1GIT-REV-LIST(1) Git Manual GIT-REV-LIST(1)
2
6 git-rev-list - Lists commit objects in reverse chronological order
7
9 git rev-list [<options>] <commit>... [[--] <path>...]
10
12 List commits that are reachable by following the parent links from the
13 given commit(s), but exclude commits that are reachable from the one(s)
14 given with a ^ in front of them. The output is given in reverse
15 chronological order by default.
16 You can think of this as a set operation. Commits reachable from any of
18 the commits given on the command line form a set, and then commits
19 reachable from any of the ones given with ^ in front are subtracted
20 from that set. The remaining commits are what comes out in the
21 command’s output. Various other options and paths parameters can be
22 used to further limit the result.
23 Thus, the following command:
25 $ git rev-list foo bar ^baz
27 means "list all the commits which are reachable from foo or bar, but
29 not from baz".
30 A special notation "<commit1>..<commit2>" can be used as a short-hand
32 for "^<commit1> <commit2>". For example, either of the following may be
33 used interchangeably:
34 $ git rev-list origin..HEAD
36 $ git rev-list HEAD ^origin
37 Another special notation is "<commit1>...<commit2>" which is useful for
39 merges. The resulting set of commits is the symmetric difference
40 between the two operands. The following two commands are equivalent:
41 $ git rev-list A B --not $(git merge-base --all A B)
43 $ git rev-list A...B
44 rev-list is a very essential Git command, since it provides the ability
46 to build and traverse commit ancestry graphs. For this reason, it has a
47 lot of different options that enables it to be used by commands as
48 different as git bisect and git repack.
49
51 Commit Limiting
52 Besides specifying a range of commits that should be listed using the
53 special notations explained in the description, additional commit
54 limiting may be applied.
55 Using more options generally further limits the output (e.g.
57 --since=<date1> limits to commits newer than <date1>, and using it with
58 --grep=<pattern> further limits to commits whose log message has a line
59 that matches <pattern>), unless otherwise noted.
60 Note that these are applied before commit ordering and formatting
62 options, such as --reverse.
63 -<number>, -n <number>, --max-count=<number>
65 Limit the number of commits to output.
66 --skip=<number>
68 Skip number commits before starting to show the commit output.
69 --since=<date>, --after=<date>
71 Show commits more recent than a specific date.
72 --until=<date>, --before=<date>
74 Show commits older than a specific date.
75 --max-age=<timestamp>, --min-age=<timestamp>
77 Limit the commits output to specified time range.
78 --author=<pattern>, --committer=<pattern>
80 Limit the commits output to ones with author/committer header lines
81 that match the specified pattern (regular expression). With more
82 than one --author=<pattern>, commits whose author matches any of
83 the given patterns are chosen (similarly for multiple
84 --committer=<pattern>).
85 --grep-reflog=<pattern>
87 Limit the commits output to ones with reflog entries that match the
88 specified pattern (regular expression). With more than one
89 --grep-reflog, commits whose reflog message matches any of the
90 given patterns are chosen. It is an error to use this option unless
91 --walk-reflogs is in use.
92 --grep=<pattern>
94 Limit the commits output to ones with log message that matches the
95 specified pattern (regular expression). With more than one
96 --grep=<pattern>, commits whose message matches any of the given
97 patterns are chosen (but see --all-match).
98 --all-match
100 Limit the commits output to ones that match all given --grep,
101 instead of ones that match at least one.
102 --invert-grep
104 Limit the commits output to ones with log message that do not match
105 the pattern specified with --grep=<pattern>.
106 -i, --regexp-ignore-case
108 Match the regular expression limiting patterns without regard to
109 letter case.
110 --basic-regexp
112 Consider the limiting patterns to be basic regular expressions;
113 this is the default.
114 -E, --extended-regexp
116 Consider the limiting patterns to be extended regular expressions
117 instead of the default basic regular expressions.
118 -F, --fixed-strings
120 Consider the limiting patterns to be fixed strings (don’t interpret
121 pattern as a regular expression).
122 -P, --perl-regexp
124 Consider the limiting patterns to be Perl-compatible regular
125 expressions.
126 Support for these types of regular expressions is an optional
128 compile-time dependency. If Git wasn’t compiled with support for
129 them providing this option will cause it to die.
130 --remove-empty
132 Stop when a given path disappears from the tree.
133 --merges
135 Print only merge commits. This is exactly the same as
136 --min-parents=2.
137 --no-merges
139 Do not print commits with more than one parent. This is exactly the
140 same as --max-parents=1.
141 --min-parents=<number>, --max-parents=<number>, --no-min-parents,
143 --no-max-parents
144 Show only commits which have at least (or at most) that many parent
145 commits. In particular, --max-parents=1 is the same as --no-merges,
146 --min-parents=2 is the same as --merges. --max-parents=0 gives all
147 root commits and --min-parents=3 all octopus merges.
148 --no-min-parents and --no-max-parents reset these limits (to no
150 limit) again. Equivalent forms are --min-parents=0 (any commit has
151 0 or more parents) and --max-parents=-1 (negative numbers denote no
152 upper limit).
153 --first-parent
155 Follow only the first parent commit upon seeing a merge commit.
156 This option can give a better overview when viewing the evolution
157 of a particular topic branch, because merges into a topic branch
158 tend to be only about adjusting to updated upstream from time to
159 time, and this option allows you to ignore the individual commits
160 brought in to your history by such a merge.
161 --not
163 Reverses the meaning of the ^ prefix (or lack thereof) for all
164 following revision specifiers, up to the next --not.
165 --all
167 Pretend as if all the refs in refs/, along with HEAD, are listed on
168 the command line as <commit>.
169 --branches[=<pattern>]
171 Pretend as if all the refs in refs/heads are listed on the command
172 line as <commit>. If <pattern> is given, limit branches to ones
173 matching given shell glob. If pattern lacks ?, *, or [, /* at the
174 end is implied.
175 --tags[=<pattern>]
177 Pretend as if all the refs in refs/tags are listed on the command
178 line as <commit>. If <pattern> is given, limit tags to ones
179 matching given shell glob. If pattern lacks ?, *, or [, /* at the
180 end is implied.
181 --remotes[=<pattern>]
183 Pretend as if all the refs in refs/remotes are listed on the
184 command line as <commit>. If <pattern> is given, limit
185 remote-tracking branches to ones matching given shell glob. If
186 pattern lacks ?, *, or [, /* at the end is implied.
187 --glob=<glob-pattern>
189 Pretend as if all the refs matching shell glob <glob-pattern> are
190 listed on the command line as <commit>. Leading refs/, is
191 automatically prepended if missing. If pattern lacks ?, *, or [, /*
192 at the end is implied.
