[git/git.git] / Documentation / revisions.txt
2 --------------------
4 A revision parameter typically, but not necessarily, names a
5 commit object. They use what is called an 'extended SHA1'
6 syntax. Here are various ways to spell object names. The
7 ones listed near the end of this list are to name trees and
8 blobs contained in a commit.
10 * The full SHA1 object name (40-byte hexadecimal string), or
11 a substring of such that is unique within the repository.
12 E.g. dae86e1950b1277e545cee180551750029cfe735 and dae86e both
13 name the same commit object if there are no other object in
14 your repository whose object name starts with dae86e.
16 * An output from 'git describe'; i.e. a closest tag, optionally
17 followed by a dash and a number of commits, followed by a dash, a
18 `g`, and an abbreviated object name.
20 * A symbolic ref name. E.g. 'master' typically means the commit
21 object referenced by refs/heads/master. If you
22 happen to have both heads/master and tags/master, you can
23 explicitly say 'heads/master' to tell git which one you mean.
24 When ambiguous, a `<name>` is disambiguated by taking the
25 first match in the following rules:
27 . if `$GIT_DIR/<name>` exists, that is what you mean (this is usually
28 useful only for `HEAD`, `FETCH_HEAD`, `ORIG_HEAD`, `MERGE_HEAD`
31 . otherwise, `refs/<name>` if exists;
33 . otherwise, `refs/tags/<name>` if exists;
35 . otherwise, `refs/heads/<name>` if exists;
37 . otherwise, `refs/remotes/<name>` if exists;
39 . otherwise, `refs/remotes/<name>/HEAD` if exists.
40 +
41 HEAD names the commit your changes in the working tree is based on.
42 FETCH_HEAD records the branch you fetched from a remote repository
43 with your last 'git fetch' invocation.
44 ORIG_HEAD is created by commands that moves your HEAD in a drastic
45 way, to record the position of the HEAD before their operation, so that
46 you can change the tip of the branch back to the state before you ran
47 them easily.
48 MERGE_HEAD records the commit(s) you are merging into your branch
49 when you run 'git merge'.
50 CHERRY_PICK_HEAD records the commit you are cherry-picking
51 when you run 'git cherry-pick'.
52 +
53 Note that any of the `refs/*` cases above may come either from
54 the `$GIT_DIR/refs` directory or from the `$GIT_DIR/packed-refs` file.
56 * A ref followed by the suffix '@' with a date specification
57 enclosed in a brace
58 pair (e.g. '\{yesterday\}', '\{1 month 2 weeks 3 days 1 hour 1
59 second ago\}' or '\{1979-02-26 18:30:00\}') to specify the value
60 of the ref at a prior point in time. This suffix may only be
61 used immediately following a ref name and the ref must have an
62 existing log ($GIT_DIR/logs/<ref>). Note that this looks up the state
63 of your *local* ref at a given time; e.g., what was in your local
64 `master` branch last week. If you want to look at commits made during
65 certain times, see `--since` and `--until`.
67 * A ref followed by the suffix '@' with an ordinal specification
68 enclosed in a brace pair (e.g. '\{1\}', '\{15\}') to specify
69 the n-th prior value of that ref. For example 'master@\{1\}'
70 is the immediate prior value of 'master' while 'master@\{5\}'
71 is the 5th prior value of 'master'. This suffix may only be used
72 immediately following a ref name and the ref must have an existing
73 log ($GIT_DIR/logs/<ref>).
75 * You can use the '@' construct with an empty ref part to get at a
76 reflog of the current branch. For example, if you are on the
77 branch 'blabla', then '@\{1\}' means the same as 'blabla@\{1\}'.
79 * The special construct '@\{-<n>\}' means the <n>th branch checked out
80 before the current one.
82 * The suffix '@\{upstream\}' to a ref (short form 'ref@\{u\}') refers to
83 the branch the ref is set to build on top of. Missing ref defaults
84 to the current branch.
86 * A suffix '{caret}' to a revision parameter (e.g. 'HEAD{caret}') means the first parent of
87 that commit object. '{caret}<n>' means the <n>th parent (i.e.
88 'rev{caret}'
89 is equivalent to 'rev{caret}1'). As a special rule,
90 'rev{caret}0' means the commit itself and is used when 'rev' is the
91 object name of a tag object that refers to a commit object.
93 * A suffix '{tilde}<n>' to a revision parameter means the commit
94 object that is the <n>th generation grand-parent of the named
95 commit object, following only the first parent. I.e. rev~3 is
96 equivalent to rev{caret}{caret}{caret} which is equivalent to
97 rev{caret}1{caret}1{caret}1. See below for a illustration of
98 the usage of this form.
