1GIT-SHOW(1) Git Manual GIT-SHOW(1)
2
6 git-show - Show various types of objects
7
9 git show [<options>] [<object>...]
10
12 Shows one or more objects (blobs, trees, tags and commits).
13 For commits it shows the log message and textual diff. It also presents
15 the merge commit in a special format as produced by git diff-tree --cc.
16 For tags, it shows the tag message and the referenced objects.
18 For trees, it shows the names (equivalent to git ls-tree with
20 --name-only).
21 For plain blobs, it shows the plain contents.
23 The command takes options applicable to the git diff-tree command to
25 control how the changes the commit introduces are shown.
26 This manual page describes only the most frequently used options.
28
30 <object>...
31 The names of objects to show (defaults to HEAD). For a more
32 complete list of ways to spell object names, see "SPECIFYING
33 REVISIONS" section in gitrevisions(7).
34 --pretty[=<format>], --format=<format>
36 Pretty-print the contents of the commit logs in a given format,
37 where <format> can be one of oneline, short, medium, full, fuller,
38 reference, email, raw, format:<string> and tformat:<string>. When
39 <format> is none of the above, and has %placeholder in it, it acts
40 as if --pretty=tformat:<format> were given.
41 See the "PRETTY FORMATS" section for some additional details for
43 each format. When =<format> part is omitted, it defaults to medium.
44 Note: you can specify the default pretty format in the repository
46 configuration (see git-config(1)).
47 --abbrev-commit
49 Instead of showing the full 40-byte hexadecimal commit object name,
50 show a prefix that names the object uniquely. "--abbrev=<n>" (which
51 also modifies diff output, if it is displayed) option can be used
52 to specify the minimum length of the prefix.
53 This should make "--pretty=oneline" a whole lot more readable for
55 people using 80-column terminals.
56 --no-abbrev-commit
58 Show the full 40-byte hexadecimal commit object name. This negates
59 --abbrev-commit, either explicit or implied by other options such
60 as "--oneline". It also overrides the log.abbrevCommit variable.
61 --oneline
63 This is a shorthand for "--pretty=oneline --abbrev-commit" used
64 together.
65 --encoding=<encoding>
67 The commit objects record the encoding used for the log message in
68 their encoding header; this option can be used to tell the command
69 to re-code the commit log message in the encoding preferred by the
70 user. For non plumbing commands this defaults to UTF-8. Note that
71 if an object claims to be encoded in X and we are outputting in X,
72 we will output the object verbatim; this means that invalid
73 sequences in the original commit may be copied to the output.
74 --expand-tabs=<n>, --expand-tabs, --no-expand-tabs
76 Perform a tab expansion (replace each tab with enough spaces to
77 fill to the next display column that is multiple of <n>) in the log
78 message before showing it in the output. --expand-tabs is a
79 short-hand for --expand-tabs=8, and --no-expand-tabs is a
80 short-hand for --expand-tabs=0, which disables tab expansion.
81 By default, tabs are expanded in pretty formats that indent the log
83 message by 4 spaces (i.e. medium, which is the default, full, and
84 fuller).
85 --notes[=<ref>]
87 Show the notes (see git-notes(1)) that annotate the commit, when
88 showing the commit log message. This is the default for git log,
89 git show and git whatchanged commands when there is no --pretty,
90 --format, or --oneline option given on the command line.
91 By default, the notes shown are from the notes refs listed in the
93 core.notesRef and notes.displayRef variables (or corresponding
94 environment overrides). See git-config(1) for more details.
95 With an optional <ref> argument, use the ref to find the notes to
97 display. The ref can specify the full refname when it begins with
98 refs/notes/; when it begins with notes/, refs/ and otherwise
99 refs/notes/ is prefixed to form a full name of the ref.
100 Multiple --notes options can be combined to control which notes are
102 being displayed. Examples: "--notes=foo" will show only notes from
103 "refs/notes/foo"; "--notes=foo --notes" will show both notes from
104 "refs/notes/foo" and from the default notes ref(s).
105 --no-notes
107 Do not show notes. This negates the above --notes option, by
108 resetting the list of notes refs from which notes are shown.
109 Options are parsed in the order given on the command line, so e.g.
110 "--notes --notes=foo --no-notes --notes=bar" will only show notes
111 from "refs/notes/bar".
112 --show-notes[=<ref>], --[no-]standard-notes
114 These options are deprecated. Use the above --notes/--no-notes
115 options instead.
116 --show-signature
118 Check the validity of a signed commit object by passing the
119 signature to gpg --verify and show the output.
120
122 If the commit is a merge, and if the pretty-format is not oneline,
123 email or raw, an additional line is inserted before the Author: line.
124 This line begins with "Merge: " and the hashes of ancestral commits are
125 printed, separated by spaces. Note that the listed commits may not
126 necessarily be the list of the direct parent commits if you have
127 limited your view of history: for example, if you are only interested
128 in changes related to a certain directory or file.
129 There are several built-in formats, and you can define additional
131 formats by setting a pretty.<name> config option to either another
132 format name, or a format: string, as described below (see git-
133 config(1)). Here are the details of the built-in formats:
134 • oneline
136 <hash> <title line>
138 This is designed to be as compact as possible.
140 • short
142 commit <hash>
144 Author: <author>
145 <title line>
147 • medium
149 commit <hash>
151 Author: <author>
152 Date: <author date>
153 <title line>
155 <full commit message>
157 • full
159 commit <hash>
161 Author: <author>
162 Commit: <committer>
163 <title line>
165 <full commit message>
167 • fuller
169 commit <hash>
171 Author: <author>
172 AuthorDate: <author date>
173 Commit: <committer>
174 CommitDate: <committer date>
175 <title line>
177 <full commit message>
179 • reference
181 <abbrev hash> (<title line>, <short author date>)
183 This format is used to refer to another commit in a commit message
185 and is the same as --pretty='format:%C(auto)%h (%s, %ad)'. By
186 default, the date is formatted with --date=short unless another
187 --date option is explicitly specified. As with any format: with
188 format placeholders, its output is not affected by other options
189 like --decorate and --walk-reflogs.
190 • email
192 From <hash> <date>
194 From: <author>
195 Date: <author date>
196 Subject: [PATCH] <title line>
197 <full commit message>
199 • mboxrd
201 Like email, but lines in the commit message starting with "From "
203 (preceded by zero or more ">") are quoted with ">" so they aren’t
204 confused as starting a new commit.
205 • raw
207 The raw format shows the entire commit exactly as stored in the
209 commit object. Notably, the hashes are displayed in full,
210 regardless of whether --abbrev or --no-abbrev are used, and parents
211 information show the true parent commits, without taking grafts or
212 history simplification into account. Note that this format affects
213 the way commits are displayed, but not the way the diff is shown
214 e.g. with git log --raw. To get full object names in a raw diff
215 format, use --no-abbrev.