193 --exclude=<glob-pattern>
195 Do not include refs matching <glob-pattern> that the next --all,
196 --branches, --tags, --remotes, or --glob would otherwise consider.
197 Repetitions of this option accumulate exclusion patterns up to the
198 next --all, --branches, --tags, --remotes, or --glob option (other
199 options or arguments do not clear accumulated patterns).
200 The patterns given should not begin with refs/heads, refs/tags, or
202 refs/remotes when applied to --branches, --tags, or --remotes,
203 respectively, and they must begin with refs/ when applied to --glob
204 or --all. If a trailing /* is intended, it must be given
205 explicitly.
206 --reflog
208 Pretend as if all objects mentioned by reflogs are listed on the
209 command line as <commit>.
210 --alternate-refs
212 Pretend as if all objects mentioned as ref tips of alternate
213 repositories were listed on the command line. An alternate
214 repository is any repository whose object directory is specified in
215 objects/info/alternates. The set of included objects may be
216 modified by core.alternateRefsCommand, etc. See git-config(1).
217 --single-worktree
219 By default, all working trees will be examined by the following
220 options when there are more than one (see git-worktree(1)): --all,
221 --reflog and --indexed-objects. This option forces them to examine
222 the current working tree only.
223 --ignore-missing
225 Upon seeing an invalid object name in the input, pretend as if the
226 bad input was not given.
227 --stdin
229 In addition to the <commit> listed on the command line, read them
230 from the standard input. If a -- separator is seen, stop reading
231 commits and start reading paths to limit the result.
232 --quiet
234 Don’t print anything to standard output. This form is primarily
235 meant to allow the caller to test the exit status to see if a range
236 of objects is fully connected (or not). It is faster than
237 redirecting stdout to /dev/null as the output does not have to be
238 formatted.
239 --cherry-mark
241 Like --cherry-pick (see below) but mark equivalent commits with =
242 rather than omitting them, and inequivalent ones with +.
243 --cherry-pick
245 Omit any commit that introduces the same change as another commit
246 on the “other side” when the set of commits are limited with
247 symmetric difference.
248 For example, if you have two branches, A and B, a usual way to list
250 all commits on only one side of them is with --left-right (see the
251 example below in the description of the --left-right option).
252 However, it shows the commits that were cherry-picked from the
253 other branch (for example, “3rd on b” may be cherry-picked from
254 branch A). With this option, such pairs of commits are excluded
255 from the output.
256 --left-only, --right-only
258 List only commits on the respective side of a symmetric difference,
259 i.e. only those which would be marked < resp. > by --left-right.
260 For example, --cherry-pick --right-only A...B omits those commits
262 from B which are in A or are patch-equivalent to a commit in A. In
263 other words, this lists the + commits from git cherry A B. More
264 precisely, --cherry-pick --right-only --no-merges gives the exact
265 list.
266 --cherry
268 A synonym for --right-only --cherry-mark --no-merges; useful to
269 limit the output to the commits on our side and mark those that
270 have been applied to the other side of a forked history with git
271 log --cherry upstream...mybranch, similar to git cherry upstream
272 mybranch.
273 -g, --walk-reflogs
275 Instead of walking the commit ancestry chain, walk reflog entries
276 from the most recent one to older ones. When this option is used
277 you cannot specify commits to exclude (that is, ^commit,
278 commit1..commit2, and commit1...commit2 notations cannot be used).
279 With --pretty format other than oneline and reference (for obvious
281 reasons), this causes the output to have two extra lines of
282 information taken from the reflog. The reflog designator in the
283 output may be shown as ref@{Nth} (where Nth is the
284 reverse-chronological index in the reflog) or as ref@{timestamp}
285 (with the timestamp for that entry), depending on a few rules:
286 1. If the starting point is specified as ref@{Nth}, show the index
288 format.
289 2. If the starting point was specified as ref@{now}, show the
291 timestamp format.
292 3. If neither was used, but --date was given on the command line,
294 show the timestamp in the format requested by --date.
295 4. Otherwise, show the index format.
297 Under --pretty=oneline, the commit message is prefixed with this
299 information on the same line. This option cannot be combined with
300 --reverse. See also git-reflog(1).
301 Under --pretty=reference, this information will not be shown at
303 all.
304 --merge
306 After a failed merge, show refs that touch files having a conflict
307 and don’t exist on all heads to merge.
308 --boundary
310 Output excluded boundary commits. Boundary commits are prefixed
311 with -.
312 --use-bitmap-index
314 Try to speed up the traversal using the pack bitmap index (if one
315 is available). Note that when traversing with --objects, trees and
316 blobs will not have their associated path printed.
317 --progress=<header>
319 Show progress reports on stderr as objects are considered. The
320 <header> text will be printed with each progress update.
321 History Simplification
323 Sometimes you are only interested in parts of the history, for example
324 the commits modifying a particular <path>. But there are two parts of
325 History Simplification, one part is selecting the commits and the other
326 is how to do it, as there are various strategies to simplify the
327 history.
328 The following options select the commits to be shown:
330 <paths>
332 Commits modifying the given <paths> are selected.
333 --simplify-by-decoration
335 Commits that are referred by some branch or tag are selected.
336 Note that extra commits can be shown to give a meaningful history.
338 The following options affect the way the simplification is performed:
340 Default mode
342 Simplifies the history to the simplest history explaining the final
343 state of the tree. Simplest because it prunes some side branches if
344 the end result is the same (i.e. merging branches with the same
345 content)
346 --show-pulls
348 Include all commits from the default mode, but also any merge
349 commits that are not TREESAME to the first parent but are TREESAME
350 to a later parent. This mode is helpful for showing the merge
351 commits that "first introduced" a change to a branch.
352 --full-history
354 Same as the default mode, but does not prune some history.
355 --dense
357 Only the selected commits are shown, plus some to have a meaningful
358 history.
359 --sparse
361 All commits in the simplified history are shown.
362 --simplify-merges
364 Additional option to --full-history to remove some needless merges
365 from the resulting history, as there are no selected commits
366 contributing to this merge.
367 --ancestry-path
369 When given a range of commits to display (e.g. commit1..commit2 or
370 commit2 ^commit1), only display commits that exist directly on the
371 ancestry chain between the commit1 and commit2, i.e. commits that
372 are both descendants of commit1, and ancestors of commit2.
373 A more detailed explanation follows.
375 Suppose you specified foo as the <paths>. We shall call commits that
377 modify foo !TREESAME, and the rest TREESAME. (In a diff filtered for
378 foo, they look different and equal, respectively.)
