1GIT-UPDATE-INDEX(1) Git Manual GIT-UPDATE-INDEX(1)
2
6 git-update-index - Register file contents in the working tree to the
7 index
8
10 git update-index
11 [--add] [--remove | --force-remove] [--replace]
12 [--refresh] [-q] [--unmerged] [--ignore-missing]
13 [(--cacheinfo <mode> <object> <file>)...]
14 [--chmod=(+|-)x]
15 [--assume-unchanged | --no-assume-unchanged]
16 [--skip-worktree | --no-skip-worktree]
17 [--ignore-submodules]
18 [--really-refresh] [--unresolve] [--again | -g]
19 [--info-only] [--index-info]
20 [-z] [--stdin]
21 [--verbose]
22 [--] [<file>...]
23
26 Modifies the index or directory cache. Each file mentioned is updated
27 into the index and any unmerged or needs updating state is cleared.
28 See also git-add(1) for a more user-friendly way to do some of the most
30 common operations on the index.
31 The way git update-index handles files it is told about can be modified
33 using the various options:
34
36 --add
37 If a specified file isn’t in the index already then it’s added.
38 Default behaviour is to ignore new files.
39 --remove
41 If a specified file is in the index but is missing then it’s
42 removed. Default behavior is to ignore removed file.
43 --refresh
45 Looks at the current index and checks to see if merges or updates
46 are needed by checking stat() information.
47 -q
49 Quiet. If --refresh finds that the index needs an update, the
50 default behavior is to error out. This option makes git
51 update-index continue anyway.
52 --ignore-submodules
54 Do not try to update submodules. This option is only respected when
55 passed before --refresh.
56 --unmerged
58 If --refresh finds unmerged changes in the index, the default
59 behavior is to error out. This option makes git update-index
60 continue anyway.
61 --ignore-missing
63 Ignores missing files during a --refresh
64 --cacheinfo <mode> <object> <path>
66 Directly insert the specified info into the index.
67 --index-info
69 Read index information from stdin.
70 --chmod=(+|-)x
72 Set the execute permissions on the updated files.
73 --assume-unchanged, --no-assume-unchanged
75 When these flags are specified, the object names recorded for the
76 paths are not updated. Instead, these options set and unset the
77 "assume unchanged" bit for the paths. When the "assume unchanged"
78 bit is on, git stops checking the working tree files for possible
79 modifications, so you need to manually unset the bit to tell git
80 when you change the working tree file. This is sometimes helpful
81 when working with a big project on a filesystem that has very slow
82 lstat(2) system call (e.g. cifs).
83 This option can be also used as a coarse file-level mechanism to
85 ignore uncommitted changes in tracked files (akin to what
86 .gitignore does for untracked files). Git will fail (gracefully) in
87 case it needs to modify this file in the index e.g. when merging in
88 a commit; thus, in case the assumed-untracked file is changed
89 upstream, you will need to handle the situation manually.
90 --really-refresh
92 Like --refresh, but checks stat information unconditionally,
93 without regard to the "assume unchanged" setting.
94 --skip-worktree, --no-skip-worktree
96 When one of these flags is specified, the object name recorded for
97 the paths are not updated. Instead, these options set and unset the
98 "skip-worktree" bit for the paths. See section "Skip-worktree bit"
99 below for more information.
100 -g, --again
102 Runs git update-index itself on the paths whose index entries are
103 different from those from the HEAD commit.
104 --unresolve
106 Restores the unmerged or needs updating state of a file during a
107 merge if it was cleared by accident.
108 --info-only
110 Do not create objects in the object database for all <file>
111 arguments that follow this flag; just insert their object IDs into
112 the index.
113 --force-remove
115 Remove the file from the index even when the working directory
116 still has such a file. (Implies --remove.)
117 --replace
119 By default, when a file path exists in the index, git update-index
120 refuses an attempt to add path/file. Similarly if a file path/file
121 exists, a file path cannot be added. With --replace flag, existing
122 entries that conflict with the entry being added are automatically
123 removed with warning messages.
