1GIT-SHOW(1) Git Manual GIT-SHOW(1)
2
6 git-show - Show various types of objects
7
9 git show [<options>] [<object>...]
10
13 Shows one or more objects (blobs, trees, tags and commits).
14 For commits it shows the log message and textual diff. It also presents
16 the merge commit in a special format as produced by git diff-tree --cc.
17 For tags, it shows the tag message and the referenced objects.
19 For trees, it shows the names (equivalent to git ls-tree with
21 --name-only).
22 For plain blobs, it shows the plain contents.
24 The command takes options applicable to the git diff-tree command to
26 control how the changes the commit introduces are shown.
27 This manual page describes only the most frequently used options.
29
31 <object>...
32 The names of objects to show (defaults to HEAD). For a more
33 complete list of ways to spell object names, see "SPECIFYING
34 REVISIONS" section in gitrevisions(7).
35 --pretty[=<format>], --format=<format>
37 Pretty-print the contents of the commit logs in a given format,
38 where <format> can be one of oneline, short, medium, full, fuller,
39 email, raw, format:<string> and tformat:<string>. When <format> is
40 none of the above, and has %placeholder in it, it acts as if
41 --pretty=tformat:<format> were given.
42 See the "PRETTY FORMATS" section for some additional details for
44 each format. When =<format> part is omitted, it defaults to medium.
45 Note: you can specify the default pretty format in the repository
47 configuration (see git-config(1)).
48 --abbrev-commit
50 Instead of showing the full 40-byte hexadecimal commit object name,
51 show only a partial prefix. Non default number of digits can be
52 specified with "--abbrev=<n>" (which also modifies diff output, if
53 it is displayed).
54 This should make "--pretty=oneline" a whole lot more readable for
56 people using 80-column terminals.
57 --no-abbrev-commit
59 Show the full 40-byte hexadecimal commit object name. This negates
60 --abbrev-commit and those options which imply it such as
61 "--oneline". It also overrides the log.abbrevCommit variable.
62 --oneline
64 This is a shorthand for "--pretty=oneline --abbrev-commit" used
65 together.
66 --encoding=<encoding>
68 The commit objects record the encoding used for the log message in
69 their encoding header; this option can be used to tell the command
70 to re-code the commit log message in the encoding preferred by the
71 user. For non plumbing commands this defaults to UTF-8. Note that
72 if an object claims to be encoded in X and we are outputting in X,
73 we will output the object verbatim; this means that invalid
74 sequences in the original commit may be copied to the output.
75 --expand-tabs=<n>, --expand-tabs, --no-expand-tabs
77 Perform a tab expansion (replace each tab with enough spaces to
78 fill to the next display column that is multiple of <n>) in the log
79 message before showing it in the output. --expand-tabs is a
80 short-hand for --expand-tabs=8, and --no-expand-tabs is a
81 short-hand for --expand-tabs=0, which disables tab expansion.
82 By default, tabs are expanded in pretty formats that indent the log
84 message by 4 spaces (i.e. medium, which is the default, full, and
85 fuller).
86 --notes[=<treeish>]
88 Show the notes (see git-notes(1)) that annotate the commit, when
89 showing the commit log message. This is the default for git log,
90 git show and git whatchanged commands when there is no --pretty,
91 --format, or --oneline option given on the command line.
92 By default, the notes shown are from the notes refs listed in the
94 core.notesRef and notes.displayRef variables (or corresponding
95 environment overrides). See git-config(1) for more details.
96 With an optional <treeish> argument, use the treeish to find the
98 notes to display. The treeish can specify the full refname when it
99 begins with refs/notes/; when it begins with notes/, refs/ and
100 otherwise refs/notes/ is prefixed to form a full name of the ref.
101 Multiple --notes options can be combined to control which notes are
103 being displayed. Examples: "--notes=foo" will show only notes from
104 "refs/notes/foo"; "--notes=foo --notes" will show both notes from
105 "refs/notes/foo" and from the default notes ref(s).
106 --no-notes
108 Do not show notes. This negates the above --notes option, by
109 resetting the list of notes refs from which notes are shown.
110 Options are parsed in the order given on the command line, so e.g.
111 "--notes --notes=foo --no-notes --notes=bar" will only show notes
112 from "refs/notes/bar".
113 --show-notes[=<treeish>], --[no-]standard-notes
115 These options are deprecated. Use the above --notes/--no-notes
116 options instead.
117 --show-signature
119 Check the validity of a signed commit object by passing the
120 signature to gpg --verify and show the output.
121
123 If the commit is a merge, and if the pretty-format is not oneline,
124 email or raw, an additional line is inserted before the Author: line.
125 This line begins with "Merge: " and the sha1s of ancestral commits are
126 printed, separated by spaces. Note that the listed commits may not
127 necessarily be the list of the direct parent commits if you have
128 limited your view of history: for example, if you are only interested
129 in changes related to a certain directory or file.
130 There are several built-in formats, and you can define additional
132 formats by setting a pretty.<name> config option to either another
133 format name, or a format: string, as described below (see git-
134 config(1)). Here are the details of the built-in formats:
135 · oneline
137 <sha1> <title line>
139 This is designed to be as compact as possible.
141 · short
143 commit <sha1>
145 Author: <author>
146 <title line>
148 · medium
150 commit <sha1>
152 Author: <author>
153 Date: <author date>
154 <title line>
156 <full commit message>
158 · full
160 commit <sha1>
162 Author: <author>
163 Commit: <committer>
164 <title line>
166 <full commit message>
168 · fuller
170 commit <sha1>
172 Author: <author>
173 AuthorDate: <author date>
174 Commit: <committer>
175 CommitDate: <committer date>
176 <title line>
178 <full commit message>
180 · email
182 From <sha1> <date>
184 From: <author>
185 Date: <author date>
186 Subject: [PATCH] <title line>
187 <full commit message>
189 · raw
191 The raw format shows the entire commit exactly as stored in the
193 commit object. Notably, the SHA-1s are displayed in full,
194 regardless of whether --abbrev or --no-abbrev are used, and parents
195 information show the true parent commits, without taking grafts or
196 history simplification into account. Note that this format affects
197 the way commits are displayed, but not the way the diff is shown
198 e.g. with git log --raw. To get full object names in a raw diff
199 format, use --no-abbrev.