379 In the following, we will always refer to the same example history to
381 illustrate the differences between simplification settings. We assume
382 that you are filtering for a file foo in this commit graph:
383 .-A---M---N---O---P---Q
385 / / / / / /
386 I B C D E Y
387 \ / / / / /
388 `-------------' X
389 The horizontal line of history A---Q is taken to be the first parent of
391 each merge. The commits are:
392 · I is the initial commit, in which foo exists with contents “asdf”,
394 and a file quux exists with contents “quux”. Initial commits are
395 compared to an empty tree, so I is !TREESAME.
396 · In A, foo contains just “foo”.
398 · B contains the same change as A. Its merge M is trivial and hence
400 TREESAME to all parents.
401 · C does not change foo, but its merge N changes it to “foobar”, so
403 it is not TREESAME to any parent.
404 · D sets foo to “baz”. Its merge O combines the strings from N and D
406 to “foobarbaz”; i.e., it is not TREESAME to any parent.
407 · E changes quux to “xyzzy”, and its merge P combines the strings to
409 “quux xyzzy”. P is TREESAME to O, but not to E.
410 · X is an independent root commit that added a new file side, and Y
412 modified it. Y is TREESAME to X. Its merge Q added side to P, and
413 Q is TREESAME to P, but not to Y.
414 rev-list walks backwards through history, including or excluding
416 commits based on whether --full-history and/or parent rewriting (via
417 --parents or --children) are used. The following settings are
418 available.
419 Default mode
421 Commits are included if they are not TREESAME to any parent (though
422 this can be changed, see --sparse below). If the commit was a
423 merge, and it was TREESAME to one parent, follow only that parent.
424 (Even if there are several TREESAME parents, follow only one of
425 them.) Otherwise, follow all parents.
426 This results in:
428 .-A---N---O
430 / / /
431 I---------D
432 Note how the rule to only follow the TREESAME parent, if one is
434 available, removed B from consideration entirely. C was considered
435 via N, but is TREESAME. Root commits are compared to an empty tree,
436 so I is !TREESAME.
437 Parent/child relations are only visible with --parents, but that
439 does not affect the commits selected in default mode, so we have
440 shown the parent lines.
441 --full-history without parent rewriting
443 This mode differs from the default in one point: always follow all
444 parents of a merge, even if it is TREESAME to one of them. Even if
445 more than one side of the merge has commits that are included, this
446 does not imply that the merge itself is! In the example, we get
447 I A B N D O P Q
449 M was excluded because it is TREESAME to both parents. E, C and B
451 were all walked, but only B was !TREESAME, so the others do not
452 appear.
453 Note that without parent rewriting, it is not really possible to
455 talk about the parent/child relationships between the commits, so
456 we show them disconnected.
457 --full-history with parent rewriting
459 Ordinary commits are only included if they are !TREESAME (though
460 this can be changed, see --sparse below).
461 Merges are always included. However, their parent list is
463 rewritten: Along each parent, prune away commits that are not
464 included themselves. This results in
465 .-A---M---N---O---P---Q
467 / / / / /
468 I B / D /
469 \ / / / /
470 `-------------'
471 Compare to --full-history without rewriting above. Note that E was
473 pruned away because it is TREESAME, but the parent list of P was
474 rewritten to contain E's parent I. The same happened for C and N,
475 and X, Y and Q.
476 In addition to the above settings, you can change whether TREESAME
478 affects inclusion:
479 --dense
481 Commits that are walked are included if they are not TREESAME to
482 any parent.
483 --sparse
485 All commits that are walked are included.
486 Note that without --full-history, this still simplifies merges: if
488 one of the parents is TREESAME, we follow only that one, so the
489 other sides of the merge are never walked.
490 --simplify-merges
492 First, build a history graph in the same way that --full-history
493 with parent rewriting does (see above).
494 Then simplify each commit C to its replacement C' in the final
496 history according to the following rules:
497 · Set C' to C.
499 · Replace each parent P of C' with its simplification P'. In the
501 process, drop parents that are ancestors of other parents or
502 that are root commits TREESAME to an empty tree, and remove
503 duplicates, but take care to never drop all parents that we are
504 TREESAME to.
505 · If after this parent rewriting, C' is a root or merge commit
507 (has zero or >1 parents), a boundary commit, or !TREESAME, it
508 remains. Otherwise, it is replaced with its only parent.
509 The effect of this is best shown by way of comparing to
511 --full-history with parent rewriting. The example turns into:
512 .-A---M---N---O
514 / / /
515 I B D
516 \ / /
517 `---------'
518 Note the major differences in N, P, and Q over --full-history:
520 · N's parent list had I removed, because it is an ancestor of the
522 other parent M. Still, N remained because it is !TREESAME.
523 · P's parent list similarly had I removed. P was then removed
525 completely, because it had one parent and is TREESAME.
526 · Q's parent list had Y simplified to X. X was then removed,
528 because it was a TREESAME root. Q was then removed completely,
529 because it had one parent and is TREESAME.
530 There is another simplification mode available:
532 --ancestry-path
534 Limit the displayed commits to those directly on the ancestry chain
535 between the “from” and “to” commits in the given commit range. I.e.
536 only display commits that are ancestor of the “to” commit and
537 descendants of the “from” commit.
538 As an example use case, consider the following commit history:
540 D---E-------F
542 / \ \
543 B---C---G---H---I---J
544 / \
545 A-------K---------------L--M
546 A regular D..M computes the set of commits that are ancestors of M,
548 but excludes the ones that are ancestors of D. This is useful to
549 see what happened to the history leading to M since D, in the sense
550 that “what does M have that did not exist in D”. The result in this
551 example would be all the commits, except A and B (and D itself, of
552 course).
553 When we want to find out what commits in M are contaminated with
555 the bug introduced by D and need fixing, however, we might want to
556 view only the subset of D..M that are actually descendants of D,
557 i.e. excluding C and K. This is exactly what the --ancestry-path
558 option does. Applied to the D..M range, it results in:
559 E-------F
561 \ \
562 G---H---I---J
563 \
564 L--M
565 Before discussing another option, --show-pulls, we need to create a new
567 example history.
