2441
@example
2442
if (foo)
2443
  @{
2444
    if (bar)
2445
      win ();
2446
    else
2447
      lose ();
2448
  @}
2449
@end example
2450
 
2451
If you have an @code{if} statement nested inside of an @code{else}
2452
statement, either write @code{else if} on one line, like this,
2453
 
2454
@example
2455
if (foo)
2456
  @dots{}
2457
else if (bar)
2458
  @dots{}
2459
@end example
2460
 
2461
@noindent
2462
with its @code{then}-part indented like the preceding @code{then}-part,
2463
or write the nested @code{if} within braces like this:
2464
 
2465
@example
2466
if (foo)
2467
  @dots{}
2468
else
2469
  @{
2470
    if (bar)
2471
      @dots{}
2472
  @}
2473
@end example
2474
 
2475
Don't declare both a structure tag and variables or typedefs in the
2476
same declaration.  Instead, declare the structure tag separately
2477
and then use it to declare the variables or typedefs.
2478
 
2479
Try to avoid assignments inside @code{if}-conditions.  For example,
2480
don't write this:
2481
 
2482
@example
2483
if ((foo = (char *) malloc (sizeof *foo)) == 0)
2484
  fatal ("virtual memory exhausted");
2485
@end example
2486
 
2487
@noindent
2488
instead, write this:
2489
 
2490
@example
2491
foo = (char *) malloc (sizeof *foo);
2492
if (foo == 0)
2493
  fatal ("virtual memory exhausted");
2494
@end example
2495
 
2496
@pindex lint
2497
Don't make the program ugly to placate @code{lint}.  Please don't insert any
2498
casts to @code{void}.  Zero without a cast is perfectly fine as a null
2499
pointer constant, except when calling a varargs function.
2500
 
2501
@node Names
2502
@section Naming Variables, Functions, and Files
2503
 
2504
@cindex names of variables, functions, and files
2505
The names of global variables and functions in a program serve as
2506
comments of a sort.  So don't choose terse names---instead, look for
2507
names that give useful information about the meaning of the variable or
2508
function.  In a GNU program, names should be English, like other
2509
comments.
2510
 
2511
Local variable names can be shorter, because they are used only within
2512
one context, where (presumably) comments explain their purpose.
2513
 
2514
Try to limit your use of abbreviations in symbol names.  It is ok to
2515
make a few abbreviations, explain what they mean, and then use them
2516
frequently, but don't use lots of obscure abbreviations.
2517
 
2518
Please use underscores to separate words in a name, so that the Emacs
2519
word commands can be useful within them.  Stick to lower case; reserve
2520
upper case for macros and @code{enum} constants, and for name-prefixes
2521
that follow a uniform convention.
2522
 
2523
For example, you should use names like @code{ignore_space_change_flag};
2524
don't use names like @code{iCantReadThis}.
2525
 
2526
Variables that indicate whether command-line options have been
2527
specified should be named after the meaning of the option, not after
2528
the option-letter.  A comment should state both the exact meaning of
2529
the option and its letter.  For example,
2530
 
2531
@example
2532
@group
2533
/* Ignore changes in horizontal whitespace (-b).  */
2534
int ignore_space_change_flag;
2535
@end group
2536
@end example
2537
 
2538
When you want to define names with constant integer values, use
2539
@code{enum} rather than @samp{#define}.  GDB knows about enumeration
2540
constants.
2541
 
2542
@cindex file-name limitations
2543
@pindex doschk
2544
You might want to make sure that none of the file names would conflict
2545
the files were loaded onto an MS-DOS file system which shortens the
2546
names.  You can use the program @code{doschk} to test for this.
2547
 
2548
Some GNU programs were designed to limit themselves to file names of 14
2549
characters or less, to avoid file name conflicts if they are read into
2550
older System V systems.  Please preserve this feature in the existing
2551
GNU programs that have it, but there is no need to do this in new GNU
2552
programs.  @code{doschk} also reports file names longer than 14
2553
characters.
2554
 
2555
@node System Portability
2556
@section Portability between System Types
2557
@cindex portability, between system types
2558
 
2559
In the Unix world, ``portability'' refers to porting to different Unix
2560
versions.  For a GNU program, this kind of portability is desirable, but
2561
not paramount.
2562
 
2563
The primary purpose of GNU software is to run on top of the GNU kernel,
2564
compiled with the GNU C compiler, on various types of @sc{cpu}.  So the
2565
kinds of portability that are absolutely necessary are quite limited.
2566
But it is important to support Linux-based GNU systems, since they
2567
are the form of GNU that is popular.
2568
 
2569
Beyond that, it is good to support the other free operating systems
2570
(*BSD), and it is nice to support other Unix-like systems if you want
2571
to.  Supporting a variety of Unix-like systems is desirable, although
2572
not paramount.  It is usually not too hard, so you may as well do it.
2573
But you don't have to consider it an obligation, if it does turn out to
2574
be hard.
2575
 
2576
@pindex autoconf
2577
The easiest way to achieve portability to most Unix-like systems is to
2578
use Autoconf.  It's unlikely that your program needs to know more
2579
information about the host platform than Autoconf can provide, simply
2580
because most of the programs that need such knowledge have already been
2581
written.
2582
 
2583
Avoid using the format of semi-internal data bases (e.g., directories)
2584
when there is a higher-level alternative (@code{readdir}).
2585
 
2586
@cindex non-@sc{posix} systems, and portability
2587
As for systems that are not like Unix, such as MSDOS, Windows, the
2588
Macintosh, VMS, and MVS, supporting them is often a lot of work.  When
2589
that is the case, it is better to spend your time adding features that
2590
will be useful on GNU and GNU/Linux, rather than on supporting other
2591
incompatible systems.
2592
 
2593
It is a good idea to define the ``feature test macro''
2594
@code{_GNU_SOURCE} when compiling your C files.  When you compile on GNU
2595
or GNU/Linux, this will enable the declarations of GNU library extension
2596
functions, and that will usually give you a compiler error message if
2597
you define the same function names in some other way in your program.
2598
(You don't have to actually @emph{use} these functions, if you prefer
2599
to make the program more portable to other systems.)
2600
 
2601
But whether or not you use these GNU extensions, you should avoid
2602
using their names for any other meanings.  Doing so would make it hard
2603
to move your code into other GNU programs.
2604
 
2605
@node CPU Portability
2606
@section Portability between @sc{cpu}s
2607
 
2608
@cindex data types, and portability
2609
@cindex portability, and data types
2610
Even GNU systems will differ because of differences among @sc{cpu}
2611
types---for example, difference in byte ordering and alignment
2612
requirements.  It is absolutely essential to handle these differences.
2613
However, don't make any effort to cater to the possibility that an
2614
@code{int} will be less than 32 bits.  We don't support 16-bit machines
2615
in GNU.
2616
 
