Merge branch 'maint'
[git/git.git] / Documentation / git-read-tree.txt
1 git-read-tree(1)
2 ================
3
4 NAME
5 ----
6 git-read-tree - Reads tree information into the index
7
8
9 SYNOPSIS
10 --------
11 'git-read-tree' (<tree-ish> | [[-m [--aggressive] | --reset | --prefix=<prefix>] [-u | -i]] [--exclude-per-directory=<gitignore>] <tree-ish1> [<tree-ish2> [<tree-ish3>]])
12
13
14 DESCRIPTION
15 -----------
16 Reads the tree information given by <tree-ish> into the index,
17 but does not actually *update* any of the files it "caches". (see:
18 gitlink:git-checkout-index[1])
19
20 Optionally, it can merge a tree into the index, perform a
21 fast-forward (i.e. 2-way) merge, or a 3-way merge, with the `-m`
22 flag. When used with `-m`, the `-u` flag causes it to also update
23 the files in the work tree with the result of the merge.
24
25 Trivial merges are done by `git-read-tree` itself. Only conflicting paths
26 will be in unmerged state when `git-read-tree` returns.
27
28 OPTIONS
29 -------
30 -m::
31 Perform a merge, not just a read. The command will
32 refuse to run if your index file has unmerged entries,
33 indicating that you have not finished previous merge you
34 started.
35
36 --reset::
37 Same as -m, except that unmerged entries are discarded
38 instead of failing.
39
40 -u::
41 After a successful merge, update the files in the work
42 tree with the result of the merge.
43
44 -i::
45 Usually a merge requires the index file as well as the
46 files in the working tree are up to date with the
47 current head commit, in order not to lose local
48 changes. This flag disables the check with the working
49 tree and is meant to be used when creating a merge of
50 trees that are not directly related to the current
51 working tree status into a temporary index file.
52
53 --aggressive::
54 Usually a three-way merge by `git-read-tree` resolves
55 the merge for really trivial cases and leaves other
56 cases unresolved in the index, so that Porcelains can
57 implement different merge policies. This flag makes the
58 command to resolve a few more cases internally:
59 +
60 * when one side removes a path and the other side leaves the path
61 unmodified. The resolution is to remove that path.
62 * when both sides remove a path. The resolution is to remove that path.
63 * when both sides adds a path identically. The resolution
64 is to add that path.
65
66 --prefix=<prefix>/::
67 Keep the current index contents, and read the contents
68 of named tree-ish under directory at `<prefix>`. The
69 original index file cannot have anything at the path
70 `<prefix>` itself, and have nothing in `<prefix>/`
71 directory. Note that the `<prefix>/` value must end
72 with a slash.
73
74 --exclude-per-directory=<gitignore>::
75 When running the command with `-u` and `-m` options, the
76 merge result may need to overwrite paths that are not
77 tracked in the current branch. The command usually
78 refuses to proceed with the merge to avoid losing such a
79 path. However this safety valve sometimes gets in the
80 way. For example, it often happens that the other
81 branch added a file that used to be a generated file in
82 your branch, and the safety valve triggers when you try
83 to switch to that branch after you ran `make` but before
84 running `make clean` to remove the generated file. This
85 option tells the command to read per-directory exclude
86 file (usually '.gitignore') and allows such an untracked
87 but explicitly ignored file to be overwritten.
88
89 <tree-ish#>::
90 The id of the tree object(s) to be read/merged.
91
92
93 Merging
94 -------
95 If `-m` is specified, `git-read-tree` can perform 3 kinds of
96 merge, a single tree merge if only 1 tree is given, a
97 fast-forward merge with 2 trees, or a 3-way merge if 3 trees are
98 provided.
99
100
101 Single Tree Merge
102 ~~~~~~~~~~~~~~~~~
103 If only 1 tree is specified, git-read-tree operates as if the user did not
104 specify `-m`, except that if the original index has an entry for a
105 given pathname, and the contents of the path matches with the tree
106 being read, the stat info from the index is used. (In other words, the
107 index's stat()s take precedence over the merged tree's).
108
109 That means that if you do a `git-read-tree -m <newtree>` followed by a
110 `git-checkout-index -f -u -a`, the `git-checkout-index` only checks out
111 the stuff that really changed.
112
113 This is used to avoid unnecessary false hits when `git-diff-files` is
114 run after `git-read-tree`.
