Merge branch 'jc/post-simplify' into tr/filter-branch
[git/git.git] / git-filter-branch.sh
CommitLineData
6f6826c5
JS
1#!/bin/sh
2#
3# Rewrite revision history
4# Copyright (c) Petr Baudis, 2006
5# Minimal changes to "port" it to core-git (c) Johannes Schindelin, 2007
6#
c401b33c
JS
7# Lets you rewrite the revision history of the current branch, creating
8# a new branch. You can specify a number of filters to modify the commits,
9# files and trees.
6f6826c5 10
16ed34ad
JS
11# The following functions will also be available in the commit filter:
12
13functions=$(cat << \EOF
b5669a05
SP
14warn () {
15 echo "$*" >&2
16}
17
6f6826c5
JS
18map()
19{
3520e1e8 20 # if it was not rewritten, take the original
c57a3494
JS
21 if test -r "$workdir/../map/$1"
22 then
23 cat "$workdir/../map/$1"
24 else
25 echo "$1"
26 fi
6f6826c5
JS
27}
28
f95eef15
JS
29# if you run 'skip_commit "$@"' in a commit filter, it will print
30# the (mapped) parents, effectively skipping the commit.
31
32skip_commit()
33{
34 shift;
35 while [ -n "$1" ];
36 do
37 shift;
38 map "$1";
39 shift;
40 done;
41}
42
8c1ce0f4
JS
43# override die(): this version puts in an extra line break, so that
44# the progress is still visible
45
46die()
47{
48 echo >&2
49 echo "$*" >&2
50 exit 1
51}
16ed34ad
JS
52EOF
53)
54
55eval "$functions"
8c1ce0f4 56
6f6826c5
JS
57# When piped a commit, output a script to set the ident of either
58# "author" or "committer
59
60set_ident () {
40a7ce64
JK
61 lid="$(echo "$1" | tr "[A-Z]" "[a-z]")"
62 uid="$(echo "$1" | tr "[a-z]" "[A-Z]")"
6f6826c5
JS
63 pick_id_script='
64 /^'$lid' /{
65 s/'\''/'\''\\'\'\''/g
66 h
67 s/^'$lid' \([^<]*\) <[^>]*> .*$/\1/
68 s/'\''/'\''\'\'\''/g
3addc94a 69 s/.*/GIT_'$uid'_NAME='\''&'\''; export GIT_'$uid'_NAME/p
6f6826c5
JS
70
71 g
72 s/^'$lid' [^<]* <\([^>]*\)> .*$/\1/
73 s/'\''/'\''\'\'\''/g
3addc94a 74 s/.*/GIT_'$uid'_EMAIL='\''&'\''; export GIT_'$uid'_EMAIL/p
6f6826c5
JS
75
76 g
77 s/^'$lid' [^<]* <[^>]*> \(.*\)$/\1/
78 s/'\''/'\''\'\'\''/g
3addc94a 79 s/.*/GIT_'$uid'_DATE='\''&'\''; export GIT_'$uid'_DATE/p
6f6826c5
JS
80
81 q
82 }
83 '
84
85 LANG=C LC_ALL=C sed -ne "$pick_id_script"
86 # Ensure non-empty id name.
3addc94a 87 echo "case \"\$GIT_${uid}_NAME\" in \"\") GIT_${uid}_NAME=\"\${GIT_${uid}_EMAIL%%@*}\" && export GIT_${uid}_NAME;; esac"
6f6826c5
JS
88}
89
7e0f1704
JS
90USAGE="[--env-filter <command>] [--tree-filter <command>] \
91[--index-filter <command>] [--parent-filter <command>] \
92[--msg-filter <command>] [--commit-filter <command>] \
93[--tag-name-filter <command>] [--subdirectory-filter <directory>] \
94[--original <namespace>] [-d <directory>] [-f | --force] \
bb8eebb9 95[<rev-list options>...]"
7e0f1704 96
8f321a39 97OPTIONS_SPEC=
7e0f1704
JS
98. git-sh-setup
99
a4661b01
PB
100if [ "$(is_bare_repository)" = false ]; then
101 git diff-files --quiet &&
38762c47 102 git diff-index --cached --quiet HEAD -- ||
46eb449c 103 die "Cannot rewrite branch(es) with a dirty working directory."
