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).
32 Suppress diff output. Useful for commands like `git show` that
33 show the patch by default, or to cancel the effect of `--patch`.
34 endif::git-format-patch[]
38 Generate diffs with <n> lines of context instead of
39 the usual three. Implies `--patch`.
40 ifndef::git-format-patch[]
42 endif::git-format-patch[]
44 ifndef::git-format-patch[]
47 Generate the diff in raw format.
48 ifdef::git-diff-core[]
50 endif::git-diff-core[]
53 For each commit, show a summary of changes using the raw diff
54 format. See the "RAW OUTPUT FORMAT" section of
55 linkgit:git-diff[1]. This is different from showing the log
56 itself in raw format, which you can achieve with
59 endif::git-format-patch[]
61 ifndef::git-format-patch[]
63 Synonym for `-p --raw`.
64 endif::git-format-patch[]
67 Enable the heuristic that shifts diff hunk boundaries to make patches
68 easier to read. This is the default.
70 --no-indent-heuristic::
71 Disable the indent heuristic.
74 Spend extra time to make sure the smallest possible
78 Generate a diff using the "patience diff" algorithm.
81 Generate a diff using the "histogram diff" algorithm.
84 Generate a diff using the "anchored diff" algorithm.
86 This option may be specified more than once.
88 If a line exists in both the source and destination, exists only once,
89 and starts with this text, this algorithm attempts to prevent it from
90 appearing as a deletion or addition in the output. It uses the "patience
91 diff" algorithm internally.
93 --diff-algorithm={patience|minimal|histogram|myers}::
94 Choose a diff algorithm. The variants are as follows:
98 The basic greedy diff algorithm. Currently, this is the default.
100 Spend extra time to make sure the smallest possible diff is
103 Use "patience diff" algorithm when generating patches.
105 This algorithm extends the patience algorithm to "support
106 low-occurrence common elements".
109 For instance, if you configured the `diff.algorithm` variable to a
110 non-default value and want to use the default one, then you
111 have to use `--diff-algorithm=default` option.
113 --stat[=<width>[,<name-width>[,<count>]]]::
114 Generate a diffstat. By default, as much space as necessary
115 will be used for the filename part, and the rest for the graph
116 part. Maximum width defaults to terminal width, or 80 columns
117 if not connected to a terminal, and can be overridden by
118 `<width>`. The width of the filename part can be limited by
119 giving another width `<name-width>` after a comma. The width
120 of the graph part can be limited by using
121 `--stat-graph-width=<width>` (affects all commands generating
122 a stat graph) or by setting `diff.statGraphWidth=<width>`
123 (does not affect `git format-patch`).
124 By giving a third parameter `<count>`, you can limit the
125 output to the first `<count>` lines, followed by `...` if
128 These parameters can also be set individually with `--stat-width=<width>`,
129 `--stat-name-width=<name-width>` and `--stat-count=<count>`.
132 Output a condensed summary of extended header information such
133 as file creations or deletions ("new" or "gone", optionally "+l"
134 if it's a symlink) and mode changes ("+x" or "-x" for adding
135 or removing executable bit respectively) in diffstat. The
136 information is put between the filename part and the graph
137 part. Implies `--stat`.
140 Similar to `--stat`, but shows number of added and
141 deleted lines in decimal notation and pathname without
142 abbreviation, to make it more machine friendly. For
143 binary files, outputs two `-` instead of saying
147 Output only the last line of the `--stat` format containing total
148 number of modified files, as well as number of added and deleted
151 -X[<param1,param2,...>]::
152 --dirstat[=<param1,param2,...>]::
153 Output the distribution of relative amount of changes for each
154 sub-directory. The behavior of `--dirstat` can be customized by
155 passing it a comma separated list of parameters.
156 The defaults are controlled by the `diff.dirstat` configuration
157 variable (see linkgit:git-config[1]).
