1GIT-DIFF-TREE(1) Git Manual GIT-DIFF-TREE(1)
2
6 git-diff-tree - Compares the content and mode of blobs found via two
7 tree objects
8
10 git diff-tree [--stdin] [-m] [-s] [-v] [--no-commit-id] [--pretty]
11 [-t] [-r] [-c | --cc] [--combined-all-paths] [--root]
12 [<common diff options>] <tree-ish> [<tree-ish>] [<path>...]
13
16 Compares the content and mode of the blobs found via two tree objects.
17 If there is only one <tree-ish> given, the commit is compared with its
19 parents (see --stdin below).
20 Note that git diff-tree can use the tree encapsulated in a commit
22 object.
23
25 -p, -u, --patch
26 Generate patch (see section on generating patches).
27 -s, --no-patch
29 Suppress diff output. Useful for commands like git show that show
30 the patch by default, or to cancel the effect of --patch.
31 -U<n>, --unified=<n>
33 Generate diffs with <n> lines of context instead of the usual
34 three. Implies --patch. Implies -p.
35 --output=<file>
37 Output to a specific file instead of stdout.
38 --output-indicator-new=<char>, --output-indicator-old=<char>,
40 --output-indicator-context=<char>
41 Specify the character used to indicate new, old or context lines in
42 the generated patch. Normally they are +, - and ' ' respectively.
43 --raw
45 Generate the diff in raw format. This is the default.
46 --patch-with-raw
48 Synonym for -p --raw.
49 --indent-heuristic
51 Enable the heuristic that shifts diff hunk boundaries to make
52 patches easier to read. This is the default.
53 --no-indent-heuristic
55 Disable the indent heuristic.
56 --minimal
58 Spend extra time to make sure the smallest possible diff is
59 produced.
60 --patience
62 Generate a diff using the "patience diff" algorithm.
63 --histogram
65 Generate a diff using the "histogram diff" algorithm.
66 --anchored=<text>
68 Generate a diff using the "anchored diff" algorithm.
69 This option may be specified more than once.
71 If a line exists in both the source and destination, exists only
73 once, and starts with this text, this algorithm attempts to prevent
74 it from appearing as a deletion or addition in the output. It uses
75 the "patience diff" algorithm internally.
76 --diff-algorithm={patience|minimal|histogram|myers}
78 Choose a diff algorithm. The variants are as follows:
79 default, myers
81 The basic greedy diff algorithm. Currently, this is the
82 default.
83 minimal
85 Spend extra time to make sure the smallest possible diff is
86 produced.
87 patience
89 Use "patience diff" algorithm when generating patches.
90 histogram
92 This algorithm extends the patience algorithm to "support
93 low-occurrence common elements".
94 For instance, if you configured the diff.algorithm variable to a
96 non-default value and want to use the default one, then you have to
97 use --diff-algorithm=default option.
98 --stat[=<width>[,<name-width>[,<count>]]]
100 Generate a diffstat. By default, as much space as necessary will be
101 used for the filename part, and the rest for the graph part.
102 Maximum width defaults to terminal width, or 80 columns if not
103 connected to a terminal, and can be overridden by <width>. The
104 width of the filename part can be limited by giving another width
105 <name-width> after a comma. The width of the graph part can be
106 limited by using --stat-graph-width=<width> (affects all commands
107 generating a stat graph) or by setting diff.statGraphWidth=<width>
108 (does not affect git format-patch). By giving a third parameter
109 <count>, you can limit the output to the first <count> lines,
110 followed by ... if there are more.
111 These parameters can also be set individually with
113 --stat-width=<width>, --stat-name-width=<name-width> and
114 --stat-count=<count>.
115 --compact-summary
117 Output a condensed summary of extended header information such as
118 file creations or deletions ("new" or "gone", optionally "+l" if
119 it’s a symlink) and mode changes ("+x" or "-x" for adding or
120 removing executable bit respectively) in diffstat. The information
121 is put between the filename part and the graph part. Implies
122 --stat.
123 --numstat
125 Similar to --stat, but shows number of added and deleted lines in
126 decimal notation and pathname without abbreviation, to make it more
127 machine friendly. For binary files, outputs two - instead of saying
128 0 0.
129 --shortstat
131 Output only the last line of the --stat format containing total
132 number of modified files, as well as number of added and deleted
133 lines.
134 -X[<param1,param2,...>], --dirstat[=<param1,param2,...>]
136 Output the distribution of relative amount of changes for each
137 sub-directory. The behavior of --dirstat can be customized by
138 passing it a comma separated list of parameters. The defaults are
139 controlled by the diff.dirstat configuration variable (see git-
140 config(1)). The following parameters are available:
141 changes
143 Compute the dirstat numbers by counting the lines that have
144 been removed from the source, or added to the destination. This
145 ignores the amount of pure code movements within a file. In
146 other words, rearranging lines in a file is not counted as much
147 as other changes. This is the default behavior when no
148 parameter is given.
149 lines
151 Compute the dirstat numbers by doing the regular line-based
152 diff analysis, and summing the removed/added line counts. (For
153 binary files, count 64-byte chunks instead, since binary files
154 have no natural concept of lines). This is a more expensive
155 --dirstat behavior than the changes behavior, but it does count
156 rearranged lines within a file as much as other changes. The
157 resulting output is consistent with what you get from the other
158 --*stat options.
159 files
161 Compute the dirstat numbers by counting the number of files
162 changed. Each changed file counts equally in the dirstat
163 analysis. This is the computationally cheapest --dirstat
164 behavior, since it does not have to look at the file contents
165 at all.
166 cumulative
168 Count changes in a child directory for the parent directory as
169 well. Note that when using cumulative, the sum of the
170 percentages reported may exceed 100%. The default
171 (non-cumulative) behavior can be specified with the
172 noncumulative parameter.
173 <limit>
175 An integer parameter specifies a cut-off percent (3% by
176 default). Directories contributing less than this percentage of
177 the changes are not shown in the output.
178 Example: The following will count changed files, while ignoring
180 directories with less than 10% of the total amount of changed
181 files, and accumulating child directory counts in the parent
182 directories: --dirstat=files,10,cumulative.
183 --cumulative
185 Synonym for --dirstat=cumulative
186 --dirstat-by-file[=<param1,param2>...]
188 Synonym for --dirstat=files,param1,param2...
189 --summary
191 Output a condensed summary of extended header information such as
192 creations, renames and mode changes.
193 --patch-with-stat
195 Synonym for -p --stat.
196 -z
198 When --raw, --numstat, --name-only or --name-status has been given,
199 do not munge pathnames and use NULs as output field terminators.
200 Without this option, pathnames with "unusual" characters are quoted
202 as explained for the configuration variable core.quotePath (see
203 git-config(1)).
204 --name-only
206 Show only names of changed files.
207 --name-status
209 Show only names and status of changed files. See the description of
210 the --diff-filter option on what the status letters mean.