115
116
117 Two Tree Merge
118 ~~~~~~~~~~~~~~
119
120 Typically, this is invoked as `git-read-tree -m $H $M`, where $H
121 is the head commit of the current repository, and $M is the head
122 of a foreign tree, which is simply ahead of $H (i.e. we are in a
123 fast forward situation).
124
125 When two trees are specified, the user is telling git-read-tree
126 the following:
127
128 1. The current index and work tree is derived from $H, but
129 the user may have local changes in them since $H;
130
131 2. The user wants to fast-forward to $M.
132
133 In this case, the `git-read-tree -m $H $M` command makes sure
134 that no local change is lost as the result of this "merge".
135 Here are the "carry forward" rules:
136
137 I (index) H M Result
138 -------------------------------------------------------
139 0 nothing nothing nothing (does not happen)
140 1 nothing nothing exists use M
141 2 nothing exists nothing remove path from index
142 3 nothing exists exists use M
143
144 clean I==H I==M
145 ------------------
146 4 yes N/A N/A nothing nothing keep index
147 5 no N/A N/A nothing nothing keep index
148
149 6 yes N/A yes nothing exists keep index
150 7 no N/A yes nothing exists keep index
151 8 yes N/A no nothing exists fail
152 9 no N/A no nothing exists fail
153
154 10 yes yes N/A exists nothing remove path from index
155 11 no yes N/A exists nothing fail
156 12 yes no N/A exists nothing fail
157 13 no no N/A exists nothing fail
158
159 clean (H=M)
160 ------
161 14 yes exists exists keep index
162 15 no exists exists keep index
163
164 clean I==H I==M (H!=M)
165 ------------------
166 16 yes no no exists exists fail
167 17 no no no exists exists fail
168 18 yes no yes exists exists keep index
169 19 no no yes exists exists keep index
170 20 yes yes no exists exists use M
171 21 no yes no exists exists fail
172
173 In all "keep index" cases, the index entry stays as in the
174 original index file. If the entry were not up to date,
175 git-read-tree keeps the copy in the work tree intact when
176 operating under the -u flag.
177
178 When this form of git-read-tree returns successfully, you can
179 see what "local changes" you made are carried forward by running
180 `git-diff-index --cached $M`. Note that this does not
181 necessarily match `git-diff-index --cached $H` would have
182 produced before such a two tree merge. This is because of cases
183 18 and 19 --- if you already had the changes in $M (e.g. maybe
184 you picked it up via e-mail in a patch form), `git-diff-index
185 --cached $H` would have told you about the change before this
186 merge, but it would not show in `git-diff-index --cached $M`
187 output after two-tree merge.
188
189
190 3-Way Merge
191 ~~~~~~~~~~~
192 Each "index" entry has two bits worth of "stage" state. stage 0 is the
193 normal one, and is the only one you'd see in any kind of normal use.
194
195 However, when you do `git-read-tree` with three trees, the "stage"
196 starts out at 1.
197
198 This means that you can do
199
200 ----------------
201 $ git-read-tree -m <tree1> <tree2> <tree3>
202 ----------------
203
204 and you will end up with an index with all of the <tree1> entries in
205 "stage1", all of the <tree2> entries in "stage2" and all of the
206 <tree3> entries in "stage3". When performing a merge of another
207 branch into the current branch, we use the common ancestor tree
208 as <tree1>, the current branch head as <tree2>, and the other
209 branch head as <tree3>.
210
211 Furthermore, `git-read-tree` has special-case logic that says: if you see
212 a file that matches in all respects in the following states, it
213 "collapses" back to "stage0":
214
215 - stage 2 and 3 are the same; take one or the other (it makes no
216 difference - the same work has been done on our branch in
217 stage 2 and their branch in stage 3)
218
219 - stage 1 and stage 2 are the same and stage 3 is different; take
220 stage 3 (our branch in stage 2 did not do anything since the
221 ancestor in stage 1 while their branch in stage 3 worked on
222 it)
223
224 - stage 1 and stage 3 are the same and stage 2 is different take
225 stage 2 (we did something while they did nothing)
226
227 The `git-write-tree` command refuses to write a nonsensical tree, and it
228 will complain about unmerged entries if it sees a single entry that is not
229 stage 0.