200 · format:<string>
202 The format:<string> format allows you to specify which information
204 you want to show. It works a little bit like printf format, with
205 the notable exception that you get a newline with %n instead of \n.
206 E.g, format:"The author of %h was %an, %ar%nThe title was >>%s<<%n"
208 would show something like this:
209 The author of fe6e0ee was Junio C Hamano, 23 hours ago
211 The title was >>t4119: test autocomputing -p<n> for traditional diff input.<<
212 The placeholders are:
214 · Placeholders that expand to a single literal character:
216 %n
218 newline
219 %%
221 a raw %
222 %x00
224 print a byte from a hex code
225 · Placeholders that affect formatting of later placeholders:
227 %Cred
229 switch color to red
230 %Cgreen
232 switch color to green
233 %Cblue
235 switch color to blue
236 %Creset
238 reset color
239 %C(...)
241 color specification, as described under Values in the
242 "CONFIGURATION FILE" section of git-config(1). By default,
243 colors are shown only when enabled for log output (by
244 color.diff, color.ui, or --color, and respecting the auto
245 settings of the former if we are going to a terminal).
246 %C(auto,...) is accepted as a historical synonym for the
247 default (e.g., %C(auto,red)). Specifying %C(always,...)
248 will show the colors even when color is not otherwise
249 enabled (though consider just using --color=always to
250 enable color for the whole output, including this format
251 and anything else git might color). auto alone (i.e.
252 %C(auto)) will turn on auto coloring on the next
253 placeholders until the color is switched again.
254 %m
256 left (<), right (>) or boundary (-) mark
257 %w([<w>[,<i1>[,<i2>]]])
259 switch line wrapping, like the -w option of git-
260 shortlog(1).
261 %<(<N>[,trunc|ltrunc|mtrunc])
263 make the next placeholder take at least N columns, padding
264 spaces on the right if necessary. Optionally truncate at
265 the beginning (ltrunc), the middle (mtrunc) or the end
266 (trunc) if the output is longer than N columns. Note that
267 truncating only works correctly with N >= 2.
268 %<|(<N>)
270 make the next placeholder take at least until Nth columns,
271 padding spaces on the right if necessary
272 %>(<N>), %>|(<N>)
274 similar to %<(<N>), %<|(<N>) respectively, but padding
275 spaces on the left
276 %>>(<N>), %>>|(<N>)
278 similar to %>(<N>), %>|(<N>) respectively, except that if
279 the next placeholder takes more spaces than given and there
280 are spaces on its left, use those spaces
281 %><(<N>), %><|(<N>)
283 similar to %<(<N>), %<|(<N>) respectively, but padding both
284 sides (i.e. the text is centered)
285 · Placeholders that expand to information extracted from the
287 commit:
288 %H
290 commit hash
291 %h
293 abbreviated commit hash
294 %T
296 tree hash
297 %t
299 abbreviated tree hash
300 %P
302 parent hashes
303 %p
305 abbreviated parent hashes
306 %an
308 author name
309 %aN
311 author name (respecting .mailmap, see git-shortlog(1) or
312 git-blame(1))
313 %ae
315 author email
316 %aE
318 author email (respecting .mailmap, see git-shortlog(1) or
319 git-blame(1))
320 %ad
322 author date (format respects --date= option)
323 %aD
325 author date, RFC2822 style
326 %ar
328 author date, relative
329 %at
331 author date, UNIX timestamp
332 %ai
334 author date, ISO 8601-like format
335 %aI
337 author date, strict ISO 8601 format
338 %cn
340 committer name
341 %cN
343 committer name (respecting .mailmap, see git-shortlog(1) or
344 git-blame(1))
345 %ce
347 committer email
348 %cE
350 committer email (respecting .mailmap, see git-shortlog(1)
351 or git-blame(1))
352 %cd
354 committer date (format respects --date= option)
355 %cD
357 committer date, RFC2822 style
358 %cr
360 committer date, relative
361 %ct
363 committer date, UNIX timestamp
364 %ci
366 committer date, ISO 8601-like format
367 %cI
369 committer date, strict ISO 8601 format
370 %d
372 ref names, like the --decorate option of git-log(1)
373 %D
375 ref names without the " (", ")" wrapping.
376 %S
378 ref name given on the command line by which the commit was
379 reached (like git log --source), only works with git log
380 %e
382 encoding
383 %s
385 subject
386 %f
388 sanitized subject line, suitable for a filename
389 %b
391 body
392 %B
394 raw body (unwrapped subject and body)
395 %N
397 commit notes
398 %GG
400 raw verification message from GPG for a signed commit
401 %G?
403 show "G" for a good (valid) signature, "B" for a bad
404 signature, "U" for a good signature with unknown validity,
405 "X" for a good signature that has expired, "Y" for a good
406 signature made by an expired key, "R" for a good signature
407 made by a revoked key, "E" if the signature cannot be
408 checked (e.g. missing key) and "N" for no signature
409 %GS
411 show the name of the signer for a signed commit
412 %GK
414 show the key used to sign a signed commit
415 %GF
417 show the fingerprint of the key used to sign a signed
418 commit
419 %GP
421 show the fingerprint of the primary key whose subkey was
422 used to sign a signed commit
423 %gD
425 reflog selector, e.g., refs/stash@{1} or refs/stash@{2
426 minutes ago}; the format follows the rules described for
427 the -g option. The portion before the @ is the refname as
428 given on the command line (so git log -g refs/heads/master
429 would yield refs/heads/master@{0}).
430 %gd
432 shortened reflog selector; same as %gD, but the refname
433 portion is shortened for human readability (so
434 refs/heads/master becomes just master).
435 %gn
437 reflog identity name
438 %gN
440 reflog identity name (respecting .mailmap, see git-
441 shortlog(1) or git-blame(1))
442 %ge
444 reflog identity email
445 %gE
447 reflog identity email (respecting .mailmap, see git-
448 shortlog(1) or git-blame(1))
449 %gs
451 reflog subject
452 %(trailers[:options])
454 display the trailers of the body as interpreted by git-
455 interpret-trailers(1). The trailers string may be followed
456 by a colon and zero or more comma-separated options:
457 · key=<K>: only show trailers with specified key.
