]>
Commit | Line | Data |
---|---|---|
2cf565c5 DG |
1 | git-read-tree(1) |
2 | ================ | |
2cf565c5 DG |
3 | |
4 | NAME | |
5 | ---- | |
5f3aa197 | 6 | git-read-tree - Reads tree information into the index |
2cf565c5 DG |
7 | |
8 | ||
9 | SYNOPSIS | |
10 | -------- | |
7791a1d9 | 11 | [verse] |
5a56da58 SB |
12 | 'git read-tree' [[-m [--trivial] [--aggressive] | --reset | --prefix=<prefix>] |
13 | [-u [--exclude-per-directory=<gitignore>] | -i]] | |
a5d07d0f | 14 | [--index-output=<file>] [--no-sparse-checkout] |
fb1bb965 | 15 | (--empty | <tree-ish1> [<tree-ish2> [<tree-ish3>]]) |
ccef66b5 | 16 | |
2cf565c5 DG |
17 | |
18 | DESCRIPTION | |
19 | ----------- | |
5f3aa197 | 20 | Reads the tree information given by <tree-ish> into the index, |
c1bdacf9 | 21 | but does not actually *update* any of the files it "caches". (see: |
5162e697 | 22 | linkgit:git-checkout-index[1]) |
2cf565c5 | 23 | |
5f3aa197 | 24 | Optionally, it can merge a tree into the index, perform a |
61f693bd JL |
25 | fast-forward (i.e. 2-way) merge, or a 3-way merge, with the `-m` |
26 | flag. When used with `-m`, the `-u` flag causes it to also update | |
ccef66b5 | 27 | the files in the work tree with the result of the merge. |
2cf565c5 | 28 | |
0b444cdb TR |
29 | Trivial merges are done by 'git read-tree' itself. Only conflicting paths |
30 | will be in unmerged state when 'git read-tree' returns. | |
2cf565c5 DG |
31 | |
32 | OPTIONS | |
33 | ------- | |
34 | -m:: | |
3f41f5a9 | 35 | Perform a merge, not just a read. The command will |
36 | refuse to run if your index file has unmerged entries, | |
37 | indicating that you have not finished previous merge you | |
38 | started. | |
ccef66b5 | 39 | |
2db0bfbc | 40 | --reset:: |
3f41f5a9 | 41 | Same as -m, except that unmerged entries are discarded |
42 | instead of failing. | |
2db0bfbc | 43 | |
ccef66b5 JH |
44 | -u:: |
45 | After a successful merge, update the files in the work | |
46 | tree with the result of the merge. | |
2cf565c5 | 47 | |
f318dd22 JH |
48 | -i:: |
49 | Usually a merge requires the index file as well as the | |
cc1a2b66 | 50 | files in the working tree to be up to date with the |
f318dd22 JH |
51 | current head commit, in order not to lose local |
52 | changes. This flag disables the check with the working | |
53 | tree and is meant to be used when creating a merge of | |
54 | trees that are not directly related to the current | |
55 | working tree status into a temporary index file. | |
56 | ||
ea5070c9 JL |
57 | -n:: |
58 | --dry-run:: | |
59 | Check if the command would error out, without updating the index | |
a58088ab | 60 | or the files in the working tree for real. |
ea5070c9 | 61 | |
22e801f2 MV |
62 | -v:: |
63 | Show the progress of checking files out. | |
64 | ||
6da08783 | 65 | --trivial:: |
0b444cdb | 66 | Restrict three-way merge by 'git read-tree' to happen |
6da08783 JN |
67 | only if there is no file-level merging required, instead |
68 | of resolving merge for trivial cases and leaving | |
69 | conflicting files unresolved in the index. | |
70 | ||
afaa8d66 | 71 | --aggressive:: |
0b444cdb | 72 | Usually a three-way merge by 'git read-tree' resolves |
afaa8d66 | 73 | the merge for really trivial cases and leaves other |
cc1a2b66 | 74 | cases unresolved in the index, so that porcelains can |
afaa8d66 | 75 | implement different merge policies. This flag makes the |
cc1a2b66 | 76 | command resolve a few more cases internally: |
afaa8d66 JH |
77 | + |
78 | * when one side removes a path and the other side leaves the path | |
79 | unmodified. The resolution is to remove that path. | |
80 | * when both sides remove a path. The resolution is to remove that path. | |