211 --submodule[=<format>]
213 Specify how differences in submodules are shown. When specifying
214 --submodule=short the short format is used. This format just shows
215 the names of the commits at the beginning and end of the range.
216 When --submodule or --submodule=log is specified, the log format is
217 used. This format lists the commits in the range like git-
218 submodule(1) summary does. When --submodule=diff is specified, the
219 diff format is used. This format shows an inline diff of the
220 changes in the submodule contents between the commit range.
221 Defaults to diff.submodule or the short format if the config option
222 is unset.
223 --color[=<when>]
225 Show colored diff. --color (i.e. without =<when>) is the same as
226 --color=always. <when> can be one of always, never, or auto.
227 --no-color
229 Turn off colored diff. It is the same as --color=never.
230 --color-moved[=<mode>]
232 Moved lines of code are colored differently. The <mode> defaults to
233 no if the option is not given and to zebra if the option with no
234 mode is given. The mode must be one of:
235 no
237 Moved lines are not highlighted.
238 default
240 Is a synonym for zebra. This may change to a more sensible mode
241 in the future.
242 plain
244 Any line that is added in one location and was removed in
245 another location will be colored with color.diff.newMoved.
246 Similarly color.diff.oldMoved will be used for removed lines
247 that are added somewhere else in the diff. This mode picks up
248 any moved line, but it is not very useful in a review to
249 determine if a block of code was moved without permutation.
250 blocks
252 Blocks of moved text of at least 20 alphanumeric characters are
253 detected greedily. The detected blocks are painted using either
254 the color.diff.{old,new}Moved color. Adjacent blocks cannot be
255 told apart.
256 zebra
258 Blocks of moved text are detected as in blocks mode. The blocks
259 are painted using either the color.diff.{old,new}Moved color or
260 color.diff.{old,new}MovedAlternative. The change between the
261 two colors indicates that a new block was detected.
262 dimmed-zebra
264 Similar to zebra, but additional dimming of uninteresting parts
265 of moved code is performed. The bordering lines of two adjacent
266 blocks are considered interesting, the rest is uninteresting.
267 dimmed_zebra is a deprecated synonym.
268 --no-color-moved
270 Turn off move detection. This can be used to override configuration
271 settings. It is the same as --color-moved=no.
272 --color-moved-ws=<modes>
274 This configures how whitespace is ignored when performing the move
275 detection for --color-moved. These modes can be given as a comma
276 separated list:
277 no
279 Do not ignore whitespace when performing move detection.
280 ignore-space-at-eol
282 Ignore changes in whitespace at EOL.
283 ignore-space-change
285 Ignore changes in amount of whitespace. This ignores whitespace
286 at line end, and considers all other sequences of one or more
287 whitespace characters to be equivalent.
288 ignore-all-space
290 Ignore whitespace when comparing lines. This ignores
291 differences even if one line has whitespace where the other
292 line has none.
293 allow-indentation-change
295 Initially ignore any whitespace in the move detection, then
296 group the moved code blocks only into a block if the change in
297 whitespace is the same per line. This is incompatible with the
298 other modes.
299 --no-color-moved-ws
301 Do not ignore whitespace when performing move detection. This can
302 be used to override configuration settings. It is the same as
303 --color-moved-ws=no.
304 --word-diff[=<mode>]
306 Show a word diff, using the <mode> to delimit changed words. By
307 default, words are delimited by whitespace; see --word-diff-regex
308 below. The <mode> defaults to plain, and must be one of:
309 color
311 Highlight changed words using only colors. Implies --color.
312 plain
314 Show words as [-removed-] and {+added+}. Makes no attempts to
315 escape the delimiters if they appear in the input, so the
316 output may be ambiguous.
317 porcelain
319 Use a special line-based format intended for script
320 consumption. Added/removed/unchanged runs are printed in the
321 usual unified diff format, starting with a +/-/` ` character at
322 the beginning of the line and extending to the end of the line.
323 Newlines in the input are represented by a tilde ~ on a line of
324 its own.
325 none
327 Disable word diff again.
328 Note that despite the name of the first mode, color is used to
330 highlight the changed parts in all modes if enabled.
331 --word-diff-regex=<regex>
333 Use <regex> to decide what a word is, instead of considering runs
334 of non-whitespace to be a word. Also implies --word-diff unless it
335 was already enabled.
336 Every non-overlapping match of the <regex> is considered a word.
338 Anything between these matches is considered whitespace and
339 ignored(!) for the purposes of finding differences. You may want to
340 append |[^[:space:]] to your regular expression to make sure that
341 it matches all non-whitespace characters. A match that contains a
342 newline is silently truncated(!) at the newline.
343 For example, --word-diff-regex=. will treat each character as a
345 word and, correspondingly, show differences character by character.
346 The regex can also be set via a diff driver or configuration
348 option, see gitattributes(5) or git-config(1). Giving it explicitly
349 overrides any diff driver or configuration setting. Diff drivers
350 override configuration settings.
351 --color-words[=<regex>]
353 Equivalent to --word-diff=color plus (if a regex was specified)
354 --word-diff-regex=<regex>.
355 --no-renames
357 Turn off rename detection, even when the configuration file gives
358 the default to do so.
359 --[no-]rename-empty
361 Whether to use empty blobs as rename source.
362 --check
364 Warn if changes introduce conflict markers or whitespace errors.
365 What are considered whitespace errors is controlled by
366 core.whitespace configuration. By default, trailing whitespaces
367 (including lines that consist solely of whitespaces) and a space
368 character that is immediately followed by a tab character inside
369 the initial indent of the line are considered whitespace errors.
370 Exits with non-zero status if problems are found. Not compatible
371 with --exit-code.
372 --ws-error-highlight=<kind>
374 Highlight whitespace errors in the context, old or new lines of the
375 diff. Multiple values are separated by comma, none resets previous
376 values, default reset the list to new and all is a shorthand for
377 old,new,context. When this option is not given, and the
378 configuration variable diff.wsErrorHighlight is not set, only
379 whitespace errors in new lines are highlighted. The whitespace
380 errors are colored with color.diff.whitespace.
381 --full-index
383 Instead of the first handful of characters, show the full pre- and
384 post-image blob object names on the "index" line when generating
385 patch format output.
386 --binary
388 In addition to --full-index, output a binary diff that can be
389 applied with git-apply. Implies --patch.
390 --abbrev[=<n>]
392 Instead of showing the full 40-byte hexadecimal object name in
393 diff-raw format output and diff-tree header lines, show only a
394 partial prefix. This is independent of the --full-index option
395 above, which controls the diff-patch output format. Non default
396 number of digits can be specified with --abbrev=<n>.