459 Matching is done case-insensitively and trailing colon
460 is optional. If option is given multiple times trailer
461 lines matching any of the keys are shown. This option
462 automatically enables the only option so that
463 non-trailer lines in the trailer block are hidden. If
464 that is not desired it can be disabled with only=false.
465 E.g., %(trailers:key=Reviewed-by) shows trailer lines
466 with key Reviewed-by.
467 · only[=val]: select whether non-trailer lines from the
469 trailer block should be included. The only keyword may
470 optionally be followed by an equal sign and one of
471 true, on, yes to omit or false, off, no to show the
472 non-trailer lines. If option is given without value it
473 is enabled. If given multiple times the last value is
474 used.
475 · separator=<SEP>: specify a separator inserted between
477 trailer lines. When this option is not given each
478 trailer line is terminated with a line feed character.
479 The string SEP may contain the literal formatting codes
480 described above. To use comma as separator one must use
481 %x2C as it would otherwise be parsed as next option. If
482 separator option is given multiple times only the last
483 one is used. E.g., %(trailers:key=Ticket,separator=%x2C
484 ) shows all trailer lines whose key is "Ticket"
485 separated by a comma and a space.
486 · unfold[=val]: make it behave as if interpret-trailer’s
488 --unfold option was given. In same way as to for only
489 it can be followed by an equal sign and explicit value.
490 E.g., %(trailers:only,unfold=true) unfolds and shows
491 all trailer lines.
492 · valueonly[=val]: skip over the key part of the trailer
494 line and only show the value part. Also this optionally
495 allows explicit value.
496 Note
498 Some placeholders may depend on other options given to the revision
499 traversal engine. For example, the %g* reflog options will insert
500 an empty string unless we are traversing reflog entries (e.g., by
501 git log -g). The %d and %D placeholders will use the "short"
502 decoration format if --decorate was not already provided on the
503 command line.
504 If you add a + (plus sign) after % of a placeholder, a line-feed is
506 inserted immediately before the expansion if and only if the
507 placeholder expands to a non-empty string.
508 If you add a - (minus sign) after % of a placeholder, all consecutive
510 line-feeds immediately preceding the expansion are deleted if and only
511 if the placeholder expands to an empty string.
512 If you add a ` ` (space) after % of a placeholder, a space is inserted
514 immediately before the expansion if and only if the placeholder expands
515 to a non-empty string.
516 · tformat:
518 The tformat: format works exactly like format:, except that it
520 provides "terminator" semantics instead of "separator" semantics.
521 In other words, each commit has the message terminator character
522 (usually a newline) appended, rather than a separator placed
523 between entries. This means that the final entry of a single-line
524 format will be properly terminated with a new line, just as the
525 "oneline" format does. For example:
526 $ git log -2 --pretty=format:%h 4da45bef \
528 | perl -pe '$_ .= " -- NO NEWLINE\n" unless /\n/'
529 4da45be
530 7134973 -- NO NEWLINE
531 $ git log -2 --pretty=tformat:%h 4da45bef \
533 | perl -pe '$_ .= " -- NO NEWLINE\n" unless /\n/'
534 4da45be
535 7134973
536 In addition, any unrecognized string that has a % in it is
538 interpreted as if it has tformat: in front of it. For example,
539 these two are equivalent:
540 $ git log -2 --pretty=tformat:%h 4da45bef
542 $ git log -2 --pretty=%h 4da45bef
543
546 -p, -u, --patch
547 Generate patch (see section on generating patches).
548 -s, --no-patch
550 Suppress diff output. Useful for commands like git show that show
551 the patch by default, or to cancel the effect of --patch.
552 -U<n>, --unified=<n>
554 Generate diffs with <n> lines of context instead of the usual
555 three. Implies --patch. Implies -p.
556 --output=<file>
558 Output to a specific file instead of stdout.
559 --output-indicator-new=<char>, --output-indicator-old=<char>,
561 --output-indicator-context=<char>
562 Specify the character used to indicate new, old or context lines in
563 the generated patch. Normally they are +, - and ' ' respectively.
564 --raw
566 For each commit, show a summary of changes using the raw diff
567 format. See the "RAW OUTPUT FORMAT" section of git-diff(1). This is
568 different from showing the log itself in raw format, which you can
569 achieve with --format=raw.
570 --patch-with-raw
572 Synonym for -p --raw.
573 --indent-heuristic
575 Enable the heuristic that shifts diff hunk boundaries to make
576 patches easier to read. This is the default.
577 --no-indent-heuristic
579 Disable the indent heuristic.
580 --minimal
582 Spend extra time to make sure the smallest possible diff is
583 produced.
584 --patience
586 Generate a diff using the "patience diff" algorithm.
587 --histogram
589 Generate a diff using the "histogram diff" algorithm.
590 --anchored=<text>
592 Generate a diff using the "anchored diff" algorithm.
593 This option may be specified more than once.
595 If a line exists in both the source and destination, exists only
597 once, and starts with this text, this algorithm attempts to prevent
598 it from appearing as a deletion or addition in the output. It uses
599 the "patience diff" algorithm internally.
600 --diff-algorithm={patience|minimal|histogram|myers}
602 Choose a diff algorithm. The variants are as follows:
603 default, myers
605 The basic greedy diff algorithm. Currently, this is the
606 default.
607 minimal
609 Spend extra time to make sure the smallest possible diff is
610 produced.
611 patience
613 Use "patience diff" algorithm when generating patches.
614 histogram
616 This algorithm extends the patience algorithm to "support
617 low-occurrence common elements".
618 For instance, if you configured the diff.algorithm variable to a
620 non-default value and want to use the default one, then you have to
621 use --diff-algorithm=default option.
622 --stat[=<width>[,<name-width>[,<count>]]]
624 Generate a diffstat. By default, as much space as necessary will be
625 used for the filename part, and the rest for the graph part.
626 Maximum width defaults to terminal width, or 80 columns if not
627 connected to a terminal, and can be overridden by <width>. The
628 width of the filename part can be limited by giving another width
629 <name-width> after a comma. The width of the graph part can be
630 limited by using --stat-graph-width=<width> (affects all commands
631 generating a stat graph) or by setting diff.statGraphWidth=<width>
632 (does not affect git format-patch). By giving a third parameter
633 <count>, you can limit the output to the first <count> lines,
634 followed by ... if there are more.
