1 // Please don't remove this comment as asciidoc behaves badly when
2 // the first non-empty line is ifdef/ifndef. The symptom is that
3 // without this comment the <git-diff-core> attribute conditionally
4 // defined below ends up being defined unconditionally.
5 // Last checked with asciidoc 7.0.2.
7 ifndef::git-format-patch[]
13 endif::git-format-patch[]
15 ifdef::git-format-patch[]
18 Generate plain patches without any diffstats.
19 endif::git-format-patch[]
21 ifndef::git-format-patch[]
25 Generate patch (see section on generating patches).
26 {git-diff? This is the default.}
27 endif::git-format-patch[]
31 Suppress diff output. Useful for commands like `git show` that
32 show the patch by default, or to cancel the effect of `--patch`.
36 Generate diffs with <n> lines of context instead of
38 ifndef::git-format-patch[]
40 endif::git-format-patch[]
42 ifndef::git-format-patch[]
44 Generate the raw format.
45 {git-diff-core? This is the default.}
46 endif::git-format-patch[]
48 ifndef::git-format-patch[]
50 Synonym for `-p --raw`.
51 endif::git-format-patch[]
54 Spend extra time to make sure the smallest possible
58 Generate a diff using the "patience diff" algorithm.
61 Generate a diff using the "histogram diff" algorithm.
63 --diff-algorithm={patience|minimal|histogram|myers}::
64 Choose a diff algorithm. The variants are as follows:
68 The basic greedy diff algorithm. Currently, this is the default.
70 Spend extra time to make sure the smallest possible diff is
73 Use "patience diff" algorithm when generating patches.
75 This algorithm extends the patience algorithm to "support
76 low-occurrence common elements".
79 For instance, if you configured diff.algorithm variable to a
80 non-default value and want to use the default one, then you
81 have to use `--diff-algorithm=default` option.
83 --stat[=<width>[,<name-width>[,<count>]]]::
84 Generate a diffstat. By default, as much space as necessary
85 will be used for the filename part, and the rest for the graph
86 part. Maximum width defaults to terminal width, or 80 columns
87 if not connected to a terminal, and can be overridden by
88 `<width>`. The width of the filename part can be limited by
89 giving another width `<name-width>` after a comma. The width
90 of the graph part can be limited by using
91 `--stat-graph-width=<width>` (affects all commands generating
92 a stat graph) or by setting `diff.statGraphWidth=<width>`
93 (does not affect `git format-patch`).
94 By giving a third parameter `<count>`, you can limit the
95 output to the first `<count>` lines, followed by `...` if
98 These parameters can also be set individually with `--stat-width=<width>`,
99 `--stat-name-width=<name-width>` and `--stat-count=<count>`.
102 Similar to `--stat`, but shows number of added and
103 deleted lines in decimal notation and pathname without
104 abbreviation, to make it more machine friendly. For
105 binary files, outputs two `-` instead of saying
109 Output only the last line of the `--stat` format containing total
110 number of modified files, as well as number of added and deleted
113 --dirstat[=<param1,param2,...>]::
114 Output the distribution of relative amount of changes for each
115 sub-directory. The behavior of `--dirstat` can be customized by
116 passing it a comma separated list of parameters.
117 The defaults are controlled by the `diff.dirstat` configuration
118 variable (see linkgit:git-config[1]).
119 The following parameters are available:
123 Compute the dirstat numbers by counting the lines that have been
124 removed from the source, or added to the destination. This ignores
125 the amount of pure code movements within a file. In other words,
126 rearranging lines in a file is not counted as much as other changes.
127 This is the default behavior when no parameter is given.
129 Compute the dirstat numbers by doing the regular line-based diff
130 analysis, and summing the removed/added line counts. (For binary
131 files, count 64-byte chunks instead, since binary files have no
132 natural concept of lines). This is a more expensive `--dirstat`
133 behavior than the `changes` behavior, but it does count rearranged
134 lines within a file as much as other changes. The resulting output
135 is consistent with what you get from the other `--*stat` options.
137 Compute the dirstat numbers by counting the number of files changed.
138 Each changed file counts equally in the dirstat analysis. This is
139 the computationally cheapest `--dirstat` behavior, since it does
140 not have to look at the file contents at all.
142 Count changes in a child directory for the parent directory as well.
143 Note that when using `cumulative`, the sum of the percentages
144 reported may exceed 100%. The default (non-cumulative) behavior can
145 be specified with the `noncumulative` parameter.