2617
Similarly, don't make any effort to cater to the possibility that
2618
@code{long} will be smaller than predefined types like @code{size_t}.
2619
For example, the following code is ok:
2620
 
2621
@example
2622
printf ("size = %lu\n", (unsigned long) sizeof array);
2623
printf ("diff = %ld\n", (long) (pointer2 - pointer1));
2624
@end example
2625
 
2626
1989 Standard C requires this to work, and we know of only one
2627
counterexample: 64-bit programs on Microsoft Windows IA-64.  We will
2628
leave it to those who want to port GNU programs to that environment
2629
to figure out how to do it.
2630
 
2631
Predefined file-size types like @code{off_t} are an exception: they are
2632
longer than @code{long} on many platforms, so code like the above won't
2633
work with them.  One way to print an @code{off_t} value portably is to
2634
print its digits yourself, one by one.
2635
 
2636
Don't assume that the address of an @code{int} object is also the
2637
address of its least-significant byte.  This is false on big-endian
2638
machines.  Thus, don't make the following mistake:
2639
 
2640
@example
2641
int c;
2642
@dots{}
2643
while ((c = getchar()) != EOF)
2644
  write(file_descriptor, &c, 1);
2645
@end example
2646
 
2647
When calling functions, you need not worry about the difference between
2648
pointers of various types, or between pointers and integers.  On most
2649
machines, there's no difference anyway.  As for the few machines where
2650
there is a difference, all of them support Standard C prototypes, so you can
2651
use prototypes (perhaps conditionalized to be active only in Standard C)
2652
to make the code work on those systems.
2653
 
2654
In certain cases, it is ok to pass integer and pointer arguments
2655
indiscriminately to the same function, and use no prototype on any
2656
system.  For example, many GNU programs have error-reporting functions
2657
that pass their arguments along to @code{printf} and friends:
2658
 
2659
@example
2660
error (s, a1, a2, a3)
2661
     char *s;
2662
     char *a1, *a2, *a3;
2663
@{
2664
  fprintf (stderr, "error: ");
2665
  fprintf (stderr, s, a1, a2, a3);
2666
@}
2667
@end example
2668
 
2669
@noindent
2670
In practice, this works on all machines, since a pointer is generally
2671
the widest possible kind of argument; it is much simpler than any
2672
``correct'' alternative.  Be sure @emph{not} to use a prototype for such
2673
functions.
2674
 
2675
If you have decided to use Standard C, then you can instead define
2676
@code{error} using @file{stdarg.h}, and pass the arguments along to
2677
@code{vfprintf}.
2678
 
2679
@cindex casting pointers to integers
2680
Avoid casting pointers to integers if you can.  Such casts greatly
2681
reduce portability, and in most programs they are easy to avoid.  In the
2682
cases where casting pointers to integers is essential---such as, a Lisp
2683
interpreter which stores type information as well as an address in one
2684
word---you'll have to make explicit provisions to handle different word
2685
sizes.  You will also need to make provision for systems in which the
2686
normal range of addresses you can get from @code{malloc} starts far away
2687
from zero.
2688
 
2689
@node System Functions
2690
@section Calling System Functions
2691
@cindex library functions, and portability
2692
@cindex portability, and library functions
2693
 
2694
C implementations differ substantially.  Standard C reduces but does
2695
not eliminate the incompatibilities; meanwhile, many GNU packages still
2696
support pre-standard compilers because this is not hard to do.  This
2697
chapter gives recommendations for how to use the more-or-less standard C
2698
library functions to avoid unnecessary loss of portability.
2699
 
2700
@itemize @bullet
2701
@item
2702
Don't use the return value of @code{sprintf}.  It returns the number of
2703
characters written on some systems, but not on all systems.
2704
 
2705
@item
2706
Be aware that @code{vfprintf} is not always available.
2707
 
2708
@item
2709
@code{main} should be declared to return type @code{int}.  It should
2710
terminate either by calling @code{exit} or by returning the integer
2711
status code; make sure it cannot ever return an undefined value.
2712
 
2713
@cindex declaration for system functions
2714
@item
2715
Don't declare system functions explicitly.
2716
 
2717
Almost any declaration for a system function is wrong on some system.
2718
To minimize conflicts, leave it to the system header files to declare
2719
system functions.  If the headers don't declare a function, let it
2720
remain undeclared.
2721
 
2722
While it may seem unclean to use a function without declaring it, in
2723
practice this works fine for most system library functions on the
2724
systems where this really happens; thus, the disadvantage is only
2725
theoretical.  By contrast, actual declarations have frequently caused
2726
actual conflicts.
2727
 
2728
@item
2729
If you must declare a system function, don't specify the argument types.
2730
Use an old-style declaration, not a Standard C prototype.  The more you
2731
specify about the function, the more likely a conflict.
2732
 
2733
@item
2734
In particular, don't unconditionally declare @code{malloc} or
2735
@code{realloc}.
2736
 
2737
Most GNU programs use those functions just once, in functions
2738
conventionally named @code{xmalloc} and @code{xrealloc}.  These
2739
functions call @code{malloc} and @code{realloc}, respectively, and
2740
check the results.
2741
 
2742
Because @code{xmalloc} and @code{xrealloc} are defined in your program,
2743
you can declare them in other files without any risk of type conflict.
2744
 
2745
On most systems, @code{int} is the same length as a pointer; thus, the
2746
calls to @code{malloc} and @code{realloc} work fine.  For the few
2747
exceptional systems (mostly 64-bit machines), you can use
2748
@strong{conditionalized} declarations of @code{malloc} and
2749
@code{realloc}---or put these declarations in configuration files
2750
specific to those systems.
2751
 
2752
@cindex string library functions
2753
@item
2754
The string functions require special treatment.  Some Unix systems have
2755
a header file @file{string.h}; others have @file{strings.h}.  Neither
2756
file name is portable.  There are two things you can do: use Autoconf to
2757
figure out which file to include, or don't include either file.
2758
 
2759
@item
2760
If you don't include either strings file, you can't get declarations for
2761
the string functions from the header file in the usual way.
2762
 
2763
That causes less of a problem than you might think.  The newer standard
2764
string functions should be avoided anyway because many systems still
2765
don't support them.  The string functions you can use are these:
2766
 
2767
@example
2768
strcpy   strncpy   strcat   strncat
2769
strlen   strcmp    strncmp
2770
strchr   strrchr
2771
@end example
2772
 
2773
The copy and concatenate functions work fine without a declaration as
2774
long as you don't use their values.  Using their values without a
2775
declaration fails on systems where the width of a pointer differs from
2776
the width of @code{int}, and perhaps in other cases.  It is trivial to
2777
avoid using their values, so do that.
2778
 
2779
The compare functions and @code{strlen} work fine without a declaration
2780
on most systems, possibly all the ones that GNU software runs on.
2781
You may find it necessary to declare them @strong{conditionally} on a
2782
few systems.
2783
 