216 • format:<string>
218 The format:<string> format allows you to specify which information
220 you want to show. It works a little bit like printf format, with
221 the notable exception that you get a newline with %n instead of \n.
222 E.g, format:"The author of %h was %an, %ar%nThe title was >>%s<<%n"
224 would show something like this:
225 The author of fe6e0ee was Junio C Hamano, 23 hours ago
227 The title was >>t4119: test autocomputing -p<n> for traditional diff input.<<
228 The placeholders are:
230 • Placeholders that expand to a single literal character:
232 %n
234 newline
235 %%
237 a raw %
238 %x00
240 print a byte from a hex code
241 • Placeholders that affect formatting of later placeholders:
243 %Cred
245 switch color to red
246 %Cgreen
248 switch color to green
249 %Cblue
251 switch color to blue
252 %Creset
254 reset color
255 %C(...)
257 color specification, as described under Values in the
258 "CONFIGURATION FILE" section of git-config(1). By default,
259 colors are shown only when enabled for log output (by
260 color.diff, color.ui, or --color, and respecting the auto
261 settings of the former if we are going to a terminal).
262 %C(auto,...) is accepted as a historical synonym for the
263 default (e.g., %C(auto,red)). Specifying %C(always,...)
264 will show the colors even when color is not otherwise
265 enabled (though consider just using --color=always to
266 enable color for the whole output, including this format
267 and anything else git might color). auto alone (i.e.
268 %C(auto)) will turn on auto coloring on the next
269 placeholders until the color is switched again.
270 %m
272 left (<), right (>) or boundary (-) mark
273 %w([<w>[,<i1>[,<i2>]]])
275 switch line wrapping, like the -w option of git-
276 shortlog(1).
277 %<(<N>[,trunc|ltrunc|mtrunc])
279 make the next placeholder take at least N columns, padding
280 spaces on the right if necessary. Optionally truncate at
281 the beginning (ltrunc), the middle (mtrunc) or the end
282 (trunc) if the output is longer than N columns. Note that
283 truncating only works correctly with N >= 2.
284 %<|(<N>)
286 make the next placeholder take at least until Nth columns,
287 padding spaces on the right if necessary
288 %>(<N>), %>|(<N>)
290 similar to %<(<N>), %<|(<N>) respectively, but padding
291 spaces on the left
292 %>>(<N>), %>>|(<N>)
294 similar to %>(<N>), %>|(<N>) respectively, except that if
295 the next placeholder takes more spaces than given and there
296 are spaces on its left, use those spaces
297 %><(<N>), %><|(<N>)
299 similar to %<(<N>), %<|(<N>) respectively, but padding both
300 sides (i.e. the text is centered)
301 • Placeholders that expand to information extracted from the
303 commit:
304 %H
306 commit hash
307 %h
309 abbreviated commit hash
310 %T
312 tree hash
313 %t
315 abbreviated tree hash
316 %P
318 parent hashes
319 %p
321 abbreviated parent hashes
322 %an
324 author name
325 %aN
327 author name (respecting .mailmap, see git-shortlog(1) or
328 git-blame(1))
329 %ae
331 author email
332 %aE
334 author email (respecting .mailmap, see git-shortlog(1) or
335 git-blame(1))
336 %al
338 author email local-part (the part before the @ sign)
339 %aL
341 author local-part (see %al) respecting .mailmap, see git-
342 shortlog(1) or git-blame(1))
343 %ad
345 author date (format respects --date= option)
346 %aD
348 author date, RFC2822 style
349 %ar
351 author date, relative
352 %at
354 author date, UNIX timestamp
355 %ai
357 author date, ISO 8601-like format
358 %aI
360 author date, strict ISO 8601 format
361 %as
363 author date, short format (YYYY-MM-DD)
364 %cn
366 committer name
367 %cN
369 committer name (respecting .mailmap, see git-shortlog(1) or
370 git-blame(1))
371 %ce
373 committer email
374 %cE
376 committer email (respecting .mailmap, see git-shortlog(1)
377 or git-blame(1))
378 %cl
380 committer email local-part (the part before the @ sign)
381 %cL
383 committer local-part (see %cl) respecting .mailmap, see
384 git-shortlog(1) or git-blame(1))
385 %cd
387 committer date (format respects --date= option)
388 %cD
390 committer date, RFC2822 style
391 %cr
393 committer date, relative
394 %ct
396 committer date, UNIX timestamp
397 %ci
399 committer date, ISO 8601-like format
400 %cI
402 committer date, strict ISO 8601 format
403 %cs
405 committer date, short format (YYYY-MM-DD)
406 %d
408 ref names, like the --decorate option of git-log(1)
409 %D
411 ref names without the " (", ")" wrapping.
412 %S
414 ref name given on the command line by which the commit was
415 reached (like git log --source), only works with git log
416 %e
418 encoding
419 %s
421 subject
422 %f
424 sanitized subject line, suitable for a filename
425 %b
427 body
428 %B
430 raw body (unwrapped subject and body)
431 %N
433 commit notes
434 %GG
436 raw verification message from GPG for a signed commit
437 %G?
439 show "G" for a good (valid) signature, "B" for a bad
440 signature, "U" for a good signature with unknown validity,
441 "X" for a good signature that has expired, "Y" for a good
442 signature made by an expired key, "R" for a good signature
443 made by a revoked key, "E" if the signature cannot be
444 checked (e.g. missing key) and "N" for no signature
445 %GS
447 show the name of the signer for a signed commit
448 %GK
450 show the key used to sign a signed commit
451 %GF
453 show the fingerprint of the key used to sign a signed
454 commit
455 %GP
457 show the fingerprint of the primary key whose subkey was
458 used to sign a signed commit
459 %GT
461 show the trust level for the key used to sign a signed
462 commit
463 %gD
465 reflog selector, e.g., refs/stash@{1} or refs/stash@{2
466 minutes ago}; the format follows the rules described for
467 the -g option. The portion before the @ is the refname as
468 given on the command line (so git log -g refs/heads/master
469 would yield refs/heads/master@{0}).
470 %gd
472 shortened reflog selector; same as %gD, but the refname
473 portion is shortened for human readability (so
474 refs/heads/master becomes just master).
475 %gn
477 reflog identity name
478 %gN
480 reflog identity name (respecting .mailmap, see git-
481 shortlog(1) or git-blame(1))
482 %ge
484 reflog identity email
485 %gE
487 reflog identity email (respecting .mailmap, see git-
488 shortlog(1) or git-blame(1))
489 %gs
491 reflog subject
492 %(trailers[:options])
494 display the trailers of the body as interpreted by git-
495 interpret-trailers(1). The trailers string may be followed
496 by a colon and zero or more comma-separated options. If any
497 option is provided multiple times the last occurance wins.