397 -B[<n>][/<m>], --break-rewrites[=[<n>][/<m>]]
399 Break complete rewrite changes into pairs of delete and create.
400 This serves two purposes:
401 It affects the way a change that amounts to a total rewrite of a
403 file not as a series of deletion and insertion mixed together with
404 a very few lines that happen to match textually as the context, but
405 as a single deletion of everything old followed by a single
406 insertion of everything new, and the number m controls this aspect
407 of the -B option (defaults to 60%). -B/70% specifies that less
408 than 30% of the original should remain in the result for Git to
409 consider it a total rewrite (i.e. otherwise the resulting patch
410 will be a series of deletion and insertion mixed together with
411 context lines).
412 When used with -M, a totally-rewritten file is also considered as
414 the source of a rename (usually -M only considers a file that
415 disappeared as the source of a rename), and the number n controls
416 this aspect of the -B option (defaults to 50%). -B20% specifies
417 that a change with addition and deletion compared to 20% or more of
418 the file’s size are eligible for being picked up as a possible
419 source of a rename to another file.
420 -M[<n>], --find-renames[=<n>]
422 Detect renames. If n is specified, it is a threshold on the
423 similarity index (i.e. amount of addition/deletions compared to the
424 file’s size). For example, -M90% means Git should consider a
425 delete/add pair to be a rename if more than 90% of the file hasn’t
426 changed. Without a % sign, the number is to be read as a fraction,
427 with a decimal point before it. I.e., -M5 becomes 0.5, and is thus
428 the same as -M50%. Similarly, -M05 is the same as -M5%. To limit
429 detection to exact renames, use -M100%. The default similarity
430 index is 50%.
431 -C[<n>], --find-copies[=<n>]
433 Detect copies as well as renames. See also --find-copies-harder. If
434 n is specified, it has the same meaning as for -M<n>.
435 --find-copies-harder
437 For performance reasons, by default, -C option finds copies only if
438 the original file of the copy was modified in the same changeset.
439 This flag makes the command inspect unmodified files as candidates
440 for the source of copy. This is a very expensive operation for
441 large projects, so use it with caution. Giving more than one -C
442 option has the same effect.
443 -D, --irreversible-delete
445 Omit the preimage for deletes, i.e. print only the header but not
446 the diff between the preimage and /dev/null. The resulting patch is
447 not meant to be applied with patch or git apply; this is solely for
448 people who want to just concentrate on reviewing the text after the
449 change. In addition, the output obviously lacks enough information
450 to apply such a patch in reverse, even manually, hence the name of
451 the option.
452 When used together with -B, omit also the preimage in the deletion
454 part of a delete/create pair.
455 -l<num>
457 The -M and -C options require O(n^2) processing time where n is the
458 number of potential rename/copy targets. This option prevents
459 rename/copy detection from running if the number of rename/copy
460 targets exceeds the specified number.
461 --diff-filter=[(A|C|D|M|R|T|U|X|B)...[*]]
463 Select only files that are Added (A), Copied (C), Deleted (D),
464 Modified (M), Renamed (R), have their type (i.e. regular file,
465 symlink, submodule, ...) changed (T), are Unmerged (U), are Unknown
466 (X), or have had their pairing Broken (B). Any combination of the
467 filter characters (including none) can be used. When *
468 (All-or-none) is added to the combination, all paths are selected
469 if there is any file that matches other criteria in the comparison;
470 if there is no file that matches other criteria, nothing is
471 selected.
472 Also, these upper-case letters can be downcased to exclude. E.g.
474 --diff-filter=ad excludes added and deleted paths.
475 Note that not all diffs can feature all types. For instance, diffs
477 from the index to the working tree can never have Added entries
478 (because the set of paths included in the diff is limited by what
479 is in the index). Similarly, copied and renamed entries cannot
480 appear if detection for those types is disabled.
481 -S<string>
483 Look for differences that change the number of occurrences of the
484 specified string (i.e. addition/deletion) in a file. Intended for
485 the scripter’s use.
486 It is useful when you’re looking for an exact block of code (like a
488 struct), and want to know the history of that block since it first
489 came into being: use the feature iteratively to feed the
490 interesting block in the preimage back into -S, and keep going
491 until you get the very first version of the block.
492 Binary files are searched as well.
494 -G<regex>
496 Look for differences whose patch text contains added/removed lines
497 that match <regex>.
498 To illustrate the difference between -S<regex> --pickaxe-regex and
500 -G<regex>, consider a commit with the following diff in the same
501 file:
502 + return !regexec(regexp, two->ptr, 1, ®match, 0);
504 ...
505 - hit = !regexec(regexp, mf2.ptr, 1, ®match, 0);
506 While git log -G"regexec\(regexp" will show this commit, git log
508 -S"regexec\(regexp" --pickaxe-regex will not (because the number of
509 occurrences of that string did not change).
510 Unless --text is supplied patches of binary files without a
512 textconv filter will be ignored.
513 See the pickaxe entry in gitdiffcore(7) for more information.
515 --find-object=<object-id>
517 Look for differences that change the number of occurrences of the
518 specified object. Similar to -S, just the argument is different in
519 that it doesn’t search for a specific string but for a specific
520 object id.
521 The object can be a blob or a submodule commit. It implies the -t
523 option in git-log to also find trees.
524 --pickaxe-all
526 When -S or -G finds a change, show all the changes in that
527 changeset, not just the files that contain the change in <string>.
528 --pickaxe-regex
530 Treat the <string> given to -S as an extended POSIX regular
531 expression to match.
532 -O<orderfile>
534 Control the order in which files appear in the output. This
535 overrides the diff.orderFile configuration variable (see git-
536 config(1)). To cancel diff.orderFile, use -O/dev/null.
537 The output order is determined by the order of glob patterns in
539 <orderfile>. All files with pathnames that match the first pattern
540 are output first, all files with pathnames that match the second
541 pattern (but not the first) are output next, and so on. All files
542 with pathnames that do not match any pattern are output last, as if
543 there was an implicit match-all pattern at the end of the file. If
544 multiple pathnames have the same rank (they match the same pattern
545 but no earlier patterns), their output order relative to each other
546 is the normal order.
547 <orderfile> is parsed as follows:
549 · Blank lines are ignored, so they can be used as separators for
551 readability.
552 · Lines starting with a hash ("#") are ignored, so they can be
554 used for comments. Add a backslash ("\") to the beginning of
555 the pattern if it starts with a hash.
556 · Each other line contains a single pattern.
558 Patterns have the same syntax and semantics as patterns used for
560 fnmatch(3) without the FNM_PATHNAME flag, except a pathname also
561 matches a pattern if removing any number of the final pathname
562 components matches the pattern. For example, the pattern "foo*bar"
563 matches "fooasdfbar" and "foo/bar/baz/asdf" but not "foobarx".
564 -R
566 Swap two inputs; that is, show differences from index or on-disk
567 file to tree contents.
568 --relative[=<path>]
570 When run from a subdirectory of the project, it can be told to
571 exclude changes outside the directory and show pathnames relative
572 to it with this option. When you are not in a subdirectory (e.g. in
573 a bare repository), you can name which subdirectory to make the
574 output relative to by giving a <path> as an argument.