a4661b01 104fi
46eb449c 105
6f6826c5 106tempdir=.git-rewrite
6f6826c5
JS
107filter_env=
108filter_tree=
109filter_index=
110filter_parent=
111filter_msg=cat
5be60078 112filter_commit='git commit-tree "$@"'
6f6826c5 113filter_tag_name=
685ef546 114filter_subdir=
dfd05e38
JS
115orig_namespace=refs/original/
116force=
822f7c73 117while :
6f6826c5
JS
118do
119 case "$1" in
120 --)
121 shift
122 break
123 ;;
dfd05e38
JS
124 --force|-f)
125 shift
126 force=t
127 continue
128 ;;
6f6826c5
JS
129 -*)
130 ;;
131 *)
132 break;
133 esac
134
135 # all switches take one argument
136 ARG="$1"
137 case "$#" in 1) usage ;; esac
138 shift
139 OPTARG="$1"
140 shift
141
142 case "$ARG" in
143 -d)
144 tempdir="$OPTARG"
145 ;;
6f6826c5
JS
146 --env-filter)
147 filter_env="$OPTARG"
148 ;;
149 --tree-filter)
150 filter_tree="$OPTARG"
151 ;;
152 --index-filter)
153 filter_index="$OPTARG"
154 ;;
155 --parent-filter)
156 filter_parent="$OPTARG"
157 ;;
158 --msg-filter)
159 filter_msg="$OPTARG"
160 ;;
161 --commit-filter)
16ed34ad 162 filter_commit="$functions; $OPTARG"
6f6826c5
JS
163 ;;
164 --tag-name-filter)
165 filter_tag_name="$OPTARG"
166 ;;
685ef546
JS
167 --subdirectory-filter)
168 filter_subdir="$OPTARG"
169 ;;
dfd05e38 170 --original)
55ced83d 171 orig_namespace=$(expr "$OPTARG/" : '\(.*[^/]\)/*$')/
dfd05e38 172 ;;
6f6826c5
JS
173 *)
174 usage
175 ;;
176 esac
177done
178
dfd05e38
JS
179case "$force" in
180t)
181 rm -rf "$tempdir"
182;;
183'')
184 test -d "$tempdir" &&
185 die "$tempdir already exists, please remove it"
186esac
af580e9c 187mkdir -p "$tempdir/t" &&
dfd05e38 188tempdir="$(cd "$tempdir"; pwd)" &&
af580e9c
JS
189cd "$tempdir/t" &&
190workdir="$(pwd)" ||
191die ""
6f6826c5 192
def16e71
BC
193# Remove tempdir on exit
194trap 'cd ../..; rm -rf "$tempdir"' 0
195
dfd05e38
JS
196# Make sure refs/original is empty
197git for-each-ref > "$tempdir"/backup-refs
198while read sha1 type name
199do
200 case "$force,$name" in
201 ,$orig_namespace*)
202 die "Namespace $orig_namespace not empty"
203 ;;
204 t,$orig_namespace*)
205 git update-ref -d "$name" $sha1
206 ;;
207 esac
208done < "$tempdir"/backup-refs
209
46eb449c
JS
210ORIG_GIT_DIR="$GIT_DIR"
211ORIG_GIT_WORK_TREE="$GIT_WORK_TREE"
212ORIG_GIT_INDEX_FILE="$GIT_INDEX_FILE"
3addc94a
JS
213GIT_WORK_TREE=.
214export GIT_DIR GIT_WORK_TREE
6f6826c5 215
418fa3a5 216# The refs should be updated if their heads were rewritten
0f047f3b 217git rev-parse --no-flags --revs-only --symbolic-full-name --default HEAD "$@" |
418fa3a5 218sed -e '/^^/d' >"$tempdir"/heads
dfd05e38
JS
219
220test -s "$tempdir"/heads ||
221 die "Which ref do you want to rewrite?"