498 The boolean options accept an optional value [=<BOOL>]. The
500 values true, false, on, off etc. are all accepted. See the
501 "boolean" sub-section in "EXAMPLES" in git-config(1). If a
502 boolean option is given with no value, it’s enabled.
503 • key=<K>: only show trailers with specified key.
505 Matching is done case-insensitively and trailing colon
506 is optional. If option is given multiple times trailer
507 lines matching any of the keys are shown. This option
508 automatically enables the only option so that
509 non-trailer lines in the trailer block are hidden. If
510 that is not desired it can be disabled with only=false.
511 E.g., %(trailers:key=Reviewed-by) shows trailer lines
512 with key Reviewed-by.
513 • only[=<BOOL>]: select whether non-trailer lines from
515 the trailer block should be included.
516 • separator=<SEP>: specify a separator inserted between
518 trailer lines. When this option is not given each
519 trailer line is terminated with a line feed character.
520 The string SEP may contain the literal formatting codes
521 described above. To use comma as separator one must use
522 %x2C as it would otherwise be parsed as next option.
523 E.g., %(trailers:key=Ticket,separator=%x2C ) shows all
524 trailer lines whose key is "Ticket" separated by a
525 comma and a space.
526 • unfold[=<BOOL>]: make it behave as if
528 interpret-trailer’s --unfold option was given. E.g.,
529 %(trailers:only,unfold=true) unfolds and shows all
530 trailer lines.
531 • keyonly[=<BOOL>]: only show the key part of the
533 trailer.
534 • valueonly[=<BOOL>]: only show the value part of the
536 trailer.
537 • key_value_separator=<SEP>: specify a separator inserted
539 between trailer lines. When this option is not given
540 each trailer key-value pair is separated by ": ".
541 Otherwise it shares the same semantics as
542 separator=<SEP> above.
543 Note
545 Some placeholders may depend on other options given to the revision
546 traversal engine. For example, the %g* reflog options will insert
547 an empty string unless we are traversing reflog entries (e.g., by
548 git log -g). The %d and %D placeholders will use the "short"
549 decoration format if --decorate was not already provided on the
550 command line.
551 If you add a + (plus sign) after % of a placeholder, a line-feed is
553 inserted immediately before the expansion if and only if the
554 placeholder expands to a non-empty string.
555 If you add a - (minus sign) after % of a placeholder, all consecutive
557 line-feeds immediately preceding the expansion are deleted if and only
558 if the placeholder expands to an empty string.
559 If you add a ` ` (space) after % of a placeholder, a space is inserted
561 immediately before the expansion if and only if the placeholder expands
562 to a non-empty string.
563 • tformat:
565 The tformat: format works exactly like format:, except that it
567 provides "terminator" semantics instead of "separator" semantics.
568 In other words, each commit has the message terminator character
569 (usually a newline) appended, rather than a separator placed
570 between entries. This means that the final entry of a single-line
571 format will be properly terminated with a new line, just as the
572 "oneline" format does. For example:
573 $ git log -2 --pretty=format:%h 4da45bef \
575 | perl -pe '$_ .= " -- NO NEWLINE\n" unless /\n/'
576 4da45be
577 7134973 -- NO NEWLINE
578 $ git log -2 --pretty=tformat:%h 4da45bef \
580 | perl -pe '$_ .= " -- NO NEWLINE\n" unless /\n/'
581 4da45be
582 7134973
583 In addition, any unrecognized string that has a % in it is
585 interpreted as if it has tformat: in front of it. For example,
586 these two are equivalent:
587 $ git log -2 --pretty=tformat:%h 4da45bef
589 $ git log -2 --pretty=%h 4da45bef
590
592 The options below can be used to change the way git show generates diff
593 output.
594 -p, -u, --patch
596 Generate patch (see section on generating patches).
597 -s, --no-patch
599 Suppress diff output. Useful for commands like git show that show
600 the patch by default, or to cancel the effect of --patch.
601 --diff-merges=(off|none|first-parent|1|separate|m|combined|c|dense-combined|cc),
603 --no-diff-merges
604 Specify diff format to be used for merge commits. Default is
605 `dense-combined` unless --first-parent is in use, in which case
606 first-parent is the default.
607 --diff-merges=(off|none), --no-diff-merges
609 Disable output of diffs for merge commits. Useful to override
610 implied value.
611 --diff-merges=first-parent, --diff-merges=1
613 This option makes merge commits show the full diff with respect
614 to the first parent only.
615 --diff-merges=separate, --diff-merges=m, -m
617 This makes merge commits show the full diff with respect to
618 each of the parents. Separate log entry and diff is generated
619 for each parent. -m doesn’t produce any output without -p.
620 --diff-merges=combined, --diff-merges=c, -c
622 With this option, diff output for a merge commit shows the
623 differences from each of the parents to the merge result
624 simultaneously instead of showing pairwise diff between a
625 parent and the result one at a time. Furthermore, it lists only
626 files which were modified from all parents. -c implies -p.
627 --diff-merges=dense-combined, --diff-merges=cc, --cc
629 With this option the output produced by --diff-merges=combined
630 is further compressed by omitting uninteresting hunks whose
631 contents in the parents have only two variants and the merge
632 result picks one of them without modification. --cc implies
633 -p.
634 --combined-all-paths
636 This flag causes combined diffs (used for merge commits) to list
637 the name of the file from all parents. It thus only has effect when
638 --diff-merges=[dense-]combined is in use, and is likely only useful
639 if filename changes are detected (i.e. when either rename or copy
640 detection have been requested).
641 -U<n>, --unified=<n>
643 Generate diffs with <n> lines of context instead of the usual
644 three. Implies --patch.
645 --output=<file>
647 Output to a specific file instead of stdout.
648 --output-indicator-new=<char>, --output-indicator-old=<char>,
650 --output-indicator-context=<char>
651 Specify the character used to indicate new, old or context lines in
652 the generated patch. Normally they are +, - and ' ' respectively.
653 --raw
655 For each commit, show a summary of changes using the raw diff
656 format. See the "RAW OUTPUT FORMAT" section of git-diff(1). This is
657 different from showing the log itself in raw format, which you can
658 achieve with --format=raw.
659 --patch-with-raw
661 Synonym for -p --raw.
662 -t
664 Show the tree objects in the diff output.
665 --indent-heuristic
667 Enable the heuristic that shifts diff hunk boundaries to make
668 patches easier to read. This is the default.
669 --no-indent-heuristic
671 Disable the indent heuristic.
672 --minimal
674 Spend extra time to make sure the smallest possible diff is
675 produced.