2784
The search functions must be declared to return @code{char *}.  Luckily,
2785
there is no variation in the data type they return.  But there is
2786
variation in their names.  Some systems give these functions the names
2787
@code{index} and @code{rindex}; other systems use the names
2788
@code{strchr} and @code{strrchr}.  Some systems support both pairs of
2789
names, but neither pair works on all systems.
2790
 
2791
You should pick a single pair of names and use it throughout your
2792
program.  (Nowadays, it is better to choose @code{strchr} and
2793
@code{strrchr} for new programs, since those are the standard
2794
names.)  Declare both of those names as functions returning @code{char
2795
*}.  On systems which don't support those names, define them as macros
2796
in terms of the other pair.  For example, here is what to put at the
2797
beginning of your file (or in a header) if you want to use the names
2798
@code{strchr} and @code{strrchr} throughout:
2799
 
2800
@example
2801
#ifndef HAVE_STRCHR
2802
#define strchr index
2803
#endif
2804
#ifndef HAVE_STRRCHR
2805
#define strrchr rindex
2806
#endif
2807
 
2808
char *strchr ();
2809
char *strrchr ();
2810
@end example
2811
@end itemize
2812
 
2813
Here we assume that @code{HAVE_STRCHR} and @code{HAVE_STRRCHR} are
2814
macros defined in systems where the corresponding functions exist.
2815
One way to get them properly defined is to use Autoconf.
2816
 
2817
@node Internationalization
2818
@section Internationalization
2819
@cindex internationalization
2820
 
2821
@pindex gettext
2822
GNU has a library called GNU gettext that makes it easy to translate the
2823
messages in a program into various languages.  You should use this
2824
library in every program.  Use English for the messages as they appear
2825
in the program, and let gettext provide the way to translate them into
2826
other languages.
2827
 
2828
Using GNU gettext involves putting a call to the @code{gettext} macro
2829
around each string that might need translation---like this:
2830
 
2831
@example
2832
printf (gettext ("Processing file `%s'..."));
2833
@end example
2834
 
2835
@noindent
2836
This permits GNU gettext to replace the string @code{"Processing file
2837
`%s'..."} with a translated version.
2838
 
2839
Once a program uses gettext, please make a point of writing calls to
2840
@code{gettext} when you add new strings that call for translation.
2841
 
2842
Using GNU gettext in a package involves specifying a @dfn{text domain
2843
name} for the package.  The text domain name is used to separate the
2844
translations for this package from the translations for other packages.
2845
Normally, the text domain name should be the same as the name of the
2846
package---for example, @samp{fileutils} for the GNU file utilities.
2847
 
2848
@cindex message text, and internationalization
2849
To enable gettext to work well, avoid writing code that makes
2850
assumptions about the structure of words or sentences.  When you want
2851
the precise text of a sentence to vary depending on the data, use two or
2852
more alternative string constants each containing a complete sentences,
2853
rather than inserting conditionalized words or phrases into a single
2854
sentence framework.
2855
 
2856
Here is an example of what not to do:
2857
 
2858
@example
2859
printf ("%d file%s processed", nfiles,
2860
        nfiles != 1 ? "s" : "");
2861
@end example
2862
 
2863
@noindent
2864
The problem with that example is that it assumes that plurals are made
2865
by adding `s'.  If you apply gettext to the format string, like this,
2866
 
2867
@example
2868
printf (gettext ("%d file%s processed"), nfiles,
2869
        nfiles != 1 ? "s" : "");
2870
@end example
2871
 
2872
@noindent
2873
the message can use different words, but it will still be forced to use
2874
`s' for the plural.  Here is a better way:
2875
 
2876
@example
2877
printf ((nfiles != 1 ? "%d files processed"
2878
         : "%d file processed"),
2879
        nfiles);
2880
@end example
2881
 
2882
@noindent
2883
This way, you can apply gettext to each of the two strings
2884
independently:
2885
 
2886
@example
2887
printf ((nfiles != 1 ? gettext ("%d files processed")
2888
         : gettext ("%d file processed")),
2889
        nfiles);
2890
@end example
2891
 
2892
@noindent
2893
This can be any method of forming the plural of the word for ``file'', and
2894
also handles languages that require agreement in the word for
2895
``processed''.
2896
 
2897
A similar problem appears at the level of sentence structure with this
2898
code:
2899
 
2900
@example
2901
printf ("#  Implicit rule search has%s been done.\n",
2902
        f->tried_implicit ? "" : " not");
2903
@end example
2904
 
2905
@noindent
2906
Adding @code{gettext} calls to this code cannot give correct results for
2907
all languages, because negation in some languages requires adding words
2908
at more than one place in the sentence.  By contrast, adding
2909
@code{gettext} calls does the job straightfowardly if the code starts
2910
out like this:
2911
 
2912
@example
2913
printf (f->tried_implicit
2914
        ? "#  Implicit rule search has been done.\n",
2915
        : "#  Implicit rule search has not been done.\n");
2916
@end example
2917
 
2918
@node Mmap
2919
@section Mmap
2920
@findex mmap
2921
 
2922
Don't assume that @code{mmap} either works on all files or fails
2923
for all files.  It may work on some files and fail on others.
2924
 
2925
The proper way to use @code{mmap} is to try it on the specific file for
2926
which you want to use it---and if @code{mmap} doesn't work, fall back on
2927
doing the job in another way using @code{read} and @code{write}.
2928
 
2929
The reason this precaution is needed is that the GNU kernel (the HURD)
2930
provides a user-extensible file system, in which there can be many
2931
different kinds of ``ordinary files.''  Many of them support
2932
@code{mmap}, but some do not.  It is important to make programs handle
2933
all these kinds of files.
2934
 
2935
@node Documentation
2936
@chapter Documenting Programs
2937
@cindex documentation
2938
 
2939
A GNU program should ideally come with full free documentation, adequate
2940
for both reference and tutorial purposes.  If the package can be
2941
programmed or extended, the documentation should cover programming or
2942
extending it, as well as just using it.
2943
 
2944
@menu
2945
* GNU Manuals::                 Writing proper manuals.
2946
* Doc Strings and Manuals::     Compiling doc strings doesn't make a manual.
2947
* Manual Structure Details::    Specific structure conventions.
2948
* License for Manuals::         Writing the distribution terms for a manual.
2949
* Manual Credits::              Giving credit to documentation contributors.
2950
* Printed Manuals::             Mentioning the printed manual.
2951
* NEWS File::                   NEWS files supplement manuals.
2952
* Change Logs::                 Recording Changes
2953
* Man Pages::                   Man pages are secondary.
2954
* Reading other Manuals::       How far you can go in learning
2955
                                from other manuals.
2956
@end menu
2957
 