635 These parameters can also be set individually with
637 --stat-width=<width>, --stat-name-width=<name-width> and
638 --stat-count=<count>.
639 --compact-summary
641 Output a condensed summary of extended header information such as
642 file creations or deletions ("new" or "gone", optionally "+l" if
643 it’s a symlink) and mode changes ("+x" or "-x" for adding or
644 removing executable bit respectively) in diffstat. The information
645 is put between the filename part and the graph part. Implies
646 --stat.
647 --numstat
649 Similar to --stat, but shows number of added and deleted lines in
650 decimal notation and pathname without abbreviation, to make it more
651 machine friendly. For binary files, outputs two - instead of saying
652 0 0.
653 --shortstat
655 Output only the last line of the --stat format containing total
656 number of modified files, as well as number of added and deleted
657 lines.
658 -X[<param1,param2,...>], --dirstat[=<param1,param2,...>]
660 Output the distribution of relative amount of changes for each
661 sub-directory. The behavior of --dirstat can be customized by
662 passing it a comma separated list of parameters. The defaults are
663 controlled by the diff.dirstat configuration variable (see git-
664 config(1)). The following parameters are available:
665 changes
667 Compute the dirstat numbers by counting the lines that have
668 been removed from the source, or added to the destination. This
669 ignores the amount of pure code movements within a file. In
670 other words, rearranging lines in a file is not counted as much
671 as other changes. This is the default behavior when no
672 parameter is given.
673 lines
675 Compute the dirstat numbers by doing the regular line-based
676 diff analysis, and summing the removed/added line counts. (For
677 binary files, count 64-byte chunks instead, since binary files
678 have no natural concept of lines). This is a more expensive
679 --dirstat behavior than the changes behavior, but it does count
680 rearranged lines within a file as much as other changes. The
681 resulting output is consistent with what you get from the other
682 --*stat options.
683 files
685 Compute the dirstat numbers by counting the number of files
686 changed. Each changed file counts equally in the dirstat
687 analysis. This is the computationally cheapest --dirstat
688 behavior, since it does not have to look at the file contents
689 at all.
690 cumulative
692 Count changes in a child directory for the parent directory as
693 well. Note that when using cumulative, the sum of the
694 percentages reported may exceed 100%. The default
695 (non-cumulative) behavior can be specified with the
696 noncumulative parameter.
697 <limit>
699 An integer parameter specifies a cut-off percent (3% by
700 default). Directories contributing less than this percentage of
701 the changes are not shown in the output.
702 Example: The following will count changed files, while ignoring
704 directories with less than 10% of the total amount of changed
705 files, and accumulating child directory counts in the parent
706 directories: --dirstat=files,10,cumulative.
707 --cumulative
709 Synonym for --dirstat=cumulative
710 --dirstat-by-file[=<param1,param2>...]
712 Synonym for --dirstat=files,param1,param2...
713 --summary
715 Output a condensed summary of extended header information such as
716 creations, renames and mode changes.
717 --patch-with-stat
719 Synonym for -p --stat.
720 -z
722 Separate the commits with NULs instead of with new newlines.
723 Also, when --raw or --numstat has been given, do not munge
725 pathnames and use NULs as output field terminators.
726 Without this option, pathnames with "unusual" characters are quoted
728 as explained for the configuration variable core.quotePath (see
729 git-config(1)).
730 --name-only
732 Show only names of changed files.
733 --name-status
735 Show only names and status of changed files. See the description of
736 the --diff-filter option on what the status letters mean.
737 --submodule[=<format>]
739 Specify how differences in submodules are shown. When specifying
740 --submodule=short the short format is used. This format just shows
741 the names of the commits at the beginning and end of the range.
742 When --submodule or --submodule=log is specified, the log format is
743 used. This format lists the commits in the range like git-
744 submodule(1) summary does. When --submodule=diff is specified, the
745 diff format is used. This format shows an inline diff of the
746 changes in the submodule contents between the commit range.
747 Defaults to diff.submodule or the short format if the config option
748 is unset.
749 --color[=<when>]
751 Show colored diff. --color (i.e. without =<when>) is the same as
752 --color=always. <when> can be one of always, never, or auto.
753 --no-color
755 Turn off colored diff. It is the same as --color=never.
756 --color-moved[=<mode>]
758 Moved lines of code are colored differently. The <mode> defaults to
759 no if the option is not given and to zebra if the option with no
760 mode is given. The mode must be one of:
761 no
763 Moved lines are not highlighted.
764 default
766 Is a synonym for zebra. This may change to a more sensible mode
767 in the future.
768 plain
770 Any line that is added in one location and was removed in
771 another location will be colored with color.diff.newMoved.
772 Similarly color.diff.oldMoved will be used for removed lines
773 that are added somewhere else in the diff. This mode picks up
774 any moved line, but it is not very useful in a review to
775 determine if a block of code was moved without permutation.
776 blocks
778 Blocks of moved text of at least 20 alphanumeric characters are
779 detected greedily. The detected blocks are painted using either
780 the color.diff.{old,new}Moved color. Adjacent blocks cannot be
781 told apart.
782 zebra
784 Blocks of moved text are detected as in blocks mode. The blocks
785 are painted using either the color.diff.{old,new}Moved color or
786 color.diff.{old,new}MovedAlternative. The change between the
787 two colors indicates that a new block was detected.
788 dimmed-zebra
790 Similar to zebra, but additional dimming of uninteresting parts
791 of moved code is performed. The bordering lines of two adjacent
792 blocks are considered interesting, the rest is uninteresting.
793 dimmed_zebra is a deprecated synonym.
794 --no-color-moved
796 Turn off move detection. This can be used to override configuration
797 settings. It is the same as --color-moved=no.
798 --color-moved-ws=<modes>
800 This configures how whitespace is ignored when performing the move
801 detection for --color-moved. These modes can be given as a comma
802 separated list:
803 no
805 Do not ignore whitespace when performing move detection.
806 ignore-space-at-eol
808 Ignore changes in whitespace at EOL.
809 ignore-space-change
811 Ignore changes in amount of whitespace. This ignores whitespace
812 at line end, and considers all other sequences of one or more
813 whitespace characters to be equivalent.