230
231 OK, this all sounds like a collection of totally nonsensical rules,
232 but it's actually exactly what you want in order to do a fast
233 merge. The different stages represent the "result tree" (stage 0, aka
234 "merged"), the original tree (stage 1, aka "orig"), and the two trees
235 you are trying to merge (stage 2 and 3 respectively).
236
237 The order of stages 1, 2 and 3 (hence the order of three
238 <tree-ish> command line arguments) are significant when you
239 start a 3-way merge with an index file that is already
240 populated. Here is an outline of how the algorithm works:
241
242 - if a file exists in identical format in all three trees, it will
243 automatically collapse to "merged" state by git-read-tree.
244
245 - a file that has _any_ difference what-so-ever in the three trees
246 will stay as separate entries in the index. It's up to "porcelain
247 policy" to determine how to remove the non-0 stages, and insert a
248 merged version.
249
250 - the index file saves and restores with all this information, so you
251 can merge things incrementally, but as long as it has entries in
252 stages 1/2/3 (i.e., "unmerged entries") you can't write the result. So
253 now the merge algorithm ends up being really simple:
254
255 * you walk the index in order, and ignore all entries of stage 0,
256 since they've already been done.
257
258 * if you find a "stage1", but no matching "stage2" or "stage3", you
259 know it's been removed from both trees (it only existed in the
260 original tree), and you remove that entry.
261
262 * if you find a matching "stage2" and "stage3" tree, you remove one
263 of them, and turn the other into a "stage0" entry. Remove any
264 matching "stage1" entry if it exists too. .. all the normal
265 trivial rules ..
266
267 You would normally use `git-merge-index` with supplied
268 `git-merge-one-file` to do this last step. The script updates
269 the files in the working tree as it merges each path and at the
270 end of a successful merge.
271
272 When you start a 3-way merge with an index file that is already
273 populated, it is assumed that it represents the state of the
274 files in your work tree, and you can even have files with
275 changes unrecorded in the index file. It is further assumed
276 that this state is "derived" from the stage 2 tree. The 3-way
277 merge refuses to run if it finds an entry in the original index
278 file that does not match stage 2.
279
280 This is done to prevent you from losing your work-in-progress
281 changes, and mixing your random changes in an unrelated merge
282 commit. To illustrate, suppose you start from what has been
283 committed last to your repository:
284
285 ----------------
286 $ JC=`git-rev-parse --verify "HEAD^0"`
287 $ git-checkout-index -f -u -a $JC
288 ----------------
289
290 You do random edits, without running git-update-index. And then
291 you notice that the tip of your "upstream" tree has advanced
292 since you pulled from him:
293
294 ----------------
295 $ git-fetch git://.... linus
296 $ LT=`cat .git/FETCH_HEAD`
297 ----------------
298
299 Your work tree is still based on your HEAD ($JC), but you have
300 some edits since. Three-way merge makes sure that you have not
301 added or modified index entries since $JC, and if you haven't,
302 then does the right thing. So with the following sequence:
303
304 ----------------
305 $ git-read-tree -m -u `git-merge-base $JC $LT` $JC $LT
306 $ git-merge-index git-merge-one-file -a
307 $ echo "Merge with Linus" | \
308 git-commit-tree `git-write-tree` -p $JC -p $LT
309 ----------------
310
311 what you would commit is a pure merge between $JC and $LT without
312 your work-in-progress changes, and your work tree would be
313 updated to the result of the merge.
314
315 However, if you have local changes in the working tree that
316 would be overwritten by this merge,`git-read-tree` will refuse
317 to run to prevent your changes from being lost.
318
319 In other words, there is no need to worry about what exists only
320 in the working tree. When you have local changes in a part of
321 the project that is not involved in the merge, your changes do
322 not interfere with the merge, and are kept intact. When they
323 *do* interfere, the merge does not even start (`git-read-tree`
324 complains loudly and fails without modifying anything). In such
325 a case, you can simply continue doing what you were in the
326 middle of doing, and when your working tree is ready (i.e. you
327 have finished your work-in-progress), attempt the merge again.
328
329
330 See Also
331 --------
332 gitlink:git-write-tree[1]; gitlink:git-ls-files[1]
333
334
335 Author
336 ------
337 Written by Linus Torvalds <torvalds@osdl.org>
338
339 Documentation
340 --------------
341 Documentation by David Greaves, Junio C Hamano and the git-list <git@vger.kernel.org>.
342
343 GIT
344 ---
345 Part of the gitlink:git[7] suite
346