222
3addc94a
JS
223GIT_INDEX_FILE="$(pwd)/../index"
224export GIT_INDEX_FILE
af580e9c 225git read-tree || die "Could not seed the index"
6f6826c5
JS
226
227ret=0
228
af580e9c
JS
229# map old->new commit ids for rewriting parents
230mkdir ../map || die "Could not create map/ directory"
6f6826c5 231
685ef546
JS
232case "$filter_subdir" in
233"")
5be60078 234 git rev-list --reverse --topo-order --default HEAD \
813b4734 235 --parents "$@"
685ef546
JS
236 ;;
237*)
5be60078 238 git rev-list --reverse --topo-order --default HEAD \
a17171b4 239 --parents "$@" -- "$filter_subdir"
af580e9c 240esac > ../revs || die "Could not get the commits"
9d6f220c 241commits=$(wc -l <../revs | tr -d " ")
6f6826c5
JS
242
243test $commits -eq 0 && die "Found nothing to rewrite"
244
dfd05e38
JS
245# Rewrite the commits
246
6f6826c5 247i=0
813b4734 248while read commit parents; do
c12764b8 249 i=$(($i+1))
5efb48b5 250 printf "\rRewrite $commit ($i/$commits)"
6f6826c5 251
685ef546
JS
252 case "$filter_subdir" in
253 "")
5be60078 254 git read-tree -i -m $commit
685ef546
JS
255 ;;
256 *)
5b044ac3
JH
257 # The commit may not have the subdirectory at all
258 err=$(git read-tree -i -m $commit:"$filter_subdir" 2>&1) || {
259 if ! git rev-parse --verify $commit:"$filter_subdir" 2>/dev/null
260 then
261 rm -f "$GIT_INDEX_FILE"
262 else
263 echo >&2 "$err"
264 false
265 fi
266 }
af580e9c 267 esac || die "Could not initialize the index"
6f6826c5 268
3addc94a
JS
269 GIT_COMMIT=$commit
270 export GIT_COMMIT
af580e9c
JS
271 git cat-file commit "$commit" >../commit ||
272 die "Cannot read commit $commit"
6f6826c5 273
8c1ce0f4
JS
274 eval "$(set_ident AUTHOR <../commit)" ||
275 die "setting author failed for commit $commit"
276 eval "$(set_ident COMMITTER <../commit)" ||
277 die "setting committer failed for commit $commit"
278 eval "$filter_env" < /dev/null ||
279 die "env filter failed: $filter_env"
6f6826c5
JS
280
281 if [ "$filter_tree" ]; then
af580e9c
JS
282 git checkout-index -f -u -a ||
283 die "Could not checkout the index"
6f6826c5
JS
284 # files that $commit removed are now still in the working tree;
285 # remove them, else they would be added again
6a589fda 286 git clean -d -q -f -x
8c1ce0f4
JS
287 eval "$filter_tree" < /dev/null ||
288 die "tree filter failed: $filter_tree"
289
1fe32cb9
JH
290 (
291 git diff-index -r --name-only $commit
292 git ls-files --others
293 ) |
294 git update-index --add --replace --remove --stdin
6f6826c5
JS
295 fi
296
8c1ce0f4
JS
297 eval "$filter_index" < /dev/null ||
298 die "index filter failed: $filter_index"
6f6826c5
JS
299
300 parentstr=
813b4734 301 for parent in $parents; do
3520e1e8
JS
302 for reparent in $(map "$parent"); do
303 parentstr="$parentstr -p $reparent"
304 done
6f6826c5
JS
305 done
306 if [ "$filter_parent" ]; then
8c1ce0f4
JS
307 parentstr="$(echo "$parentstr" | eval "$filter_parent")" ||
308 die "parent filter failed: $filter_parent"
6f6826c5
JS
309 fi
310
311 sed -e '1,/^$/d' <../commit | \
8c1ce0f4
JS
312 eval "$filter_msg" > ../message ||
313 die "msg filter failed: $filter_msg"
4bf9f27d 314 @SHELL_PATH@ -c "$filter_commit" "git commit-tree" \
8c1ce0f4 315 $(git write-tree) $parentstr < ../message > ../map/$commit
6f6826c5
JS
316done <../revs
317
dfd05e38
JS
318# In case of a subdirectory filter, it is possible that a specified head
319# is not in the set of rewritten commits, because it was pruned by the
a0e46390
TR
320# revision walker. Fix it by mapping these heads to the unique nearest
321# ancestor that survived the pruning.
dfd05e38 322
a0e46390
TR
323if test "$filter_subdir"
324then
325 while read ref
dfd05e38 326 do
a0e46390
TR
327 sha1=$(git rev-parse "$ref"^0)
328 test -f "$workdir"/../map/$sha1 && continue
329 ancestor=$(git rev-list -1 $ref -- "$filter_subdir")
330 test "$ancestor" && echo $(map $ancestor) >> "$workdir"/../map/$sha1
331 done < "$tempdir"/heads
332fi
dfd05e38
JS
333
334# Finally update the refs
335
336_x40='[0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f]'
337_x40="$_x40$_x40$_x40$_x40$_x40$_x40$_x40$_x40"
dfd05e38
JS
338echo
339while read ref
340do
341 # avoid rewriting a ref twice
342 test -f "$orig_namespace$ref" && continue
343
344 sha1=$(git rev-parse "$ref"^0)
345 rewritten=$(map $sha1)
346
347 test $sha1 = "$rewritten" &&
348 warn "WARNING: Ref '$ref' is unchanged" &&
349 continue
350
351 case "$rewritten" in
352 '')
353 echo "Ref '$ref' was deleted"
354 git update-ref -m "filter-branch: delete" -d "$ref" $sha1 ||
355 die "Could not delete $ref"
98409060 356 ;;
dfd05e38
JS
357 $_x40)
358 echo "Ref '$ref' was rewritten"
261044e8
TR
359 if ! git update-ref -m "filter-branch: rewrite" \
360 "$ref" $rewritten $sha1 2>/dev/null; then
361 if test $(git cat-file -t "$ref") = tag; then
362 if test -z "$filter_tag_name"; then
363 warn "WARNING: You said to rewrite tagged commits, but not the corresponding tag."