124 --stdin
126 Instead of taking list of paths from the command line, read list of
127 paths from the standard input. Paths are separated by LF (i.e. one
128 path per line) by default.
129 --verbose
131 Report what is being added and removed from index.
132 -z
134 Only meaningful with --stdin or --index-info; paths are separated
135 with NUL character instead of LF.
136 --
138 Do not interpret any more arguments as options.
139 <file>
141 Files to act on. Note that files beginning with . are discarded.
142 This includes ./file and dir/./file. If you don’t want this, then
143 use cleaner names. The same applies to directories ending / and
144 paths with //
145
147 --refresh does not calculate a new sha1 file or bring the index
148 up-to-date for mode/content changes. But what it does do is to
149 "re-match" the stat information of a file with the index, so that you
150 can refresh the index for a file that hasn’t been changed but where the
151 stat entry is out of date.
152 For example, you’d want to do this after doing a git read-tree, to link
154 up the stat index details with the proper files.
155
157 --cacheinfo is used to register a file that is not in the current
158 working directory. This is useful for minimum-checkout merging.
159 To pretend you have a file with mode and sha1 at path, say:
161 $ git update-index --cacheinfo mode sha1 path
163 --info-only is used to register files without placing them in the
166 object database. This is useful for status-only repositories.
167 Both --cacheinfo and --info-only behave similarly: the index is updated
169 but the object database isn’t. --cacheinfo is useful when the object is
170 in the database but the file isn’t available locally. --info-only is
171 useful when the file is available, but you do not wish to update the
172 object database.
173
175 --index-info is a more powerful mechanism that lets you feed multiple
176 entry definitions from the standard input, and designed specifically
177 for scripts. It can take inputs of three formats:
178 1. mode SP sha1 TAB path
180 The first format is what "git-apply --index-info" reports, and used
182 to reconstruct a partial tree that is used for phony merge base
183 tree when falling back on 3-way merge.
184 2. mode SP type SP sha1 TAB path
186 The second format is to stuff git ls-tree output into the index
188 file.
189 3. mode SP sha1 SP stage TAB path
191 This format is to put higher order stages into the index file and
193 matches git ls-files --stage output.
194 To place a higher stage entry to the index, the path should first be
196 removed by feeding a mode=0 entry for the path, and then feeding
197 necessary input lines in the third format.
198 For example, starting with this index:
200 $ git ls-files -s
202 100644 8a1218a1024a212bb3db30becd860315f9f3ac52 0 frotz
203 you can feed the following input to --index-info:
206 $ git update-index --index-info
208 0 0000000000000000000000000000000000000000 frotz
209 100644 8a1218a1024a212bb3db30becd860315f9f3ac52 1 frotz
210 100755 8a1218a1024a212bb3db30becd860315f9f3ac52 2 frotz
211 The first line of the input feeds 0 as the mode to remove the path; the
214 SHA1 does not matter as long as it is well formatted. Then the second
215 and third line feeds stage 1 and stage 2 entries for that path. After
216 the above, we would end up with this:
217 $ git ls-files -s
219 100644 8a1218a1024a212bb3db30becd860315f9f3ac52 1 frotz
220 100755 8a1218a1024a212bb3db30becd860315f9f3ac52 2 frotz
221
224 Many operations in git depend on your filesystem to have an efficient
225 lstat(2) implementation, so that st_mtime information for working tree
226 files can be cheaply checked to see if the file contents have changed
227 from the version recorded in the index file. Unfortunately, some
228 filesystems have inefficient lstat(2). If your filesystem is one of
229 them, you can set "assume unchanged" bit to paths you have not changed
230 to cause git not to do this check. Note that setting this bit on a path
231 does not mean git will check the contents of the file to see if it has
232 changed — it makes git to omit any checking and assume it has not
233 changed. When you make changes to working tree files, you have to
234 explicitly tell git about it by dropping "assume unchanged" bit, either
235 before or after you modify them.