cc1a2b66 | 81 | * when both sides add a path identically. The resolution |
afaa8d66 JH |
82 | is to add that path. |
83 | ||
f4c6f2d3 JH |
84 | --prefix=<prefix>/:: |
85 | Keep the current index contents, and read the contents | |
5c951ef4 CB |
86 | of the named tree-ish under the directory at `<prefix>`. |
87 | The command will refuse to overwrite entries that already | |
88 | existed in the original index file. Note that the `<prefix>/` | |
89 | value must end with a slash. | |
f4c6f2d3 | 90 | |
22f741da JH |
91 | --exclude-per-directory=<gitignore>:: |
92 | When running the command with `-u` and `-m` options, the | |
93 | merge result may need to overwrite paths that are not | |
94 | tracked in the current branch. The command usually | |
95 | refuses to proceed with the merge to avoid losing such a | |
96 | path. However this safety valve sometimes gets in the | |
97 | way. For example, it often happens that the other | |
98 | branch added a file that used to be a generated file in | |
99 | your branch, and the safety valve triggers when you try | |
100 | to switch to that branch after you ran `make` but before | |
101 | running `make clean` to remove the generated file. This | |
102 | option tells the command to read per-directory exclude | |
103 | file (usually '.gitignore') and allows such an untracked | |
104 | but explicitly ignored file to be overwritten. | |
f4c6f2d3 | 105 | |
5e7f56ac JH |
106 | --index-output=<file>:: |
107 | Instead of writing the results out to `$GIT_INDEX_FILE`, | |
108 | write the resulting index in the named file. While the | |
109 | command is operating, the original index file is locked | |
110 | with the same mechanism as usual. The file must allow | |
111 | to be rename(2)ed into from a temporary file that is | |
112 | created next to the usual index file; typically this | |
113 | means it needs to be on the same filesystem as the index | |
114 | file itself, and you need write permission to the | |
115 | directories the index file and index output file are | |
116 | located in. | |
117 | ||
a5d07d0f NTND |
118 | --no-sparse-checkout:: |
119 | Disable sparse checkout support even if `core.sparseCheckout` | |
120 | is true. | |
121 | ||
fb1bb965 JK |
122 | --empty:: |
123 | Instead of reading tree object(s) into the index, just empty | |
124 | it. | |
125 | ||
2cf565c5 DG |
126 | <tree-ish#>:: |
127 | The id of the tree object(s) to be read/merged. | |
128 | ||
129 | ||
130 | Merging | |
131 | ------- | |
0b444cdb | 132 | If `-m` is specified, 'git read-tree' can perform 3 kinds of |
ccef66b5 JH |
133 | merge, a single tree merge if only 1 tree is given, a |
134 | fast-forward merge with 2 trees, or a 3-way merge if 3 trees are | |
2cf565c5 DG |
135 | provided. |
136 | ||
ccef66b5 | 137 | |
2cf565c5 DG |
138 | Single Tree Merge |
139 | ~~~~~~~~~~~~~~~~~ | |
0b444cdb | 140 | If only 1 tree is specified, 'git read-tree' operates as if the user did not |
61f693bd | 141 | specify `-m`, except that if the original index has an entry for a |
73252839 | 142 | given pathname, and the contents of the path match with the tree |
5f3aa197 LS |
143 | being read, the stat info from the index is used. (In other words, the |
144 | index's stat()s take precedence over the merged tree's). | |
2cf565c5 | 145 | |
b1889c36 | 146 | That means that if you do a `git read-tree -m <newtree>` followed by a |
0b444cdb | 147 | `git checkout-index -f -u -a`, the 'git checkout-index' only checks out |
2cf565c5 DG |
148 | the stuff that really changed. |
149 | ||
0b444cdb TR |
150 | This is used to avoid unnecessary false hits when 'git diff-files' is |
151 | run after 'git read-tree'. | |
2cf565c5 | 152 | |
c8596009 JH |
153 | |
154 | Two Tree Merge | |
155 | ~~~~~~~~~~~~~~ | |
156 | ||