2958
@node GNU Manuals
2959
@section GNU Manuals
2960
 
2961
The preferred document format for the GNU system is the Texinfo
2962
formatting language.  Every GNU package should (ideally) have
2963
documentation in Texinfo both for reference and for learners.  Texinfo
2964
makes it possible to produce a good quality formatted book, using
2965
@TeX{}, and to generate an Info file.  It is also possible to generate
2966
HTML output from Texinfo source.  See the Texinfo manual, either the
2967
hardcopy, or the on-line version available through @code{info} or the
2968
Emacs Info subsystem (@kbd{C-h i}).
2969
 
2970
Nowadays some other formats such as Docbook and Sgmltexi can be
2971
converted automatically into Texinfo.  It is ok to produce the Texinfo
2972
documentation by conversion this way, as long as it gives good results.
2973
 
2974
Programmers often find it most natural to structure the documentation
2975
following the structure of the implementation, which they know.  But
2976
this structure is not necessarily good for explaining how to use the
2977
program; it may be irrelevant and confusing for a user.
2978
 
2979
At every level, from the sentences in a paragraph to the grouping of
2980
topics into separate manuals, the right way to structure documentation
2981
is according to the concepts and questions that a user will have in mind
2982
when reading it.  Sometimes this structure of ideas matches the
2983
structure of the implementation of the software being documented---but
2984
often they are different.  Often the most important part of learning to
2985
write good documentation is learning to notice when you are structuring
2986
the documentation like the implementation, and think about better
2987
alternatives.
2988
 
2989
For example, each program in the GNU system probably ought to be
2990
documented in one manual; but this does not mean each program should
2991
have its own manual.  That would be following the structure of the
2992
implementation, rather than the structure that helps the user
2993
understand.
2994
 
2995
Instead, each manual should cover a coherent @emph{topic}.  For example,
2996
instead of a manual for @code{diff} and a manual for @code{diff3}, we
2997
have one manual for ``comparison of files'' which covers both of those
2998
programs, as well as @code{cmp}.  By documenting these programs
2999
together, we can make the whole subject clearer.
3000
 
3001
The manual which discusses a program should certainly document all of
3002
the program's command-line options and all of its commands.  It should
3003
give examples of their use.  But don't organize the manual as a list of
3004
features.  Instead, organize it logically, by subtopics.  Address the
3005
questions that a user will ask when thinking about the job that the
3006
program does.
3007
 
3008
In general, a GNU manual should serve both as tutorial and reference.
3009
It should be set up for convenient access to each topic through Info,
3010
and for reading straight through (appendixes aside).  A GNU manual
3011
should give a good introduction to a beginner reading through from the
3012
start, and should also provide all the details that hackers want.
3013
The Bison manual is a good example of this---please take a look at it
3014
to see what we mean.
3015
 
3016
That is not as hard as it first sounds.  Arrange each chapter as a
3017
logical breakdown of its topic, but order the sections, and write their
3018
text, so that reading the chapter straight through makes sense.  Do
3019
likewise when structuring the book into chapters, and when structuring a
3020
section into paragraphs.  The watchword is, @emph{at each point, address
3021
the most fundamental and important issue raised by the preceding text.}
3022
 
3023
If necessary, add extra chapters at the beginning of the manual which
3024
are purely tutorial and cover the basics of the subject.  These provide
3025
the framework for a beginner to understand the rest of the manual.  The
3026
Bison manual provides a good example of how to do this.
3027
 
3028
To serve as a reference, a manual should have an Index that list all the
3029
functions, variables, options, and important concepts that are part of
3030
the program.  One combined Index should do for a short manual, but
3031
sometimes for a complex package it is better to use multiple indices.
3032
The Texinfo manual includes advice on preparing good index entries, see
3033
@ref{Index Entries, , Making Index Entries, texinfo, The GNU Texinfo
3034
Manual}, and see @ref{Indexing Commands, , Defining the Entries of an
3035
Index, texinfo, The GNU Texinfo manual}.
3036
 
3037
Don't use Unix man pages as a model for how to write GNU documentation;
3038
most of them are terse, badly structured, and give inadequate
3039
explanation of the underlying concepts.  (There are, of course, some
3040
exceptions.)  Also, Unix man pages use a particular format which is
3041
different from what we use in GNU manuals.
3042
 
3043
Please include an email address in the manual for where to report
3044
bugs @emph{in the manual}.
3045
 
3046
Please do not use the term ``pathname'' that is used in Unix
3047
documentation; use ``file name'' (two words) instead.  We use the term
3048
``path'' only for search paths, which are lists of directory names.
3049
 
3050
Please do not use the term ``illegal'' to refer to erroneous input to a
3051
computer program.  Please use ``invalid'' for this, and reserve the term
3052
``illegal'' for activities punishable by law.
3053
 
3054
@node Doc Strings and Manuals
3055
@section Doc Strings and Manuals
3056
 
3057
Some programming systems, such as Emacs, provide a documentation string
3058
for each function, command or variable.  You may be tempted to write a
3059
reference manual by compiling the documentation strings and writing a
3060
little additional text to go around them---but you must not do it.  That
3061
approach is a fundamental mistake.  The text of well-written
3062
documentation strings will be entirely wrong for a manual.
3063
 
3064
A documentation string needs to stand alone---when it appears on the
3065
screen, there will be no other text to introduce or explain it.
3066
Meanwhile, it can be rather informal in style.
3067
 
3068
The text describing a function or variable in a manual must not stand
3069
alone; it appears in the context of a section or subsection.  Other text
3070
at the beginning of the section should explain some of the concepts, and
3071
should often make some general points that apply to several functions or
3072
variables.  The previous descriptions of functions and variables in the
3073
section will also have given information about the topic.  A description
3074
written to stand alone would repeat some of that information; this
3075
redundance looks bad.  Meanwhile, the informality that is acceptable in
3076
a documentation string is totally unacceptable in a manual.
3077
 
3078
The only good way to use documentation strings in writing a good manual
3079
is to use them as a source of information for writing good text.
3080
 
3081
@node Manual Structure Details
3082
@section Manual Structure Details
3083
@cindex manual structure
3084
 
3085
The title page of the manual should state the version of the programs or
3086
packages documented in the manual.  The Top node of the manual should
3087
also contain this information.  If the manual is changing more
3088
frequently than or independent of the program, also state a version
3089
number for the manual in both of these places.
3090
 
3091
Each program documented in the manual should have a node named
3092
@samp{@var{program} Invocation} or @samp{Invoking @var{program}}.  This
3093
node (together with its subnodes, if any) should describe the program's
3094
command line arguments and how to run it (the sort of information people
3095
would look in a man page for).  Start with an @samp{@@example}
3096
containing a template for all the options and arguments that the program
3097
uses.
3098
 
3099
Alternatively, put a menu item in some menu whose item name fits one of
3100
the above patterns.  This identifies the node which that item points to
3101
as the node for this purpose, regardless of the node's actual name.
3102
 
3103
The @samp{--usage} feature of the Info reader looks for such a node
3104
or menu item in order to find the relevant text, so it is essential
3105
for every Texinfo file to have one.
3106
 