158 The following parameters are available:
162 Compute the dirstat numbers by counting the lines that have been
163 removed from the source, or added to the destination. This ignores
164 the amount of pure code movements within a file. In other words,
165 rearranging lines in a file is not counted as much as other changes.
166 This is the default behavior when no parameter is given.
168 Compute the dirstat numbers by doing the regular line-based diff
169 analysis, and summing the removed/added line counts. (For binary
170 files, count 64-byte chunks instead, since binary files have no
171 natural concept of lines). This is a more expensive `--dirstat`
172 behavior than the `changes` behavior, but it does count rearranged
173 lines within a file as much as other changes. The resulting output
174 is consistent with what you get from the other `--*stat` options.
176 Compute the dirstat numbers by counting the number of files changed.
177 Each changed file counts equally in the dirstat analysis. This is
178 the computationally cheapest `--dirstat` behavior, since it does
179 not have to look at the file contents at all.
181 Count changes in a child directory for the parent directory as well.
182 Note that when using `cumulative`, the sum of the percentages
183 reported may exceed 100%. The default (non-cumulative) behavior can
184 be specified with the `noncumulative` parameter.
186 An integer parameter specifies a cut-off percent (3% by default).
187 Directories contributing less than this percentage of the changes
188 are not shown in the output.
191 Example: The following will count changed files, while ignoring
192 directories with less than 10% of the total amount of changed files,
193 and accumulating child directory counts in the parent directories:
194 `--dirstat=files,10,cumulative`.
197 Synonym for --dirstat=cumulative
199 --dirstat-by-file[=<param1,param2>...]::
200 Synonym for --dirstat=files,param1,param2...
203 Output a condensed summary of extended header information
204 such as creations, renames and mode changes.
206 ifndef::git-format-patch[]
208 Synonym for `-p --stat`.
209 endif::git-format-patch[]
211 ifndef::git-format-patch[]
215 Separate the commits with NULs instead of with new newlines.
217 Also, when `--raw` or `--numstat` has been given, do not munge
218 pathnames and use NULs as output field terminators.
221 When `--raw`, `--numstat`, `--name-only` or `--name-status` has been
222 given, do not munge pathnames and use NULs as output field terminators.
225 Without this option, pathnames with "unusual" characters are quoted as
226 explained for the configuration variable `core.quotePath` (see
227 linkgit:git-config[1]).
230 Show only names of changed files.
233 Show only names and status of changed files. See the description
234 of the `--diff-filter` option on what the status letters mean.
236 --submodule[=<format>]::
237 Specify how differences in submodules are shown. When specifying
238 `--submodule=short` the 'short' format is used. This format just
239 shows the names of the commits at the beginning and end of the range.
240 When `--submodule` or `--submodule=log` is specified, the 'log'
241 format is used. This format lists the commits in the range like
242 linkgit:git-submodule[1] `summary` does. When `--submodule=diff`
243 is specified, the 'diff' format is used. This format shows an
244 inline diff of the changes in the submodule contents between the
245 commit range. Defaults to `diff.submodule` or the 'short' format
246 if the config option is unset.
250 `--color` (i.e. without '=<when>') is the same as `--color=always`.
251 '<when>' can be one of `always`, `never`, or `auto`.
253 It can be changed by the `color.ui` and `color.diff`
254 configuration settings.
258 Turn off colored diff.
260 This can be used to override configuration settings.
262 It is the same as `--color=never`.
264 --color-moved[=<mode>]::
265 Moved lines of code are colored differently.
267 It can be changed by the `diff.colorMoved` configuration setting.
269 The <mode> defaults to 'no' if the option is not given
270 and to 'zebra' if the option with no mode is given.
271 The mode must be one of:
275 Moved lines are not highlighted.
277 Is a synonym for `zebra`. This may change to a more sensible mode
280 Any line that is added in one location and was removed
281 in another location will be colored with 'color.diff.newMoved'.
282 Similarly 'color.diff.oldMoved' will be used for removed lines
283 that are added somewhere else in the diff. This mode picks up any
284 moved line, but it is not very useful in a review to determine
285 if a block of code was moved without permutation.