b1889c36 | 157 | Typically, this is invoked as `git read-tree -m $H $M`, where $H |
c8596009 JH |
158 | is the head commit of the current repository, and $M is the head |
159 | of a foreign tree, which is simply ahead of $H (i.e. we are in a | |
a75d7b54 | 160 | fast-forward situation). |
c8596009 | 161 | |
0b444cdb | 162 | When two trees are specified, the user is telling 'git read-tree' |
c8596009 JH |
163 | the following: |
164 | ||
df8baa42 | 165 | 1. The current index and work tree is derived from $H, but |
73252839 | 166 | the user may have local changes in them since $H. |
c8596009 | 167 | |
df8baa42 | 168 | 2. The user wants to fast-forward to $M. |
c8596009 | 169 | |
b1889c36 | 170 | In this case, the `git read-tree -m $H $M` command makes sure |
c8596009 | 171 | that no local change is lost as the result of this "merge". |
73252839 MG |
172 | Here are the "carry forward" rules, where "I" denotes the index, |
173 | "clean" means that index and work tree coincide, and "exists"/"nothing" | |
174 | refer to the presence of a path in the specified commit: | |
c8596009 | 175 | |
73252839 | 176 | I H M Result |
c8596009 | 177 | ------------------------------------------------------- |
71928f7f MG |
178 | 0 nothing nothing nothing (does not happen) |
179 | 1 nothing nothing exists use M | |
180 | 2 nothing exists nothing remove path from index | |
73252839 | 181 | 3 nothing exists exists, use M if "initial checkout", |
55218834 | 182 | H == M keep index otherwise |
73252839 | 183 | exists, fail |
55218834 | 184 | H != M |
c8596009 JH |
185 | |
186 | clean I==H I==M | |
187 | ------------------ | |
71928f7f MG |
188 | 4 yes N/A N/A nothing nothing keep index |
189 | 5 no N/A N/A nothing nothing keep index | |
c8596009 | 190 | |
71928f7f MG |
191 | 6 yes N/A yes nothing exists keep index |
192 | 7 no N/A yes nothing exists keep index | |
193 | 8 yes N/A no nothing exists fail | |
194 | 9 no N/A no nothing exists fail | |
c8596009 | 195 | |
5f3aa197 | 196 | 10 yes yes N/A exists nothing remove path from index |
c8596009 JH |
197 | 11 no yes N/A exists nothing fail |
198 | 12 yes no N/A exists nothing fail | |
199 | 13 no no N/A exists nothing fail | |
200 | ||
73252839 | 201 | clean (H==M) |
c8596009 JH |
202 | ------ |
203 | 14 yes exists exists keep index | |
204 | 15 no exists exists keep index | |
205 | ||
206 | clean I==H I==M (H!=M) | |
207 | ------------------ | |
208 | 16 yes no no exists exists fail | |
209 | 17 no no no exists exists fail | |
210 | 18 yes no yes exists exists keep index | |
211 | 19 no no yes exists exists keep index | |
212 | 20 yes yes no exists exists use M | |
213 | 21 no yes no exists exists fail | |
214 | ||
5f3aa197 | 215 | In all "keep index" cases, the index entry stays as in the |
73252839 | 216 | original index file. If the entry is not up to date, |
0b444cdb | 217 | 'git read-tree' keeps the copy in the work tree intact when |
c8596009 JH |
218 | operating under the -u flag. |
219 | ||
0b444cdb | 220 | When this form of 'git read-tree' returns successfully, you can |
73252839 | 221 | see which of the "local changes" that you made were carried forward by running |
b1889c36 | 222 | `git diff-index --cached $M`. Note that this does not |
73252839 | 223 | necessarily match what `git diff-index --cached $H` would have |
c8596009 JH |
224 | produced before such a two tree merge. This is because of cases |
225 | 18 and 19 --- if you already had the changes in $M (e.g. maybe | |
b1889c36 | 226 | you picked it up via e-mail in a patch form), `git diff-index |
61f693bd | 227 | --cached $H` would have told you about the change before this |
b1889c36 | 228 | merge, but it would not show in `git diff-index --cached $M` |