575 -a, --text
577 Treat all files as text.
578 --ignore-cr-at-eol
580 Ignore carriage-return at the end of line when doing a comparison.
581 --ignore-space-at-eol
583 Ignore changes in whitespace at EOL.
584 -b, --ignore-space-change
586 Ignore changes in amount of whitespace. This ignores whitespace at
587 line end, and considers all other sequences of one or more
588 whitespace characters to be equivalent.
589 -w, --ignore-all-space
591 Ignore whitespace when comparing lines. This ignores differences
592 even if one line has whitespace where the other line has none.
593 --ignore-blank-lines
595 Ignore changes whose lines are all blank.
596 --inter-hunk-context=<lines>
598 Show the context between diff hunks, up to the specified number of
599 lines, thereby fusing hunks that are close to each other. Defaults
600 to diff.interHunkContext or 0 if the config option is unset.
601 -W, --function-context
603 Show whole surrounding functions of changes.
604 --exit-code
606 Make the program exit with codes similar to diff(1). That is, it
607 exits with 1 if there were differences and 0 means no differences.
608 --quiet
610 Disable all output of the program. Implies --exit-code.
611 --ext-diff
613 Allow an external diff helper to be executed. If you set an
614 external diff driver with gitattributes(5), you need to use this
615 option with git-log(1) and friends.
616 --no-ext-diff
618 Disallow external diff drivers.
619 --textconv, --no-textconv
621 Allow (or disallow) external text conversion filters to be run when
622 comparing binary files. See gitattributes(5) for details. Because
623 textconv filters are typically a one-way conversion, the resulting
624 diff is suitable for human consumption, but cannot be applied. For
625 this reason, textconv filters are enabled by default only for git-
626 diff(1) and git-log(1), but not for git-format-patch(1) or diff
627 plumbing commands.
628 --ignore-submodules[=<when>]
630 Ignore changes to submodules in the diff generation. <when> can be
631 either "none", "untracked", "dirty" or "all", which is the default.
632 Using "none" will consider the submodule modified when it either
633 contains untracked or modified files or its HEAD differs from the
634 commit recorded in the superproject and can be used to override any
635 settings of the ignore option in git-config(1) or gitmodules(5).
636 When "untracked" is used submodules are not considered dirty when
637 they only contain untracked content (but they are still scanned for
638 modified content). Using "dirty" ignores all changes to the work
639 tree of submodules, only changes to the commits stored in the
640 superproject are shown (this was the behavior until 1.7.0). Using
641 "all" hides all changes to submodules.
642 --src-prefix=<prefix>
644 Show the given source prefix instead of "a/".
645 --dst-prefix=<prefix>
647 Show the given destination prefix instead of "b/".
648 --no-prefix
650 Do not show any source or destination prefix.
651 --line-prefix=<prefix>
653 Prepend an additional prefix to every line of output.
654 --ita-invisible-in-index
656 By default entries added by "git add -N" appear as an existing
657 empty file in "git diff" and a new file in "git diff --cached".
658 This option makes the entry appear as a new file in "git diff" and
659 non-existent in "git diff --cached". This option could be reverted
660 with --ita-visible-in-index. Both options are experimental and
661 could be removed in future.
662 For more detailed explanation on these common options, see also
664 gitdiffcore(7).
665 <tree-ish>
667 The id of a tree object.
668 <path>...
670 If provided, the results are limited to a subset of files matching
671 one of the provided pathspecs.
672 -r
674 recurse into sub-trees
675 -t
677 show tree entry itself as well as subtrees. Implies -r.
678 --root
680 When --root is specified the initial commit will be shown as a big
681 creation event. This is equivalent to a diff against the NULL tree.
682 --stdin
684 When --stdin is specified, the command does not take <tree-ish>
685 arguments from the command line. Instead, it reads lines containing
686 either two <tree>, one <commit>, or a list of <commit> from its
687 standard input. (Use a single space as separator.)
688 When two trees are given, it compares the first tree with the
690 second. When a single commit is given, it compares the commit with
691 its parents. The remaining commits, when given, are used as if they
692 are parents of the first commit.
693 When comparing two trees, the ID of both trees (separated by a
695 space and terminated by a newline) is printed before the
696 difference. When comparing commits, the ID of the first (or only)
697 commit, followed by a newline, is printed.
698 The following flags further affect the behavior when comparing
700 commits (but not trees).
701 -m
703 By default, git diff-tree --stdin does not show differences for
704 merge commits. With this flag, it shows differences to that commit
705 from all of its parents. See also -c.
706 -s
708 By default, git diff-tree --stdin shows differences, either in
709 machine-readable form (without -p) or in patch form (with -p). This
710 output can be suppressed. It is only useful with -v flag.
711 -v
713 This flag causes git diff-tree --stdin to also show the commit
714 message before the differences.
715 --pretty[=<format>], --format=<format>
717 Pretty-print the contents of the commit logs in a given format,
718 where <format> can be one of oneline, short, medium, full, fuller,
719 email, raw, format:<string> and tformat:<string>. When <format> is
720 none of the above, and has %placeholder in it, it acts as if
721 --pretty=tformat:<format> were given.
722 See the "PRETTY FORMATS" section for some additional details for
724 each format. When =<format> part is omitted, it defaults to medium.
725 Note: you can specify the default pretty format in the repository
727 configuration (see git-config(1)).
728 --abbrev-commit
730 Instead of showing the full 40-byte hexadecimal commit object name,
731 show only a partial prefix. Non default number of digits can be
732 specified with "--abbrev=<n>" (which also modifies diff output, if
733 it is displayed).
734 This should make "--pretty=oneline" a whole lot more readable for
736 people using 80-column terminals.
737 --no-abbrev-commit
739 Show the full 40-byte hexadecimal commit object name. This negates
740 --abbrev-commit and those options which imply it such as
741 "--oneline". It also overrides the log.abbrevCommit variable.
742 --oneline
744 This is a shorthand for "--pretty=oneline --abbrev-commit" used
745 together.
746 --encoding=<encoding>
748 The commit objects record the encoding used for the log message in
749 their encoding header; this option can be used to tell the command
750 to re-code the commit log message in the encoding preferred by the
751 user. For non plumbing commands this defaults to UTF-8. Note that
752 if an object claims to be encoded in X and we are outputting in X,
753 we will output the object verbatim; this means that invalid
754 sequences in the original commit may be copied to the output.
755 --expand-tabs=<n>, --expand-tabs, --no-expand-tabs
757 Perform a tab expansion (replace each tab with enough spaces to
758 fill to the next display column that is multiple of <n>) in the log
759 message before showing it in the output. --expand-tabs is a
760 short-hand for --expand-tabs=8, and --no-expand-tabs is a
761 short-hand for --expand-tabs=0, which disables tab expansion.
762 By default, tabs are expanded in pretty formats that indent the log
764 message by 4 spaces (i.e. medium, which is the default, full, and
765 fuller).
