1GITREVISIONS(7) Git Manual GITREVISIONS(7)
2
6 gitrevisions - Specifying revisions and ranges for Git
7
9 gitrevisions
10
12 Many Git commands take revision parameters as arguments. Depending on
13 the command, they denote a specific commit or, for commands which walk
14 the revision graph (such as git-log(1)), all commits which are
15 reachable from that commit. For commands that walk the revision graph
16 one can also specify a range of revisions explicitly.
17 In addition, some Git commands (such as git-show(1) and git-push(1))
19 can also take revision parameters which denote other objects than
20 commits, e.g. blobs ("files") or trees ("directories of files").
21
23 A revision parameter <rev> typically, but not necessarily, names a
24 commit object. It uses what is called an extended SHA-1 syntax. Here
25 are various ways to spell object names. The ones listed near the end of
26 this list name trees and blobs contained in a commit.
27 Note
29 This document shows the "raw" syntax as seen by git. The shell and
30 other UIs might require additional quoting to protect special
31 characters and to avoid word splitting.
32 <sha1>, e.g. dae86e1950b1277e545cee180551750029cfe735, dae86e
34 The full SHA-1 object name (40-byte hexadecimal string), or a
35 leading substring that is unique within the repository. E.g.
36 dae86e1950b1277e545cee180551750029cfe735 and dae86e both name the
37 same commit object if there is no other object in your repository
38 whose object name starts with dae86e.
39 <describeOutput>, e.g. v1.7.4.2-679-g3bee7fb
41 Output from git describe; i.e. a closest tag, optionally followed
42 by a dash and a number of commits, followed by a dash, a g, and an
43 abbreviated object name.
44 <refname>, e.g. master, heads/master, refs/heads/master
46 A symbolic ref name. E.g. master typically means the commit object
47 referenced by refs/heads/master. If you happen to have both
48 heads/master and tags/master, you can explicitly say heads/master
49 to tell Git which one you mean. When ambiguous, a <refname> is
50 disambiguated by taking the first match in the following rules:
51 1. If $GIT_DIR/<refname> exists, that is what you mean (this is
53 usually useful only for HEAD, FETCH_HEAD, ORIG_HEAD, MERGE_HEAD
54 and CHERRY_PICK_HEAD);
55 2. otherwise, refs/<refname> if it exists;
57 3. otherwise, refs/tags/<refname> if it exists;
59 4. otherwise, refs/heads/<refname> if it exists;
61 5. otherwise, refs/remotes/<refname> if it exists;
63 6. otherwise, refs/remotes/<refname>/HEAD if it exists.
65 HEAD names the commit on which you based the changes in the
67 working tree. FETCH_HEAD records the branch which you fetched
68 from a remote repository with your last git fetch invocation.
69 ORIG_HEAD is created by commands that move your HEAD in a
70 drastic way, to record the position of the HEAD before their
71 operation, so that you can easily change the tip of the branch
72 back to the state before you ran them. MERGE_HEAD records the
73 commit(s) which you are merging into your branch when you run
74 git merge. CHERRY_PICK_HEAD records the commit which you are
75 cherry-picking when you run git cherry-pick.
76 Note that any of the refs/* cases above may come either from
78 the $GIT_DIR/refs directory or from the $GIT_DIR/packed-refs
79 file. While the ref name encoding is unspecified, UTF-8 is
80 preferred as some output processing may assume ref names in
81 UTF-8.
82 @
84 @ alone is a shortcut for HEAD.
85 [<refname>]@{<date>}, e.g. master@{yesterday}, HEAD@{5 minutes ago}
87 A ref followed by the suffix @ with a date specification enclosed
88 in a brace pair (e.g. {yesterday}, {1 month 2 weeks 3 days 1 hour
89 1 second ago} or {1979-02-26 18:30:00}) specifies the value of the
90 ref at a prior point in time. This suffix may only be used
91 immediately following a ref name and the ref must have an existing
92 log ($GIT_DIR/logs/<ref>). Note that this looks up the state of
93 your local ref at a given time; e.g., what was in your local master
94 branch last week. If you want to look at commits made during
95 certain times, see --since and --until.