287 Blocks of moved text of at least 20 alphanumeric characters
288 are detected greedily. The detected blocks are
289 painted using either the 'color.diff.{old,new}Moved' color.
290 Adjacent blocks cannot be told apart.
292 Blocks of moved text are detected as in 'blocks' mode. The blocks
293 are painted using either the 'color.diff.{old,new}Moved' color or
294 'color.diff.{old,new}MovedAlternative'. The change between
295 the two colors indicates that a new block was detected.
297 Similar to 'zebra', but additional dimming of uninteresting parts
298 of moved code is performed. The bordering lines of two adjacent
299 blocks are considered interesting, the rest is uninteresting.
300 `dimmed_zebra` is a deprecated synonym.
303 --color-moved-ws=<modes>::
304 This configures how white spaces are ignored when performing the
305 move detection for `--color-moved`.
307 It can be set by the `diff.colorMovedWS` configuration setting.
309 These modes can be given as a comma separated list:
312 ignore-space-at-eol::
313 Ignore changes in whitespace at EOL.
314 ignore-space-change::
315 Ignore changes in amount of whitespace. This ignores whitespace
316 at line end, and considers all other sequences of one or
317 more whitespace characters to be equivalent.
319 Ignore whitespace when comparing lines. This ignores differences
320 even if one line has whitespace where the other line has none.
321 allow-indentation-change::
322 Initially ignore any white spaces in the move detection, then
323 group the moved code blocks only into a block if the change in
324 whitespace is the same per line. This is incompatible with the
328 --word-diff[=<mode>]::
329 Show a word diff, using the <mode> to delimit changed words.
330 By default, words are delimited by whitespace; see
331 `--word-diff-regex` below. The <mode> defaults to 'plain', and
336 Highlight changed words using only colors. Implies `--color`.
338 Show words as `[-removed-]` and `{+added+}`. Makes no
339 attempts to escape the delimiters if they appear in the input,
340 so the output may be ambiguous.
342 Use a special line-based format intended for script
343 consumption. Added/removed/unchanged runs are printed in the
344 usual unified diff format, starting with a `+`/`-`/` `
345 character at the beginning of the line and extending to the
346 end of the line. Newlines in the input are represented by a
347 tilde `~` on a line of its own.
349 Disable word diff again.
352 Note that despite the name of the first mode, color is used to
353 highlight the changed parts in all modes if enabled.
355 --word-diff-regex=<regex>::
356 Use <regex> to decide what a word is, instead of considering
357 runs of non-whitespace to be a word. Also implies
358 `--word-diff` unless it was already enabled.
360 Every non-overlapping match of the
361 <regex> is considered a word. Anything between these matches is
362 considered whitespace and ignored(!) for the purposes of finding
363 differences. You may want to append `|[^[:space:]]` to your regular
364 expression to make sure that it matches all non-whitespace characters.
365 A match that contains a newline is silently truncated(!) at the
368 For example, `--word-diff-regex=.` will treat each character as a word
369 and, correspondingly, show differences character by character.
371 The regex can also be set via a diff driver or configuration option, see
372 linkgit:gitattributes[5] or linkgit:git-config[1]. Giving it explicitly
373 overrides any diff driver or configuration setting. Diff drivers
374 override configuration settings.
376 --color-words[=<regex>]::
377 Equivalent to `--word-diff=color` plus (if a regex was
378 specified) `--word-diff-regex=<regex>`.
379 endif::git-format-patch[]
382 Turn off rename detection, even when the configuration
383 file gives the default to do so.
385 ifndef::git-format-patch[]
387 Warn if changes introduce conflict markers or whitespace errors.
388 What are considered whitespace errors is controlled by `core.whitespace`
389 configuration. By default, trailing whitespaces (including
390 lines that consist solely of whitespaces) and a space character
391 that is immediately followed by a tab character inside the
392 initial indent of the line are considered whitespace errors.