766 --notes[=<treeish>]
768 Show the notes (see git-notes(1)) that annotate the commit, when
769 showing the commit log message. This is the default for git log,
770 git show and git whatchanged commands when there is no --pretty,
771 --format, or --oneline option given on the command line.
772 By default, the notes shown are from the notes refs listed in the
774 core.notesRef and notes.displayRef variables (or corresponding
775 environment overrides). See git-config(1) for more details.
776 With an optional <treeish> argument, use the treeish to find the
778 notes to display. The treeish can specify the full refname when it
779 begins with refs/notes/; when it begins with notes/, refs/ and
780 otherwise refs/notes/ is prefixed to form a full name of the ref.
781 Multiple --notes options can be combined to control which notes are
783 being displayed. Examples: "--notes=foo" will show only notes from
784 "refs/notes/foo"; "--notes=foo --notes" will show both notes from
785 "refs/notes/foo" and from the default notes ref(s).
786 --no-notes
788 Do not show notes. This negates the above --notes option, by
789 resetting the list of notes refs from which notes are shown.
790 Options are parsed in the order given on the command line, so e.g.
791 "--notes --notes=foo --no-notes --notes=bar" will only show notes
792 from "refs/notes/bar".
793 --show-notes[=<treeish>], --[no-]standard-notes
795 These options are deprecated. Use the above --notes/--no-notes
796 options instead.
797 --show-signature
799 Check the validity of a signed commit object by passing the
800 signature to gpg --verify and show the output.
801 --no-commit-id
803 git diff-tree outputs a line with the commit ID when applicable.
804 This flag suppressed the commit ID output.
805 -c
807 This flag changes the way a merge commit is displayed (which means
808 it is useful only when the command is given one <tree-ish>, or
809 --stdin). It shows the differences from each of the parents to the
810 merge result simultaneously instead of showing pairwise diff
811 between a parent and the result one at a time (which is what the -m
812 option does). Furthermore, it lists only files which were modified
813 from all parents.
814 --cc
816 This flag changes the way a merge commit patch is displayed, in a
817 similar way to the -c option. It implies the -c and -p options and
818 further compresses the patch output by omitting uninteresting hunks
819 whose the contents in the parents have only two variants and the
820 merge result picks one of them without modification. When all hunks
821 are uninteresting, the commit itself and the commit log message is
822 not shown, just like in any other "empty diff" case.
823 --combined-all-paths
825 This flag causes combined diffs (used for merge commits) to list
826 the name of the file from all parents. It thus only has effect when
827 -c or --cc are specified, and is likely only useful if filename
828 changes are detected (i.e. when either rename or copy detection
829 have been requested).
830 --always
832 Show the commit itself and the commit log message even if the diff
833 itself is empty.
834
836 If the commit is a merge, and if the pretty-format is not oneline,
837 email or raw, an additional line is inserted before the Author: line.
838 This line begins with "Merge: " and the sha1s of ancestral commits are
839 printed, separated by spaces. Note that the listed commits may not
840 necessarily be the list of the direct parent commits if you have
841 limited your view of history: for example, if you are only interested
842 in changes related to a certain directory or file.
843 There are several built-in formats, and you can define additional
845 formats by setting a pretty.<name> config option to either another
846 format name, or a format: string, as described below (see git-
847 config(1)). Here are the details of the built-in formats:
848 · oneline
850 <sha1> <title line>
852 This is designed to be as compact as possible.
854 · short
856 commit <sha1>
858 Author: <author>
859 <title line>
861 · medium
863 commit <sha1>
865 Author: <author>
866 Date: <author date>
867 <title line>
869 <full commit message>
871 · full
873 commit <sha1>
875 Author: <author>
876 Commit: <committer>
877 <title line>
879 <full commit message>
881 · fuller
883 commit <sha1>
885 Author: <author>
886 AuthorDate: <author date>
887 Commit: <committer>
888 CommitDate: <committer date>
889 <title line>
891 <full commit message>
893 · email
895 From <sha1> <date>
897 From: <author>
898 Date: <author date>
899 Subject: [PATCH] <title line>
900 <full commit message>
902 · raw
904 The raw format shows the entire commit exactly as stored in the
906 commit object. Notably, the SHA-1s are displayed in full,
907 regardless of whether --abbrev or --no-abbrev are used, and parents
908 information show the true parent commits, without taking grafts or
909 history simplification into account. Note that this format affects
910 the way commits are displayed, but not the way the diff is shown
911 e.g. with git log --raw. To get full object names in a raw diff
912 format, use --no-abbrev.
913 · format:<string>
915 The format:<string> format allows you to specify which information
917 you want to show. It works a little bit like printf format, with
918 the notable exception that you get a newline with %n instead of \n.
919 E.g, format:"The author of %h was %an, %ar%nThe title was >>%s<<%n"
921 would show something like this:
922 The author of fe6e0ee was Junio C Hamano, 23 hours ago
924 The title was >>t4119: test autocomputing -p<n> for traditional diff input.<<
925 The placeholders are:
927 · Placeholders that expand to a single literal character:
929 %n
931 newline
932 %%
934 a raw %
935 %x00
937 print a byte from a hex code
938 · Placeholders that affect formatting of later placeholders:
940 %Cred
942 switch color to red
943 %Cgreen
945 switch color to green
946 %Cblue
948 switch color to blue
949 %Creset
951 reset color
952 %C(...)
954 color specification, as described under Values in the
955 "CONFIGURATION FILE" section of git-config(1). By default,
956 colors are shown only when enabled for log output (by
957 color.diff, color.ui, or --color, and respecting the auto
958 settings of the former if we are going to a terminal).
959 %C(auto,...) is accepted as a historical synonym for the
960 default (e.g., %C(auto,red)). Specifying %C(always,...)
961 will show the colors even when color is not otherwise
962 enabled (though consider just using --color=always to
963 enable color for the whole output, including this format
964 and anything else git might color). auto alone (i.e.
965 %C(auto)) will turn on auto coloring on the next
966 placeholders until the color is switched again.
967 %m
969 left (<), right (>) or boundary (-) mark
970 %w([<w>[,<i1>[,<i2>]]])
972 switch line wrapping, like the -w option of git-
973 shortlog(1).
974 %<(<N>[,trunc|ltrunc|mtrunc])
976 make the next placeholder take at least N columns, padding
977 spaces on the right if necessary. Optionally truncate at
978 the beginning (ltrunc), the middle (mtrunc) or the end
979 (trunc) if the output is longer than N columns. Note that
980 truncating only works correctly with N >= 2.