568 A common problem users face when looking at simplified history is that
570 a commit they know changed a file somehow does not appear in the file’s
571 simplified history. Let’s demonstrate a new example and show how
572 options such as --full-history and --simplify-merges works in that
573 case:
574 .-A---M-----C--N---O---P
576 / / \ \ \/ / /
577 I B \ R-'`-Z' /
578 \ / \/ /
579 \ / /\ /
580 `---X--' `---Y--'
581 For this example, suppose I created file.txt which was modified by A,
583 B, and X in different ways. The single-parent commits C, Z, and Y do
584 not change file.txt. The merge commit M was created by resolving the
585 merge conflict to include both changes from A and B and hence is not
586 TREESAME to either. The merge commit R, however, was created by
587 ignoring the contents of file.txt at M and taking only the contents of
588 file.txt at X. Hence, R is TREESAME to X but not M. Finally, the
589 natural merge resolution to create N is to take the contents of
590 file.txt at R, so N is TREESAME to R but not C. The merge commits O and
591 P are TREESAME to their first parents, but not to their second parents,
592 Z and Y respectively.
593 When using the default mode, N and R both have a TREESAME parent, so
595 those edges are walked and the others are ignored. The resulting
596 history graph is:
597 I---X
599 When using --full-history, Git walks every edge. This will discover the
601 commits A and B and the merge M, but also will reveal the merge commits
602 O and P. With parent rewriting, the resulting graph is:
603 .-A---M--------N---O---P
605 / / \ \ \/ / /
606 I B \ R-'`--' /
607 \ / \/ /
608 \ / /\ /
609 `---X--' `------'
610 Here, the merge commits O and P contribute extra noise, as they did not
612 actually contribute a change to file.txt. They only merged a topic that
613 was based on an older version of file.txt. This is a common issue in
614 repositories using a workflow where many contributors work in parallel
615 and merge their topic branches along a single trunk: manu unrelated
616 merges appear in the --full-history results.
617 When using the --simplify-merges option, the commits O and P disappear
619 from the results. This is because the rewritten second parents of O and
620 P are reachable from their first parents. Those edges are removed and
621 then the commits look like single-parent commits that are TREESAME to
622 their parent. This also happens to the commit N, resulting in a history
623 view as follows:
624 .-A---M--.
626 / / \
627 I B R
628 \ / /
629 \ / /
630 `---X--'
631 In this view, we see all of the important single-parent changes from A,
633 B, and X. We also see the carefully-resolved merge M and the
634 not-so-carefully-resolved merge R. This is usually enough information
635 to determine why the commits A and B "disappeared" from history in the
636 default view. However, there are a few issues with this approach.
637 The first issue is performance. Unlike any previous option, the
639 --simplify-merges option requires walking the entire commit history
640 before returning a single result. This can make the option difficult to
641 use for very large repositories.
642 The second issue is one of auditing. When many contributors are working
644 on the same repository, it is important which merge commits introduced
645 a change into an important branch. The problematic merge R above is not
646 likely to be the merge commit that was used to merge into an important
647 branch. Instead, the merge N was used to merge R and X into the
648 important branch. This commit may have information about why the change
649 X came to override the changes from A and B in its commit message.
650 --show-pulls
652 In addition to the commits shown in the default history, show each
653 merge commit that is not TREESAME to its first parent but is
654 TREESAME to a later parent.
655 When a merge commit is included by --show-pulls, the merge is
657 treated as if it "pulled" the change from another branch. When
658 using --show-pulls on this example (and no other options) the
659 resulting graph is:
660 I---X---R---N
662 Here, the merge commits R and N are included because they pulled
664 the commits X and R into the base branch, respectively. These
665 merges are the reason the commits A and B do not appear in the
666 default history.
667 When --show-pulls is paired with --simplify-merges, the graph
669 includes all of the necessary information:
670 .-A---M--. N
672 / / \ /
673 I B R
674 \ / /
675 \ / /
676 `---X--'
677 Notice that since M is reachable from R, the edge from N to M was
679 simplified away. However, N still appears in the history as an
680 important commit because it "pulled" the change R into the main
681 branch.
682 The --simplify-by-decoration option allows you to view only the big
684 picture of the topology of the history, by omitting commits that are
685 not referenced by tags. Commits are marked as !TREESAME (in other
686 words, kept after history simplification rules described above) if (1)
687 they are referenced by tags, or (2) they change the contents of the
688 paths given on the command line. All other commits are marked as
689 TREESAME (subject to be simplified away).
690 Bisection Helpers
692 --bisect
693 Limit output to the one commit object which is roughly halfway
694 between included and excluded commits. Note that the bad bisection
695 ref refs/bisect/bad is added to the included commits (if it exists)
696 and the good bisection refs refs/bisect/good-* are added to the
697 excluded commits (if they exist). Thus, supposing there are no refs
698 in refs/bisect/, if
699 $ git rev-list --bisect foo ^bar ^baz
701 outputs midpoint, the output of the two commands
703 $ git rev-list foo ^midpoint
705 $ git rev-list midpoint ^bar ^baz
706 would be of roughly the same length. Finding the change which
708 introduces a regression is thus reduced to a binary search:
709 repeatedly generate and test new 'midpoint’s until the commit chain
710 is of length one.
711 --bisect-vars
713 This calculates the same as --bisect, except that refs in
714 refs/bisect/ are not used, and except that this outputs text ready
715 to be eval’ed by the shell. These lines will assign the name of the
716 midpoint revision to the variable bisect_rev, and the expected
717 number of commits to be tested after bisect_rev is tested to
718 bisect_nr, the expected number of commits to be tested if
719 bisect_rev turns out to be good to bisect_good, the expected number
720 of commits to be tested if bisect_rev turns out to be bad to
721 bisect_bad, and the number of commits we are bisecting right now to
722 bisect_all.
723 --bisect-all
725 This outputs all the commit objects between the included and
726 excluded commits, ordered by their distance to the included and
727 excluded commits. Refs in refs/bisect/ are not used. The farthest
728 from them is displayed first. (This is the only one displayed by
729 --bisect.)
730 This is useful because it makes it easy to choose a good commit to
732 test when you want to avoid to test some of them for some reason
733 (they may not compile for example).
734 This option can be used along with --bisect-vars, in this case,
736 after all the sorted commit objects, there will be the same text as
737 if --bisect-vars had been used alone.
738 Commit Ordering
740 By default, the commits are shown in reverse chronological order.
741 --date-order
743 Show no parents before all of its children are shown, but otherwise
744 show commits in the commit timestamp order.
745 --author-date-order
747 Show no parents before all of its children are shown, but otherwise
748 show commits in the author timestamp order.
749 --topo-order
751 Show no parents before all of its children are shown, and avoid
752 showing commits on multiple lines of history intermixed.
753 For example, in a commit history like this:
755 ---1----2----4----7
757 \ \
758 3----5----6----8---
759 where the numbers denote the order of commit timestamps, git
761 rev-list and friends with --date-order show the commits in the
762 timestamp order: 8 7 6 5 4 3 2 1.