147 An integer parameter specifies a cut-off percent (3% by default).
148 Directories contributing less than this percentage of the changes
149 are not shown in the output.
152 Example: The following will count changed files, while ignoring
153 directories with less than 10% of the total amount of changed files,
154 and accumulating child directory counts in the parent directories:
155 `--dirstat=files,10,cumulative`.
158 Output a condensed summary of extended header information
159 such as creations, renames and mode changes.
161 ifndef::git-format-patch[]
163 Synonym for `-p --stat`.
164 endif::git-format-patch[]
166 ifndef::git-format-patch[]
170 Separate the commits with NULs instead of with new newlines.
172 Also, when `--raw` or `--numstat` has been given, do not munge
173 pathnames and use NULs as output field terminators.
176 When `--raw`, `--numstat`, `--name-only` or `--name-status` has been
177 given, do not munge pathnames and use NULs as output field terminators.
180 Without this option, each pathname output will have TAB, LF, double quotes,
181 and backslash characters replaced with `\t`, `\n`, `\"`, and `\\`,
182 respectively, and the pathname will be enclosed in double quotes if
183 any of those replacements occurred.
186 Show only names of changed files.
189 Show only names and status of changed files. See the description
190 of the `--diff-filter` option on what the status letters mean.
192 --submodule[=<format>]::
193 Specify how differences in submodules are shown. When `--submodule`
194 or `--submodule=log` is given, the 'log' format is used. This format lists
195 the commits in the range like linkgit:git-submodule[1] `summary` does.
196 Omitting the `--submodule` option or specifying `--submodule=short`,
197 uses the 'short' format. This format just shows the names of the commits
198 at the beginning and end of the range. Can be tweaked via the
199 `diff.submodule` configuration variable.
203 `--color` (i.e. without '=<when>') is the same as `--color=always`.
204 '<when>' can be one of `always`, `never`, or `auto`.
206 It can be changed by the `color.ui` and `color.diff`
207 configuration settings.
211 Turn off colored diff.
213 This can be used to override configuration settings.
215 It is the same as `--color=never`.
217 --word-diff[=<mode>]::
218 Show a word diff, using the <mode> to delimit changed words.
219 By default, words are delimited by whitespace; see
220 `--word-diff-regex` below. The <mode> defaults to 'plain', and
225 Highlight changed words using only colors. Implies `--color`.
227 Show words as `[-removed-]` and `{+added+}`. Makes no
228 attempts to escape the delimiters if they appear in the input,
229 so the output may be ambiguous.
231 Use a special line-based format intended for script
232 consumption. Added/removed/unchanged runs are printed in the
233 usual unified diff format, starting with a `+`/`-`/` `
234 character at the beginning of the line and extending to the
235 end of the line. Newlines in the input are represented by a
236 tilde `~` on a line of its own.
238 Disable word diff again.
241 Note that despite the name of the first mode, color is used to
242 highlight the changed parts in all modes if enabled.
244 --word-diff-regex=<regex>::
245 Use <regex> to decide what a word is, instead of considering
246 runs of non-whitespace to be a word. Also implies
247 `--word-diff` unless it was already enabled.
249 Every non-overlapping match of the
250 <regex> is considered a word. Anything between these matches is
251 considered whitespace and ignored(!) for the purposes of finding
252 differences. You may want to append `|[^[:space:]]` to your regular
253 expression to make sure that it matches all non-whitespace characters.
254 A match that contains a newline is silently truncated(!) at the
257 The regex can also be set via a diff driver or configuration option, see
258 linkgit:gitattributes[1] or linkgit:git-config[1]. Giving it explicitly
259 overrides any diff driver or configuration setting. Diff drivers
260 override configuration settings.
262 --color-words[=<regex>]::
263 Equivalent to `--word-diff=color` plus (if a regex was
264 specified) `--word-diff-regex=<regex>`.
265 endif::git-format-patch[]
268 Turn off rename detection, even when the configuration
269 file gives the default to do so.
271 ifndef::git-format-patch[]
273 Warn if changes introduce whitespace errors. What are
274 considered whitespace errors is controlled by `core.whitespace`
275 configuration. By default, trailing whitespaces (including
276 lines that solely consist of whitespaces) and a space character
277 that is immediately followed by a tab character inside the
278 initial indent of the line are considered whitespace errors.