814 ignore-all-space
816 Ignore whitespace when comparing lines. This ignores
817 differences even if one line has whitespace where the other
818 line has none.
819 allow-indentation-change
821 Initially ignore any whitespace in the move detection, then
822 group the moved code blocks only into a block if the change in
823 whitespace is the same per line. This is incompatible with the
824 other modes.
825 --no-color-moved-ws
827 Do not ignore whitespace when performing move detection. This can
828 be used to override configuration settings. It is the same as
829 --color-moved-ws=no.
830 --word-diff[=<mode>]
832 Show a word diff, using the <mode> to delimit changed words. By
833 default, words are delimited by whitespace; see --word-diff-regex
834 below. The <mode> defaults to plain, and must be one of:
835 color
837 Highlight changed words using only colors. Implies --color.
838 plain
840 Show words as [-removed-] and {+added+}. Makes no attempts to
841 escape the delimiters if they appear in the input, so the
842 output may be ambiguous.
843 porcelain
845 Use a special line-based format intended for script
846 consumption. Added/removed/unchanged runs are printed in the
847 usual unified diff format, starting with a +/-/` ` character at
848 the beginning of the line and extending to the end of the line.
849 Newlines in the input are represented by a tilde ~ on a line of
850 its own.
851 none
853 Disable word diff again.
854 Note that despite the name of the first mode, color is used to
856 highlight the changed parts in all modes if enabled.
857 --word-diff-regex=<regex>
859 Use <regex> to decide what a word is, instead of considering runs
860 of non-whitespace to be a word. Also implies --word-diff unless it
861 was already enabled.
862 Every non-overlapping match of the <regex> is considered a word.
864 Anything between these matches is considered whitespace and
865 ignored(!) for the purposes of finding differences. You may want to
866 append |[^[:space:]] to your regular expression to make sure that
867 it matches all non-whitespace characters. A match that contains a
868 newline is silently truncated(!) at the newline.
869 For example, --word-diff-regex=. will treat each character as a
871 word and, correspondingly, show differences character by character.
872 The regex can also be set via a diff driver or configuration
874 option, see gitattributes(5) or git-config(1). Giving it explicitly
875 overrides any diff driver or configuration setting. Diff drivers
876 override configuration settings.
877 --color-words[=<regex>]
879 Equivalent to --word-diff=color plus (if a regex was specified)
880 --word-diff-regex=<regex>.
881 --no-renames
883 Turn off rename detection, even when the configuration file gives
884 the default to do so.
885 --[no-]rename-empty
887 Whether to use empty blobs as rename source.
888 --check
890 Warn if changes introduce conflict markers or whitespace errors.
891 What are considered whitespace errors is controlled by
892 core.whitespace configuration. By default, trailing whitespaces
893 (including lines that consist solely of whitespaces) and a space
894 character that is immediately followed by a tab character inside
895 the initial indent of the line are considered whitespace errors.
896 Exits with non-zero status if problems are found. Not compatible
897 with --exit-code.
898 --ws-error-highlight=<kind>
900 Highlight whitespace errors in the context, old or new lines of the
901 diff. Multiple values are separated by comma, none resets previous
902 values, default reset the list to new and all is a shorthand for
903 old,new,context. When this option is not given, and the
904 configuration variable diff.wsErrorHighlight is not set, only
905 whitespace errors in new lines are highlighted. The whitespace
906 errors are colored with color.diff.whitespace.
907 --full-index
909 Instead of the first handful of characters, show the full pre- and
910 post-image blob object names on the "index" line when generating
911 patch format output.
912 --binary
914 In addition to --full-index, output a binary diff that can be
915 applied with git-apply. Implies --patch.
916 --abbrev[=<n>]
918 Instead of showing the full 40-byte hexadecimal object name in
919 diff-raw format output and diff-tree header lines, show only a
920 partial prefix. This is independent of the --full-index option
921 above, which controls the diff-patch output format. Non default
922 number of digits can be specified with --abbrev=<n>.
923 -B[<n>][/<m>], --break-rewrites[=[<n>][/<m>]]
925 Break complete rewrite changes into pairs of delete and create.
926 This serves two purposes:
927 It affects the way a change that amounts to a total rewrite of a
929 file not as a series of deletion and insertion mixed together with
930 a very few lines that happen to match textually as the context, but
931 as a single deletion of everything old followed by a single
932 insertion of everything new, and the number m controls this aspect
933 of the -B option (defaults to 60%). -B/70% specifies that less
934 than 30% of the original should remain in the result for Git to
935 consider it a total rewrite (i.e. otherwise the resulting patch
936 will be a series of deletion and insertion mixed together with
937 context lines).
938 When used with -M, a totally-rewritten file is also considered as
940 the source of a rename (usually -M only considers a file that
941 disappeared as the source of a rename), and the number n controls
942 this aspect of the -B option (defaults to 50%). -B20% specifies
943 that a change with addition and deletion compared to 20% or more of
944 the file’s size are eligible for being picked up as a possible
945 source of a rename to another file.
946 -M[<n>], --find-renames[=<n>]
948 If generating diffs, detect and report renames for each commit. For
949 following files across renames while traversing history, see
950 --follow. If n is specified, it is a threshold on the similarity
951 index (i.e. amount of addition/deletions compared to the file’s
952 size). For example, -M90% means Git should consider a delete/add
953 pair to be a rename if more than 90% of the file hasn’t changed.
954 Without a % sign, the number is to be read as a fraction, with a
955 decimal point before it. I.e., -M5 becomes 0.5, and is thus the
956 same as -M50%. Similarly, -M05 is the same as -M5%. To limit
957 detection to exact renames, use -M100%. The default similarity
958 index is 50%.
959 -C[<n>], --find-copies[=<n>]
961 Detect copies as well as renames. See also --find-copies-harder. If
962 n is specified, it has the same meaning as for -M<n>.
963 --find-copies-harder
965 For performance reasons, by default, -C option finds copies only if
966 the original file of the copy was modified in the same changeset.
967 This flag makes the command inspect unmodified files as candidates
968 for the source of copy. This is a very expensive operation for
969 large projects, so use it with caution. Giving more than one -C
970 option has the same effect.