763 With --topo-order, they would show 8 6 5 3 7 4 2 1 (or 8 7 4 2 6 5
765 3 1); some older commits are shown before newer ones in order to
766 avoid showing the commits from two parallel development track mixed
767 together.
768 --reverse
770 Output the commits chosen to be shown (see Commit Limiting section
771 above) in reverse order. Cannot be combined with --walk-reflogs.
772 Object Traversal
774 These options are mostly targeted for packing of Git repositories.
775 --objects
777 Print the object IDs of any object referenced by the listed
778 commits. --objects foo ^bar thus means “send me all object IDs
779 which I need to download if I have the commit object bar but not
780 foo”.
781 --in-commit-order
783 Print tree and blob ids in order of the commits. The tree and blob
784 ids are printed after they are first referenced by a commit.
785 --objects-edge
787 Similar to --objects, but also print the IDs of excluded commits
788 prefixed with a “-” character. This is used by git-pack-objects(1)
789 to build a “thin” pack, which records objects in deltified form
790 based on objects contained in these excluded commits to reduce
791 network traffic.
792 --objects-edge-aggressive
794 Similar to --objects-edge, but it tries harder to find excluded
795 commits at the cost of increased time. This is used instead of
796 --objects-edge to build “thin” packs for shallow repositories.
797 --indexed-objects
799 Pretend as if all trees and blobs used by the index are listed on
800 the command line. Note that you probably want to use --objects,
801 too.
802 --unpacked
804 Only useful with --objects; print the object IDs that are not in
805 packs.
806 --object-names
808 Only useful with --objects; print the names of the object IDs that
809 are found. This is the default behavior.
810 --no-object-names
812 Only useful with --objects; does not print the names of the object
813 IDs that are found. This inverts --object-names. This flag allows
814 the output to be more easily parsed by commands such as git-cat-
815 file(1).
816 --filter=<filter-spec>
818 Only useful with one of the --objects*; omits objects (usually
819 blobs) from the list of printed objects. The <filter-spec> may be
820 one of the following:
821 The form --filter=blob:none omits all blobs.
823 The form --filter=blob:limit=<n>[kmg] omits blobs larger than n
825 bytes or units. n may be zero. The suffixes k, m, and g can be used
826 to name units in KiB, MiB, or GiB. For example, blob:limit=1k is
827 the same as blob:limit=1024.
828 The form --filter=sparse:oid=<blob-ish> uses a sparse-checkout
830 specification contained in the blob (or blob-expression) <blob-ish>
831 to omit blobs that would not be not required for a sparse checkout
832 on the requested refs.
833 The form --filter=tree:<depth> omits all blobs and trees whose
835 depth from the root tree is >= <depth> (minimum depth if an object
836 is located at multiple depths in the commits traversed). <depth>=0
837 will not include any trees or blobs unless included explicitly in
838 the command-line (or standard input when --stdin is used).
839 <depth>=1 will include only the tree and blobs which are referenced
840 directly by a commit reachable from <commit> or an explicitly-given
841 object. <depth>=2 is like <depth>=1 while also including trees and
842 blobs one more level removed from an explicitly-given commit or
843 tree.
844 Note that the form --filter=sparse:path=<path> that wants to read
846 from an arbitrary path on the filesystem has been dropped for
847 security reasons.
848 Multiple --filter= flags can be specified to combine filters. Only
850 objects which are accepted by every filter are included.
851 The form --filter=combine:<filter1>+<filter2>+...<filterN> can also
853 be used to combined several filters, but this is harder than just
854 repeating the --filter flag and is usually not necessary. Filters
855 are joined by + and individual filters are %-encoded (i.e.
856 URL-encoded). Besides the + and % characters, the following
857 characters are reserved and also must be encoded:
858 ~!@#$^&*()[]{}\;",<>?'` as well as all characters with ASCII code
859 <= 0x20, which includes space and newline.
860 Other arbitrary characters can also be encoded. For instance,
862 combine:tree:3+blob:none and combine:tree%3A3+blob%3Anone are
863 equivalent.
864 --no-filter
866 Turn off any previous --filter= argument.
867 --filter-print-omitted
869 Only useful with --filter=; prints a list of the objects omitted by
870 the filter. Object IDs are prefixed with a “~” character.
871 --missing=<missing-action>
873 A debug option to help with future "partial clone" development.
874 This option specifies how missing objects are handled.
875 The form --missing=error requests that rev-list stop with an error
877 if a missing object is encountered. This is the default action.
878 The form --missing=allow-any will allow object traversal to
880 continue if a missing object is encountered. Missing objects will
881 silently be omitted from the results.
882 The form --missing=allow-promisor is like allow-any, but will only
884 allow object traversal to continue for EXPECTED promisor missing
885 objects. Unexpected missing objects will raise an error.
886 The form --missing=print is like allow-any, but will also print a
888 list of the missing objects. Object IDs are prefixed with a “?”
889 character.
890 --exclude-promisor-objects
892 (For internal use only.) Prefilter object traversal at promisor
893 boundary. This is used with partial clone. This is stronger than
894 --missing=allow-promisor because it limits the traversal, rather
895 than just silencing errors about missing objects.
896 --no-walk[=(sorted|unsorted)]
898 Only show the given commits, but do not traverse their ancestors.
899 This has no effect if a range is specified. If the argument
900 unsorted is given, the commits are shown in the order they were
901 given on the command line. Otherwise (if sorted or no argument was
902 given), the commits are shown in reverse chronological order by
903 commit time. Cannot be combined with --graph.
904 --do-walk
906 Overrides a previous --no-walk.
907 Commit Formatting
909 Using these options, git-rev-list(1) will act similar to the more
910 specialized family of commit log tools: git-log(1), git-show(1), and
911 git-whatchanged(1)
912 --pretty[=<format>], --format=<format>
914 Pretty-print the contents of the commit logs in a given format,
915 where <format> can be one of oneline, short, medium, full, fuller,
916 reference, email, raw, format:<string> and tformat:<string>. When
917 <format> is none of the above, and has %placeholder in it, it acts
918 as if --pretty=tformat:<format> were given.
919 See the "PRETTY FORMATS" section for some additional details for
921 each format. When =<format> part is omitted, it defaults to medium.
922 Note: you can specify the default pretty format in the repository
924 configuration (see git-config(1)).
925 --abbrev-commit
927 Instead of showing the full 40-byte hexadecimal commit object name,
928 show a prefix that names the object uniquely. "--abbrev=<n>" (which
929 also modifies diff output, if it is displayed) option can be used
930 to specify the minimum length of the prefix.