3107
If one manual describes several programs, it should have such a node for
3108
each program described in the manual.
3109
 
3110
@node License for Manuals
3111
@section License for Manuals
3112
@cindex license for manuals
3113
 
3114
Please use the GNU Free Documentation License for all GNU manuals that
3115
are more than a few pages long.  Likewise for a collection of short
3116
documents---you only need one copy of the GNU FDL for the whole
3117
collection.  For a single short document, you can use a very permissive
3118
non-copyleft license, to avoid taking up space with a long license.
3119
 
3120
See @uref{http://www.gnu.org/copyleft/fdl-howto.html} for more explanation
3121
of how to employ the GFDL.
3122
 
3123
Note that it is not obligatory to include a copy of the GNU GPL or GNU
3124
LGPL in a manual whose license is neither the GPL nor the LGPL.  It can
3125
be a good idea to include the program's license in a large manual; in a
3126
short manual, whose size would be increased considerably by including
3127
the program's license, it is probably better not to include it.
3128
 
3129
@node Manual Credits
3130
@section Manual Credits
3131
@cindex credits for manuals
3132
 
3133
Please credit the principal human writers of the manual as the authors,
3134
on the title page of the manual.  If a company sponsored the work, thank
3135
the company in a suitable place in the manual, but do not cite the
3136
company as an author.
3137
 
3138
@node Printed Manuals
3139
@section Printed Manuals
3140
 
3141
The FSF publishes some GNU manuals in printed form.  To encourage sales
3142
of these manuals, the on-line versions of the manual should mention at
3143
the very start that the printed manual is available and should point at
3144
information for getting it---for instance, with a link to the page
3145
@url{http://www.gnu.org/order/order.html}.  This should not be included
3146
in the printed manual, though, because there it is redundant.
3147
 
3148
It is also useful to explain in the on-line forms of the manual how the
3149
user can print out the manual from the sources.
3150
 
3151
@node NEWS File
3152
@section The NEWS File
3153
@cindex @file{NEWS} file
3154
 
3155
In addition to its manual, the package should have a file named
3156
@file{NEWS} which contains a list of user-visible changes worth
3157
mentioning.  In each new release, add items to the front of the file and
3158
identify the version they pertain to.  Don't discard old items; leave
3159
them in the file after the newer items.  This way, a user upgrading from
3160
any previous version can see what is new.
3161
 
3162
If the @file{NEWS} file gets very long, move some of the older items
3163
into a file named @file{ONEWS} and put a note at the end referring the
3164
user to that file.
3165
 
3166
@node Change Logs
3167
@section Change Logs
3168
@cindex change logs
3169
 
3170
Keep a change log to describe all the changes made to program source
3171
files.  The purpose of this is so that people investigating bugs in the
3172
future will know about the changes that might have introduced the bug.
3173
Often a new bug can be found by looking at what was recently changed.
3174
More importantly, change logs can help you eliminate conceptual
3175
inconsistencies between different parts of a program, by giving you a
3176
history of how the conflicting concepts arose and who they came from.
3177
 
3178
@menu
3179
* Change Log Concepts::
3180
* Style of Change Logs::
3181
* Simple Changes::
3182
* Conditional Changes::
3183
* Indicating the Part Changed::
3184
@end menu
3185
 
3186
@node Change Log Concepts
3187
@subsection Change Log Concepts
3188
 
3189
You can think of the change log as a conceptual ``undo list'' which
3190
explains how earlier versions were different from the current version.
3191
People can see the current version; they don't need the change log
3192
to tell them what is in it.  What they want from a change log is a
3193
clear explanation of how the earlier version differed.
3194
 
3195
The change log file is normally called @file{ChangeLog} and covers an
3196
entire directory.  Each directory can have its own change log, or a
3197
directory can use the change log of its parent directory--it's up to
3198
you.
3199
 
3200
Another alternative is to record change log information with a version
3201
control system such as RCS or CVS.  This can be converted automatically
3202
to a @file{ChangeLog} file using @code{rcs2log}; in Emacs, the command
3203
@kbd{C-x v a} (@code{vc-update-change-log}) does the job.
3204
 
3205
There's no need to describe the full purpose of the changes or how they
3206
work together.  If you think that a change calls for explanation, you're
3207
probably right.  Please do explain it---but please put the explanation
3208
in comments in the code, where people will see it whenever they see the
3209
code.  For example, ``New function'' is enough for the change log when
3210
you add a function, because there should be a comment before the
3211
function definition to explain what it does.
3212
 
3213
However, sometimes it is useful to write one line to describe the
3214
overall purpose of a batch of changes.
3215
 
3216
The easiest way to add an entry to @file{ChangeLog} is with the Emacs
3217
command @kbd{M-x add-change-log-entry}.  An entry should have an
3218
asterisk, the name of the changed file, and then in parentheses the name
3219
of the changed functions, variables or whatever, followed by a colon.
3220
Then describe the changes you made to that function or variable.
3221
 
3222
@node Style of Change Logs
3223
@subsection Style of Change Logs
3224
@cindex change logs, style
3225
 
3226
Here are some simple examples of change log entries, starting with the
3227
header line that says who made the change and when, followed by
3228
descriptions of specific changes.  (These examples are drawn from Emacs
3229
and GCC.)
3230
 
3231
@example
3232
1998-08-17  Richard Stallman  <rms@@gnu.org>
3233
 
3234
* register.el (insert-register): Return nil.
3235
(jump-to-register): Likewise.
3236
 
3237
* sort.el (sort-subr): Return nil.
3238
 
3239
* tex-mode.el (tex-bibtex-file, tex-file, tex-region):
3240
Restart the tex shell if process is gone or stopped.
3241
(tex-shell-running): New function.
3242
 
3243
* expr.c (store_one_arg): Round size up for move_block_to_reg.
3244
(expand_call): Round up when emitting USE insns.
3245
* stmt.c (assign_parms): Round size up for move_block_from_reg.
3246
@end example
3247
 
3248
It's important to name the changed function or variable in full.  Don't
3249
abbreviate function or variable names, and don't combine them.
3250
Subsequent maintainers will often search for a function name to find all
3251
the change log entries that pertain to it; if you abbreviate the name,
3252
they won't find it when they search.
3253
 
3254
For example, some people are tempted to abbreviate groups of function
3255
names by writing @samp{* register.el (@{insert,jump-to@}-register)};
3256
this is not a good idea, since searching for @code{jump-to-register} or
3257
@code{insert-register} would not find that entry.
3258
 
3259
Separate unrelated change log entries with blank lines.  When two
3260
entries represent parts of the same change, so that they work together,
3261
then don't put blank lines between them.  Then you can omit the file
3262
name and the asterisk when successive entries are in the same file.
3263
 
3264
Break long lists of function names by closing continued lines with
3265
@samp{)}, rather than @samp{,}, and opening the continuation with
3266
@samp{(} as in this example:
3267
 