73252839 | 229 | output after the two-tree merge. |
c8596009 | 230 | |
73252839 | 231 | Case 3 is slightly tricky and needs explanation. The result from this |
55218834 | 232 | rule logically should be to remove the path if the user staged the removal |
79fd4cc7 | 233 | of the path and then switching to a new branch. That however will prevent |
55218834 | 234 | the initial checkout from happening, so the rule is modified to use M (new |
73252839 | 235 | tree) only when the content of the index is empty. Otherwise the removal |
55218834 | 236 | of the path is kept as long as $H and $M are the same. |
c8596009 | 237 | |
2cf565c5 DG |
238 | 3-Way Merge |
239 | ~~~~~~~~~~~ | |
240 | Each "index" entry has two bits worth of "stage" state. stage 0 is the | |
241 | normal one, and is the only one you'd see in any kind of normal use. | |
242 | ||
0b444cdb | 243 | However, when you do 'git read-tree' with three trees, the "stage" |
2cf565c5 DG |
244 | starts out at 1. |
245 | ||
246 | This means that you can do | |
247 | ||
61f693bd | 248 | ---------------- |
b1889c36 | 249 | $ git read-tree -m <tree1> <tree2> <tree3> |
61f693bd | 250 | ---------------- |
2cf565c5 DG |
251 | |
252 | and you will end up with an index with all of the <tree1> entries in | |
253 | "stage1", all of the <tree2> entries in "stage2" and all of the | |
bb6d7b89 JH |
254 | <tree3> entries in "stage3". When performing a merge of another |
255 | branch into the current branch, we use the common ancestor tree | |
256 | as <tree1>, the current branch head as <tree2>, and the other | |
257 | branch head as <tree3>. | |
2cf565c5 | 258 | |
0b444cdb | 259 | Furthermore, 'git read-tree' has special-case logic that says: if you see |
2cf565c5 DG |
260 | a file that matches in all respects in the following states, it |
261 | "collapses" back to "stage0": | |
262 | ||
263 | - stage 2 and 3 are the same; take one or the other (it makes no | |
bb6d7b89 JH |
264 | difference - the same work has been done on our branch in |
265 | stage 2 and their branch in stage 3) | |
2cf565c5 DG |
266 | |
267 | - stage 1 and stage 2 are the same and stage 3 is different; take | |
bb6d7b89 JH |
268 | stage 3 (our branch in stage 2 did not do anything since the |
269 | ancestor in stage 1 while their branch in stage 3 worked on | |
270 | it) | |
2cf565c5 DG |
271 | |
272 | - stage 1 and stage 3 are the same and stage 2 is different take | |
bb6d7b89 | 273 | stage 2 (we did something while they did nothing) |
2cf565c5 | 274 | |
0b444cdb | 275 | The 'git write-tree' command refuses to write a nonsensical tree, and it |
2cf565c5 DG |
276 | will complain about unmerged entries if it sees a single entry that is not |
277 | stage 0. | |
278 | ||
abda1ef5 | 279 | OK, this all sounds like a collection of totally nonsensical rules, |
2cf565c5 DG |
280 | but it's actually exactly what you want in order to do a fast |
281 | merge. The different stages represent the "result tree" (stage 0, aka | |
282 | "merged"), the original tree (stage 1, aka "orig"), and the two trees | |
283 | you are trying to merge (stage 2 and 3 respectively). | |
284 | ||
ccef66b5 | 285 | The order of stages 1, 2 and 3 (hence the order of three |
06ab60c0 | 286 | <tree-ish> command-line arguments) are significant when you |
ccef66b5 JH |
287 | start a 3-way merge with an index file that is already |
288 | populated. Here is an outline of how the algorithm works: | |
2cf565c5 DG |
289 | |
290 | - if a file exists in identical format in all three trees, it will | |
0b444cdb | 291 | automatically collapse to "merged" state by 'git read-tree'. |
2cf565c5 DG |
292 | |
293 | - a file that has _any_ difference what-so-ever in the three trees | |
2c6e4771 | 294 | will stay as separate entries in the index. It's up to "porcelain |
