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 Generate diffs with <n> lines of context instead of
33 ifndef::git-format-patch[]
35 endif::git-format-patch[]
37 ifndef::git-format-patch[]
39 Generate the raw format.
40 {git-diff-core? This is the default.}
41 endif::git-format-patch[]
43 ifndef::git-format-patch[]
45 Synonym for `-p --raw`.
46 endif::git-format-patch[]
49 Spend extra time to make sure the smallest possible
53 Generate a diff using the "patience diff" algorithm.
56 Generate a diff using the "histogram diff" algorithm.
58 --diff-algorithm={patience|minimal|histogram|myers}::
59 Choose a diff algorithm. The variants are as follows:
63 The basic greedy diff algorithm. Currently, this is the default.
65 Spend extra time to make sure the smallest possible diff is
68 Use "patience diff" algorithm when generating patches.
70 This algorithm extends the patience algorithm to "support
71 low-occurrence common elements".
74 For instance, if you configured diff.algorithm variable to a
75 non-default value and want to use the default one, then you
76 have to use `--diff-algorithm=default` option.
78 --stat[=<width>[,<name-width>[,<count>]]]::
79 Generate a diffstat. By default, as much space as necessary
80 will be used for the filename part, and the rest for the graph
81 part. Maximum width defaults to terminal width, or 80 columns
82 if not connected to a terminal, and can be overridden by
83 `<width>`. The width of the filename part can be limited by
84 giving another width `<name-width>` after a comma. The width
85 of the graph part can be limited by using
86 `--stat-graph-width=<width>` (affects all commands generating
87 a stat graph) or by setting `diff.statGraphWidth=<width>`
88 (does not affect `git format-patch`).
89 By giving a third parameter `<count>`, you can limit the
90 output to the first `<count>` lines, followed by `...` if
93 These parameters can also be set individually with `--stat-width=<width>`,
94 `--stat-name-width=<name-width>` and `--stat-count=<count>`.
97 Similar to `--stat`, but shows number of added and
98 deleted lines in decimal notation and pathname without
99 abbreviation, to make it more machine friendly. For
100 binary files, outputs two `-` instead of saying
104 Output only the last line of the `--stat` format containing total
105 number of modified files, as well as number of added and deleted
108 --dirstat[=<param1,param2,...>]::
109 Output the distribution of relative amount of changes for each
110 sub-directory. The behavior of `--dirstat` can be customized by
111 passing it a comma separated list of parameters.
112 The defaults are controlled by the `diff.dirstat` configuration
113 variable (see linkgit:git-config[1]).
114 The following parameters are available:
118 Compute the dirstat numbers by counting the lines that have been
119 removed from the source, or added to the destination. This ignores
120 the amount of pure code movements within a file. In other words,
121 rearranging lines in a file is not counted as much as other changes.
122 This is the default behavior when no parameter is given.
124 Compute the dirstat numbers by doing the regular line-based diff
125 analysis, and summing the removed/added line counts. (For binary
126 files, count 64-byte chunks instead, since binary files have no
127 natural concept of lines). This is a more expensive `--dirstat`
128 behavior than the `changes` behavior, but it does count rearranged
129 lines within a file as much as other changes. The resulting output
130 is consistent with what you get from the other `--*stat` options.
132 Compute the dirstat numbers by counting the number of files changed.
133 Each changed file counts equally in the dirstat analysis. This is
134 the computationally cheapest `--dirstat` behavior, since it does
135 not have to look at the file contents at all.
137 Count changes in a child directory for the parent directory as well.
138 Note that when using `cumulative`, the sum of the percentages
139 reported may exceed 100%. The default (non-cumulative) behavior can
140 be specified with the `noncumulative` parameter.
142 An integer parameter specifies a cut-off percent (3% by default).
143 Directories contributing less than this percentage of the changes
144 are not shown in the output.
147 Example: The following will count changed files, while ignoring
148 directories with less than 10% of the total amount of changed files,
149 and accumulating child directory counts in the parent directories:
150 `--dirstat=files,10,cumulative`.
153 Output a condensed summary of extended header information
154 such as creations, renames and mode changes.