3268
@example
3269
* keyboard.c (menu_bar_items, tool_bar_items)
3270
(Fexecute_extended_command): Deal with `keymap' property.
3271
@end example
3272
 
3273
@node Simple Changes
3274
@subsection Simple Changes
3275
 
3276
Certain simple kinds of changes don't need much detail in the change
3277
log.
3278
 
3279
When you change the calling sequence of a function in a simple fashion,
3280
and you change all the callers of the function to use the new calling
3281
sequence, there is no need to make individual entries for all the
3282
callers that you changed.  Just write in the entry for the function
3283
being called, ``All callers changed''---like this:
3284
 
3285
@example
3286
* keyboard.c (Fcommand_execute): New arg SPECIAL.
3287
All callers changed.
3288
@end example
3289
 
3290
When you change just comments or doc strings, it is enough to write an
3291
entry for the file, without mentioning the functions.  Just ``Doc
3292
fixes'' is enough for the change log.
3293
 
3294
There's no need to make change log entries for documentation files.
3295
This is because documentation is not susceptible to bugs that are hard
3296
to fix.  Documentation does not consist of parts that must interact in a
3297
precisely engineered fashion.  To correct an error, you need not know
3298
the history of the erroneous passage; it is enough to compare what the
3299
documentation says with the way the program actually works.
3300
 
3301
@node Conditional Changes
3302
@subsection Conditional Changes
3303
@cindex conditional changes, and change logs
3304
@cindex change logs, conditional changes
3305
 
3306
C programs often contain compile-time @code{#if} conditionals.  Many
3307
changes are conditional; sometimes you add a new definition which is
3308
entirely contained in a conditional.  It is very useful to indicate in
3309
the change log the conditions for which the change applies.
3310
 
3311
Our convention for indicating conditional changes is to use square
3312
brackets around the name of the condition.
3313
 
3314
Here is a simple example, describing a change which is conditional but
3315
does not have a function or entity name associated with it:
3316
 
3317
@example
3318
* xterm.c [SOLARIS2]: Include string.h.
3319
@end example
3320
 
3321
Here is an entry describing a new definition which is entirely
3322
conditional.  This new definition for the macro @code{FRAME_WINDOW_P} is
3323
used only when @code{HAVE_X_WINDOWS} is defined:
3324
 
3325
@example
3326
* frame.h [HAVE_X_WINDOWS] (FRAME_WINDOW_P): Macro defined.
3327
@end example
3328
 
3329
Here is an entry for a change within the function @code{init_display},
3330
whose definition as a whole is unconditional, but the changes themselves
3331
are contained in a @samp{#ifdef HAVE_LIBNCURSES} conditional:
3332
 
3333
@example
3334
* dispnew.c (init_display) [HAVE_LIBNCURSES]: If X, call tgetent.
3335
@end example
3336
 
3337
Here is an entry for a change that takes affect only when
3338
a certain macro is @emph{not} defined:
3339
 
3340
@example
3341
(gethostname) [!HAVE_SOCKETS]: Replace with winsock version.
3342
@end example
3343
 
3344
@node Indicating the Part Changed
3345
@subsection Indicating the Part Changed
3346
 
3347
Indicate the part of a function which changed by using angle brackets
3348
enclosing an indication of what the changed part does.  Here is an entry
3349
for a change in the part of the function @code{sh-while-getopts} that
3350
deals with @code{sh} commands:
3351
 
3352
@example
3353
* progmodes/sh-script.el (sh-while-getopts) <sh>: Handle case that
3354
user-specified option string is empty.
3355
@end example
3356
 
3357
 
3358
@node Man Pages
3359
@section Man Pages
3360
@cindex man pages
3361
 
3362
In the GNU project, man pages are secondary.  It is not necessary or
3363
expected for every GNU program to have a man page, but some of them do.
3364
It's your choice whether to include a man page in your program.
3365
 
3366
When you make this decision, consider that supporting a man page
3367
requires continual effort each time the program is changed.  The time
3368
you spend on the man page is time taken away from more useful work.
3369
 
3370
For a simple program which changes little, updating the man page may be
3371
a small job.  Then there is little reason not to include a man page, if
3372
you have one.
3373
 
3374
For a large program that changes a great deal, updating a man page may
3375
be a substantial burden.  If a user offers to donate a man page, you may
3376
find this gift costly to accept.  It may be better to refuse the man
3377
page unless the same person agrees to take full responsibility for
3378
maintaining it---so that you can wash your hands of it entirely.  If
3379
this volunteer later ceases to do the job, then don't feel obliged to
3380
pick it up yourself; it may be better to withdraw the man page from the
3381
distribution until someone else agrees to update it.
3382
 
3383
When a program changes only a little, you may feel that the
3384
discrepancies are small enough that the man page remains useful without
3385
updating.  If so, put a prominent note near the beginning of the man
3386
page explaining that you don't maintain it and that the Texinfo manual
3387
is more authoritative.  The note should say how to access the Texinfo
3388
documentation.
3389
 
3390
@node Reading other Manuals
3391
@section Reading other Manuals
3392
 
3393
There may be non-free books or documentation files that describe the
3394
program you are documenting.
3395
 
3396
It is ok to use these documents for reference, just as the author of a
3397
new algebra textbook can read other books on algebra.  A large portion
3398
of any non-fiction book consists of facts, in this case facts about how
3399
a certain program works, and these facts are necessarily the same for
3400
everyone who writes about the subject.  But be careful not to copy your
3401
outline structure, wording, tables or examples from preexisting non-free
3402
documentation.  Copying from free documentation may be ok; please check
3403
with the FSF about the individual case.
3404
 
3405
@node Managing Releases
3406
@chapter The Release Process
3407
@cindex releasing
3408
 
3409
Making a release is more than just bundling up your source files in a
3410
tar file and putting it up for FTP.  You should set up your software so
3411
that it can be configured to run on a variety of systems.  Your Makefile
3412
should conform to the GNU standards described below, and your directory
3413
layout should also conform to the standards discussed below.  Doing so
3414
makes it easy to include your package into the larger framework of
3415
all GNU software.
3416
 
3417
@menu
3418
* Configuration::               How Configuration Should Work
3419
* Makefile Conventions::        Makefile Conventions
3420
* Releases::                    Making Releases
3421
@end menu
3422
 
3423
@node Configuration
3424
@section How Configuration Should Work
3425
@cindex program configuration
3426
 
3427
@pindex configure
3428
Each GNU distribution should come with a shell script named
3429
@code{configure}.  This script is given arguments which describe the
3430
kind of machine and system you want to compile the program for.
3431
 
3432
The @code{configure} script must record the configuration options so
3433
that they affect compilation.
3434
 
3435
One way to do this is to make a link from a standard name such as
3436
@file{config.h} to the proper configuration file for the chosen system.
3437
If you use this technique, the distribution should @emph{not} contain a
3438
file named @file{config.h}.  This is so that people won't be able to
3439
build the program without configuring it first.
3440
 