676 --patience
678 Generate a diff using the "patience diff" algorithm.
679 --histogram
681 Generate a diff using the "histogram diff" algorithm.
682 --anchored=<text>
684 Generate a diff using the "anchored diff" algorithm.
685 This option may be specified more than once.
687 If a line exists in both the source and destination, exists only
689 once, and starts with this text, this algorithm attempts to prevent
690 it from appearing as a deletion or addition in the output. It uses
691 the "patience diff" algorithm internally.
692 --diff-algorithm={patience|minimal|histogram|myers}
694 Choose a diff algorithm. The variants are as follows:
695 default, myers
697 The basic greedy diff algorithm. Currently, this is the
698 default.
699 minimal
701 Spend extra time to make sure the smallest possible diff is
702 produced.
703 patience
705 Use "patience diff" algorithm when generating patches.
706 histogram
708 This algorithm extends the patience algorithm to "support
709 low-occurrence common elements".
710 For instance, if you configured the diff.algorithm variable to a
712 non-default value and want to use the default one, then you have to
713 use --diff-algorithm=default option.
714 --stat[=<width>[,<name-width>[,<count>]]]
716 Generate a diffstat. By default, as much space as necessary will be
717 used for the filename part, and the rest for the graph part.
718 Maximum width defaults to terminal width, or 80 columns if not
719 connected to a terminal, and can be overridden by <width>. The
720 width of the filename part can be limited by giving another width
721 <name-width> after a comma. The width of the graph part can be
722 limited by using --stat-graph-width=<width> (affects all commands
723 generating a stat graph) or by setting diff.statGraphWidth=<width>
724 (does not affect git format-patch). By giving a third parameter
725 <count>, you can limit the output to the first <count> lines,
726 followed by ... if there are more.
727 These parameters can also be set individually with
729 --stat-width=<width>, --stat-name-width=<name-width> and
730 --stat-count=<count>.
731 --compact-summary
733 Output a condensed summary of extended header information such as
734 file creations or deletions ("new" or "gone", optionally "+l" if
735 it’s a symlink) and mode changes ("+x" or "-x" for adding or
736 removing executable bit respectively) in diffstat. The information
737 is put between the filename part and the graph part. Implies
738 --stat.
739 --numstat
741 Similar to --stat, but shows number of added and deleted lines in
742 decimal notation and pathname without abbreviation, to make it more
743 machine friendly. For binary files, outputs two - instead of saying
744 0 0.
745 --shortstat
747 Output only the last line of the --stat format containing total
748 number of modified files, as well as number of added and deleted
749 lines.
750 -X[<param1,param2,...>], --dirstat[=<param1,param2,...>]
752 Output the distribution of relative amount of changes for each
753 sub-directory. The behavior of --dirstat can be customized by
754 passing it a comma separated list of parameters. The defaults are
755 controlled by the diff.dirstat configuration variable (see git-
756 config(1)). The following parameters are available:
757 changes
759 Compute the dirstat numbers by counting the lines that have
760 been removed from the source, or added to the destination. This
761 ignores the amount of pure code movements within a file. In
762 other words, rearranging lines in a file is not counted as much
763 as other changes. This is the default behavior when no
764 parameter is given.
765 lines
767 Compute the dirstat numbers by doing the regular line-based
768 diff analysis, and summing the removed/added line counts. (For
769 binary files, count 64-byte chunks instead, since binary files
770 have no natural concept of lines). This is a more expensive
771 --dirstat behavior than the changes behavior, but it does count
772 rearranged lines within a file as much as other changes. The
773 resulting output is consistent with what you get from the other
774 --*stat options.
775 files
777 Compute the dirstat numbers by counting the number of files
778 changed. Each changed file counts equally in the dirstat
779 analysis. This is the computationally cheapest --dirstat
780 behavior, since it does not have to look at the file contents
781 at all.
782 cumulative
784 Count changes in a child directory for the parent directory as
785 well. Note that when using cumulative, the sum of the
786 percentages reported may exceed 100%. The default
787 (non-cumulative) behavior can be specified with the
788 noncumulative parameter.
789 <limit>
791 An integer parameter specifies a cut-off percent (3% by
792 default). Directories contributing less than this percentage of
793 the changes are not shown in the output.
794 Example: The following will count changed files, while ignoring
796 directories with less than 10% of the total amount of changed
797 files, and accumulating child directory counts in the parent
798 directories: --dirstat=files,10,cumulative.
799 --cumulative
801 Synonym for --dirstat=cumulative
802 --dirstat-by-file[=<param1,param2>...]
804 Synonym for --dirstat=files,param1,param2...
805 --summary
807 Output a condensed summary of extended header information such as
808 creations, renames and mode changes.
809 --patch-with-stat
811 Synonym for -p --stat.
812 -z
814 Separate the commits with NULs instead of with new newlines.
815 Also, when --raw or --numstat has been given, do not munge
817 pathnames and use NULs as output field terminators.
818 Without this option, pathnames with "unusual" characters are quoted
820 as explained for the configuration variable core.quotePath (see
821 git-config(1)).
822 --name-only
824 Show only names of changed files.
825 --name-status
827 Show only names and status of changed files. See the description of
828 the --diff-filter option on what the status letters mean.
829 --submodule[=<format>]
831 Specify how differences in submodules are shown. When specifying
832 --submodule=short the short format is used. This format just shows
833 the names of the commits at the beginning and end of the range.
834 When --submodule or --submodule=log is specified, the log format is
835 used. This format lists the commits in the range like git-
836 submodule(1) summary does. When --submodule=diff is specified, the
837 diff format is used. This format shows an inline diff of the
838 changes in the submodule contents between the commit range.
839 Defaults to diff.submodule or the short format if the config option
840 is unset.
841 --color[=<when>]
843 Show colored diff. --color (i.e. without =<when>) is the same as
844 --color=always. <when> can be one of always, never, or auto.
845 --no-color
847 Turn off colored diff. It is the same as --color=never.
848 --color-moved[=<mode>]
850 Moved lines of code are colored differently. The <mode> defaults to
851 no if the option is not given and to zebra if the option with no
852 mode is given. The mode must be one of:
853 no
855 Moved lines are not highlighted.
856 default
858 Is a synonym for zebra. This may change to a more sensible mode
859 in the future.
860 plain
862 Any line that is added in one location and was removed in
863 another location will be colored with color.diff.newMoved.
864 Similarly color.diff.oldMoved will be used for removed lines
865 that are added somewhere else in the diff. This mode picks up
866 any moved line, but it is not very useful in a review to
867 determine if a block of code was moved without permutation.
868 blocks
870 Blocks of moved text of at least 20 alphanumeric characters are
871 detected greedily. The detected blocks are painted using either
872 the color.diff.{old,new}Moved color. Adjacent blocks cannot be
873 told apart.