393 Exits with non-zero status if problems are found. Not compatible
396 --ws-error-highlight=<kind>::
397 Highlight whitespace errors in the `context`, `old` or `new`
398 lines of the diff. Multiple values are separated by comma,
399 `none` resets previous values, `default` reset the list to
400 `new` and `all` is a shorthand for `old,new,context`. When
401 this option is not given, and the configuration variable
402 `diff.wsErrorHighlight` is not set, only whitespace errors in
403 `new` lines are highlighted. The whitespace errors are colored
404 with `color.diff.whitespace`.
406 endif::git-format-patch[]
409 Instead of the first handful of characters, show the full
410 pre- and post-image blob object names on the "index"
411 line when generating patch format output.
414 In addition to `--full-index`, output a binary diff that
415 can be applied with `git-apply`.
418 Instead of showing the full 40-byte hexadecimal object
419 name in diff-raw format output and diff-tree header
420 lines, show only a partial prefix. This is
421 independent of the `--full-index` option above, which controls
422 the diff-patch output format. Non default number of
423 digits can be specified with `--abbrev=<n>`.
426 --break-rewrites[=[<n>][/<m>]]::
427 Break complete rewrite changes into pairs of delete and
428 create. This serves two purposes:
430 It affects the way a change that amounts to a total rewrite of a file
431 not as a series of deletion and insertion mixed together with a very
432 few lines that happen to match textually as the context, but as a
433 single deletion of everything old followed by a single insertion of
434 everything new, and the number `m` controls this aspect of the -B
435 option (defaults to 60%). `-B/70%` specifies that less than 30% of the
436 original should remain in the result for Git to consider it a total
437 rewrite (i.e. otherwise the resulting patch will be a series of
438 deletion and insertion mixed together with context lines).
440 When used with -M, a totally-rewritten file is also considered as the
441 source of a rename (usually -M only considers a file that disappeared
442 as the source of a rename), and the number `n` controls this aspect of
443 the -B option (defaults to 50%). `-B20%` specifies that a change with
444 addition and deletion compared to 20% or more of the file's size are
445 eligible for being picked up as a possible source of a rename to
449 --find-renames[=<n>]::
454 If generating diffs, detect and report renames for each commit.
455 For following files across renames while traversing history, see
458 If `n` is specified, it is a threshold on the similarity
459 index (i.e. amount of addition/deletions compared to the
460 file's size). For example, `-M90%` means Git should consider a
461 delete/add pair to be a rename if more than 90% of the file
462 hasn't changed. Without a `%` sign, the number is to be read as
463 a fraction, with a decimal point before it. I.e., `-M5` becomes
464 0.5, and is thus the same as `-M50%`. Similarly, `-M05` is
465 the same as `-M5%`. To limit detection to exact renames, use
466 `-M100%`. The default similarity index is 50%.
469 --find-copies[=<n>]::
470 Detect copies as well as renames. See also `--find-copies-harder`.
471 If `n` is specified, it has the same meaning as for `-M<n>`.
473 --find-copies-harder::
474 For performance reasons, by default, `-C` option finds copies only
475 if the original file of the copy was modified in the same
476 changeset. This flag makes the command
477 inspect unmodified files as candidates for the source of
478 copy. This is a very expensive operation for large
479 projects, so use it with caution. Giving more than one
480 `-C` option has the same effect.
483 --irreversible-delete::
484 Omit the preimage for deletes, i.e. print only the header but not
485 the diff between the preimage and `/dev/null`. The resulting patch
486 is not meant to be applied with `patch` or `git apply`; this is
487 solely for people who want to just concentrate on reviewing the
488 text after the change. In addition, the output obviously lacks
489 enough information to apply such a patch in reverse, even manually,
490 hence the name of the option.
492 When used together with `-B`, omit also the preimage in the deletion part
493 of a delete/create pair.
