Mercurial > hg > Applications > mh

comparison miscellany/patch-2.0.12u8/patch.man @ 0:bce86c4163a3

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Initial revision
author kono
date Mon, 18 Apr 2005 23:46:02 +0900
parents
children 848621fd0092 441a2190cfae
comparison
equal deleted inserted replaced
-1:000000000000 0:bce86c4163a3
1 .\" -*- nroff -*-
2 .rn '' }`
3 '\" $Header$
4 '\"
5 '\" $Log$
6 '\" Revision 1.1 2005/04/18 14:46:05 kono
7 '\" Initial revision
8 '\"
9 '\" Revision 2.0.1.2 88/06/22 20:47:18 lwall
10 '\" patch12: now avoids Bell System Logo
11 '\"
12 '\" Revision 2.0.1.1 88/06/03 15:12:51 lwall
13 '\" patch10: -B switch was contributed.
14 '\"
15 '\" Revision 2.0 86/09/17 15:39:09 lwall
16 '\" Baseline for netwide release.
17 '\"
18 '\" Revision 1.4 86/08/01 19:23:22 lwall
19 '\" Documented -v, -p, -F.
20 '\" Added notes to patch senders.
21 '\"
22 '\" Revision 1.3 85/03/26 15:11:06 lwall
23 '\" Frozen.
24 '\"
25 '\" Revision 1.2.1.4 85/03/12 16:14:27 lwall
26 '\" Documented -p.
27 '\"
28 '\" Revision 1.2.1.3 85/03/12 16:09:41 lwall
29 '\" Documented -D.
30 '\"
31 '\" Revision 1.2.1.2 84/12/05 11:06:55 lwall
32 '\" Added -l switch, and noted bistability bug.
33 '\"
34 '\" Revision 1.2.1.1 84/12/04 17:23:39 lwall
35 '\" Branch for sdcrdcf changes.
36 '\"
37 '\" Revision 1.2 84/12/04 17:22:02 lwall
38 '\" Baseline version.
39 '\"
40 .de Sh
41 .br
42 .ne 5
43 .PP
44 \fB\\$1\fR
45 .PP
46 ..
47 .de Sp
48 .if t .sp .5v
49 .if n .sp
50 ..
51 '\"
52 '\" Set up \*(-- to give an unbreakable dash;
53 '\" string Tr holds user defined translation string.
54 '\" Bell System Logo is used as a dummy character.
55 '\"
56 '\" Shut up a groff -ww warning.
57 .if \n(.g .if !dTr .ds Tr
58 .ie n \{\
59 .tr \(*W-\*(Tr
60 .ds -- \(*W-
61 .if (\n(.H=4u)&(1m=24u) .ds -- \(*W\h'-12u'\(*W\h'-12u'-\" diablo 10 pitch
62 .if (\n(.H=4u)&(1m=20u) .ds -- \(*W\h'-12u'\(*W\h'-8u'-\" diablo 12 pitch
63 .ds L" ""
64 .ds R" ""
65 .ds L' '
66 .ds R' '
67 'br \}
68 .el \{\
69 .ds -- \(em\|
70 .tr \*(Tr
71 .ds L" ``
72 .ds R" ''
73 .ds L' `
74 .ds R' '
75 'br\}
76 .TH PATCH 1 LOCAL
77 .SH NAME
78 patch - apply a diff file to an original
79 .SH SYNOPSIS
80 .B patch
81 [options] [origfile [patchfile]] [+ [options] [origfile]]...
82 .sp
83 but usually just
84 .sp
85 .B patch
86 <patchfile
87 .SH DESCRIPTION
88 .I Patch
89 will take a patch file containing any of the four forms of difference
90 listing produced by the
91 .I diff
92 program and apply those differences to an original file, producing a patched
93 version.
94 By default, the patched version is put in place of the original, with
95 the original file backed up to the same name with the
96 extension \*(L".orig\*(R" (\*(L"~\*(R" on systems that do not
97 support long filenames), or as specified by the
98 .BR -b ,
99 .BR -B ,
100 or
101 .B -V
102 switches.