874 zebra
876 Blocks of moved text are detected as in blocks mode. The blocks
877 are painted using either the color.diff.{old,new}Moved color or
878 color.diff.{old,new}MovedAlternative. The change between the
879 two colors indicates that a new block was detected.
880 dimmed-zebra
882 Similar to zebra, but additional dimming of uninteresting parts
883 of moved code is performed. The bordering lines of two adjacent
884 blocks are considered interesting, the rest is uninteresting.
885 dimmed_zebra is a deprecated synonym.
886 --no-color-moved
888 Turn off move detection. This can be used to override configuration
889 settings. It is the same as --color-moved=no.
890 --color-moved-ws=<modes>
892 This configures how whitespace is ignored when performing the move
893 detection for --color-moved. These modes can be given as a comma
894 separated list:
895 no
897 Do not ignore whitespace when performing move detection.
898 ignore-space-at-eol
900 Ignore changes in whitespace at EOL.
901 ignore-space-change
903 Ignore changes in amount of whitespace. This ignores whitespace
904 at line end, and considers all other sequences of one or more
905 whitespace characters to be equivalent.
906 ignore-all-space
908 Ignore whitespace when comparing lines. This ignores
909 differences even if one line has whitespace where the other
910 line has none.
911 allow-indentation-change
913 Initially ignore any whitespace in the move detection, then
914 group the moved code blocks only into a block if the change in
915 whitespace is the same per line. This is incompatible with the
916 other modes.
917 --no-color-moved-ws
919 Do not ignore whitespace when performing move detection. This can
920 be used to override configuration settings. It is the same as
921 --color-moved-ws=no.
922 --word-diff[=<mode>]
924 Show a word diff, using the <mode> to delimit changed words. By
925 default, words are delimited by whitespace; see --word-diff-regex
926 below. The <mode> defaults to plain, and must be one of:
927 color
929 Highlight changed words using only colors. Implies --color.
930 plain
932 Show words as [-removed-] and {+added+}. Makes no attempts to
933 escape the delimiters if they appear in the input, so the
934 output may be ambiguous.
935 porcelain
937 Use a special line-based format intended for script
938 consumption. Added/removed/unchanged runs are printed in the
939 usual unified diff format, starting with a +/-/` ` character at
940 the beginning of the line and extending to the end of the line.
941 Newlines in the input are represented by a tilde ~ on a line of
942 its own.
943 none
945 Disable word diff again.
946 Note that despite the name of the first mode, color is used to
948 highlight the changed parts in all modes if enabled.
949 --word-diff-regex=<regex>
951 Use <regex> to decide what a word is, instead of considering runs
952 of non-whitespace to be a word. Also implies --word-diff unless it
953 was already enabled.
954 Every non-overlapping match of the <regex> is considered a word.
956 Anything between these matches is considered whitespace and
957 ignored(!) for the purposes of finding differences. You may want to
958 append |[^[:space:]] to your regular expression to make sure that
959 it matches all non-whitespace characters. A match that contains a
960 newline is silently truncated(!) at the newline.
961 For example, --word-diff-regex=. will treat each character as a
963 word and, correspondingly, show differences character by character.
964 The regex can also be set via a diff driver or configuration
966 option, see gitattributes(5) or git-config(1). Giving it explicitly
967 overrides any diff driver or configuration setting. Diff drivers
968 override configuration settings.
969 --color-words[=<regex>]
971 Equivalent to --word-diff=color plus (if a regex was specified)
972 --word-diff-regex=<regex>.
973 --no-renames
975 Turn off rename detection, even when the configuration file gives
976 the default to do so.
977 --[no-]rename-empty
979 Whether to use empty blobs as rename source.
980 --check
982 Warn if changes introduce conflict markers or whitespace errors.
983 What are considered whitespace errors is controlled by
984 core.whitespace configuration. By default, trailing whitespaces
985 (including lines that consist solely of whitespaces) and a space
986 character that is immediately followed by a tab character inside
987 the initial indent of the line are considered whitespace errors.
988 Exits with non-zero status if problems are found. Not compatible
989 with --exit-code.
990 --ws-error-highlight=<kind>
992 Highlight whitespace errors in the context, old or new lines of the
993 diff. Multiple values are separated by comma, none resets previous
994 values, default reset the list to new and all is a shorthand for
995 old,new,context. When this option is not given, and the
996 configuration variable diff.wsErrorHighlight is not set, only
997 whitespace errors in new lines are highlighted. The whitespace
998 errors are colored with color.diff.whitespace.
999 --full-index
1001 Instead of the first handful of characters, show the full pre- and
1002 post-image blob object names on the "index" line when generating
1003 patch format output.
1004 --binary
1006 In addition to --full-index, output a binary diff that can be
1007 applied with git-apply. Implies --patch.
1008 --abbrev[=<n>]
1010 Instead of showing the full 40-byte hexadecimal object name in
1011 diff-raw format output and diff-tree header lines, show the
1012 shortest prefix that is at least <n> hexdigits long that uniquely
1013 refers the object. In diff-patch output format, --full-index takes
1014 higher precedence, i.e. if --full-index is specified, full blob
1015 names will be shown regardless of --abbrev. Non default number of
1016 digits can be specified with --abbrev=<n>.
1017 -B[<n>][/<m>], --break-rewrites[=[<n>][/<m>]]
1019 Break complete rewrite changes into pairs of delete and create.
1020 This serves two purposes:
1021 It affects the way a change that amounts to a total rewrite of a
1023 file not as a series of deletion and insertion mixed together with
1024 a very few lines that happen to match textually as the context, but
1025 as a single deletion of everything old followed by a single
1026 insertion of everything new, and the number m controls this aspect
1027 of the -B option (defaults to 60%). -B/70% specifies that less
1028 than 30% of the original should remain in the result for Git to
1029 consider it a total rewrite (i.e. otherwise the resulting patch
1030 will be a series of deletion and insertion mixed together with
1031 context lines).
1032 When used with -M, a totally-rewritten file is also considered as
1034 the source of a rename (usually -M only considers a file that
1035 disappeared as the source of a rename), and the number n controls
1036 this aspect of the -B option (defaults to 50%). -B20% specifies
1037 that a change with addition and deletion compared to 20% or more of
1038 the file’s size are eligible for being picked up as a possible
1039 source of a rename to another file.
1040 -M[<n>], --find-renames[=<n>]
1042 If generating diffs, detect and report renames for each commit. For
1043 following files across renames while traversing history, see
1044 --follow. If n is specified, it is a threshold on the similarity
1045 index (i.e. amount of addition/deletions compared to the file’s
1046 size). For example, -M90% means Git should consider a delete/add
1047 pair to be a rename if more than 90% of the file hasn’t changed.