981 %<|(<N>)
983 make the next placeholder take at least until Nth columns,
984 padding spaces on the right if necessary
985 %>(<N>), %>|(<N>)
987 similar to %<(<N>), %<|(<N>) respectively, but padding
988 spaces on the left
989 %>>(<N>), %>>|(<N>)
991 similar to %>(<N>), %>|(<N>) respectively, except that if
992 the next placeholder takes more spaces than given and there
993 are spaces on its left, use those spaces
994 %><(<N>), %><|(<N>)
996 similar to %<(<N>), %<|(<N>) respectively, but padding both
997 sides (i.e. the text is centered)
998 · Placeholders that expand to information extracted from the
1000 commit:
1001 %H
1003 commit hash
1004 %h
1006 abbreviated commit hash
1007 %T
1009 tree hash
1010 %t
1012 abbreviated tree hash
1013 %P
1015 parent hashes
1016 %p
1018 abbreviated parent hashes
1019 %an
1021 author name
1022 %aN
1024 author name (respecting .mailmap, see git-shortlog(1) or
1025 git-blame(1))
1026 %ae
1028 author email
1029 %aE
1031 author email (respecting .mailmap, see git-shortlog(1) or
1032 git-blame(1))
1033 %ad
1035 author date (format respects --date= option)
1036 %aD
1038 author date, RFC2822 style
1039 %ar
1041 author date, relative
1042 %at
1044 author date, UNIX timestamp
1045 %ai
1047 author date, ISO 8601-like format
1048 %aI
1050 author date, strict ISO 8601 format
1051 %cn
1053 committer name
1054 %cN
1056 committer name (respecting .mailmap, see git-shortlog(1) or
1057 git-blame(1))
1058 %ce
1060 committer email
1061 %cE
1063 committer email (respecting .mailmap, see git-shortlog(1)
1064 or git-blame(1))
1065 %cd
1067 committer date (format respects --date= option)
1068 %cD
1070 committer date, RFC2822 style
1071 %cr
1073 committer date, relative
1074 %ct
1076 committer date, UNIX timestamp
1077 %ci
1079 committer date, ISO 8601-like format
1080 %cI
1082 committer date, strict ISO 8601 format
1083 %d
1085 ref names, like the --decorate option of git-log(1)
1086 %D
1088 ref names without the " (", ")" wrapping.
1089 %S
1091 ref name given on the command line by which the commit was
1092 reached (like git log --source), only works with git log
1093 %e
1095 encoding
1096 %s
1098 subject
1099 %f
1101 sanitized subject line, suitable for a filename
1102 %b
1104 body
1105 %B
1107 raw body (unwrapped subject and body)
1108 %N
1110 commit notes
1111 %GG
1113 raw verification message from GPG for a signed commit
1114 %G?
1116 show "G" for a good (valid) signature, "B" for a bad
1117 signature, "U" for a good signature with unknown validity,
1118 "X" for a good signature that has expired, "Y" for a good
1119 signature made by an expired key, "R" for a good signature
1120 made by a revoked key, "E" if the signature cannot be
1121 checked (e.g. missing key) and "N" for no signature
1122 %GS
1124 show the name of the signer for a signed commit
1125 %GK
1127 show the key used to sign a signed commit
1128 %GF
1130 show the fingerprint of the key used to sign a signed
1131 commit
1132 %GP
1134 show the fingerprint of the primary key whose subkey was
1135 used to sign a signed commit
1136 %gD
1138 reflog selector, e.g., refs/stash@{1} or refs/stash@{2
1139 minutes ago}; the format follows the rules described for
1140 the -g option. The portion before the @ is the refname as
1141 given on the command line (so git log -g refs/heads/master
1142 would yield refs/heads/master@{0}).
1143 %gd
1145 shortened reflog selector; same as %gD, but the refname
1146 portion is shortened for human readability (so
1147 refs/heads/master becomes just master).
1148 %gn
1150 reflog identity name
1151 %gN
1153 reflog identity name (respecting .mailmap, see git-
1154 shortlog(1) or git-blame(1))
1155 %ge
1157 reflog identity email
1158 %gE
1160 reflog identity email (respecting .mailmap, see git-
1161 shortlog(1) or git-blame(1))
1162 %gs
1164 reflog subject
1165 %(trailers[:options])
1167 display the trailers of the body as interpreted by git-
1168 interpret-trailers(1). The trailers string may be followed
1169 by a colon and zero or more comma-separated options:
1170 · key=<K>: only show trailers with specified key.
1172 Matching is done case-insensitively and trailing colon
1173 is optional. If option is given multiple times trailer
1174 lines matching any of the keys are shown. This option
1175 automatically enables the only option so that
1176 non-trailer lines in the trailer block are hidden. If
1177 that is not desired it can be disabled with only=false.
1178 E.g., %(trailers:key=Reviewed-by) shows trailer lines
1179 with key Reviewed-by.
1180 · only[=val]: select whether non-trailer lines from the
1182 trailer block should be included. The only keyword may
1183 optionally be followed by an equal sign and one of
1184 true, on, yes to omit or false, off, no to show the
1185 non-trailer lines. If option is given without value it
1186 is enabled. If given multiple times the last value is
1187 used.
1188 · separator=<SEP>: specify a separator inserted between
1190 trailer lines. When this option is not given each
1191 trailer line is terminated with a line feed character.
1192 The string SEP may contain the literal formatting codes
1193 described above. To use comma as separator one must use
1194 %x2C as it would otherwise be parsed as next option. If
1195 separator option is given multiple times only the last
1196 one is used. E.g., %(trailers:key=Ticket,separator=%x2C
1197 ) shows all trailer lines whose key is "Ticket"
1198 separated by a comma and a space.
1199 · unfold[=val]: make it behave as if interpret-trailer’s
1201 --unfold option was given. In same way as to for only
1202 it can be followed by an equal sign and explicit value.
1203 E.g., %(trailers:only,unfold=true) unfolds and shows
1204 all trailer lines.
1205 · valueonly[=val]: skip over the key part of the trailer
1207 line and only show the value part. Also this optionally
1208 allows explicit value.
1209 Note
1211 Some placeholders may depend on other options given to the revision
1212 traversal engine. For example, the %g* reflog options will insert
1213 an empty string unless we are traversing reflog entries (e.g., by
1214 git log -g). The %d and %D placeholders will use the "short"
1215 decoration format if --decorate was not already provided on the
1216 command line.
1217 If you add a + (plus sign) after % of a placeholder, a line-feed is
1219 inserted immediately before the expansion if and only if the
1220 placeholder expands to a non-empty string.
1221 If you add a - (minus sign) after % of a placeholder, all consecutive
1223 line-feeds immediately preceding the expansion are deleted if and only
1224 if the placeholder expands to an empty string.
1225 If you add a ` ` (space) after % of a placeholder, a space is inserted
1227 immediately before the expansion if and only if the placeholder expands
1228 to a non-empty string.
1229 · tformat:
1231 The tformat: format works exactly like format:, except that it
1233 provides "terminator" semantics instead of "separator" semantics.