2cf565c5 | 295 | policy" to determine how to remove the non-0 stages, and insert a |
ccef66b5 | 296 | merged version. |
2cf565c5 DG |
297 | |
298 | - the index file saves and restores with all this information, so you | |
299 | can merge things incrementally, but as long as it has entries in | |
abda1ef5 | 300 | stages 1/2/3 (i.e., "unmerged entries") you can't write the result. So |
2cf565c5 DG |
301 | now the merge algorithm ends up being really simple: |
302 | ||
303 | * you walk the index in order, and ignore all entries of stage 0, | |
304 | since they've already been done. | |
305 | ||
306 | * if you find a "stage1", but no matching "stage2" or "stage3", you | |
307 | know it's been removed from both trees (it only existed in the | |
308 | original tree), and you remove that entry. | |
309 | ||
310 | * if you find a matching "stage2" and "stage3" tree, you remove one | |
311 | of them, and turn the other into a "stage0" entry. Remove any | |
312 | matching "stage1" entry if it exists too. .. all the normal | |
313 | trivial rules .. | |
314 | ||
0b444cdb TR |
315 | You would normally use 'git merge-index' with supplied |
316 | 'git merge-one-file' to do this last step. The script updates | |
bb6d7b89 JH |
317 | the files in the working tree as it merges each path and at the |
318 | end of a successful merge. | |
ccef66b5 JH |
319 | |
320 | When you start a 3-way merge with an index file that is already | |
321 | populated, it is assumed that it represents the state of the | |
322 | files in your work tree, and you can even have files with | |
323 | changes unrecorded in the index file. It is further assumed | |
324 | that this state is "derived" from the stage 2 tree. The 3-way | |
325 | merge refuses to run if it finds an entry in the original index | |
326 | file that does not match stage 2. | |
327 | ||
328 | This is done to prevent you from losing your work-in-progress | |
bb6d7b89 JH |
329 | changes, and mixing your random changes in an unrelated merge |
330 | commit. To illustrate, suppose you start from what has been | |
37425065 | 331 | committed last to your repository: |
ccef66b5 | 332 | |
61f693bd | 333 | ---------------- |
b1889c36 JN |
334 | $ JC=`git rev-parse --verify "HEAD^0"` |
335 | $ git checkout-index -f -u -a $JC | |
61f693bd | 336 | ---------------- |
ccef66b5 | 337 | |
0b444cdb | 338 | You do random edits, without running 'git update-index'. And then |
ccef66b5 JH |
339 | you notice that the tip of your "upstream" tree has advanced |
340 | since you pulled from him: | |
341 | ||
61f693bd | 342 | ---------------- |
b1889c36 | 343 | $ git fetch git://.... linus |
96890f4c | 344 | $ LT=`git rev-parse FETCH_HEAD` |
61f693bd | 345 | ---------------- |
ccef66b5 JH |
346 | |
347 | Your work tree is still based on your HEAD ($JC), but you have | |
348 | some edits since. Three-way merge makes sure that you have not | |
5f3aa197 | 349 | added or modified index entries since $JC, and if you haven't, |
ccef66b5 JH |
350 | then does the right thing. So with the following sequence: |
351 | ||
61f693bd | 352 | ---------------- |
b1889c36 JN |
353 | $ git read-tree -m -u `git merge-base $JC $LT` $JC $LT |
354 | $ git merge-index git-merge-one-file -a | |
61f693bd | 355 | $ echo "Merge with Linus" | \ |
b1889c36 | 356 | git commit-tree `git write-tree` -p $JC -p $LT |
61f693bd | 357 | ---------------- |
ccef66b5 | 358 | |
61f693bd | 359 | what you would commit is a pure merge between $JC and $LT without |
ccef66b5 JH |
360 | your work-in-progress changes, and your work tree would be |
361 | updated to the result of the merge. | |
362 | ||
bb6d7b89 | 363 | However, if you have local changes in the working tree that |