496 The `-M` and `-C` options require O(n^2) processing time where n
497 is the number of potential rename/copy targets. This
498 option prevents rename/copy detection from running if
499 the number of rename/copy targets exceeds the specified
502 ifndef::git-format-patch[]
503 --diff-filter=[(A|C|D|M|R|T|U|X|B)...[*]]::
504 Select only files that are Added (`A`), Copied (`C`),
505 Deleted (`D`), Modified (`M`), Renamed (`R`), have their
506 type (i.e. regular file, symlink, submodule, ...) changed (`T`),
507 are Unmerged (`U`), are
508 Unknown (`X`), or have had their pairing Broken (`B`).
509 Any combination of the filter characters (including none) can be used.
510 When `*` (All-or-none) is added to the combination, all
511 paths are selected if there is any file that matches
512 other criteria in the comparison; if there is no file
513 that matches other criteria, nothing is selected.
515 Also, these upper-case letters can be downcased to exclude. E.g.
516 `--diff-filter=ad` excludes added and deleted paths.
518 Note that not all diffs can feature all types. For instance, diffs
519 from the index to the working tree can never have Added entries
520 (because the set of paths included in the diff is limited by what is in
521 the index). Similarly, copied and renamed entries cannot appear if
522 detection for those types is disabled.
525 Look for differences that change the number of occurrences of
526 the specified string (i.e. addition/deletion) in a file.
527 Intended for the scripter's use.
529 It is useful when you're looking for an exact block of code (like a
530 struct), and want to know the history of that block since it first
531 came into being: use the feature iteratively to feed the interesting
532 block in the preimage back into `-S`, and keep going until you get the
533 very first version of the block.
535 Binary files are searched as well.
538 Look for differences whose patch text contains added/removed
539 lines that match <regex>.
541 To illustrate the difference between `-S<regex> --pickaxe-regex` and
542 `-G<regex>`, consider a commit with the following diff in the same
546 + return !regexec(regexp, two->ptr, 1, ®match, 0);
548 - hit = !regexec(regexp, mf2.ptr, 1, ®match, 0);
551 While `git log -G"regexec\(regexp"` will show this commit, `git log
552 -S"regexec\(regexp" --pickaxe-regex` will not (because the number of
553 occurrences of that string did not change).
555 Unless `--text` is supplied patches of binary files without a textconv
556 filter will be ignored.
558 See the 'pickaxe' entry in linkgit:gitdiffcore[7] for more
561 --find-object=<object-id>::
562 Look for differences that change the number of occurrences of
563 the specified object. Similar to `-S`, just the argument is different
564 in that it doesn't search for a specific string but for a specific
567 The object can be a blob or a submodule commit. It implies the `-t` option in
568 `git-log` to also find trees.
571 When `-S` or `-G` finds a change, show all the changes in that
572 changeset, not just the files that contain the change
576 Treat the <string> given to `-S` as an extended POSIX regular
579 endif::git-format-patch[]
582 Control the order in which files appear in the output.
583 This overrides the `diff.orderFile` configuration variable
584 (see linkgit:git-config[1]). To cancel `diff.orderFile`,
587 The output order is determined by the order of glob patterns in
589 All files with pathnames that match the first pattern are output
590 first, all files with pathnames that match the second pattern (but not
591 the first) are output next, and so on.
592 All files with pathnames that do not match any pattern are output
593 last, as if there was an implicit match-all pattern at the end of the
595 If multiple pathnames have the same rank (they match the same pattern
596 but no earlier patterns), their output order relative to each other is
599 <orderfile> is parsed as follows:
602 - Blank lines are ignored, so they can be used as separators for
605 - Lines starting with a hash ("`#`") are ignored, so they can be used
606 for comments. Add a backslash ("`\`") to the beginning of the
607 pattern if it starts with a hash.
609 - Each other line contains a single pattern.