1234 In other words, each commit has the message terminator character
1235 (usually a newline) appended, rather than a separator placed
1236 between entries. This means that the final entry of a single-line
1237 format will be properly terminated with a new line, just as the
1238 "oneline" format does. For example:
1239 $ git log -2 --pretty=format:%h 4da45bef \
1241 | perl -pe '$_ .= " -- NO NEWLINE\n" unless /\n/'
1242 4da45be
1243 7134973 -- NO NEWLINE
1244 $ git log -2 --pretty=tformat:%h 4da45bef \
1246 | perl -pe '$_ .= " -- NO NEWLINE\n" unless /\n/'
1247 4da45be
1248 7134973
1249 In addition, any unrecognized string that has a % in it is
1251 interpreted as if it has tformat: in front of it. For example,
1252 these two are equivalent:
1253 $ git log -2 --pretty=tformat:%h 4da45bef
1255 $ git log -2 --pretty=%h 4da45bef
1256
1259 The raw output format from "git-diff-index", "git-diff-tree",
1260 "git-diff-files" and "git diff --raw" are very similar.
1261 These commands all compare two sets of things; what is compared
1263 differs:
1264 git-diff-index <tree-ish>
1266 compares the <tree-ish> and the files on the filesystem.
1267 git-diff-index --cached <tree-ish>
1269 compares the <tree-ish> and the index.
1270 git-diff-tree [-r] <tree-ish-1> <tree-ish-2> [<pattern>...]
1272 compares the trees named by the two arguments.
1273 git-diff-files [<pattern>...]
1275 compares the index and the files on the filesystem.
1276 The "git-diff-tree" command begins its output by printing the hash of
1278 what is being compared. After that, all the commands print one output
1279 line per changed file.
1280 An output line is formatted this way:
1282 in-place edit :100644 100644 bcd1234 0123456 M file0
1284 copy-edit :100644 100644 abcd123 1234567 C68 file1 file2
1285 rename-edit :100644 100644 abcd123 1234567 R86 file1 file3
1286 create :000000 100644 0000000 1234567 A file4
1287 delete :100644 000000 1234567 0000000 D file5
1288 unmerged :000000 000000 0000000 0000000 U file6
1289 That is, from the left to the right:
1292 1. a colon.
1294 2. mode for "src"; 000000 if creation or unmerged.
1296 3. a space.
1298 4. mode for "dst"; 000000 if deletion or unmerged.
1300 5. a space.
1302 6. sha1 for "src"; 0{40} if creation or unmerged.
1304 7. a space.
1306 8. sha1 for "dst"; 0{40} if creation, unmerged or "look at work tree".
1308 9. a space.
1310 10. status, followed by optional "score" number.
1312 11. a tab or a NUL when -z option is used.
1314 12. path for "src"
1316 13. a tab or a NUL when -z option is used; only exists for C or R.
1318 14. path for "dst"; only exists for C or R.
1320 15. an LF or a NUL when -z option is used, to terminate the record.
1322 Possible status letters are:
1324 · A: addition of a file
1326 · C: copy of a file into a new one
1328 · D: deletion of a file
1330 · M: modification of the contents or mode of a file
1332 · R: renaming of a file
1334 · T: change in the type of the file
1336 · U: file is unmerged (you must complete the merge before it can be
1338 committed)
1339 · X: "unknown" change type (most probably a bug, please report it)
1341 Status letters C and R are always followed by a score (denoting the
1343 percentage of similarity between the source and target of the move or
1344 copy). Status letter M may be followed by a score (denoting the
1345 percentage of dissimilarity) for file rewrites.
1346 <sha1> is shown as all 0’s if a file is new on the filesystem and it is
1348 out of sync with the index.
1349 Example:
1351 :100644 100644 5be4a4a 0000000 M file.c
1353 Without the -z option, pathnames with "unusual" characters are quoted
1356 as explained for the configuration variable core.quotePath (see git-
1357 config(1)). Using -z the filename is output verbatim and the line is
1358 terminated by a NUL byte.
1359
1361 "git-diff-tree", "git-diff-files" and "git-diff --raw" can take -c or
1362 --cc option to generate diff output also for merge commits. The output
1363 differs from the format described above in the following way:
1364 1. there is a colon for each parent
1366 2. there are more "src" modes and "src" sha1
1368 3. status is concatenated status characters for each parent
1370 4. no optional "score" number
1372 5. tab-separated pathname(s) of the file
1374 For -c and --cc, only the destination or final path is shown even if
1376 the file was renamed on any side of history. With --combined-all-paths,
1377 the name of the path in each parent is shown followed by the name of
1378 the path in the merge commit.
1379 Examples for -c and --cc without --combined-all-paths:
1381 ::100644 100644 100644 fabadb8 cc95eb0 4866510 MM desc.c
1383 ::100755 100755 100755 52b7a2d 6d1ac04 d2ac7d7 RM bar.sh
1384 ::100644 100644 100644 e07d6c5 9042e82 ee91881 RR phooey.c
1385 Examples when --combined-all-paths added to either -c or --cc:
1388 ::100644 100644 100644 fabadb8 cc95eb0 4866510 MM desc.c desc.c desc.c
1390 ::100755 100755 100755 52b7a2d 6d1ac04 d2ac7d7 RM foo.sh bar.sh bar.sh
1391 ::100644 100644 100644 e07d6c5 9042e82 ee91881 RR fooey.c fuey.c phooey.c
1392 Note that combined diff lists only files which were modified from all
1395 parents.
1396
1398 Running git-diff(1), git-log(1), git-show(1), git-diff-index(1), git-
1399 diff-tree(1), or git-diff-files(1) with the -p option produces patch
1400 text. You can customize the creation of patch text via the
1401 GIT_EXTERNAL_DIFF and the GIT_DIFF_OPTS environment variables.
1402 What the -p option produces is slightly different from the traditional
1404 diff format:
1405 1. It is preceded with a "git diff" header that looks like this:
1407 diff --git a/file1 b/file2
1409 The a/ and b/ filenames are the same unless rename/copy is
1411 involved. Especially, even for a creation or a deletion, /dev/null
1412 is not used in place of the a/ or b/ filenames.
1413 When rename/copy is involved, file1 and file2 show the name of the
1415 source file of the rename/copy and the name of the file that
1416 rename/copy produces, respectively.
1417 2. It is followed by one or more extended header lines:
1419 old mode <mode>
1421 new mode <mode>
1422 deleted file mode <mode>
1423 new file mode <mode>
1424 copy from <path>
1425 copy to <path>
1426 rename from <path>
1427 rename to <path>
1428 similarity index <number>
1429 dissimilarity index <number>
1430 index <hash>..<hash> <mode>
1431 File modes are printed as 6-digit octal numbers including the file
1433 type and file permission bits.
1434 Path names in extended headers do not include the a/ and b/
1436 prefixes.
1437 The similarity index is the percentage of unchanged lines, and the
1439 dissimilarity index is the percentage of changed lines. It is a
1440 rounded down integer, followed by a percent sign. The similarity
1441 index value of 100% is thus reserved for two equal files, while
1442 100% dissimilarity means that no line from the old file made it
1443 into the new one.