1048 Without a % sign, the number is to be read as a fraction, with a
1049 decimal point before it. I.e., -M5 becomes 0.5, and is thus the
1050 same as -M50%. Similarly, -M05 is the same as -M5%. To limit
1051 detection to exact renames, use -M100%. The default similarity
1052 index is 50%.
1053 -C[<n>], --find-copies[=<n>]
1055 Detect copies as well as renames. See also --find-copies-harder. If
1056 n is specified, it has the same meaning as for -M<n>.
1057 --find-copies-harder
1059 For performance reasons, by default, -C option finds copies only if
1060 the original file of the copy was modified in the same changeset.
1061 This flag makes the command inspect unmodified files as candidates
1062 for the source of copy. This is a very expensive operation for
1063 large projects, so use it with caution. Giving more than one -C
1064 option has the same effect.
1065 -D, --irreversible-delete
1067 Omit the preimage for deletes, i.e. print only the header but not
1068 the diff between the preimage and /dev/null. The resulting patch is
1069 not meant to be applied with patch or git apply; this is solely for
1070 people who want to just concentrate on reviewing the text after the
1071 change. In addition, the output obviously lacks enough information
1072 to apply such a patch in reverse, even manually, hence the name of
1073 the option.
1074 When used together with -B, omit also the preimage in the deletion
1076 part of a delete/create pair.
1077 -l<num>
1079 The -M and -C options require O(n^2) processing time where n is the
1080 number of potential rename/copy targets. This option prevents
1081 rename/copy detection from running if the number of rename/copy
1082 targets exceeds the specified number.
1083 --diff-filter=[(A|C|D|M|R|T|U|X|B)...[*]]
1085 Select only files that are Added (A), Copied (C), Deleted (D),
1086 Modified (M), Renamed (R), have their type (i.e. regular file,
1087 symlink, submodule, ...) changed (T), are Unmerged (U), are Unknown
1088 (X), or have had their pairing Broken (B). Any combination of the
1089 filter characters (including none) can be used. When *
1090 (All-or-none) is added to the combination, all paths are selected
1091 if there is any file that matches other criteria in the comparison;
1092 if there is no file that matches other criteria, nothing is
1093 selected.
1094 Also, these upper-case letters can be downcased to exclude. E.g.
1096 --diff-filter=ad excludes added and deleted paths.
1097 Note that not all diffs can feature all types. For instance, diffs
1099 from the index to the working tree can never have Added entries
1100 (because the set of paths included in the diff is limited by what
1101 is in the index). Similarly, copied and renamed entries cannot
1102 appear if detection for those types is disabled.
1103 -S<string>
1105 Look for differences that change the number of occurrences of the
1106 specified string (i.e. addition/deletion) in a file. Intended for
1107 the scripter’s use.
1108 It is useful when you’re looking for an exact block of code (like a
1110 struct), and want to know the history of that block since it first
1111 came into being: use the feature iteratively to feed the
1112 interesting block in the preimage back into -S, and keep going
1113 until you get the very first version of the block.
1114 Binary files are searched as well.
1116 -G<regex>
1118 Look for differences whose patch text contains added/removed lines
1119 that match <regex>.
1120 To illustrate the difference between -S<regex> --pickaxe-regex and
1122 -G<regex>, consider a commit with the following diff in the same
1123 file:
1124 + return frotz(nitfol, two->ptr, 1, 0);
1126 ...
1127 - hit = frotz(nitfol, mf2.ptr, 1, 0);
1128 While git log -G"frotz\(nitfol" will show this commit, git log
1130 -S"frotz\(nitfol" --pickaxe-regex will not (because the number of
1131 occurrences of that string did not change).
1132 Unless --text is supplied patches of binary files without a
1134 textconv filter will be ignored.
1135 See the pickaxe entry in gitdiffcore(7) for more information.
1137 --find-object=<object-id>
1139 Look for differences that change the number of occurrences of the
1140 specified object. Similar to -S, just the argument is different in
1141 that it doesn’t search for a specific string but for a specific
1142 object id.
1143 The object can be a blob or a submodule commit. It implies the -t
1145 option in git-log to also find trees.
1146 --pickaxe-all
1148 When -S or -G finds a change, show all the changes in that
1149 changeset, not just the files that contain the change in <string>.
1150 --pickaxe-regex
1152 Treat the <string> given to -S as an extended POSIX regular
1153 expression to match.
1154 -O<orderfile>
1156 Control the order in which files appear in the output. This
1157 overrides the diff.orderFile configuration variable (see git-
1158 config(1)). To cancel diff.orderFile, use -O/dev/null.
1159 The output order is determined by the order of glob patterns in
1161 <orderfile>. All files with pathnames that match the first pattern
1162 are output first, all files with pathnames that match the second
1163 pattern (but not the first) are output next, and so on. All files
1164 with pathnames that do not match any pattern are output last, as if
1165 there was an implicit match-all pattern at the end of the file. If
1166 multiple pathnames have the same rank (they match the same pattern
1167 but no earlier patterns), their output order relative to each other
1168 is the normal order.
1169 <orderfile> is parsed as follows:
1171 • Blank lines are ignored, so they can be used as separators for
1173 readability.
1174 • Lines starting with a hash ("#") are ignored, so they can be
1176 used for comments. Add a backslash ("\") to the beginning of
1177 the pattern if it starts with a hash.
1178 • Each other line contains a single pattern.
1180 Patterns have the same syntax and semantics as patterns used for
1182 fnmatch(3) without the FNM_PATHNAME flag, except a pathname also
1183 matches a pattern if removing any number of the final pathname
1184 components matches the pattern. For example, the pattern "foo*bar"
1185 matches "fooasdfbar" and "foo/bar/baz/asdf" but not "foobarx".
1186 --skip-to=<file>, --rotate-to=<file>
1188 Discard the files before the named <file> from the output (i.e.
1189 skip to), or move them to the end of the output (i.e. rotate to).
1190 These were invented primarily for use of the git difftool command,
1191 and may not be very useful otherwise.
1192 -R
1194 Swap two inputs; that is, show differences from index or on-disk
1195 file to tree contents.
1196 --relative[=<path>], --no-relative
1198 When run from a subdirectory of the project, it can be told to
1199 exclude changes outside the directory and show pathnames relative
1200 to it with this option. When you are not in a subdirectory (e.g. in
1201 a bare repository), you can name which subdirectory to make the
1202 output relative to by giving a <path> as an argument.
1203 --no-relative can be used to countermand both diff.relative config
1204 option and previous --relative.