971 -D, --irreversible-delete
973 Omit the preimage for deletes, i.e. print only the header but not
974 the diff between the preimage and /dev/null. The resulting patch is
975 not meant to be applied with patch or git apply; this is solely for
976 people who want to just concentrate on reviewing the text after the
977 change. In addition, the output obviously lacks enough information
978 to apply such a patch in reverse, even manually, hence the name of
979 the option.
980 When used together with -B, omit also the preimage in the deletion
982 part of a delete/create pair.
983 -l<num>
985 The -M and -C options require O(n^2) processing time where n is the
986 number of potential rename/copy targets. This option prevents
987 rename/copy detection from running if the number of rename/copy
988 targets exceeds the specified number.
989 --diff-filter=[(A|C|D|M|R|T|U|X|B)...[*]]
991 Select only files that are Added (A), Copied (C), Deleted (D),
992 Modified (M), Renamed (R), have their type (i.e. regular file,
993 symlink, submodule, ...) changed (T), are Unmerged (U), are Unknown
994 (X), or have had their pairing Broken (B). Any combination of the
995 filter characters (including none) can be used. When *
996 (All-or-none) is added to the combination, all paths are selected
997 if there is any file that matches other criteria in the comparison;
998 if there is no file that matches other criteria, nothing is
999 selected.
1000 Also, these upper-case letters can be downcased to exclude. E.g.
1002 --diff-filter=ad excludes added and deleted paths.
1003 Note that not all diffs can feature all types. For instance, diffs
1005 from the index to the working tree can never have Added entries
1006 (because the set of paths included in the diff is limited by what
1007 is in the index). Similarly, copied and renamed entries cannot
1008 appear if detection for those types is disabled.
1009 -S<string>
1011 Look for differences that change the number of occurrences of the
1012 specified string (i.e. addition/deletion) in a file. Intended for
1013 the scripter’s use.
1014 It is useful when you’re looking for an exact block of code (like a
1016 struct), and want to know the history of that block since it first
1017 came into being: use the feature iteratively to feed the
1018 interesting block in the preimage back into -S, and keep going
1019 until you get the very first version of the block.
1020 Binary files are searched as well.
1022 -G<regex>
1024 Look for differences whose patch text contains added/removed lines
1025 that match <regex>.
1026 To illustrate the difference between -S<regex> --pickaxe-regex and
1028 -G<regex>, consider a commit with the following diff in the same
1029 file:
1030 + return !regexec(regexp, two->ptr, 1, ®match, 0);
1032 ...
1033 - hit = !regexec(regexp, mf2.ptr, 1, ®match, 0);
1034 While git log -G"regexec\(regexp" will show this commit, git log
1036 -S"regexec\(regexp" --pickaxe-regex will not (because the number of
1037 occurrences of that string did not change).
1038 Unless --text is supplied patches of binary files without a
1040 textconv filter will be ignored.
1041 See the pickaxe entry in gitdiffcore(7) for more information.
1043 --find-object=<object-id>
1045 Look for differences that change the number of occurrences of the
1046 specified object. Similar to -S, just the argument is different in
1047 that it doesn’t search for a specific string but for a specific
1048 object id.
1049 The object can be a blob or a submodule commit. It implies the -t
1051 option in git-log to also find trees.
1052 --pickaxe-all
1054 When -S or -G finds a change, show all the changes in that
1055 changeset, not just the files that contain the change in <string>.
1056 --pickaxe-regex
1058 Treat the <string> given to -S as an extended POSIX regular
1059 expression to match.
1060 -O<orderfile>
1062 Control the order in which files appear in the output. This
1063 overrides the diff.orderFile configuration variable (see git-
1064 config(1)). To cancel diff.orderFile, use -O/dev/null.
1065 The output order is determined by the order of glob patterns in
1067 <orderfile>. All files with pathnames that match the first pattern
1068 are output first, all files with pathnames that match the second
1069 pattern (but not the first) are output next, and so on. All files
1070 with pathnames that do not match any pattern are output last, as if
1071 there was an implicit match-all pattern at the end of the file. If
1072 multiple pathnames have the same rank (they match the same pattern
1073 but no earlier patterns), their output order relative to each other
1074 is the normal order.
1075 <orderfile> is parsed as follows:
1077 · Blank lines are ignored, so they can be used as separators for
1079 readability.
1080 · Lines starting with a hash ("#") are ignored, so they can be
1082 used for comments. Add a backslash ("\") to the beginning of
1083 the pattern if it starts with a hash.
1084 · Each other line contains a single pattern.
1086 Patterns have the same syntax and semantics as patterns used for
1088 fnmatch(3) without the FNM_PATHNAME flag, except a pathname also
1089 matches a pattern if removing any number of the final pathname
1090 components matches the pattern. For example, the pattern "foo*bar"
1091 matches "fooasdfbar" and "foo/bar/baz/asdf" but not "foobarx".
1092 -R
1094 Swap two inputs; that is, show differences from index or on-disk
1095 file to tree contents.
1096 --relative[=<path>]
1098 When run from a subdirectory of the project, it can be told to
1099 exclude changes outside the directory and show pathnames relative
1100 to it with this option. When you are not in a subdirectory (e.g. in
1101 a bare repository), you can name which subdirectory to make the
1102 output relative to by giving a <path> as an argument.
1103 -a, --text
1105 Treat all files as text.
1106 --ignore-cr-at-eol
1108 Ignore carriage-return at the end of line when doing a comparison.
1109 --ignore-space-at-eol
1111 Ignore changes in whitespace at EOL.
1112 -b, --ignore-space-change
1114 Ignore changes in amount of whitespace. This ignores whitespace at
1115 line end, and considers all other sequences of one or more
1116 whitespace characters to be equivalent.
1117 -w, --ignore-all-space
1119 Ignore whitespace when comparing lines. This ignores differences
1120 even if one line has whitespace where the other line has none.
1121 --ignore-blank-lines
1123 Ignore changes whose lines are all blank.
1124 --inter-hunk-context=<lines>
1126 Show the context between diff hunks, up to the specified number of
1127 lines, thereby fusing hunks that are close to each other. Defaults
1128 to diff.interHunkContext or 0 if the config option is unset.