364 warn "WARNING: Perhaps use '--tag-name-filter cat' to rewrite the tag."
365 fi
366 else
367 die "Could not rewrite $ref"
368 fi
369 fi
98409060 370 ;;
dfd05e38
JS
371 *)
372 # NEEDSWORK: possibly add -Werror, making this an error
373 warn "WARNING: '$ref' was rewritten into multiple commits:"
374 warn "$rewritten"
375 warn "WARNING: Ref '$ref' points to the first one now."
376 rewritten=$(echo "$rewritten" | head -n 1)
377 git update-ref -m "filter-branch: rewrite to first" \
378 "$ref" $rewritten $sha1 ||
379 die "Could not rewrite $ref"
380 ;;
381 esac
382 git update-ref -m "filter-branch: backup" "$orig_namespace$ref" $sha1
dfd05e38
JS
383done < "$tempdir"/heads
384
385# TODO: This should possibly go, with the semantics that all positive given
386# refs are updated, and their original heads stored in refs/original/
387# Filter tags
6f6826c5
JS
388
389if [ "$filter_tag_name" ]; then
5be60078 390 git for-each-ref --format='%(objectname) %(objecttype) %(refname)' refs/tags |
6f6826c5
JS
391 while read sha1 type ref; do
392 ref="${ref#refs/tags/}"
393 # XXX: Rewrite tagged trees as well?
394 if [ "$type" != "commit" -a "$type" != "tag" ]; then
395 continue;
396 fi
397
398 if [ "$type" = "tag" ]; then
399 # Dereference to a commit
400 sha1t="$sha1"
5be60078 401 sha1="$(git rev-parse "$sha1"^{commit} 2>/dev/null)" || continue
6f6826c5
JS
402 fi
403
404 [ -f "../map/$sha1" ] || continue
405 new_sha1="$(cat "../map/$sha1")"
3addc94a
JS
406 GIT_COMMIT="$sha1"
407 export GIT_COMMIT
8c1ce0f4
JS
408 new_ref="$(echo "$ref" | eval "$filter_tag_name")" ||
409 die "tag name filter failed: $filter_tag_name"
6f6826c5
JS
410
411 echo "$ref -> $new_ref ($sha1 -> $new_sha1)"
412
413 if [ "$type" = "tag" ]; then
1bf6551e
BC
414 new_sha1=$(git cat-file tag "$ref" |
415 sed -n \
416 -e "1,/^$/{
417 s/^object .*/object $new_sha1/
418 s/^type .*/type commit/
419 s/^tag .*/tag $new_ref/
420 }" \
421 -e '/^-----BEGIN PGP SIGNATURE-----/q' \
422 -e 'p' |
423 git mktag) ||
424 die "Could not create new tag object for $ref"
425 if git cat-file tag "$ref" | \
426 grep '^-----BEGIN PGP SIGNATURE-----' >/dev/null 2>&1
427 then
428 warn "gpg signature stripped from tag object $sha1t"
429 fi
6f6826c5
JS
430 fi
431
af580e9c
JS
432 git update-ref "refs/tags/$new_ref" "$new_sha1" ||
433 die "Could not write tag $new_ref"
6f6826c5
JS
434 done
435fi
436
437cd ../..
438rm -rf "$tempdir"
6f6826c5 439
def16e71
BC
440trap - 0
441
a4661b01
PB
442if [ "$(is_bare_repository)" = false ]; then
443 unset GIT_DIR GIT_WORK_TREE GIT_INDEX_FILE
444 test -z "$ORIG_GIT_DIR" || {
445 GIT_DIR="$ORIG_GIT_DIR" && export GIT_DIR
446 }
447 test -z "$ORIG_GIT_WORK_TREE" || {
448 GIT_WORK_TREE="$ORIG_GIT_WORK_TREE" &&
449 export GIT_WORK_TREE
450 }
451 test -z "$ORIG_GIT_INDEX_FILE" || {
452 GIT_INDEX_FILE="$ORIG_GIT_INDEX_FILE" &&
453 export GIT_INDEX_FILE
454 }
455 git read-tree -u -m HEAD
456fi
46eb449c 457
6f6826c5 458exit $ret