1205 -a, --text
1207 Treat all files as text.
1208 --ignore-cr-at-eol
1210 Ignore carriage-return at the end of line when doing a comparison.
1211 --ignore-space-at-eol
1213 Ignore changes in whitespace at EOL.
1214 -b, --ignore-space-change
1216 Ignore changes in amount of whitespace. This ignores whitespace at
1217 line end, and considers all other sequences of one or more
1218 whitespace characters to be equivalent.
1219 -w, --ignore-all-space
1221 Ignore whitespace when comparing lines. This ignores differences
1222 even if one line has whitespace where the other line has none.
1223 --ignore-blank-lines
1225 Ignore changes whose lines are all blank.
1226 -I<regex>, --ignore-matching-lines=<regex>
1228 Ignore changes whose all lines match <regex>. This option may be
1229 specified more than once.
1230 --inter-hunk-context=<lines>
1232 Show the context between diff hunks, up to the specified number of
1233 lines, thereby fusing hunks that are close to each other. Defaults
1234 to diff.interHunkContext or 0 if the config option is unset.
1235 -W, --function-context
1237 Show whole function as context lines for each change. The function
1238 names are determined in the same way as git diff works out patch
1239 hunk headers (see Defining a custom hunk-header in
1240 gitattributes(5)).
1241 --ext-diff
1243 Allow an external diff helper to be executed. If you set an
1244 external diff driver with gitattributes(5), you need to use this
1245 option with git-log(1) and friends.
1246 --no-ext-diff
1248 Disallow external diff drivers.
1249 --textconv, --no-textconv
1251 Allow (or disallow) external text conversion filters to be run when
1252 comparing binary files. See gitattributes(5) for details. Because
1253 textconv filters are typically a one-way conversion, the resulting
1254 diff is suitable for human consumption, but cannot be applied. For
1255 this reason, textconv filters are enabled by default only for git-
1256 diff(1) and git-log(1), but not for git-format-patch(1) or diff
1257 plumbing commands.
1258 --ignore-submodules[=<when>]
1260 Ignore changes to submodules in the diff generation. <when> can be
1261 either "none", "untracked", "dirty" or "all", which is the default.
1262 Using "none" will consider the submodule modified when it either
1263 contains untracked or modified files or its HEAD differs from the
1264 commit recorded in the superproject and can be used to override any
1265 settings of the ignore option in git-config(1) or gitmodules(5).
1266 When "untracked" is used submodules are not considered dirty when
1267 they only contain untracked content (but they are still scanned for
1268 modified content). Using "dirty" ignores all changes to the work
1269 tree of submodules, only changes to the commits stored in the
1270 superproject are shown (this was the behavior until 1.7.0). Using
1271 "all" hides all changes to submodules.
1272 --src-prefix=<prefix>
1274 Show the given source prefix instead of "a/".
1275 --dst-prefix=<prefix>
1277 Show the given destination prefix instead of "b/".
1278 --no-prefix
1280 Do not show any source or destination prefix.
1281 --line-prefix=<prefix>
1283 Prepend an additional prefix to every line of output.
1284 --ita-invisible-in-index
1286 By default entries added by "git add -N" appear as an existing
1287 empty file in "git diff" and a new file in "git diff --cached".
1288 This option makes the entry appear as a new file in "git diff" and
1289 non-existent in "git diff --cached". This option could be reverted
1290 with --ita-visible-in-index. Both options are experimental and
1291 could be removed in future.
1292 For more detailed explanation on these common options, see also
1294 gitdiffcore(7).
1295
1297 Running git-diff(1), git-log(1), git-show(1), git-diff-index(1), git-
1298 diff-tree(1), or git-diff-files(1) with the -p option produces patch
1299 text. You can customize the creation of patch text via the
1300 GIT_EXTERNAL_DIFF and the GIT_DIFF_OPTS environment variables (see
1301 git(1)).
1302 What the -p option produces is slightly different from the traditional
1304 diff format:
1305 1. It is preceded with a "git diff" header that looks like this:
1307 diff --git a/file1 b/file2
1309 The a/ and b/ filenames are the same unless rename/copy is
1311 involved. Especially, even for a creation or a deletion, /dev/null
1312 is not used in place of the a/ or b/ filenames.
1313 When rename/copy is involved, file1 and file2 show the name of the
1315 source file of the rename/copy and the name of the file that
1316 rename/copy produces, respectively.
1317 2. It is followed by one or more extended header lines:
1319 old mode <mode>
1321 new mode <mode>
1322 deleted file mode <mode>
1323 new file mode <mode>
1324 copy from <path>
1325 copy to <path>
1326 rename from <path>
1327 rename to <path>
1328 similarity index <number>
1329 dissimilarity index <number>
1330 index <hash>..<hash> <mode>
1331 File modes are printed as 6-digit octal numbers including the file
1333 type and file permission bits.
1334 Path names in extended headers do not include the a/ and b/
1336 prefixes.
1337 The similarity index is the percentage of unchanged lines, and the
1339 dissimilarity index is the percentage of changed lines. It is a
1340 rounded down integer, followed by a percent sign. The similarity
1341 index value of 100% is thus reserved for two equal files, while
1342 100% dissimilarity means that no line from the old file made it
1343 into the new one.
1344 The index line includes the blob object names before and after the
1346 change. The <mode> is included if the file mode does not change;
1347 otherwise, separate lines indicate the old and the new mode.
1348 3. Pathnames with "unusual" characters are quoted as explained for the
1350 configuration variable core.quotePath (see git-config(1)).
1351 4. All the file1 files in the output refer to files before the commit,
1353 and all the file2 files refer to files after the commit. It is
1354 incorrect to apply each change to each file sequentially. For
1355 example, this patch will swap a and b:
1356 diff --git a/a b/b
1358 rename from a
1359 rename to b
1360 diff --git a/b b/a
1361 rename from b
1362 rename to a
1363
1365 Any diff-generating command can take the -c or --cc option to produce a
1366 combined diff when showing a merge. This is the default format when
1367 showing merges with git-diff(1) or git-show(1). Note also that you can
1368 give suitable --diff-merges option to any of these commands to force
1369 generation of diffs in specific format.