96 <refname>@{<n>}, e.g. master@{1}
98 A ref followed by the suffix @ with an ordinal specification
99 enclosed in a brace pair (e.g. {1}, {15}) specifies the n-th prior
100 value of that ref. For example master@{1} is the immediate prior
101 value of master while master@{5} is the 5th prior value of master.
102 This suffix may only be used immediately following a ref name and
103 the ref must have an existing log ($GIT_DIR/logs/<refname>).
104 @{<n>}, e.g. @{1}
106 You can use the @ construct with an empty ref part to get at a
107 reflog entry of the current branch. For example, if you are on
108 branch blabla then @{1} means the same as blabla@{1}.
109 @{-<n>}, e.g. @{-1}
111 The construct @{-<n>} means the <n>th branch/commit checked out
112 before the current one.
113 [<branchname>]@{upstream}, e.g. master@{upstream}, @{u}
115 A branch B may be set up to build on top of a branch X (configured
116 with branch.<name>.merge) at a remote R (configured with
117 branch.<name>.remote). B@{u} refers to the remote-tracking branch
118 for the branch X taken from remote R, typically found at
119 refs/remotes/R/X.
120 [<branchname>]@{push}, e.g. master@{push}, @{push}
122 The suffix @{push} reports the branch "where we would push to" if
123 git push were run while branchname was checked out (or the current
124 HEAD if no branchname is specified). Like for @{upstream}, we
125 report the remote-tracking branch that corresponds to that branch
126 at the remote.
127 Here’s an example to make it more clear:
129 $ git config push.default current
131 $ git config remote.pushdefault myfork
132 $ git switch -c mybranch origin/master
133 $ git rev-parse --symbolic-full-name @{upstream}
135 refs/remotes/origin/master
136 $ git rev-parse --symbolic-full-name @{push}
138 refs/remotes/myfork/mybranch
139 Note in the example that we set up a triangular workflow, where we
141 pull from one location and push to another. In a non-triangular
142 workflow, @{push} is the same as @{upstream}, and there is no need
143 for it.
144 This suffix is also accepted when spelled in uppercase, and means
146 the same thing no matter the case.
147 <rev>^[<n>], e.g. HEAD^, v1.5.1^0
149 A suffix ^ to a revision parameter means the first parent of that
150 commit object. ^<n> means the <n>th parent (i.e. <rev>^ is
151 equivalent to <rev>^1). As a special rule, <rev>^0 means the commit
152 itself and is used when <rev> is the object name of a tag object
153 that refers to a commit object.
154 <rev>~[<n>], e.g. HEAD~, master~3
156 A suffix ~ to a revision parameter means the first parent of that
157 commit object. A suffix ~<n> to a revision parameter means the
158 commit object that is the <n>th generation ancestor of the named
159 commit object, following only the first parents. I.e. <rev>~3 is
160 equivalent to <rev>^^^ which is equivalent to <rev>^1^1^1. See
161 below for an illustration of the usage of this form.
162 <rev>^{<type>}, e.g. v0.99.8^{commit}
164 A suffix ^ followed by an object type name enclosed in brace pair
165 means dereference the object at <rev> recursively until an object
166 of type <type> is found or the object cannot be dereferenced
167 anymore (in which case, barf). For example, if <rev> is a
168 commit-ish, <rev>^{commit} describes the corresponding commit
169 object. Similarly, if <rev> is a tree-ish, <rev>^{tree} describes
170 the corresponding tree object. <rev>^0 is a short-hand for
171 <rev>^{commit}.
172 <rev>^{object} can be used to make sure <rev> names an object that
174 exists, without requiring <rev> to be a tag, and without
175 dereferencing <rev>; because a tag is already an object, it does
176 not have to be dereferenced even once to get to an object.
177 <rev>^{tag} can be used to ensure that <rev> identifies an existing
179 tag object.
180 <rev>^{}, e.g. v0.99.8^{}
182 A suffix ^ followed by an empty brace pair means the object could
183 be a tag, and dereference the tag recursively until a non-tag
184 object is found.