931 This should make "--pretty=oneline" a whole lot more readable for
933 people using 80-column terminals.
934 --no-abbrev-commit
936 Show the full 40-byte hexadecimal commit object name. This negates
937 --abbrev-commit, either explicit or implied by other options such
938 as "--oneline". It also overrides the log.abbrevCommit variable.
939 --oneline
941 This is a shorthand for "--pretty=oneline --abbrev-commit" used
942 together.
943 --encoding=<encoding>
945 The commit objects record the encoding used for the log message in
946 their encoding header; this option can be used to tell the command
947 to re-code the commit log message in the encoding preferred by the
948 user. For non plumbing commands this defaults to UTF-8. Note that
949 if an object claims to be encoded in X and we are outputting in X,
950 we will output the object verbatim; this means that invalid
951 sequences in the original commit may be copied to the output.
952 --expand-tabs=<n>, --expand-tabs, --no-expand-tabs
954 Perform a tab expansion (replace each tab with enough spaces to
955 fill to the next display column that is multiple of <n>) in the log
956 message before showing it in the output. --expand-tabs is a
957 short-hand for --expand-tabs=8, and --no-expand-tabs is a
958 short-hand for --expand-tabs=0, which disables tab expansion.
959 By default, tabs are expanded in pretty formats that indent the log
961 message by 4 spaces (i.e. medium, which is the default, full, and
962 fuller).
963 --show-signature
965 Check the validity of a signed commit object by passing the
966 signature to gpg --verify and show the output.
967 --relative-date
969 Synonym for --date=relative.
970 --date=<format>
972 Only takes effect for dates shown in human-readable format, such as
973 when using --pretty. log.date config variable sets a default value
974 for the log command’s --date option. By default, dates are shown in
975 the original time zone (either committer’s or author’s). If -local
976 is appended to the format (e.g., iso-local), the user’s local time
977 zone is used instead.
978 --date=relative shows dates relative to the current time, e.g. “2
980 hours ago”. The -local option has no effect for --date=relative.
981 --date=local is an alias for --date=default-local.
983 --date=iso (or --date=iso8601) shows timestamps in a ISO 8601-like
985 format. The differences to the strict ISO 8601 format are:
986 · a space instead of the T date/time delimiter
988 · a space between time and time zone
990 · no colon between hours and minutes of the time zone
992 --date=iso-strict (or --date=iso8601-strict) shows timestamps in
994 strict ISO 8601 format.
995 --date=rfc (or --date=rfc2822) shows timestamps in RFC 2822 format,
997 often found in email messages.
998 --date=short shows only the date, but not the time, in YYYY-MM-DD
1000 format.
1001 --date=raw shows the date as seconds since the epoch (1970-01-01
1003 00:00:00 UTC), followed by a space, and then the timezone as an
1004 offset from UTC (a + or - with four digits; the first two are
1005 hours, and the second two are minutes). I.e., as if the timestamp
1006 were formatted with strftime("%s %z")). Note that the -local option
1007 does not affect the seconds-since-epoch value (which is always
1008 measured in UTC), but does switch the accompanying timezone value.
1009 --date=human shows the timezone if the timezone does not match the
1011 current time-zone, and doesn’t print the whole date if that matches
1012 (ie skip printing year for dates that are "this year", but also
1013 skip the whole date itself if it’s in the last few days and we can
1014 just say what weekday it was). For older dates the hour and minute
1015 is also omitted.
1016 --date=unix shows the date as a Unix epoch timestamp (seconds since
1018 1970). As with --raw, this is always in UTC and therefore -local
1019 has no effect.
1020 --date=format:... feeds the format ... to your system strftime,
1022 except for %z and %Z, which are handled internally. Use
1023 --date=format:%c to show the date in your system locale’s preferred
1024 format. See the strftime manual for a complete list of format
1025 placeholders. When using -local, the correct syntax is
1026 --date=format-local:....
1027 --date=default is the default format, and is similar to
1029 --date=rfc2822, with a few exceptions:
1030 · there is no comma after the day-of-week
1032 · the time zone is omitted when the local time zone is used
1034 --header
1036 Print the contents of the commit in raw-format; each record is
1037 separated with a NUL character.
1038 --parents
1040 Print also the parents of the commit (in the form "commit parent...
1041 "). Also enables parent rewriting, see History Simplification
1042 above.
1043 --children
1045 Print also the children of the commit (in the form "commit child...
1046 "). Also enables parent rewriting, see History Simplification
1047 above.
1048 --timestamp
1050 Print the raw commit timestamp.
1051 --left-right
1053 Mark which side of a symmetric difference a commit is reachable
1054 from. Commits from the left side are prefixed with < and those from
1055 the right with >. If combined with --boundary, those commits are
1056 prefixed with -.
1057 For example, if you have this topology:
1059 y---b---b branch B
1061 / \ /
1062 / .
1063 / / \
1064 o---x---a---a branch A
1065 you would get an output like this:
1067 $ git rev-list --left-right --boundary --pretty=oneline A...B
1069 >bbbbbbb... 3rd on b
1071 >bbbbbbb... 2nd on b
1072 <aaaaaaa... 3rd on a
1073 <aaaaaaa... 2nd on a
1074 -yyyyyyy... 1st on b
1075 -xxxxxxx... 1st on a
1076 --graph
1078 Draw a text-based graphical representation of the commit history on
1079 the left hand side of the output. This may cause extra lines to be
1080 printed in between commits, in order for the graph history to be
1081 drawn properly. Cannot be combined with --no-walk.
1082 This enables parent rewriting, see History Simplification above.
1084 This implies the --topo-order option by default, but the
1086 --date-order option may also be specified.
1087 --show-linear-break[=<barrier>]
1089 When --graph is not used, all history branches are flattened which
1090 can make it hard to see that the two consecutive commits do not
1091 belong to a linear branch. This option puts a barrier in between
1092 them in that case. If <barrier> is specified, it is the string that
1093 will be shown instead of the default one.
1094 --count
1096 Print a number stating how many commits would have been listed, and
1097 suppress all other output. When used together with --left-right,
1098 instead print the counts for left and right commits, separated by a
1099 tab. When used together with --cherry-mark, omit patch equivalent
1100 commits from these counts and print the count for equivalent
1101 commits separated by a tab.
1102
1104 If the commit is a merge, and if the pretty-format is not oneline,
1105 email or raw, an additional line is inserted before the Author: line.