1370 A "combined diff" format looks like this:
1372 diff --combined describe.c
1374 index fabadb8,cc95eb0..4866510
1375 --- a/describe.c
1376 +++ b/describe.c
1377 @@@ -98,20 -98,12 +98,20 @@@
1378 return (a_date > b_date) ? -1 : (a_date == b_date) ? 0 : 1;
1379 }
1380 - static void describe(char *arg)
1382 -static void describe(struct commit *cmit, int last_one)
1383 ++static void describe(char *arg, int last_one)
1384 {
1385 + unsigned char sha1[20];
1386 + struct commit *cmit;
1387 struct commit_list *list;
1388 static int initialized = 0;
1389 struct commit_name *n;
1390 + if (get_sha1(arg, sha1) < 0)
1392 + usage(describe_usage);
1393 + cmit = lookup_commit_reference(sha1);
1394 + if (!cmit)
1395 + usage(describe_usage);
1396 +
1397 if (!initialized) {
1398 initialized = 1;
1399 for_each_ref(get_name);
1400 1. It is preceded with a "git diff" header, that looks like this (when
1402 the -c option is used):
1403 diff --combined file
1405 or like this (when the --cc option is used):
1407 diff --cc file
1409 2. It is followed by one or more extended header lines (this example
1411 shows a merge with two parents):
1412 index <hash>,<hash>..<hash>
1414 mode <mode>,<mode>..<mode>
1415 new file mode <mode>
1416 deleted file mode <mode>,<mode>
1417 The mode <mode>,<mode>..<mode> line appears only if at least one of
1419 the <mode> is different from the rest. Extended headers with
1420 information about detected contents movement (renames and copying
1421 detection) are designed to work with diff of two <tree-ish> and are
1422 not used by combined diff format.
1423 3. It is followed by two-line from-file/to-file header
1425 --- a/file
1427 +++ b/file
1428 Similar to two-line header for traditional unified diff format,
1430 /dev/null is used to signal created or deleted files.
1431 However, if the --combined-all-paths option is provided, instead of
1433 a two-line from-file/to-file you get a N+1 line from-file/to-file
1434 header, where N is the number of parents in the merge commit
1435 --- a/file
1437 --- a/file
1438 --- a/file
1439 +++ b/file
1440 This extended format can be useful if rename or copy detection is
1442 active, to allow you to see the original name of the file in
1443 different parents.
1444 4. Chunk header format is modified to prevent people from accidentally
1446 feeding it to patch -p1. Combined diff format was created for
1447 review of merge commit changes, and was not meant to be applied.
1448 The change is similar to the change in the extended index header:
1449 @@@ <from-file-range> <from-file-range> <to-file-range> @@@
1451 There are (number of parents + 1) @ characters in the chunk header
1453 for combined diff format.
1454 Unlike the traditional unified diff format, which shows two files A and
1456 B with a single column that has - (minus — appears in A but removed in
1457 B), + (plus — missing in A but added to B), or " " (space — unchanged)
1458 prefix, this format compares two or more files file1, file2,... with
1459 one file X, and shows how X differs from each of fileN. One column for
1460 each of fileN is prepended to the output line to note how X’s line is
1461 different from it.
1462 A - character in the column N means that the line appears in fileN but
1464 it does not appear in the result. A + character in the column N means
1465 that the line appears in the result, and fileN does not have that line
1466 (in other words, the line was added, from the point of view of that
1467 parent).
1468 In the above example output, the function signature was changed from
1470 both files (hence two - removals from both file1 and file2, plus ++ to
1471 mean one line that was added does not appear in either file1 or file2).
1472 Also eight other lines are the same from file1 but do not appear in
1473 file2 (hence prefixed with +).
1474 When shown by git diff-tree -c, it compares the parents of a merge
1476 commit with the merge result (i.e. file1..fileN are the parents). When
1477 shown by git diff-files -c, it compares the two unresolved merge
1478 parents with the working tree file (i.e. file1 is stage 2 aka "our
1479 version", file2 is stage 3 aka "their version").
1480
1482 git show v1.0.0
1483 Shows the tag v1.0.0, along with the object the tags points at.
1484 git show v1.0.0^{tree}
1486 Shows the tree pointed to by the tag v1.0.0.
1487 git show -s --format=%s v1.0.0^{commit}
1489 Shows the subject of the commit pointed to by the tag v1.0.0.
1490 git show next~10:Documentation/README
1492 Shows the contents of the file Documentation/README as they were
1493 current in the 10th last commit of the branch next.
1494 git show master:Makefile master:t/Makefile
1496 Concatenates the contents of said Makefiles in the head of the
1497 branch master.
1498
1500 Git is to some extent character encoding agnostic.
1501 • The contents of the blob objects are uninterpreted sequences of
1503 bytes. There is no encoding translation at the core level.
1504 • Path names are encoded in UTF-8 normalization form C. This applies
1506 to tree objects, the index file, ref names, as well as path names
1507 in command line arguments, environment variables and config files
1508 (.git/config (see git-config(1)), gitignore(5), gitattributes(5)
1509 and gitmodules(5)).
1510 Note that Git at the core level treats path names simply as
1512 sequences of non-NUL bytes, there are no path name encoding
1513 conversions (except on Mac and Windows). Therefore, using non-ASCII
1514 path names will mostly work even on platforms and file systems that
1515 use legacy extended ASCII encodings. However, repositories created
1516 on such systems will not work properly on UTF-8-based systems (e.g.
1517 Linux, Mac, Windows) and vice versa. Additionally, many Git-based
1518 tools simply assume path names to be UTF-8 and will fail to display
1519 other encodings correctly.
1520 • Commit log messages are typically encoded in UTF-8, but other
1522 extended ASCII encodings are also supported. This includes
1523 ISO-8859-x, CP125x and many others, but not UTF-16/32, EBCDIC and
1524 CJK multi-byte encodings (GBK, Shift-JIS, Big5, EUC-x, CP9xx etc.).
1525 Although we encourage that the commit log messages are encoded in
1527 UTF-8, both the core and Git Porcelain are designed not to force UTF-8
1528 on projects. If all participants of a particular project find it more
1529 convenient to use legacy encodings, Git does not forbid it. However,
1530 there are a few things to keep in mind.
1531 1. git commit and git commit-tree issues a warning if the commit log
1533 message given to it does not look like a valid UTF-8 string, unless
1534 you explicitly say your project uses a legacy encoding. The way to
1535 say this is to have i18n.commitEncoding in .git/config file, like
1536 this:
1537 [i18n]
1539 commitEncoding = ISO-8859-1
1540 Commit objects created with the above setting record the value of
1542 i18n.commitEncoding in its encoding header. This is to help other
1543 people who look at them later. Lack of this header implies that the
1544 commit log message is encoded in UTF-8.
1545 2. git log, git show, git blame and friends look at the encoding
1547 header of a commit object, and try to re-code the log message into
1548 UTF-8 unless otherwise specified. You can specify the desired
1549 output encoding with i18n.logOutputEncoding in .git/config file,
1550 like this:
1551 [i18n]
1553 logOutputEncoding = ISO-8859-1
1554 If you do not have this configuration variable, the value of
1556 i18n.commitEncoding is used instead.
1557 Note that we deliberately chose not to re-code the commit log message
1559 when a commit is made to force UTF-8 at the commit object level,
1560 because re-coding to UTF-8 is not necessarily a reversible operation.
1561
1563 Part of the git(1) suite
1564Git 2.31.1 2021-03-26 GIT-SHOW(1)