1129 -W, --function-context
1131 Show whole surrounding functions of changes.
1132 --ext-diff
1134 Allow an external diff helper to be executed. If you set an
1135 external diff driver with gitattributes(5), you need to use this
1136 option with git-log(1) and friends.
1137 --no-ext-diff
1139 Disallow external diff drivers.
1140 --textconv, --no-textconv
1142 Allow (or disallow) external text conversion filters to be run when
1143 comparing binary files. See gitattributes(5) for details. Because
1144 textconv filters are typically a one-way conversion, the resulting
1145 diff is suitable for human consumption, but cannot be applied. For
1146 this reason, textconv filters are enabled by default only for git-
1147 diff(1) and git-log(1), but not for git-format-patch(1) or diff
1148 plumbing commands.
1149 --ignore-submodules[=<when>]
1151 Ignore changes to submodules in the diff generation. <when> can be
1152 either "none", "untracked", "dirty" or "all", which is the default.
1153 Using "none" will consider the submodule modified when it either
1154 contains untracked or modified files or its HEAD differs from the
1155 commit recorded in the superproject and can be used to override any
1156 settings of the ignore option in git-config(1) or gitmodules(5).
1157 When "untracked" is used submodules are not considered dirty when
1158 they only contain untracked content (but they are still scanned for
1159 modified content). Using "dirty" ignores all changes to the work
1160 tree of submodules, only changes to the commits stored in the
1161 superproject are shown (this was the behavior until 1.7.0). Using
1162 "all" hides all changes to submodules.
1163 --src-prefix=<prefix>
1165 Show the given source prefix instead of "a/".
1166 --dst-prefix=<prefix>
1168 Show the given destination prefix instead of "b/".
1169 --no-prefix
1171 Do not show any source or destination prefix.
1172 --line-prefix=<prefix>
1174 Prepend an additional prefix to every line of output.
1175 --ita-invisible-in-index
1177 By default entries added by "git add -N" appear as an existing
1178 empty file in "git diff" and a new file in "git diff --cached".
1179 This option makes the entry appear as a new file in "git diff" and
1180 non-existent in "git diff --cached". This option could be reverted
1181 with --ita-visible-in-index. Both options are experimental and
1182 could be removed in future.
1183 For more detailed explanation on these common options, see also
1185 gitdiffcore(7).
1186
1188 Running git-diff(1), git-log(1), git-show(1), git-diff-index(1), git-
1189 diff-tree(1), or git-diff-files(1) with the -p option produces patch
1190 text. You can customize the creation of patch text via the
1191 GIT_EXTERNAL_DIFF and the GIT_DIFF_OPTS environment variables.
1192 What the -p option produces is slightly different from the traditional
1194 diff format:
1195 1. It is preceded with a "git diff" header that looks like this:
1197 diff --git a/file1 b/file2
1199 The a/ and b/ filenames are the same unless rename/copy is
1201 involved. Especially, even for a creation or a deletion, /dev/null
1202 is not used in place of the a/ or b/ filenames.
1203 When rename/copy is involved, file1 and file2 show the name of the
1205 source file of the rename/copy and the name of the file that
1206 rename/copy produces, respectively.
1207 2. It is followed by one or more extended header lines:
1209 old mode <mode>
1211 new mode <mode>
1212 deleted file mode <mode>
1213 new file mode <mode>
1214 copy from <path>
1215 copy to <path>
1216 rename from <path>
1217 rename to <path>
1218 similarity index <number>
1219 dissimilarity index <number>
1220 index <hash>..<hash> <mode>
1221 File modes are printed as 6-digit octal numbers including the file
1223 type and file permission bits.
1224 Path names in extended headers do not include the a/ and b/
1226 prefixes.
1227 The similarity index is the percentage of unchanged lines, and the
1229 dissimilarity index is the percentage of changed lines. It is a
1230 rounded down integer, followed by a percent sign. The similarity
1231 index value of 100% is thus reserved for two equal files, while
1232 100% dissimilarity means that no line from the old file made it
1233 into the new one.
1234 The index line includes the blob object names before and after the
1236 change. The <mode> is included if the file mode does not change;
1237 otherwise, separate lines indicate the old and the new mode.
1238 3. Pathnames with "unusual" characters are quoted as explained for the
1240 configuration variable core.quotePath (see git-config(1)).
1241 4. All the file1 files in the output refer to files before the commit,
1243 and all the file2 files refer to files after the commit. It is
1244 incorrect to apply each change to each file sequentially. For
1245 example, this patch will swap a and b:
1246 diff --git a/a b/b
1248 rename from a
1249 rename to b
1250 diff --git a/b b/a
1251 rename from b
1252 rename to a
1253
1255 Any diff-generating command can take the -c or --cc option to produce a
1256 combined diff when showing a merge. This is the default format when
1257 showing merges with git-diff(1) or git-show(1). Note also that you can
1258 give the -m option to any of these commands to force generation of
1259 diffs with individual parents of a merge.
1260 A "combined diff" format looks like this:
1262 diff --combined describe.c
1264 index fabadb8,cc95eb0..4866510
1265 --- a/describe.c
1266 +++ b/describe.c
1267 @@@ -98,20 -98,12 +98,20 @@@
1268 return (a_date > b_date) ? -1 : (a_date == b_date) ? 0 : 1;
1269 }
1270 - static void describe(char *arg)
1272 -static void describe(struct commit *cmit, int last_one)
1273 ++static void describe(char *arg, int last_one)
1274 {
1275 + unsigned char sha1[20];
1276 + struct commit *cmit;
1277 struct commit_list *list;
1278 static int initialized = 0;
1279 struct commit_name *n;
1280 + if (get_sha1(arg, sha1) < 0)
1282 + usage(describe_usage);
1283 + cmit = lookup_commit_reference(sha1);
1284 + if (!cmit)
1285 + usage(describe_usage);
1286 +
1287 if (!initialized) {
1288 initialized = 1;
1289 for_each_ref(get_name);
1290 1. It is preceded with a "git diff" header, that looks like this (when
1294 the -c option is used):
1295 diff --combined file
1297 or like this (when the --cc option is used):
1299 diff --cc file
1301 2. It is followed by one or more extended header lines (this example
1303 shows a merge with two parents):
1304 index <hash>,<hash>..<hash>
1306 mode <mode>,<mode>..<mode>
1307 new file mode <mode>
1308 deleted file mode <mode>,<mode>
1309 The mode <mode>,<mode>..<mode> line appears only if at least one of
1311 the <mode> is different from the rest. Extended headers with
1312 information about detected contents movement (renames and copying
1313 detection) are designed to work with diff of two <tree-ish> and are
1314 not used by combined diff format.
