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