1106 This line begins with "Merge: " and the hashes of ancestral commits are
1107 printed, separated by spaces. Note that the listed commits may not
1108 necessarily be the list of the direct parent commits if you have
1109 limited your view of history: for example, if you are only interested
1110 in changes related to a certain directory or file.
1111 There are several built-in formats, and you can define additional
1113 formats by setting a pretty.<name> config option to either another
1114 format name, or a format: string, as described below (see git-
1115 config(1)). Here are the details of the built-in formats:
1116 · oneline
1118 <hash> <title line>
1120 This is designed to be as compact as possible.
1122 · short
1124 commit <hash>
1126 Author: <author>
1127 <title line>
1129 · medium
1131 commit <hash>
1133 Author: <author>
1134 Date: <author date>
1135 <title line>
1137 <full commit message>
1139 · full
1141 commit <hash>
1143 Author: <author>
1144 Commit: <committer>
1145 <title line>
1147 <full commit message>
1149 · fuller
1151 commit <hash>
1153 Author: <author>
1154 AuthorDate: <author date>
1155 Commit: <committer>
1156 CommitDate: <committer date>
1157 <title line>
1159 <full commit message>
1161 · reference
1163 <abbrev hash> (<title line>, <short author date>)
1165 This format is used to refer to another commit in a commit message
1167 and is the same as --pretty='format:%C(auto)%h (%s, %ad)'. By
1168 default, the date is formatted with --date=short unless another
1169 --date option is explicitly specified. As with any format: with
1170 format placeholders, its output is not affected by other options
1171 like --decorate and --walk-reflogs.
1172 · email
1174 From <hash> <date>
1176 From: <author>
1177 Date: <author date>
1178 Subject: [PATCH] <title line>
1179 <full commit message>
1181 · mboxrd
1183 Like email, but lines in the commit message starting with "From "
1185 (preceded by zero or more ">") are quoted with ">" so they aren’t
1186 confused as starting a new commit.
1187 · raw
1189 The raw format shows the entire commit exactly as stored in the
1191 commit object. Notably, the hashes are displayed in full,
1192 regardless of whether --abbrev or --no-abbrev are used, and parents
1193 information show the true parent commits, without taking grafts or
1194 history simplification into account. Note that this format affects
1195 the way commits are displayed, but not the way the diff is shown
1196 e.g. with git log --raw. To get full object names in a raw diff
1197 format, use --no-abbrev.
1198 · format:<string>
1200 The format:<string> format allows you to specify which information
1202 you want to show. It works a little bit like printf format, with
1203 the notable exception that you get a newline with %n instead of \n.
1204 E.g, format:"The author of %h was %an, %ar%nThe title was >>%s<<%n"
1206 would show something like this:
1207 The author of fe6e0ee was Junio C Hamano, 23 hours ago
1209 The title was >>t4119: test autocomputing -p<n> for traditional diff input.<<
1210 The placeholders are:
1212 · Placeholders that expand to a single literal character:
1214 %n
1216 newline
1217 %%
1219 a raw %
1220 %x00
1222 print a byte from a hex code
1223 · Placeholders that affect formatting of later placeholders:
1225 %Cred
1227 switch color to red
1228 %Cgreen
1230 switch color to green
1231 %Cblue
1233 switch color to blue
1234 %Creset
1236 reset color
1237 %C(...)
1239 color specification, as described under Values in the
1240 "CONFIGURATION FILE" section of git-config(1). By default,
1241 colors are shown only when enabled for log output (by
1242 color.diff, color.ui, or --color, and respecting the auto
1243 settings of the former if we are going to a terminal).
1244 %C(auto,...) is accepted as a historical synonym for the
1245 default (e.g., %C(auto,red)). Specifying %C(always,...)
1246 will show the colors even when color is not otherwise
1247 enabled (though consider just using --color=always to
1248 enable color for the whole output, including this format
1249 and anything else git might color). auto alone (i.e.
1250 %C(auto)) will turn on auto coloring on the next
1251 placeholders until the color is switched again.
1252 %m
1254 left (<), right (>) or boundary (-) mark
1255 %w([<w>[,<i1>[,<i2>]]])
1257 switch line wrapping, like the -w option of git-
1258 shortlog(1).
1259 %<(<N>[,trunc|ltrunc|mtrunc])
1261 make the next placeholder take at least N columns, padding
1262 spaces on the right if necessary. Optionally truncate at
1263 the beginning (ltrunc), the middle (mtrunc) or the end
1264 (trunc) if the output is longer than N columns. Note that
1265 truncating only works correctly with N >= 2.
1266 %<|(<N>)
1268 make the next placeholder take at least until Nth columns,
1269 padding spaces on the right if necessary
1270 %>(<N>), %>|(<N>)
1272 similar to %<(<N>), %<|(<N>) respectively, but padding
1273 spaces on the left
1274 %>>(<N>), %>>|(<N>)
1276 similar to %>(<N>), %>|(<N>) respectively, except that if
1277 the next placeholder takes more spaces than given and there
1278 are spaces on its left, use those spaces
1279 %><(<N>), %><|(<N>)
1281 similar to %<(<N>), %<|(<N>) respectively, but padding both
1282 sides (i.e. the text is centered)
1283 · Placeholders that expand to information extracted from the
1285 commit:
1286 %H
1288 commit hash
1289 %h
1291 abbreviated commit hash
1292 %T
1294 tree hash
1295 %t
1297 abbreviated tree hash
1298 %P
1300 parent hashes
1301 %p
1303 abbreviated parent hashes
1304 %an
1306 author name
1307 %aN
1309 author name (respecting .mailmap, see git-shortlog(1) or
1310 git-blame(1))
1311 %ae
1313 author email
1314 %aE
1316 author email (respecting .mailmap, see git-shortlog(1) or
1317 git-blame(1))
1318 %al
1320 author email local-part (the part before the @ sign)
1321 %aL
1323 author local-part (see %al) respecting .mailmap, see git-
1324 shortlog(1) or git-blame(1))
1325 %ad
1327 author date (format respects --date= option)
1328 %aD
1330 author date, RFC2822 style
1331 %ar
1333 author date, relative
1334 %at
1336 author date, UNIX timestamp
1337 %ai
1339 author date, ISO 8601-like format
1340 %aI
1342 author date, strict ISO 8601 format
1343 %as
1345 author date, short format (YYYY-MM-DD)
1346 %cn
1348 committer name
1349 %cN
1351 committer name (respecting .mailmap, see git-shortlog(1) or
1352 git-blame(1))
1353 %ce
1355 committer email
1356 %cE
1358 committer email (respecting .mailmap, see git-shortlog(1)
1359 or git-blame(1))
1360 %cl
1362 committer email local-part (the part before the @ sign)
1363 %cL
1365 committer local-part (see %cl) respecting .mailmap, see
1366 git-shortlog(1) or git-blame(1))
1367 %cd
1369 committer date (format respects --date= option)
1370 %cD
1372 committer date, RFC2822 style
1373 %cr
1375 committer date, relative
1376 %ct
1378 committer date, UNIX timestamp
1379 %ci
1381 committer date, ISO 8601-like format
1382 %cI
1384 committer date, strict ISO 8601 format
1385 %cs
1387 committer date, short format (YYYY-MM-DD)
1388 %d
1390 ref names, like the --decorate option of git-log(1)
1391 %D
1393 ref names without the " (", ")" wrapping.