236 In order to set "assume unchanged" bit, use --assume-unchanged option.
238 To unset, use --no-assume-unchanged.
239 The command looks at core.ignorestat configuration variable. When this
241 is true, paths updated with git update-index paths... and paths updated
242 with other git commands that update both index and working tree (e.g.
243 git apply --index, git checkout-index -u, and git read-tree -u) are
244 automatically marked as "assume unchanged". Note that "assume
245 unchanged" bit is not set if git update-index --refresh finds the
246 working tree file matches the index (use git update-index
247 --really-refresh if you want to mark them as "assume unchanged").
248
250 To update and refresh only the files already checked out:
251 $ git checkout-index -n -f -a && git update-index --ignore-missing --refresh
253 On an inefficient filesystem with core.ignorestat set
256 $ git update-index --really-refresh [1m(1)
258 $ git update-index --no-assume-unchanged foo.c [1m(2)
259 $ git diff --name-only [1m(3)
260 $ edit foo.c
261 $ git diff --name-only [1m(4)
262 M foo.c
263 $ git update-index foo.c [1m(5)
264 $ git diff --name-only [1m(6)
265 $ edit foo.c
266 $ git diff --name-only [1m(7)
267 $ git update-index --no-assume-unchanged foo.c [1m(8)
268 $ git diff --name-only [1m(9)
269 M foo.c
270 1. forces lstat(2) to set "assume unchanged" bits for paths that
272 match index.
273 2. mark the path to be edited.
274 3. this does lstat(2) and finds index matches the path.
275 4. this does lstat(2) and finds index does not match the path.
276 5. registering the new version to index sets "assume unchanged"
277 bit.
278 6. and it is assumed unchanged.
279 7. even after you edit it.
280 8. you can tell about the change after the fact.
281 9. now it checks with lstat(2) and finds it has been changed.
282
284 Skip-worktree bit can be defined in one (long) sentence: When reading
285 an entry, if it is marked as skip-worktree, then Git pretends its
286 working directory version is up to date and read the index version
287 instead.
288 To elaborate, "reading" means checking for file existence, reading file
290 attributes or file content. The working directory version may be
291 present or absent. If present, its content may match against the index
292 version or not. Writing is not affected by this bit, content safety is
293 still first priority. Note that Git can update working directory file,
294 that is marked skip-worktree, if it is safe to do so (i.e. working
295 directory version matches index version)
296 Although this bit looks similar to assume-unchanged bit, its goal is
298 different from assume-unchanged bit’s. Skip-worktree also takes
299 precedence over assume-unchanged bit when both are set.
300
302 The command honors core.filemode configuration variable. If your
303 repository is on a filesystem whose executable bits are unreliable,
304 this should be set to false (see git-config(1)). This causes the
305 command to ignore differences in file modes recorded in the index and
306 the file mode on the filesystem if they differ only on executable bit.
307 On such an unfortunate filesystem, you may need to use git update-index
308 --chmod=.
309 Quite similarly, if core.symlinks configuration variable is set to
311 false (see git-config(1)), symbolic links are checked out as plain
312 files, and this command does not modify a recorded file mode from
313 symbolic link to regular file.
314 The command looks at core.ignorestat configuration variable. See Using
316 "assume unchanged" bit section above.
317 The command also looks at core.trustctime configuration variable. It
319 can be useful when the inode change time is regularly modified by
320 something outside Git (file system crawlers and backup systems use
321 ctime for marking files processed) (see git-config(1)).
322
324 git-config(1), git-add(1)
325
327 Written by Linus Torvalds <torvalds@osdl.org[1]>
328
330 Documentation by David Greaves, Junio C Hamano and the git-list
331 <git@vger.kernel.org[2]>.
332
334 Part of the git(1) suite
335
337 1. torvalds@osdl.org
338 mailto:torvalds@osdl.org
339 2. git@vger.kernel.org
341 mailto:git@vger.kernel.org
342Git 1.7.4.4 04/11/2011 GIT-UPDATE-INDEX(1)