0b444cdb | 364 | would be overwritten by this merge, 'git read-tree' will refuse |
bb6d7b89 JH |
365 | to run to prevent your changes from being lost. |
366 | ||
367 | In other words, there is no need to worry about what exists only | |
368 | in the working tree. When you have local changes in a part of | |
369 | the project that is not involved in the merge, your changes do | |
370 | not interfere with the merge, and are kept intact. When they | |
0b444cdb | 371 | *do* interfere, the merge does not even start ('git read-tree' |
bb6d7b89 JH |
372 | complains loudly and fails without modifying anything). In such |
373 | a case, you can simply continue doing what you were in the | |
374 | middle of doing, and when your working tree is ready (i.e. you | |
375 | have finished your work-in-progress), attempt the merge again. | |
376 | ||
2cf565c5 | 377 | |
ed5336a7 NTND |
378 | Sparse checkout |
379 | --------------- | |
380 | ||
cc1a2b66 MG |
381 | "Sparse checkout" allows populating the working directory sparsely. |
382 | It uses the skip-worktree bit (see linkgit:git-update-index[1]) to tell | |
383 | Git whether a file in the working directory is worth looking at. | |
ed5336a7 | 384 | |
cc1a2b66 MG |
385 | 'git read-tree' and other merge-based commands ('git merge', 'git |
386 | checkout'...) can help maintaining the skip-worktree bitmap and working | |
ed5336a7 | 387 | directory update. `$GIT_DIR/info/sparse-checkout` is used to |
cc1a2b66 MG |
388 | define the skip-worktree reference bitmap. When 'git read-tree' needs |
389 | to update the working directory, it resets the skip-worktree bit in the index | |
ed5336a7 | 390 | based on this file, which uses the same syntax as .gitignore files. |
1f1f575e MG |
391 | If an entry matches a pattern in this file, skip-worktree will not be |
392 | set on that entry. Otherwise, skip-worktree will be set. | |
ed5336a7 NTND |
393 | |
394 | Then it compares the new skip-worktree value with the previous one. If | |
1f1f575e MG |
395 | skip-worktree turns from set to unset, it will add the corresponding |
396 | file back. If it turns from unset to set, that file will be removed. | |
ed5336a7 NTND |
397 | |
398 | While `$GIT_DIR/info/sparse-checkout` is usually used to specify what | |
cc1a2b66 MG |
399 | files are in, you can also specify what files are _not_ in, using |
400 | negate patterns. For example, to remove the file `unwanted`: | |
ed5336a7 NTND |
401 | |
402 | ---------------- | |
5e821231 | 403 | /* |
ed5336a7 NTND |
404 | !unwanted |
405 | ---------------- | |
406 | ||
cc1a2b66 | 407 | Another tricky thing is fully repopulating the working directory when you |
ed5336a7 | 408 | no longer want sparse checkout. You cannot just disable "sparse |
cc1a2b66 MG |
409 | checkout" because skip-worktree bits are still in the index and your working |
410 | directory is still sparsely populated. You should re-populate the working | |
ed5336a7 NTND |
411 | directory with the `$GIT_DIR/info/sparse-checkout` file content as |
412 | follows: | |
413 | ||
414 | ---------------- | |
5e821231 | 415 | /* |
ed5336a7 NTND |
416 | ---------------- |
417 | ||
cc1a2b66 MG |
418 | Then you can disable sparse checkout. Sparse checkout support in 'git |
419 | read-tree' and similar commands is disabled by default. You need to | |
08aefc9e NTND |
420 | turn `core.sparseCheckout` on in order to have sparse checkout |
421 | support. | |
ed5336a7 NTND |
422 | |
423 | ||
56ae8df5 | 424 | SEE ALSO |
c1bdacf9 | 425 | -------- |
5162e697 DM |
426 | linkgit:git-write-tree[1]; linkgit:git-ls-files[1]; |
427 | linkgit:gitignore[5] | |
2cf565c5 | 428 | |
2cf565c5 DG |
429 | GIT |
430 | --- | |
9e1f0a85 | 431 | Part of the linkgit:git[1] suite |