103 The extension used for making backup files may also be specified in the
104 .B SIMPLE_BACKUP_SUFFIX
105 environment variable, which is overridden by above switches.
106 .PP
107 If the backup file already exists,
108 .B patch
109 creates a new backup file name by changing the first lowercase letter
110 in the last component of the file's name into uppercase. If there are
111 no more lowercase letters in the name, it removes the first character
112 from the name. It repeats this process until it comes up with a
113 backup file that does not already exist.
114 .PP
115 You may also specify where you want the output to go with a
116 .B -o
117 switch; if that file already exists, it is backed up first.
118 .PP
119 If
120 .I patchfile
121 is omitted, or is a hyphen, the patch will be read from standard input.
122 .PP
123 Upon startup, patch will attempt to determine the type of the diff listing,
124 unless over-ruled by a
125 .BR -c ,
126 .BR -e ,
127 .BR -n ,
128 or
129 .B -u
130 switch.
131 Context diffs (old-style, new-style, and unified) and
132 normal diffs are applied by the
133 .I patch
134 program itself, while ed diffs are simply fed to the
135 .I ed
136 editor via a pipe.
137 .PP
138 .I Patch
139 will try to skip any leading garbage, apply the diff,
140 and then skip any trailing garbage.
141 Thus you could feed an article or message containing a
142 diff listing to
143 .IR patch ,
144 and it should work.
145 If the entire diff is indented by a consistent amount,
146 this will be taken into account.
147 .PP
148 With context diffs, and to a lesser extent with normal diffs,
149 .I patch
150 can detect when the line numbers mentioned in the patch are incorrect,
151 and will attempt to find the correct place to apply each hunk of the patch.
152 As a first guess, it takes the line number mentioned for the hunk, plus or
153 minus any offset used in applying the previous hunk.
154 If that is not the correct place,
155 .I patch
156 will scan both forwards and backwards for a set of lines matching the context
157 given in the hunk.
158 First
159 .I patch
160 looks for a place where all lines of the context match.
161 If no such place is found, and it's a context diff, and the maximum fuzz factor
162 is set to 1 or more, then another scan takes place ignoring the first and last
163 line of context.
164 If that fails, and the maximum fuzz factor is set to 2 or more,
165 the first two and last two lines of context are ignored,
166 and another scan is made.
167 (The default maximum fuzz factor is 2.)
168 If
169 .I patch
170 cannot find a place to install that hunk of the patch, it will put the
171 hunk out to a reject file, which normally is the name of the output file
172 plus \*(L".rej\*(R" (\*(L"#\*(R" on systems that do not support
173 long filenames).
174 (Note that the rejected hunk will come out in context diff form whether the
175 input patch was a context diff or a normal diff.
176 If the input was a normal diff, many of the contexts will simply be null.)
177 The line numbers on the hunks in the reject file may be different than
178 in the patch file: they reflect the approximate location patch thinks the
179 failed hunks belong in the new file rather than the old one.
180 .PP
181 As each hunk is completed, you will be told whether the hunk succeeded or
182 failed, and which line (in the new file)
183 .I patch
184 thought the hunk should go on.
185 If this is different from the line number specified in the diff you will
186 be told the offset.
187 A single large offset MAY be an indication that a hunk was installed in the
188 wrong place.
189 You will also be told if a fuzz factor was used to make the match, in which
190 case you should also be slightly suspicious.
191 .PP
192 If no original file is specified on the command line,
193 .I patch
194 will try to figure out from the leading garbage what the name of the file
195 to edit is.
196 In the header of a context diff, the filename is found from lines beginning
197 with \*(L"***\*(R" or \*(L"---\*(R", with the shortest name of an existing
198 file winning.
199 Only context diffs have lines like that, but if there is an \*(L"Index:\*(R"
200 line in the leading garbage,
201 .I patch
202 will try to use the filename from that line.
203 The context diff header takes precedence over an Index line.
204 If no filename can be intuited from the leading garbage, you will be asked
205 for the name of the file to patch.