156 ifndef::git-format-patch[]
158 Synonym for `-p --stat`.
159 endif::git-format-patch[]
161 ifndef::git-format-patch[]
165 Separate the commits with NULs instead of with new newlines.
167 Also, when `--raw` or `--numstat` has been given, do not munge
168 pathnames and use NULs as output field terminators.
171 When `--raw`, `--numstat`, `--name-only` or `--name-status` has been
172 given, do not munge pathnames and use NULs as output field terminators.
175 Without this option, each pathname output will have TAB, LF, double quotes,
176 and backslash characters replaced with `\t`, `\n`, `\"`, and `\\`,
177 respectively, and the pathname will be enclosed in double quotes if
178 any of those replacements occurred.
181 Show only names of changed files.
184 Show only names and status of changed files. See the description
185 of the `--diff-filter` option on what the status letters mean.
187 --submodule[=<format>]::
188 Specify how differences in submodules are shown. When `--submodule`
189 or `--submodule=log` is given, the 'log' format is used. This format lists
190 the commits in the range like linkgit:git-submodule[1] `summary` does.
191 Omitting the `--submodule` option or specifying `--submodule=short`,
192 uses the 'short' format. This format just shows the names of the commits
193 at the beginning and end of the range. Can be tweaked via the
194 `diff.submodule` configuration variable.
198 `--color` (i.e. without '=<when>') is the same as `--color=always`.
199 '<when>' can be one of `always`, `never`, or `auto`.
201 It can be changed by the `color.ui` and `color.diff`
202 configuration settings.
206 Turn off colored diff.
208 This can be used to override configuration settings.
210 It is the same as `--color=never`.
212 --word-diff[=<mode>]::
213 Show a word diff, using the <mode> to delimit changed words.
214 By default, words are delimited by whitespace; see
215 `--word-diff-regex` below. The <mode> defaults to 'plain', and
220 Highlight changed words using only colors. Implies `--color`.
222 Show words as `[-removed-]` and `{+added+}`. Makes no
223 attempts to escape the delimiters if they appear in the input,
224 so the output may be ambiguous.
226 Use a special line-based format intended for script
227 consumption. Added/removed/unchanged runs are printed in the
228 usual unified diff format, starting with a `+`/`-`/` `
229 character at the beginning of the line and extending to the
230 end of the line. Newlines in the input are represented by a
231 tilde `~` on a line of its own.
233 Disable word diff again.
236 Note that despite the name of the first mode, color is used to
237 highlight the changed parts in all modes if enabled.
239 --word-diff-regex=<regex>::
240 Use <regex> to decide what a word is, instead of considering
241 runs of non-whitespace to be a word. Also implies
242 `--word-diff` unless it was already enabled.
244 Every non-overlapping match of the
245 <regex> is considered a word. Anything between these matches is
246 considered whitespace and ignored(!) for the purposes of finding
247 differences. You may want to append `|[^[:space:]]` to your regular
248 expression to make sure that it matches all non-whitespace characters.
249 A match that contains a newline is silently truncated(!) at the
252 The regex can also be set via a diff driver or configuration option, see
253 linkgit:gitattributes[1] or linkgit:git-config[1]. Giving it explicitly
254 overrides any diff driver or configuration setting. Diff drivers
255 override configuration settings.
257 --color-words[=<regex>]::
258 Equivalent to `--word-diff=color` plus (if a regex was
259 specified) `--word-diff-regex=<regex>`.
260 endif::git-format-patch[]
263 Turn off rename detection, even when the configuration
264 file gives the default to do so.
266 ifndef::git-format-patch[]
268 Warn if changes introduce whitespace errors. What are
269 considered whitespace errors is controlled by `core.whitespace`
270 configuration. By default, trailing whitespaces (including
271 lines that solely consist of whitespaces) and a space character
272 that is immediately followed by a tab character inside the
273 initial indent of the line are considered whitespace errors.
274 Exits with non-zero status if problems are found. Not compatible
276 endif::git-format-patch[]
279 Instead of the first handful of characters, show the full
280 pre- and post-image blob object names on the "index"
281 line when generating patch format output.
284 In addition to `--full-index`, output a binary diff that
285 can be applied with `git-apply`.