279 Exits with non-zero status if problems are found. Not compatible
281 endif::git-format-patch[]
284 Instead of the first handful of characters, show the full
285 pre- and post-image blob object names on the "index"
286 line when generating patch format output.
289 In addition to `--full-index`, output a binary diff that
290 can be applied with `git-apply`.
293 Instead of showing the full 40-byte hexadecimal object
294 name in diff-raw format output and diff-tree header
295 lines, show only a partial prefix. This is
296 independent of the `--full-index` option above, which controls
297 the diff-patch output format. Non default number of
298 digits can be specified with `--abbrev=<n>`.
301 --break-rewrites[=[<n>][/<m>]]::
302 Break complete rewrite changes into pairs of delete and
303 create. This serves two purposes:
305 It affects the way a change that amounts to a total rewrite of a file
306 not as a series of deletion and insertion mixed together with a very
307 few lines that happen to match textually as the context, but as a
308 single deletion of everything old followed by a single insertion of
309 everything new, and the number `m` controls this aspect of the -B
310 option (defaults to 60%). `-B/70%` specifies that less than 30% of the
311 original should remain in the result for Git to consider it a total
312 rewrite (i.e. otherwise the resulting patch will be a series of
313 deletion and insertion mixed together with context lines).
315 When used with -M, a totally-rewritten file is also considered as the
316 source of a rename (usually -M only considers a file that disappeared
317 as the source of a rename), and the number `n` controls this aspect of
318 the -B option (defaults to 50%). `-B20%` specifies that a change with
319 addition and deletion compared to 20% or more of the file's size are
320 eligible for being picked up as a possible source of a rename to
324 --find-renames[=<n>]::
329 If generating diffs, detect and report renames for each commit.
330 For following files across renames while traversing history, see
333 If `n` is specified, it is a threshold on the similarity
334 index (i.e. amount of addition/deletions compared to the
335 file's size). For example, `-M90%` means Git should consider a
336 delete/add pair to be a rename if more than 90% of the file
337 hasn't changed. Without a `%` sign, the number is to be read as
338 a fraction, with a decimal point before it. I.e., `-M5` becomes
339 0.5, and is thus the same as `-M50%`. Similarly, `-M05` is
340 the same as `-M5%`. To limit detection to exact renames, use
341 `-M100%`. The default similarity index is 50%.
344 --find-copies[=<n>]::
345 Detect copies as well as renames. See also `--find-copies-harder`.
346 If `n` is specified, it has the same meaning as for `-M<n>`.
348 --find-copies-harder::
349 For performance reasons, by default, `-C` option finds copies only
350 if the original file of the copy was modified in the same
351 changeset. This flag makes the command
352 inspect unmodified files as candidates for the source of
353 copy. This is a very expensive operation for large
354 projects, so use it with caution. Giving more than one
355 `-C` option has the same effect.
358 --irreversible-delete::
359 Omit the preimage for deletes, i.e. print only the header but not
360 the diff between the preimage and `/dev/null`. The resulting patch
361 is not meant to be applied with `patch` nor `git apply`; this is
362 solely for people who want to just concentrate on reviewing the
363 text after the change. In addition, the output obviously lack
364 enough information to apply such a patch in reverse, even manually,
365 hence the name of the option.
367 When used together with `-B`, omit also the preimage in the deletion part
368 of a delete/create pair.
371 The `-M` and `-C` options require O(n^2) processing time where n
372 is the number of potential rename/copy targets. This
373 option prevents rename/copy detection from running if
374 the number of rename/copy targets exceeds the specified
377 ifndef::git-format-patch[]
378 --diff-filter=[(A|C|D|M|R|T|U|X|B)...[*]]::
379 Select only files that are Added (`A`), Copied (`C`),
380 Deleted (`D`), Modified (`M`), Renamed (`R`), have their
381 type (i.e. regular file, symlink, submodule, ...) changed (`T`),
382 are Unmerged (`U`), are
383 Unknown (`X`), or have had their pairing Broken (`B`).
384 Any combination of the filter characters (including none) can be used.
385 When `*` (All-or-none) is added to the combination, all
386 paths are selected if there is any file that matches
387 other criteria in the comparison; if there is no file
388 that matches other criteria, nothing is selected.
391 Look for differences that change the number of occurrences of
392 the specified string (i.e. addition/deletion) in a file.