3441
Another thing that @code{configure} can do is to edit the Makefile.  If
3442
you do this, the distribution should @emph{not} contain a file named
3443
@file{Makefile}.  Instead, it should include a file @file{Makefile.in} which
3444
contains the input used for editing.  Once again, this is so that people
3445
won't be able to build the program without configuring it first.
3446
 
3447
If @code{configure} does write the @file{Makefile}, then @file{Makefile}
3448
should have a target named @file{Makefile} which causes @code{configure}
3449
to be rerun, setting up the same configuration that was set up last
3450
time.  The files that @code{configure} reads should be listed as
3451
dependencies of @file{Makefile}.
3452
 
3453
All the files which are output from the @code{configure} script should
3454
have comments at the beginning explaining that they were generated
3455
automatically using @code{configure}.  This is so that users won't think
3456
of trying to edit them by hand.
3457
 
3458
The @code{configure} script should write a file named @file{config.status}
3459
which describes which configuration options were specified when the
3460
program was last configured.  This file should be a shell script which,
3461
if run, will recreate the same configuration.
3462
 
3463
The @code{configure} script should accept an option of the form
3464
@samp{--srcdir=@var{dirname}} to specify the directory where sources are found
3465
(if it is not the current directory).  This makes it possible to build
3466
the program in a separate directory, so that the actual source directory
3467
is not modified.
3468
 
3469
If the user does not specify @samp{--srcdir}, then @code{configure} should
3470
check both @file{.} and @file{..} to see if it can find the sources.  If
3471
it finds the sources in one of these places, it should use them from
3472
there.  Otherwise, it should report that it cannot find the sources, and
3473
should exit with nonzero status.
3474
 
3475
Usually the easy way to support @samp{--srcdir} is by editing a
3476
definition of @code{VPATH} into the Makefile.  Some rules may need to
3477
refer explicitly to the specified source directory.  To make this
3478
possible, @code{configure} can add to the Makefile a variable named
3479
@code{srcdir} whose value is precisely the specified directory.
3480
 
3481
The @code{configure} script should also take an argument which specifies the
3482
type of system to build the program for.  This argument should look like
3483
this:
3484
 
3485
@example
3486
@var{cpu}-@var{company}-@var{system}
3487
@end example
3488
 
3489
For example, a Sun 3 might be @samp{m68k-sun-sunos4.1}.
3490
 
3491
The @code{configure} script needs to be able to decode all plausible
3492
alternatives for how to describe a machine.  Thus, @samp{sun3-sunos4.1}
3493
would be a valid alias.  For many programs, @samp{vax-dec-ultrix} would
3494
be an alias for @samp{vax-dec-bsd}, simply because the differences
3495
between Ultrix and @sc{bsd} are rarely noticeable, but a few programs
3496
might need to distinguish them.
3497
@c Real 4.4BSD now runs on some Suns.
3498
 
3499
There is a shell script called @file{config.sub} that you can use
3500
as a subroutine to validate system types and canonicalize aliases.
3501
 
3502
@cindex optional features, configure-time
3503
Other options are permitted to specify in more detail the software
3504
or hardware present on the machine, and include or exclude optional
3505
parts of the package:
3506
 
3507
@table @samp
3508
@item --enable-@var{feature}@r{[}=@var{parameter}@r{]}
3509
Configure the package to build and install an optional user-level
3510
facility called @var{feature}.  This allows users to choose which
3511
optional features to include.  Giving an optional @var{parameter} of
3512
@samp{no} should omit @var{feature}, if it is built by default.
3513
 
3514
No @samp{--enable} option should @strong{ever} cause one feature to
3515
replace another.  No @samp{--enable} option should ever substitute one
3516
useful behavior for another useful behavior.  The only proper use for
3517
@samp{--enable} is for questions of whether to build part of the program
3518
or exclude it.
3519
 
3520
@item --with-@var{package}
3521
@c @r{[}=@var{parameter}@r{]}
3522
The package @var{package} will be installed, so configure this package
3523
to work with @var{package}.
3524
 
3525
@c  Giving an optional @var{parameter} of
3526
@c @samp{no} should omit @var{package}, if it is used by default.
3527
 
3528
Possible values of @var{package} include
3529
@samp{gnu-as} (or @samp{gas}), @samp{gnu-ld}, @samp{gnu-libc},
3530
@samp{gdb},
3531
@samp{x},
3532
and
3533
@samp{x-toolkit}.
3534
 
3535
Do not use a @samp{--with} option to specify the file name to use to
3536
find certain files.  That is outside the scope of what @samp{--with}
3537
options are for.
3538
@end table
3539
 
3540
All @code{configure} scripts should accept all of these ``detail''
3541
options, whether or not they make any difference to the particular
3542
package at hand.  In particular, they should accept any option that
3543
starts with @samp{--with-} or @samp{--enable-}.  This is so users will
3544
be able to configure an entire GNU source tree at once with a single set
3545
of options.
3546
 
3547
You will note that the categories @samp{--with-} and @samp{--enable-}
3548
are narrow: they @strong{do not} provide a place for any sort of option
3549
you might think of.  That is deliberate.  We want to limit the possible
3550
configuration options in GNU software.  We do not want GNU programs to
3551
have idiosyncratic configuration options.
3552
 
3553
Packages that perform part of the compilation process may support
3554
cross-compilation.  In such a case, the host and target machines for the
3555
program may be different.
3556
 
3557
The @code{configure} script should normally treat the specified type of
3558
system as both the host and the target, thus producing a program which
3559
works for the same type of machine that it runs on.
3560
 
3561
To configure a cross-compiler, cross-assembler, or what have you, you
3562
should specify a target different from the host, using the configure
3563
option @samp{--target=@var{targettype}}.  The syntax for
3564
@var{targettype} is the same as for the host type.  So the command would
3565
look like this:
3566
 
3567
@example
3568
./configure @var{hosttype} --target=@var{targettype}
3569
@end example
3570
 
3571
Programs for which cross-operation is not meaningful need not accept the
3572
@samp{--target} option, because configuring an entire operating system for
3573
cross-operation is not a meaningful operation.
3574
 
3575
Bootstrapping a cross-compiler requires compiling it on a machine other
3576
than the host it will run on.  Compilation packages accept a
3577
configuration option @samp{--build=@var{buildtype}} for specifying the
3578
configuration on which you will compile them, but the configure script
3579
should normally guess the build machine type (using
3580
@file{config.guess}), so this option is probably not necessary.  The
3581
host and target types normally default from the build type, so in
3582
bootstrapping a cross-compiler you must specify them both explicitly.
3583
 
3584
Some programs have ways of configuring themselves automatically.  If
3585
your program is set up to do this, your @code{configure} script can simply
3586
ignore most of its arguments.
3587
 