100 * A suffix '{caret}' followed by an object type name enclosed in
101 brace pair (e.g. `v0.99.8{caret}\{commit\}`) means the object
102 could be a tag, and dereference the tag recursively until an
103 object of that type is found or the object cannot be
104 dereferenced anymore (in which case, barf). `rev{caret}0`
105 introduced earlier is a short-hand for `rev{caret}\{commit\}`.
107 * A suffix '{caret}' followed by an empty brace pair
108 (e.g. `v0.99.8{caret}\{\}`) means the object could be a tag,
109 and dereference the tag recursively until a non-tag object is
110 found.
112 * A suffix '{caret}' to a revision parameter followed by a brace
113 pair that contains a text led by a slash (e.g. `HEAD^{/fix nasty bug}`):
114 this is the same as `:/fix nasty bug` syntax below except that
115 it returns the youngest matching commit which is reachable from
116 the ref before '{caret}'.
118 * A colon, followed by a slash, followed by a text (e.g. `:/fix nasty bug`): this names
119 a commit whose commit message matches the specified regular expression.
120 This name returns the youngest matching commit which is
121 reachable from any ref. If the commit message starts with a
122 '!', you have to repeat that; the special sequence ':/!',
123 followed by something else than '!' is reserved for now.
124 The regular expression can match any part of the commit message. To
125 match messages starting with a string, one can use e.g. `:/^foo`.
127 * A suffix ':' followed by a path (e.g. `HEAD:README`); this names the blob or tree
128 at the given path in the tree-ish object named by the part
129 before the colon.
130 ':path' (with an empty part before the colon, e.g. `:README`)
131 is a special case of the syntax described next: content
132 recorded in the index at the given path.
133 A path starting with './' or '../' is relative to current working directory.
134 The given path will be converted to be relative to working tree's root directory.
135 This is most useful to address a blob or tree from a commit or tree that has
136 the same tree structure with the working tree.
138 * A colon, optionally followed by a stage number (0 to 3) and a
139 colon, followed by a path (e.g. `:0:README`); this names a blob object in the
140 index at the given path. Missing stage number (and the colon
141 that follows it, e.g. `:README`) names a stage 0 entry. During a merge, stage
142 1 is the common ancestor, stage 2 is the target branch's version
143 (typically the current branch), and stage 3 is the version from
144 the branch being merged.
146 Here is an illustration, by Jon Loeliger. Both commit nodes B
147 and C are parents of commit node A. Parent commits are ordered
148 left-to-right.
150 ........................................
151 G H I J
152 \ / \ /
153 D E F
154 \ | / \
155 \ | / |
156 \|/ |
157 B C
158 \ /
159 \ /
160 A
161 ........................................
163 A = = A^0
164 B = A^ = A^1 = A~1
165 C = A^2 = A^2
166 D = A^^ = A^1^1 = A~2
167 E = B^2 = A^^2
168 F = B^3 = A^^3
169 G = A^^^ = A^1^1^1 = A~3
170 H = D^2 = B^^2 = A^^^2 = A~2^2
171 I = F^ = B^3^ = A^^3^
172 J = F^2 = B^3^2 = A^^3^2
176 -----------------
178 History traversing commands such as 'git log' operate on a set
179 of commits, not just a single commit. To these commands,
180 specifying a single revision with the notation described in the
181 previous section means the set of commits reachable from that
182 commit, following the commit ancestry chain.
184 To exclude commits reachable from a commit, a prefix `{caret}`
185 notation is used. E.g. `{caret}r1 r2` means commits reachable
186 from `r2` but exclude the ones reachable from `r1`.
188 This set operation appears so often that there is a shorthand
189 for it. When you have two commits `r1` and `r2` (named according
190 to the syntax explained in SPECIFYING REVISIONS above), you can ask
191 for commits that are reachable from r2 excluding those that are reachable
192 from r1 by `{caret}r1 r2` and it can be written as `r1..r2`.
194 A similar notation `r1\...r2` is called symmetric difference
195 of `r1` and `r2` and is defined as
196 `r1 r2 --not $(git merge-base --all r1 r2)`.
197 It is the set of commits that are reachable from either one of
198 `r1` or `r2` but not from both.
200 Two other shorthands for naming a set that is formed by a commit
201 and its parent commits exist. The `r1{caret}@` notation means all
202 parents of `r1`. `r1{caret}!` includes commit `r1` but excludes
203 all of its parents.
205 Here are a handful of examples:
207 D G H D
208 D F G H I J D F
209 ^G D H D
210 ^D B E I J F B
211 B...C G H D E B C
212 ^D B C E I J F B C
213 C^@ I J F
214 F^! D G H D F