393 Intended for the scripter's use.
395 It is useful when you're looking for an exact block of code (like a
396 struct), and want to know the history of that block since it first
397 came into being: use the feature iteratively to feed the interesting
398 block in the preimage back into `-S`, and keep going until you get the
399 very first version of the block.
402 Look for differences whose patch text contains added/removed
403 lines that match <regex>.
405 To illustrate the difference between `-S<regex> --pickaxe-regex` and
406 `-G<regex>`, consider a commit with the following diff in the same
410 + return !regexec(regexp, two->ptr, 1, ®match, 0);
412 - hit = !regexec(regexp, mf2.ptr, 1, ®match, 0);
415 While `git log -G"regexec\(regexp"` will show this commit, `git log
416 -S"regexec\(regexp" --pickaxe-regex` will not (because the number of
417 occurrences of that string did not change).
419 See the 'pickaxe' entry in linkgit:gitdiffcore[7] for more
423 When `-S` or `-G` finds a change, show all the changes in that
424 changeset, not just the files that contain the change
428 Treat the <string> given to `-S` as an extended POSIX regular
430 endif::git-format-patch[]
433 Output the patch in the order specified in the
434 <orderfile>, which has one shell glob pattern per line.
435 This overrides the `diff.orderfile` configuration variable
436 (see linkgit:git-config[1]). To cancel `diff.orderfile`,
439 ifndef::git-format-patch[]
441 Swap two inputs; that is, show differences from index or
442 on-disk file to tree contents.
444 --relative[=<path>]::
445 When run from a subdirectory of the project, it can be
446 told to exclude changes outside the directory and show
447 pathnames relative to it with this option. When you are
448 not in a subdirectory (e.g. in a bare repository), you
449 can name which subdirectory to make the output relative
450 to by giving a <path> as an argument.
451 endif::git-format-patch[]
455 Treat all files as text.
457 --ignore-space-at-eol::
458 Ignore changes in whitespace at EOL.
461 --ignore-space-change::
462 Ignore changes in amount of whitespace. This ignores whitespace
463 at line end, and considers all other sequences of one or
464 more whitespace characters to be equivalent.
468 Ignore whitespace when comparing lines. This ignores
469 differences even if one line has whitespace where the other
472 --ignore-blank-lines::
473 Ignore changes whose lines are all blank.
475 --inter-hunk-context=<lines>::
476 Show the context between diff hunks, up to the specified number
477 of lines, thereby fusing hunks that are close to each other.
481 Show whole surrounding functions of changes.
483 ifndef::git-format-patch[]
486 Make the program exit with codes similar to diff(1).
487 That is, it exits with 1 if there were differences and
488 0 means no differences.
491 Disable all output of the program. Implies `--exit-code`.
493 endif::git-format-patch[]
496 Allow an external diff helper to be executed. If you set an
497 external diff driver with linkgit:gitattributes[5], you need
498 to use this option with linkgit:git-log[1] and friends.
501 Disallow external diff drivers.
505 Allow (or disallow) external text conversion filters to be run
506 when comparing binary files. See linkgit:gitattributes[5] for
507 details. Because textconv filters are typically a one-way
508 conversion, the resulting diff is suitable for human
509 consumption, but cannot be applied. For this reason, textconv
510 filters are enabled by default only for linkgit:git-diff[1] and
511 linkgit:git-log[1], but not for linkgit:git-format-patch[1] or
512 diff plumbing commands.
514 --ignore-submodules[=<when>]::
515 Ignore changes to submodules in the diff generation. <when> can be
516 either "none", "untracked", "dirty" or "all", which is the default.
517 Using "none" will consider the submodule modified when it either contains
518 untracked or modified files or its HEAD differs from the commit recorded
519 in the superproject and can be used to override any settings of the
520 'ignore' option in linkgit:git-config[1] or linkgit:gitmodules[5]. When
521 "untracked" is used submodules are not considered dirty when they only
522 contain untracked content (but they are still scanned for modified
523 content). Using "dirty" ignores all changes to the work tree of submodules,
524 only changes to the commits stored in the superproject are shown (this was
525 the behavior until 1.7.0). Using "all" hides all changes to submodules.
527 --src-prefix=<prefix>::
528 Show the given source prefix instead of "a/".
530 --dst-prefix=<prefix>::
531 Show the given destination prefix instead of "b/".
534 Do not show any source or destination prefix.
536 For more detailed explanation on these common options, see also
537 linkgit:gitdiffcore[7].