206 .PP
207 If the original file cannot be found or is read-only, but a suitable
208 SCCS or RCS file is handy,
209 .I patch
210 will attempt to get or check out the file.
211 .PP
212 Additionally, if the leading garbage contains a \*(L"Prereq: \*(R" line,
213 .I patch
214 will take the first word from the prerequisites line (normally a version
215 number) and check the input file to see if that word can be found.
216 If not,
217 .I patch
218 will ask for confirmation before proceeding.
219 .PP
220 The upshot of all this is that you should be able to say, while in a news
221 interface, the following:
222 .Sp
223 | patch -d /usr/src/local/blurfl
224 .Sp
225 and patch a file in the blurfl directory directly from the article containing
226 the patch.
227 .PP
228 If the patch file contains more than one patch,
229 .I patch
230 will try to apply each of them as if they came from separate patch files.
231 This means, among other things, that it is assumed that the name of the file
232 to patch must be determined for each diff listing,
233 and that the garbage before each diff listing will
234 be examined for interesting things such as filenames and revision level, as
235 mentioned previously.
236 You can give switches (and another original file name) for the second and
237 subsequent patches by separating the corresponding argument lists
238 by a \*(L'+\*(R'.
239 (The argument list for a second or subsequent patch may not specify a new
240 patch file, however.)
241 .PP
242 .I Patch
243 recognizes the following switches:
244 .TP 5
245 .B \-b
246 causes the next argument to be interpreted as the backup extension, to be
247 used in place of \*(L".orig\*(R" or \*(L"~\*(R".
248 .TP 5
249 .B \-B
250 causes the next argument to be interpreted as a prefix to the backup file
251 name. If this argument is specified any argument from -b will be ignored.
252 .TP 5
253 .B \-c
254 forces
255 .I patch
256 to interpret the patch file as a context diff.
257 .TP 5
258 .B \-d
259 causes
260 .I patch
261 to interpret the next argument as a directory, and cd to it before doing
262 anything else.
263 .TP 5
264 .B \-D
265 causes
266 .I patch
267 to use the "#ifdef...#endif" construct to mark changes.
268 The argument following will be used as the differentiating symbol.
269 Note that, unlike the C compiler, there must be a space between the
270 .B \-D
271 and the argument.
272 .TP 5
273 .B \-e
274 forces
275 .I patch
276 to interpret the patch file as an ed script.
277 .TP 5
278 .B \-E
279 causes
280 .I patch
281 to remove output files that are empty after the patches have been applied.
282 .TP 5
283 .B \-f
284 forces
285 .I patch
286 to assume that the user knows exactly what he or she is doing, and to not
287 ask any questions. It assumes the following: skip patches for which a
288 file to patch can't be found; patch files even though they have the
289 wrong version for the ``Prereq:'' line in the patch; and assume that
290 patches are not reversed even if they look like they are.
291 This option does not suppress commentary; use
292 .B \-s
293 for that.
294 .TP 5
295 .B \-t
296 similar to
297 .BR \-f ,
298 in that it suppresses questions, but makes some different assumptions:
299 skip patches for which a file to patch can't be found (the same as \fB\-f\fP);
300 skip patches for which the file has the wrong version for the ``Prereq:'' line
301 in the patch; and assume that patches are reversed if they look like
302 they are.
303 .TP 5
304 .B \-F<number>
305 sets the maximum fuzz factor.
306 This switch only applies to context diffs, and causes
307 .I patch
308 to ignore up to that many lines in looking for places to install a hunk.
309 Note that a larger fuzz factor increases the odds of a faulty patch.
310 The default fuzz factor is 2, and it may not be set to more than
311 the number of lines of context in the context diff, ordinarily 3.
312 .TP 5
313 .B \-l
314 causes the pattern matching to be done loosely, in case the tabs and
315 spaces have been munged in your input file.
316 Any sequence of whitespace in the pattern line will match any sequence
317 in the input file.