185 <rev>^{/<text>}, e.g. HEAD^{/fix nasty bug}
187 A suffix ^ to a revision parameter, followed by a brace pair that
188 contains a text led by a slash, is the same as the :/fix nasty bug
189 syntax below except that it returns the youngest matching commit
190 which is reachable from the <rev> before ^.
191 :/<text>, e.g. :/fix nasty bug
193 A colon, followed by a slash, followed by a text, names a commit
194 whose commit message matches the specified regular expression. This
195 name returns the youngest matching commit which is reachable from
196 any ref, including HEAD. The regular expression can match any part
197 of the commit message. To match messages starting with a string,
198 one can use e.g. :/^foo. The special sequence :/! is reserved for
199 modifiers to what is matched. :/!-foo performs a negative match,
200 while :/!!foo matches a literal ! character, followed by foo. Any
201 other sequence beginning with :/! is reserved for now. Depending
202 on the given text, the shell’s word splitting rules might require
203 additional quoting.
204 <rev>:<path>, e.g. HEAD:README, master:./README
206 A suffix : followed by a path names the blob or tree at the given
207 path in the tree-ish object named by the part before the colon. A
208 path starting with ./ or ../ is relative to the current working
209 directory. The given path will be converted to be relative to the
210 working tree’s root directory. This is most useful to address a
211 blob or tree from a commit or tree that has the same tree structure
212 as the working tree.
213 :[<n>:]<path>, e.g. :0:README, :README
215 A colon, optionally followed by a stage number (0 to 3) and a
216 colon, followed by a path, names a blob object in the index at the
217 given path. A missing stage number (and the colon that follows it)
218 names a stage 0 entry. During a merge, stage 1 is the common
219 ancestor, stage 2 is the target branch’s version (typically the
220 current branch), and stage 3 is the version from the branch which
221 is being merged.
222 Here is an illustration, by Jon Loeliger. Both commit nodes B and C are
224 parents of commit node A. Parent commits are ordered left-to-right.
225 G H I J
227 \ / \ /
228 D E F
229 \ | / \
230 \ | / |
231 \|/ |
232 B C
233 \ /
234 \ /
235 A
236 A = = A^0
238 B = A^ = A^1 = A~1
239 C = = A^2
240 D = A^^ = A^1^1 = A~2
241 E = B^2 = A^^2
242 F = B^3 = A^^3
243 G = A^^^ = A^1^1^1 = A~3
244 H = D^2 = B^^2 = A^^^2 = A~2^2
245 I = F^ = B^3^ = A^^3^
246 J = F^2 = B^3^2 = A^^3^2
247
249 History traversing commands such as git log operate on a set of
250 commits, not just a single commit.
251 For these commands, specifying a single revision, using the notation
253 described in the previous section, means the set of commits reachable
254 from the given commit.
255 Specifying several revisions means the set of commits reachable from
257 any of the given commits.
258 A commit’s reachable set is the commit itself and the commits in its
260 ancestry chain.
261 There are several notations to specify a set of connected commits
263 (called a "revision range"), illustrated below.
264 Commit Exclusions
266 ^<rev> (caret) Notation
267 To exclude commits reachable from a commit, a prefix ^ notation is
268 used. E.g. ^r1 r2 means commits reachable from r2 but exclude the
269 ones reachable from r1 (i.e. r1 and its ancestors).
270 Dotted Range Notations
272 The .. (two-dot) Range Notation
273 The ^r1 r2 set operation appears so often that there is a shorthand
274 for it. When you have two commits r1 and r2 (named according to the
275 syntax explained in SPECIFYING REVISIONS above), you can ask for
276 commits that are reachable from r2 excluding those that are
277 reachable from r1 by ^r1 r2 and it can be written as r1..r2.
278 The ... (three-dot) Symmetric Difference Notation
280 A similar notation r1...r2 is called symmetric difference of r1 and
281 r2 and is defined as r1 r2 --not $(git merge-base --all r1 r2). It
282 is the set of commits that are reachable from either one of r1
283 (left side) or r2 (right side) but not from both.
284 In these two shorthand notations, you can omit one end and let it
286 default to HEAD. For example, origin.. is a shorthand for origin..HEAD
287 and asks "What did I do since I forked from the origin branch?"