288 Instead of showing the full 40-byte hexadecimal object
289 name in diff-raw format output and diff-tree header
290 lines, show only a partial prefix. This is
291 independent of the `--full-index` option above, which controls
292 the diff-patch output format. Non default number of
293 digits can be specified with `--abbrev=<n>`.
296 --break-rewrites[=[<n>][/<m>]]::
297 Break complete rewrite changes into pairs of delete and
298 create. This serves two purposes:
300 It affects the way a change that amounts to a total rewrite of a file
301 not as a series of deletion and insertion mixed together with a very
302 few lines that happen to match textually as the context, but as a
303 single deletion of everything old followed by a single insertion of
304 everything new, and the number `m` controls this aspect of the -B
305 option (defaults to 60%). `-B/70%` specifies that less than 30% of the
306 original should remain in the result for Git to consider it a total
307 rewrite (i.e. otherwise the resulting patch will be a series of
308 deletion and insertion mixed together with context lines).
310 When used with -M, a totally-rewritten file is also considered as the
311 source of a rename (usually -M only considers a file that disappeared
312 as the source of a rename), and the number `n` controls this aspect of
313 the -B option (defaults to 50%). `-B20%` specifies that a change with
314 addition and deletion compared to 20% or more of the file's size are
315 eligible for being picked up as a possible source of a rename to
319 --find-renames[=<n>]::
324 If generating diffs, detect and report renames for each commit.
325 For following files across renames while traversing history, see
328 If `n` is specified, it is a threshold on the similarity
329 index (i.e. amount of addition/deletions compared to the
330 file's size). For example, `-M90%` means Git should consider a
331 delete/add pair to be a rename if more than 90% of the file
332 hasn't changed. Without a `%` sign, the number is to be read as
333 a fraction, with a decimal point before it. I.e., `-M5` becomes
334 0.5, and is thus the same as `-M50%`. Similarly, `-M05` is
335 the same as `-M5%`. To limit detection to exact renames, use
336 `-M100%`. The default similarity index is 50%.
339 --find-copies[=<n>]::
340 Detect copies as well as renames. See also `--find-copies-harder`.
341 If `n` is specified, it has the same meaning as for `-M<n>`.
343 --find-copies-harder::
344 For performance reasons, by default, `-C` option finds copies only
345 if the original file of the copy was modified in the same
346 changeset. This flag makes the command
347 inspect unmodified files as candidates for the source of
348 copy. This is a very expensive operation for large
349 projects, so use it with caution. Giving more than one
350 `-C` option has the same effect.
353 --irreversible-delete::
354 Omit the preimage for deletes, i.e. print only the header but not
355 the diff between the preimage and `/dev/null`. The resulting patch
356 is not meant to be applied with `patch` nor `git apply`; this is
357 solely for people who want to just concentrate on reviewing the
358 text after the change. In addition, the output obviously lack
359 enough information to apply such a patch in reverse, even manually,
360 hence the name of the option.
362 When used together with `-B`, omit also the preimage in the deletion part
363 of a delete/create pair.
366 The `-M` and `-C` options require O(n^2) processing time where n
367 is the number of potential rename/copy targets. This
368 option prevents rename/copy detection from running if
369 the number of rename/copy targets exceeds the specified
372 ifndef::git-format-patch[]
373 --diff-filter=[(A|C|D|M|R|T|U|X|B)...[*]]::
374 Select only files that are Added (`A`), Copied (`C`),
375 Deleted (`D`), Modified (`M`), Renamed (`R`), have their
376 type (i.e. regular file, symlink, submodule, ...) changed (`T`),
377 are Unmerged (`U`), are
378 Unknown (`X`), or have had their pairing Broken (`B`).
379 Any combination of the filter characters (including none) can be used.
380 When `*` (All-or-none) is added to the combination, all
381 paths are selected if there is any file that matches
382 other criteria in the comparison; if there is no file
383 that matches other criteria, nothing is selected.
386 Look for differences that change the number of occurrences of
387 the specified string (i.e. addition/deletion) in a file.
388 Intended for the scripter's use.
390 It is useful when you're looking for an exact block of code (like a
391 struct), and want to know the history of that block since it first
392 came into being: use the feature iteratively to feed the interesting
393 block in the preimage back into `-S`, and keep going until you get the
394 very first version of the block.