1444 The index line includes the blob object names before and after the
1446 change. The <mode> is included if the file mode does not change;
1447 otherwise, separate lines indicate the old and the new mode.
1448 3. Pathnames with "unusual" characters are quoted as explained for the
1450 configuration variable core.quotePath (see git-config(1)).
1451 4. All the file1 files in the output refer to files before the commit,
1453 and all the file2 files refer to files after the commit. It is
1454 incorrect to apply each change to each file sequentially. For
1455 example, this patch will swap a and b:
1456 diff --git a/a b/b
1458 rename from a
1459 rename to b
1460 diff --git a/b b/a
1461 rename from b
1462 rename to a
1463
1465 Any diff-generating command can take the -c or --cc option to produce a
1466 combined diff when showing a merge. This is the default format when
1467 showing merges with git-diff(1) or git-show(1). Note also that you can
1468 give the -m option to any of these commands to force generation of
1469 diffs with individual parents of a merge.
1470 A "combined diff" format looks like this:
1472 diff --combined describe.c
1474 index fabadb8,cc95eb0..4866510
1475 --- a/describe.c
1476 +++ b/describe.c
1477 @@@ -98,20 -98,12 +98,20 @@@
1478 return (a_date > b_date) ? -1 : (a_date == b_date) ? 0 : 1;
1479 }
1480 - static void describe(char *arg)
1482 -static void describe(struct commit *cmit, int last_one)
1483 ++static void describe(char *arg, int last_one)
1484 {
1485 + unsigned char sha1[20];
1486 + struct commit *cmit;
1487 struct commit_list *list;
1488 static int initialized = 0;
1489 struct commit_name *n;
1490 + if (get_sha1(arg, sha1) < 0)
1492 + usage(describe_usage);
1493 + cmit = lookup_commit_reference(sha1);
1494 + if (!cmit)
1495 + usage(describe_usage);
1496 +
1497 if (!initialized) {
1498 initialized = 1;
1499 for_each_ref(get_name);
1500 1. It is preceded with a "git diff" header, that looks like this (when
1504 the -c option is used):
1505 diff --combined file
1507 or like this (when the --cc option is used):
1509 diff --cc file
1511 2. It is followed by one or more extended header lines (this example
1513 shows a merge with two parents):
1514 index <hash>,<hash>..<hash>
1516 mode <mode>,<mode>..<mode>
1517 new file mode <mode>
1518 deleted file mode <mode>,<mode>
1519 The mode <mode>,<mode>..<mode> line appears only if at least one of
1521 the <mode> is different from the rest. Extended headers with
1522 information about detected contents movement (renames and copying
1523 detection) are designed to work with diff of two <tree-ish> and are
1524 not used by combined diff format.
1525 3. It is followed by two-line from-file/to-file header
1527 --- a/file
1529 +++ b/file
1530 Similar to two-line header for traditional unified diff format,
1532 /dev/null is used to signal created or deleted files.
1533 However, if the --combined-all-paths option is provided, instead of
1535 a two-line from-file/to-file you get a N+1 line from-file/to-file
1536 header, where N is the number of parents in the merge commit
1537 --- a/file
1539 --- a/file
1540 --- a/file
1541 +++ b/file
1542 This extended format can be useful if rename or copy detection is
1544 active, to allow you to see the original name of the file in
1545 different parents.
1546 4. Chunk header format is modified to prevent people from accidentally
1548 feeding it to patch -p1. Combined diff format was created for
1549 review of merge commit changes, and was not meant to be applied.
1550 The change is similar to the change in the extended index header:
1551 @@@ <from-file-range> <from-file-range> <to-file-range> @@@
1553 There are (number of parents + 1) @ characters in the chunk header
1555 for combined diff format.
1556 Unlike the traditional unified diff format, which shows two files A and
1558 B with a single column that has - (minus — appears in A but removed in
1559 B), + (plus — missing in A but added to B), or " " (space — unchanged)
1560 prefix, this format compares two or more files file1, file2,... with
1561 one file X, and shows how X differs from each of fileN. One column for
1562 each of fileN is prepended to the output line to note how X’s line is
1563 different from it.
1564 A - character in the column N means that the line appears in fileN but
1566 it does not appear in the result. A + character in the column N means
1567 that the line appears in the result, and fileN does not have that line
1568 (in other words, the line was added, from the point of view of that
1569 parent).
1570 In the above example output, the function signature was changed from
1572 both files (hence two - removals from both file1 and file2, plus ++ to
1573 mean one line that was added does not appear in either file1 or file2).
1574 Also eight other lines are the same from file1 but do not appear in
1575 file2 (hence prefixed with +).
1576 When shown by git diff-tree -c, it compares the parents of a merge
1578 commit with the merge result (i.e. file1..fileN are the parents). When
1579 shown by git diff-files -c, it compares the two unresolved merge
1580 parents with the working tree file (i.e. file1 is stage 2 aka "our
1581 version", file2 is stage 3 aka "their version").
1582
1584 The --summary option describes newly added, deleted, renamed and copied
1585 files. The --stat option adds diffstat(1) graph to the output. These
1586 options can be combined with other options, such as -p, and are meant
1587 for human consumption.
1588 When showing a change that involves a rename or a copy, --stat output
1590 formats the pathnames compactly by combining common prefix and suffix
1591 of the pathnames. For example, a change that moves arch/i386/Makefile
1592 to arch/x86/Makefile while modifying 4 lines will be shown like this:
1593 arch/{i386 => x86}/Makefile | 4 +--
1595 The --numstat option gives the diffstat(1) information but is designed
1598 for easier machine consumption. An entry in --numstat output looks like
1599 this:
1600 1 2 README
1602 3 1 arch/{i386 => x86}/Makefile
1603 That is, from left to right:
1606 1. the number of added lines;
1608 2. a tab;
1610 3. the number of deleted lines;
1612 4. a tab;
1614 5. pathname (possibly with rename/copy information);
1616 6. a newline.
1618 When -z output option is in effect, the output is formatted this way:
1620 1 2 README NUL
1622 3 1 NUL arch/i386/Makefile NUL arch/x86/Makefile NUL
1623 That is:
1626 1. the number of added lines;
1628 2. a tab;
1630 3. the number of deleted lines;
1632 4. a tab;
1634 5. a NUL (only exists if renamed/copied);
1636 6. pathname in preimage;
1638 7. a NUL (only exists if renamed/copied);
1640 8. pathname in postimage (only exists if renamed/copied);
1642 9. a NUL.
1644 The extra NUL before the preimage path in renamed case is to allow
1646 scripts that read the output to tell if the current record being read
1647 is a single-path record or a rename/copy record without reading ahead.
1648 After reading added and deleted lines, reading up to NUL would yield
1649 the pathname, but if that is NUL, the record will show two paths.
1650
1652 Part of the git(1) suite
1653Git 2.24.1 12/10/2019 GIT-DIFF-TREE(1)