288 Similarly, ..origin is a shorthand for HEAD..origin and asks "What did
289 the origin do since I forked from them?" Note that .. would mean
290 HEAD..HEAD which is an empty range that is both reachable and
291 unreachable from HEAD.
292 Commands that are specifically designed to take two distinct ranges
294 (e.g. "git range-diff R1 R2" to compare two ranges) do exist, but they
295 are exceptions. Unless otherwise noted, all "git" commands that operate
296 on a set of commits work on a single revision range. In other words,
297 writing two "two-dot range notation" next to each other, e.g.
298 $ git log A..B C..D
300 does not specify two revision ranges for most commands. Instead it will
302 name a single connected set of commits, i.e. those that are reachable
303 from either B or D but are reachable from neither A or C. In a linear
304 history like this:
305 ---A---B---o---o---C---D
307 because A and B are reachable from C, the revision range specified by
309 these two dotted ranges is a single commit D.
310 Other <rev>^ Parent Shorthand Notations
312 Three other shorthands exist, particularly useful for merge commits,
313 for naming a set that is formed by a commit and its parent commits.
314 The r1^@ notation means all parents of r1.
316 The r1^! notation includes commit r1 but excludes all of its parents.
318 By itself, this notation denotes the single commit r1.
319 The <rev>^-[<n>] notation includes <rev> but excludes the <n>th parent
321 (i.e. a shorthand for <rev>^<n>..<rev>), with <n> = 1 if not given.
322 This is typically useful for merge commits where you can just pass
323 <commit>^- to get all the commits in the branch that was merged in
324 merge commit <commit> (including <commit> itself).
325 While <rev>^<n> was about specifying a single commit parent, these
327 three notations also consider its parents. For example you can say
328 HEAD^2^@, however you cannot say HEAD^@^2.
329
331 <rev>
332 Include commits that are reachable from <rev> (i.e. <rev> and its
333 ancestors).
334 ^<rev>
336 Exclude commits that are reachable from <rev> (i.e. <rev> and its
337 ancestors).
338 <rev1>..<rev2>
340 Include commits that are reachable from <rev2> but exclude those
341 that are reachable from <rev1>. When either <rev1> or <rev2> is
342 omitted, it defaults to HEAD.
343 <rev1>...<rev2>
345 Include commits that are reachable from either <rev1> or <rev2> but
346 exclude those that are reachable from both. When either <rev1> or
347 <rev2> is omitted, it defaults to HEAD.
348 <rev>^@, e.g. HEAD^@
350 A suffix ^ followed by an at sign is the same as listing all
351 parents of <rev> (meaning, include anything reachable from its
352 parents, but not the commit itself).
353 <rev>^!, e.g. HEAD^!
355 A suffix ^ followed by an exclamation mark is the same as giving
356 commit <rev> and all its parents prefixed with ^ to exclude them
357 (and their ancestors).
358 <rev>^-<n>, e.g. HEAD^-, HEAD^-2
360 Equivalent to <rev>^<n>..<rev>, with <n> = 1 if not given.
361 Here are a handful of examples using the Loeliger illustration above,
363 with each step in the notation’s expansion and selection carefully
364 spelt out:
365 Args Expanded arguments Selected commits
367 D G H D
368 D F G H I J D F
369 ^G D H D
370 ^D B E I J F B
371 ^D B C E I J F B C
372 C I J F C
373 B..C = ^B C C
374 B...C = B ^F C G H D E B C
375 B^- = B^..B
376 = ^B^1 B E I J F B
377 C^@ = C^1
378 = F I J F
379 B^@ = B^1 B^2 B^3
380 = D E F D G H E F I J
381 C^! = C ^C^@
382 = C ^C^1
383 = C ^F C
384 B^! = B ^B^@
385 = B ^B^1 ^B^2 ^B^3
386 = B ^D ^E ^F B
387 F^! D = F ^I ^J D G H D F
388
390 git-rev-parse(1)
391
393 Part of the git(1) suite
394Git 2.39.1 2023-01-13 GITREVISIONS(7)