318 Normal characters must still match exactly.
319 Each line of the context must still match a line in the input file.
320 .TP 5
321 .B \-n
322 forces
323 .I patch
324 to interpret the patch file as a normal diff.
325 .TP 5
326 .B \-N
327 causes
328 .I patch
329 to ignore patches that it thinks are reversed or already applied.
330 See also
331 .B \-R .
332 .TP 5
333 .B \-o
334 causes the next argument to be interpreted as the output file name.
335 .TP 5
336 .B \-p<number>
337 sets the pathname strip count,
338 which controls how pathnames found in the patch file are treated, in case
339 the you keep your files in a different directory than the person who sent
340 out the patch.
341 The strip count specifies how many slashes are to be stripped from
342 the front of the pathname.
343 (Any intervening directory names also go away.)
344 For example, supposing the filename in the patch file was
345 .sp
346 /u/howard/src/blurfl/blurfl.c
347 .sp
348 setting
349 .B \-p
350 or
351 .B \-p0
352 gives the entire pathname unmodified,
353 .B \-p1
354 gives
355 .sp
356 u/howard/src/blurfl/blurfl.c
357 .sp
358 without the leading slash,
359 .B \-p4
360 gives
361 .sp
362 blurfl/blurfl.c
363 .sp
364 and not specifying
365 .B \-p
366 at all just gives you "blurfl.c", unless all of the directories in the
367 leading path (u/howard/src/blurfl) exist and that path is relative,
368 in which case you get the entire pathname unmodified.
369 Whatever you end up with is looked for either in the current directory,
370 or the directory specified by the
371 .B \-d
372 switch.
373 .TP 5
374 .B \-r
375 causes the next argument to be interpreted as the reject file name.
376 .TP 5
377 .B \-R
378 tells
379 .I patch
380 that this patch was created with the old and new files swapped.
381 (Yes, I'm afraid that does happen occasionally, human nature being what it
382 is.)
383 .I Patch
384 will attempt to swap each hunk around before applying it.
385 Rejects will come out in the swapped format.
386 The
387 .B \-R
388 switch will not work with ed diff scripts because there is too little
389 information to reconstruct the reverse operation.
390 .Sp
391 If the first hunk of a patch fails,
392 .I patch
393 will reverse the hunk to see if it can be applied that way.
394 If it can, you will be asked if you want to have the
395 .B \-R
396 switch set.
397 If it can't, the patch will continue to be applied normally.
398 (Note: this method cannot detect a reversed patch if it is a normal diff
399 and if the first command is an append (i.e. it should have been a delete)
400 since appends always succeed, due to the fact that a null context will match
401 anywhere.
402 Luckily, most patches add or change lines rather than delete them, so most
403 reversed normal diffs will begin with a delete, which will fail, triggering
404 the heuristic.)
405 .TP 5
406 .B \-s
407 makes
408 .I patch
409 do its work silently, unless an error occurs.
410 .TP 5
411 .B \-S
412 causes
413 .I patch
414 to ignore this patch from the patch file, but continue on looking
415 for the next patch in the file.
416 Thus
417 .sp
418 patch -S + -S + <patchfile
419 .sp
420 will ignore the first and second of three patches.
421 .TP 5
422 .B \-u
423 forces
424 .I patch
425 to interpret the patch file as a unified context diff (a unidiff).
426 .TP 5
427 .B \-v
428 causes
429 .I patch
430 to print out its revision header and patch level.
431 .TP 5
432 .B \-V
433 causes the next argument to be interpreted as a method for creating
434 backup file names. The type of backups made can also be given in the
435 .B VERSION_CONTROL
436 environment variable, which is overridden by this option.
437 The
438 .B -B
439 option overrides this option, causing the prefix to always be used for
440 making backup file names.
441 The value of the
442 .B VERSION_CONTROL
443 environment variable and the argument to the
444 .B -V
445 option are like the GNU
446 Emacs `version-control' variable; they also recognize synonyms that
447 are more descriptive. The valid values are (unique abbreviations are
448 accepted):
449 .RS
450 .TP
451 `t' or `numbered'
452 Always make numbered backups.
453 .TP
454 `nil' or `existing'
455 Make numbered backups of files that already
456 have them, simple backups of the others.