1315 3. It is followed by two-line from-file/to-file header
1317 --- a/file
1319 +++ b/file
1320 Similar to two-line header for traditional unified diff format,
1322 /dev/null is used to signal created or deleted files.
1323 However, if the --combined-all-paths option is provided, instead of
1325 a two-line from-file/to-file you get a N+1 line from-file/to-file
1326 header, where N is the number of parents in the merge commit
1327 --- a/file
1329 --- a/file
1330 --- a/file
1331 +++ b/file
1332 This extended format can be useful if rename or copy detection is
1334 active, to allow you to see the original name of the file in
1335 different parents.
1336 4. Chunk header format is modified to prevent people from accidentally
1338 feeding it to patch -p1. Combined diff format was created for
1339 review of merge commit changes, and was not meant to be applied.
1340 The change is similar to the change in the extended index header:
1341 @@@ <from-file-range> <from-file-range> <to-file-range> @@@
1343 There are (number of parents + 1) @ characters in the chunk header
1345 for combined diff format.
1346 Unlike the traditional unified diff format, which shows two files A and
1348 B with a single column that has - (minus — appears in A but removed in
1349 B), + (plus — missing in A but added to B), or " " (space — unchanged)
1350 prefix, this format compares two or more files file1, file2,... with
1351 one file X, and shows how X differs from each of fileN. One column for
1352 each of fileN is prepended to the output line to note how X’s line is
1353 different from it.
1354 A - character in the column N means that the line appears in fileN but
1356 it does not appear in the result. A + character in the column N means
1357 that the line appears in the result, and fileN does not have that line
1358 (in other words, the line was added, from the point of view of that
1359 parent).
1360 In the above example output, the function signature was changed from
1362 both files (hence two - removals from both file1 and file2, plus ++ to
1363 mean one line that was added does not appear in either file1 or file2).
1364 Also eight other lines are the same from file1 but do not appear in
1365 file2 (hence prefixed with +).
1366 When shown by git diff-tree -c, it compares the parents of a merge
1368 commit with the merge result (i.e. file1..fileN are the parents). When
1369 shown by git diff-files -c, it compares the two unresolved merge
1370 parents with the working tree file (i.e. file1 is stage 2 aka "our
1371 version", file2 is stage 3 aka "their version").
1372
1374 git show v1.0.0
1375 Shows the tag v1.0.0, along with the object the tags points at.
1376 git show v1.0.0^{tree}
1378 Shows the tree pointed to by the tag v1.0.0.
1379 git show -s --format=%s v1.0.0^{commit}
1381 Shows the subject of the commit pointed to by the tag v1.0.0.
1382 git show next~10:Documentation/README
1384 Shows the contents of the file Documentation/README as they were
1385 current in the 10th last commit of the branch next.
1386 git show master:Makefile master:t/Makefile
1388 Concatenates the contents of said Makefiles in the head of the
1389 branch master.
1390
1392 Git is to some extent character encoding agnostic.
1393 · The contents of the blob objects are uninterpreted sequences of
1395 bytes. There is no encoding translation at the core level.
1396 · Path names are encoded in UTF-8 normalization form C. This applies
1398 to tree objects, the index file, ref names, as well as path names
1399 in command line arguments, environment variables and config files
1400 (.git/config (see git-config(1)), gitignore(5), gitattributes(5)
1401 and gitmodules(5)).
1402 Note that Git at the core level treats path names simply as
1404 sequences of non-NUL bytes, there are no path name encoding
1405 conversions (except on Mac and Windows). Therefore, using non-ASCII
1406 path names will mostly work even on platforms and file systems that
1407 use legacy extended ASCII encodings. However, repositories created
1408 on such systems will not work properly on UTF-8-based systems (e.g.
1409 Linux, Mac, Windows) and vice versa. Additionally, many Git-based
1410 tools simply assume path names to be UTF-8 and will fail to display
1411 other encodings correctly.
1412 · Commit log messages are typically encoded in UTF-8, but other
1414 extended ASCII encodings are also supported. This includes
1415 ISO-8859-x, CP125x and many others, but not UTF-16/32, EBCDIC and
1416 CJK multi-byte encodings (GBK, Shift-JIS, Big5, EUC-x, CP9xx etc.).
1417 Although we encourage that the commit log messages are encoded in
1419 UTF-8, both the core and Git Porcelain are designed not to force UTF-8
1420 on projects. If all participants of a particular project find it more
1421 convenient to use legacy encodings, Git does not forbid it. However,
1422 there are a few things to keep in mind.
1423 1. git commit and git commit-tree issues a warning if the commit log
1425 message given to it does not look like a valid UTF-8 string, unless
1426 you explicitly say your project uses a legacy encoding. The way to
1427 say this is to have i18n.commitencoding in .git/config file, like
1428 this:
1429 [i18n]
1431 commitEncoding = ISO-8859-1
1432 Commit objects created with the above setting record the value of
1434 i18n.commitEncoding in its encoding header. This is to help other
1435 people who look at them later. Lack of this header implies that the
1436 commit log message is encoded in UTF-8.
1437 2. git log, git show, git blame and friends look at the encoding
1439 header of a commit object, and try to re-code the log message into
1440 UTF-8 unless otherwise specified. You can specify the desired
1441 output encoding with i18n.logOutputEncoding in .git/config file,
1442 like this:
1443 [i18n]
1445 logOutputEncoding = ISO-8859-1
1446 If you do not have this configuration variable, the value of
1448 i18n.commitEncoding is used instead.
1449 Note that we deliberately chose not to re-code the commit log message
1451 when a commit is made to force UTF-8 at the commit object level,
1452 because re-coding to UTF-8 is not necessarily a reversible operation.
1453
1455 Part of the git(1) suite
1456Git 2.24.1 12/10/2019 GIT-SHOW(1)