1394 %S
1396 ref name given on the command line by which the commit was
1397 reached (like git log --source), only works with git log
1398 %e
1400 encoding
1401 %s
1403 subject
1404 %f
1406 sanitized subject line, suitable for a filename
1407 %b
1409 body
1410 %B
1412 raw body (unwrapped subject and body)
1413 %GG
1415 raw verification message from GPG for a signed commit
1416 %G?
1418 show "G" for a good (valid) signature, "B" for a bad
1419 signature, "U" for a good signature with unknown validity,
1420 "X" for a good signature that has expired, "Y" for a good
1421 signature made by an expired key, "R" for a good signature
1422 made by a revoked key, "E" if the signature cannot be
1423 checked (e.g. missing key) and "N" for no signature
1424 %GS
1426 show the name of the signer for a signed commit
1427 %GK
1429 show the key used to sign a signed commit
1430 %GF
1432 show the fingerprint of the key used to sign a signed
1433 commit
1434 %GP
1436 show the fingerprint of the primary key whose subkey was
1437 used to sign a signed commit
1438 %GT
1440 show the trust level for the key used to sign a signed
1441 commit
1442 %gD
1444 reflog selector, e.g., refs/stash@{1} or refs/stash@{2
1445 minutes ago}; the format follows the rules described for
1446 the -g option. The portion before the @ is the refname as
1447 given on the command line (so git log -g refs/heads/master
1448 would yield refs/heads/master@{0}).
1449 %gd
1451 shortened reflog selector; same as %gD, but the refname
1452 portion is shortened for human readability (so
1453 refs/heads/master becomes just master).
1454 %gn
1456 reflog identity name
1457 %gN
1459 reflog identity name (respecting .mailmap, see git-
1460 shortlog(1) or git-blame(1))
1461 %ge
1463 reflog identity email
1464 %gE
1466 reflog identity email (respecting .mailmap, see git-
1467 shortlog(1) or git-blame(1))
1468 %gs
1470 reflog subject
1471 %(trailers[:options])
1473 display the trailers of the body as interpreted by git-
1474 interpret-trailers(1). The trailers string may be followed
1475 by a colon and zero or more comma-separated options:
1476 · key=<K>: only show trailers with specified key.
1478 Matching is done case-insensitively and trailing colon
1479 is optional. If option is given multiple times trailer
1480 lines matching any of the keys are shown. This option
1481 automatically enables the only option so that
1482 non-trailer lines in the trailer block are hidden. If
1483 that is not desired it can be disabled with only=false.
1484 E.g., %(trailers:key=Reviewed-by) shows trailer lines
1485 with key Reviewed-by.
1486 · only[=val]: select whether non-trailer lines from the
1488 trailer block should be included. The only keyword may
1489 optionally be followed by an equal sign and one of
1490 true, on, yes to omit or false, off, no to show the
1491 non-trailer lines. If option is given without value it
1492 is enabled. If given multiple times the last value is
1493 used.
1494 · separator=<SEP>: specify a separator inserted between
1496 trailer lines. When this option is not given each
1497 trailer line is terminated with a line feed character.
1498 The string SEP may contain the literal formatting codes
1499 described above. To use comma as separator one must use
1500 %x2C as it would otherwise be parsed as next option. If
1501 separator option is given multiple times only the last
1502 one is used. E.g., %(trailers:key=Ticket,separator=%x2C
1503 ) shows all trailer lines whose key is "Ticket"
1504 separated by a comma and a space.
1505 · unfold[=val]: make it behave as if interpret-trailer’s
1507 --unfold option was given. In same way as to for only
1508 it can be followed by an equal sign and explicit value.
1509 E.g., %(trailers:only,unfold=true) unfolds and shows
1510 all trailer lines.
1511 · valueonly[=val]: skip over the key part of the trailer
1513 line and only show the value part. Also this optionally
1514 allows explicit value.
1515 Note
1517 Some placeholders may depend on other options given to the revision
1518 traversal engine. For example, the %g* reflog options will insert
1519 an empty string unless we are traversing reflog entries (e.g., by
1520 git log -g). The %d and %D placeholders will use the "short"
1521 decoration format if --decorate was not already provided on the
1522 command line.
1523 If you add a + (plus sign) after % of a placeholder, a line-feed is
1525 inserted immediately before the expansion if and only if the
1526 placeholder expands to a non-empty string.
1527 If you add a - (minus sign) after % of a placeholder, all consecutive
1529 line-feeds immediately preceding the expansion are deleted if and only
1530 if the placeholder expands to an empty string.
1531 If you add a ` ` (space) after % of a placeholder, a space is inserted
1533 immediately before the expansion if and only if the placeholder expands
1534 to a non-empty string.
1535 · tformat:
1537 The tformat: format works exactly like format:, except that it
1539 provides "terminator" semantics instead of "separator" semantics.
1540 In other words, each commit has the message terminator character
1541 (usually a newline) appended, rather than a separator placed
1542 between entries. This means that the final entry of a single-line
1543 format will be properly terminated with a new line, just as the
1544 "oneline" format does. For example:
1545 $ git log -2 --pretty=format:%h 4da45bef \
1547 | perl -pe '$_ .= " -- NO NEWLINE\n" unless /\n/'
1548 4da45be
1549 7134973 -- NO NEWLINE
1550 $ git log -2 --pretty=tformat:%h 4da45bef \
1552 | perl -pe '$_ .= " -- NO NEWLINE\n" unless /\n/'
1553 4da45be
1554 7134973
1555 In addition, any unrecognized string that has a % in it is
1557 interpreted as if it has tformat: in front of it. For example,
1558 these two are equivalent:
1559 $ git log -2 --pretty=tformat:%h 4da45bef
1561 $ git log -2 --pretty=%h 4da45bef
1562
1564 Part of the git(1) suite
1565Git 2.30.2 2021-03-08 GIT-REV-LIST(1)