3588
@comment The makefile standards are in a separate file that is also
3589
@comment included by make.texinfo.  Done by roland@gnu.ai.mit.edu on 1/6/93.
3590
@comment For this document, turn chapters into sections, etc.
3591
@lowersections
3592
@include make-stds.texi
3593
@raisesections
3594
 
3595
@node Releases
3596
@section Making Releases
3597
@cindex packaging
3598
 
3599
Package the distribution of @code{Foo version 69.96} up in a gzipped tar
3600
file with the name @file{foo-69.96.tar.gz}.  It should unpack into a
3601
subdirectory named @file{foo-69.96}.
3602
 
3603
Building and installing the program should never modify any of the files
3604
contained in the distribution.  This means that all the files that form
3605
part of the program in any way must be classified into @dfn{source
3606
files} and @dfn{non-source files}.  Source files are written by humans
3607
and never changed automatically; non-source files are produced from
3608
source files by programs under the control of the Makefile.
3609
 
3610
@cindex @file{README} file
3611
The distribution should contain a file named @file{README} which gives
3612
the name of the package, and a general description of what it does.  It
3613
is also good to explain the purpose of each of the first-level
3614
subdirectories in the package, if there are any.  The @file{README} file
3615
should either state the version number of the package, or refer to where
3616
in the package it can be found.
3617
 
3618
The @file{README} file should refer to the file @file{INSTALL}, which
3619
should contain an explanation of the installation procedure.
3620
 
3621
The @file{README} file should also refer to the file which contains the
3622
copying conditions.  The GNU GPL, if used, should be in a file called
3623
@file{COPYING}.  If the GNU LGPL is used, it should be in a file called
3624
@file{COPYING.LIB}.
3625
 
3626
Naturally, all the source files must be in the distribution.  It is okay
3627
to include non-source files in the distribution, provided they are
3628
up-to-date and machine-independent, so that building the distribution
3629
normally will never modify them.  We commonly include non-source files
3630
produced by Bison, @code{lex}, @TeX{}, and @code{makeinfo}; this helps avoid
3631
unnecessary dependencies between our distributions, so that users can
3632
install whichever packages they want to install.
3633
 
3634
Non-source files that might actually be modified by building and
3635
installing the program should @strong{never} be included in the
3636
distribution.  So if you do distribute non-source files, always make
3637
sure they are up to date when you make a new distribution.
3638
 
3639
Make sure that the directory into which the distribution unpacks (as
3640
well as any subdirectories) are all world-writable (octal mode 777).
3641
This is so that old versions of @code{tar} which preserve the
3642
ownership and permissions of the files from the tar archive will be
3643
able to extract all the files even if the user is unprivileged.
3644
 
3645
Make sure that all the files in the distribution are world-readable.
3646
 
3647
Make sure that no file name in the distribution is more than 14
3648
characters long.  Likewise, no file created by building the program
3649
should have a name longer than 14 characters.  The reason for this is
3650
that some systems adhere to a foolish interpretation of the @sc{posix}
3651
standard, and refuse to open a longer name, rather than truncating as
3652
they did in the past.
3653
 
3654
Don't include any symbolic links in the distribution itself.  If the tar
3655
file contains symbolic links, then people cannot even unpack it on
3656
systems that don't support symbolic links.  Also, don't use multiple
3657
names for one file in different directories, because certain file
3658
systems cannot handle this and that prevents unpacking the
3659
distribution.
3660
 
3661
Try to make sure that all the file names will be unique on MS-DOS.  A
3662
name on MS-DOS consists of up to 8 characters, optionally followed by a
3663
period and up to three characters.  MS-DOS will truncate extra
3664
characters both before and after the period.  Thus,
3665
@file{foobarhacker.c} and @file{foobarhacker.o} are not ambiguous; they
3666
are truncated to @file{foobarha.c} and @file{foobarha.o}, which are
3667
distinct.
3668
 
3669
@cindex @file{texinfo.tex}, in a distribution
3670
Include in your distribution a copy of the @file{texinfo.tex} you used
3671
to test print any @file{*.texinfo} or @file{*.texi} files.
3672
 
3673
Likewise, if your program uses small GNU software packages like regex,
3674
getopt, obstack, or termcap, include them in the distribution file.
3675
Leaving them out would make the distribution file a little smaller at
3676
the expense of possible inconvenience to a user who doesn't know what
3677
other files to get.
3678
 
3679
@node References
3680
@chapter References to Non-Free Software and Documentation
3681
@cindex references to non-free material
3682
 
3683
A GNU program should not recommend use of any non-free program.  We
3684
can't stop some people from writing proprietary programs, or stop
3685
other people from using them, but we can and should avoid helping to
3686
advertise them to new potential customers.  Proprietary software is a
3687
social and ethical problem, and the point of GNU is to solve that
3688
problem.
3689
 
3690
When a non-free program or system is well known, you can mention it in
3691
passing---that is harmless, since users who might want to use it
3692
probably already know about it.  For instance, it is fine to explain
3693
how to build your package on top of some non-free operating system, or
3694
how to use it together with some widely used non-free program.
3695
 
3696
However, you should give only the necessary information to help those
3697
who already use the non-free program to use your program with
3698
it---don't give, or refer to, any further information about the
3699
proprietary program, and don't imply that the proprietary program
3700
enhances your program, or that its existence is in any way a good
3701
thing.  The goal should be that people already using the proprietary
3702
program will get the advice they need about how to use your free
3703
program, while people who don't already use the proprietary program
3704
will not see anything to lead them to take an interest in it.
3705
 
3706
If a non-free program or system is obscure in your program's domain,
3707
your program should not mention or support it at all, since doing so
3708
would tend to popularize the non-free program more than it popularizes
3709
your program.  (You cannot hope to find many additional users among
3710
the users of Foobar if the users of Foobar are few.)
3711
 
3712
A GNU package should not refer the user to any non-free documentation
3713
for free software.  Free documentation that can be included in free
3714
operating systems is essential for completing the GNU system, so it is
3715
a major focus of the GNU Project; to recommend use of documentation
3716
that we are not allowed to use in GNU would undermine the efforts to
3717
get documentation that we can include.  So GNU packages should never
3718
recommend non-free documentation.
3719
 
3720
@node Copying This Manual
3721
@appendix Copying This Manual
3722
 
3723
@menu
3724
* GNU Free Documentation License::  License for copying this manual
3725
@end menu
3726
 
3727
@include fdl.texi
3728
 
3729
@node Index
3730
@unnumbered Index
3731
@printindex cp
3732
 
3733
@contents
3734
 
3735
@bye
3736
@c Local variables:
3737
@c eval: (add-hook 'write-file-hooks 'time-stamp)
3738
@c time-stamp-start: "@set lastupdate "
3739
@c time-stamp-end: "$"
3740
@c time-stamp-format: "%:b %:d, %:y"
3741
@c compile-command: "make just-standards"
3742
@c End:

powered by: WebSVN 2.1.0

© copyright 1999-2024 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.