457 This is the default.
458 .TP
459 `never' or `simple'
460 Always make simple backups.
461 .RE
462 .TP 5
463 .B \-x<number>
464 sets internal debugging flags, and is of interest only to
465 .I patch
466 patchers.
467 .SH AUTHOR
468 Larry Wall <lwall@netlabs.com>
469 .br
470 with many other contributors.
471 .SH ENVIRONMENT
472 .TP
473 .B TMPDIR
474 Directory to put temporary files in; default is /tmp.
475 .TP
476 .B SIMPLE_BACKUP_SUFFIX
477 Extension to use for backup file names instead of \*(L".orig\*(R" or
478 \*(L"~\*(R".
479 .TP
480 .B VERSION_CONTROL
481 Selects when numbered backup files are made.
482 .SH FILES
483 $TMPDIR/patch*
484 .SH SEE ALSO
485 diff(1)
486 .SH NOTES FOR PATCH SENDERS
487 There are several things you should bear in mind if you are going to
488 be sending out patches.
489 First, you can save people a lot of grief by keeping a patchlevel.h file
490 which is patched to increment the patch level as the first diff in the
491 patch file you send out.
492 If you put a Prereq: line in with the patch, it won't let them apply
493 patches out of order without some warning.
494 Second, make sure you've specified the filenames right, either in a
495 context diff header, or with an Index: line.
496 If you are patching something in a subdirectory, be sure to tell the patch
497 user to specify a
498 .B \-p
499 switch as needed.
500 Third, you can create a file by sending out a diff that compares a
501 null file to the file you want to create.
502 This will only work if the file you want to create doesn't exist already in
503 the target directory.
504 Fourth, take care not to send out reversed patches, since it makes people wonder
505 whether they already applied the patch.
506 Fifth, while you may be able to get away with putting 582 diff listings into
507 one file, it is probably wiser to group related patches into separate files in
508 case something goes haywire.
509 .SH DIAGNOSTICS
510 Too many to list here, but generally indicative that
511 .I patch
512 couldn't parse your patch file.
513 .PP
514 The message \*(L"Hmm...\*(R" indicates that there is unprocessed text in
515 the patch file and that
516 .I patch
517 is attempting to intuit whether there is a patch in that text and, if so,
518 what kind of patch it is.
519 .PP
520 .I Patch
521 will exit with a non-zero status if any reject files were created.
522 When applying a set of patches in a loop it behooves you to check this
523 exit status so you don't apply a later patch to a partially patched file.
524 .SH CAVEATS
525 .I Patch
526 cannot tell if the line numbers are off in an ed script, and can only detect
527 bad line numbers in a normal diff when it finds a \*(L"change\*(R" or
528 a \*(L"delete\*(R" command.
529 A context diff using fuzz factor 3 may have the same problem.
530 Until a suitable interactive interface is added, you should probably do
531 a context diff in these cases to see if the changes made sense.
532 Of course, compiling without errors is a pretty good indication that the patch
533 worked, but not always.
534 .PP
535 .I Patch
536 usually produces the correct results, even when it has to do a lot of
537 guessing.
538 However, the results are guaranteed to be correct only when the patch is
539 applied to exactly the same version of the file that the patch was
540 generated from.
541 .SH BUGS
542 Could be smarter about partial matches, excessively \&deviant offsets and
543 swapped code, but that would take an extra pass.
544 .PP
545 If code has been duplicated (for instance with #ifdef OLDCODE ... #else ...
546 #endif),
547 .I patch
548 is incapable of patching both versions, and, if it works at all, will likely
549 patch the wrong one, and tell you that it succeeded to boot.
550 .PP
551 If you apply a patch you've already applied,
552 .I patch
553 will think it is a reversed patch, and offer to un-apply the patch.
554 This could be construed as a feature.
555 .rn }` ''