397 Look for differences whose patch text contains added/removed
398 lines that match <regex>.
400 To illustrate the difference between `-S<regex> --pickaxe-regex` and
401 `-G<regex>`, consider a commit with the following diff in the same
405 + return !regexec(regexp, two->ptr, 1, ®match, 0);
407 - hit = !regexec(regexp, mf2.ptr, 1, ®match, 0);
410 While `git log -G"regexec\(regexp"` will show this commit, `git log
411 -S"regexec\(regexp" --pickaxe-regex` will not (because the number of
412 occurrences of that string did not change).
414 See the 'pickaxe' entry in linkgit:gitdiffcore[7] for more
418 When `-S` or `-G` finds a change, show all the changes in that
419 changeset, not just the files that contain the change
423 Treat the <string> given to `-S` as an extended POSIX regular
425 endif::git-format-patch[]
428 Output the patch in the order specified in the
429 <orderfile>, which has one shell glob pattern per line.
431 ifndef::git-format-patch[]
433 Swap two inputs; that is, show differences from index or
434 on-disk file to tree contents.
436 --relative[=<path>]::
437 When run from a subdirectory of the project, it can be
438 told to exclude changes outside the directory and show
439 pathnames relative to it with this option. When you are
440 not in a subdirectory (e.g. in a bare repository), you
441 can name which subdirectory to make the output relative
442 to by giving a <path> as an argument.
443 endif::git-format-patch[]
447 Treat all files as text.
449 --ignore-space-at-eol::
450 Ignore changes in whitespace at EOL.
453 --ignore-space-change::
454 Ignore changes in amount of whitespace. This ignores whitespace
455 at line end, and considers all other sequences of one or
456 more whitespace characters to be equivalent.
460 Ignore whitespace when comparing lines. This ignores
461 differences even if one line has whitespace where the other
464 --ignore-blank-lines::
465 Ignore changes whose lines are all blank.
467 --inter-hunk-context=<lines>::
468 Show the context between diff hunks, up to the specified number
469 of lines, thereby fusing hunks that are close to each other.
473 Show whole surrounding functions of changes.
475 ifndef::git-format-patch[]
478 Make the program exit with codes similar to diff(1).
479 That is, it exits with 1 if there were differences and
480 0 means no differences.
483 Disable all output of the program. Implies `--exit-code`.
485 endif::git-format-patch[]
488 Allow an external diff helper to be executed. If you set an
489 external diff driver with linkgit:gitattributes[5], you need
490 to use this option with linkgit:git-log[1] and friends.
493 Disallow external diff drivers.
497 Allow (or disallow) external text conversion filters to be run
498 when comparing binary files. See linkgit:gitattributes[5] for
499 details. Because textconv filters are typically a one-way
500 conversion, the resulting diff is suitable for human
501 consumption, but cannot be applied. For this reason, textconv
502 filters are enabled by default only for linkgit:git-diff[1] and
503 linkgit:git-log[1], but not for linkgit:git-format-patch[1] or
504 diff plumbing commands.
506 --ignore-submodules[=<when>]::
507 Ignore changes to submodules in the diff generation. <when> can be
508 either "none", "untracked", "dirty" or "all", which is the default.
509 Using "none" will consider the submodule modified when it either contains
510 untracked or modified files or its HEAD differs from the commit recorded
511 in the superproject and can be used to override any settings of the
512 'ignore' option in linkgit:git-config[1] or linkgit:gitmodules[5]. When
513 "untracked" is used submodules are not considered dirty when they only
514 contain untracked content (but they are still scanned for modified
515 content). Using "dirty" ignores all changes to the work tree of submodules,
516 only changes to the commits stored in the superproject are shown (this was
517 the behavior until 1.7.0). Using "all" hides all changes to submodules.
519 --src-prefix=<prefix>::
520 Show the given source prefix instead of "a/".
522 --dst-prefix=<prefix>::
523 Show the given destination prefix instead of "b/".
526 Do not show any source or destination prefix.
528 For more detailed explanation on these common options, see also
529 linkgit:gitdiffcore[7].