612 Patterns have the same syntax and semantics as patterns used for
613 fnmatch(3) without the FNM_PATHNAME flag, except a pathname also
614 matches a pattern if removing any number of the final pathname
615 components matches the pattern. For example, the pattern "`foo*bar`"
616 matches "`fooasdfbar`" and "`foo/bar/baz/asdf`" but not "`foobarx`".
618 ifndef::git-format-patch[]
620 Swap two inputs; that is, show differences from index or
621 on-disk file to tree contents.
623 --relative[=<path>]::
624 When run from a subdirectory of the project, it can be
625 told to exclude changes outside the directory and show
626 pathnames relative to it with this option. When you are
627 not in a subdirectory (e.g. in a bare repository), you
628 can name which subdirectory to make the output relative
629 to by giving a <path> as an argument.
630 endif::git-format-patch[]
634 Treat all files as text.
637 Ignore carriage-return at the end of line when doing a comparison.
639 --ignore-space-at-eol::
640 Ignore changes in whitespace at EOL.
643 --ignore-space-change::
644 Ignore changes in amount of whitespace. This ignores whitespace
645 at line end, and considers all other sequences of one or
646 more whitespace characters to be equivalent.
650 Ignore whitespace when comparing lines. This ignores
651 differences even if one line has whitespace where the other
654 --ignore-blank-lines::
655 Ignore changes whose lines are all blank.
657 --inter-hunk-context=<lines>::
658 Show the context between diff hunks, up to the specified number
659 of lines, thereby fusing hunks that are close to each other.
660 Defaults to `diff.interHunkContext` or 0 if the config option
665 Show whole surrounding functions of changes.
667 ifndef::git-format-patch[]
670 Make the program exit with codes similar to diff(1).
671 That is, it exits with 1 if there were differences and
672 0 means no differences.
675 Disable all output of the program. Implies `--exit-code`.
677 endif::git-format-patch[]
680 Allow an external diff helper to be executed. If you set an
681 external diff driver with linkgit:gitattributes[5], you need
682 to use this option with linkgit:git-log[1] and friends.
685 Disallow external diff drivers.
689 Allow (or disallow) external text conversion filters to be run
690 when comparing binary files. See linkgit:gitattributes[5] for
691 details. Because textconv filters are typically a one-way
692 conversion, the resulting diff is suitable for human
693 consumption, but cannot be applied. For this reason, textconv
694 filters are enabled by default only for linkgit:git-diff[1] and
695 linkgit:git-log[1], but not for linkgit:git-format-patch[1] or
696 diff plumbing commands.
698 --ignore-submodules[=<when>]::
699 Ignore changes to submodules in the diff generation. <when> can be
700 either "none", "untracked", "dirty" or "all", which is the default.
701 Using "none" will consider the submodule modified when it either contains
702 untracked or modified files or its HEAD differs from the commit recorded
703 in the superproject and can be used to override any settings of the
704 'ignore' option in linkgit:git-config[1] or linkgit:gitmodules[5]. When
705 "untracked" is used submodules are not considered dirty when they only
706 contain untracked content (but they are still scanned for modified
707 content). Using "dirty" ignores all changes to the work tree of submodules,
708 only changes to the commits stored in the superproject are shown (this was
709 the behavior until 1.7.0). Using "all" hides all changes to submodules.
711 --src-prefix=<prefix>::
712 Show the given source prefix instead of "a/".
714 --dst-prefix=<prefix>::
715 Show the given destination prefix instead of "b/".
718 Do not show any source or destination prefix.
720 --line-prefix=<prefix>::
721 Prepend an additional prefix to every line of output.
723 --ita-invisible-in-index::
724 By default entries added by "git add -N" appear as an existing
725 empty file in "git diff" and a new file in "git diff --cached".
726 This option makes the entry appear as a new file in "git diff"
727 and non-existent in "git diff --cached". This option could be
728 reverted with `--ita-visible-in-index`. Both options are
729 experimental and could be removed in future.
731 For more detailed explanation on these common options, see also
732 linkgit:gitdiffcore[7].