git-filter-branch.sh

   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 #
   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.
  10
  11 # The following functions will also be available in the commit filter:
  12
  13 functions=$(cat << \EOF
  14 warn () {
  15         echo "$*" >&2
  16 }
  17
  18 map()
  19 {
  20         # if it was not rewritten, take the original
  21         if test -r "$workdir/../map/$1"
  22         then
  23                 cat "$workdir/../map/$1"
  24         else
  25                 echo "$1"
  26         fi
  27 }
  28
  29 # if you run 'skip_commit "$@"' in a commit filter, it will print
  30 # the (mapped) parents, effectively skipping the commit.
  31
  32 skip_commit()
  33 {
  34         shift;
  35         while [ -n "$1" ];
  36         do
  37                 shift;
  38                 map "$1";
  39                 shift;
  40         done;
  41 }
  42
  43 # override die(): this version puts in an extra line break, so that
  44 # the progress is still visible
  45
  46 die()
  47 {
  48         echo >&2
  49         echo "$*" >&2
  50         exit 1
  51 }
  52 EOF
  53 )
  54
  55 eval "$functions"
  56
  57 # When piped a commit, output a script to set the ident of either
  58 # "author" or "committer
  59
  60 set_ident () {
  61         lid="$(echo "$1" | tr "[A-Z]" "[a-z]")"
  62         uid="$(echo "$1" | tr "[a-z]" "[A-Z]")"
  63         pick_id_script='
  64                 /^'$lid' /{
  65                         s/'\''/'\''\\'\'\''/g
  66                         h
  67                         s/^'$lid' \([^<]*\) <[^>]*> .*$/\1/
  68                         s/'\''/'\''\'\'\''/g
  69                         s/.*/GIT_'$uid'_NAME='\''&'\''; export GIT_'$uid'_NAME/p
  70
  71                         g
  72                         s/^'$lid' [^<]* <\([^>]*\)> .*$/\1/
  73                         s/'\''/'\''\'\'\''/g
  74                         s/.*/GIT_'$uid'_EMAIL='\''&'\''; export GIT_'$uid'_EMAIL/p
  75
  76                         g
  77                         s/^'$lid' [^<]* <[^>]*> \(.*\)$/\1/
  78                         s/'\''/'\''\'\'\''/g
  79                         s/.*/GIT_'$uid'_DATE='\''&'\''; export GIT_'$uid'_DATE/p
  80
  81                         q
  82                 }
  83         '
  84
  85         LANG=C LC_ALL=C sed -ne "$pick_id_script"
  86         # Ensure non-empty id name.
  87         echo "case \"\$GIT_${uid}_NAME\" in \"\") GIT_${uid}_NAME=\"\${GIT_${uid}_EMAIL%%@*}\" && export GIT_${uid}_NAME;; esac"
  88 }
  89
  90 USAGE="[--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] \
  95 [<rev-list options>...]"
  96
  97 OPTIONS_SPEC=
  98 . git-sh-setup
  99
 100 if [ "$(is_bare_repository)" = false ]; then
 101         git diff-files --quiet &&
 102         git diff-index --cached --quiet HEAD -- ||
 103         die "Cannot rewrite branch(es) with a dirty working directory."
 104 fi
 105
 106 tempdir=.git-rewrite
 107 filter_env=
 108 filter_tree=
 109 filter_index=
 110 filter_parent=
 111 filter_msg=cat
 112 filter_commit='git commit-tree "$@"'
 113 filter_tag_name=
 114 filter_subdir=
 115 orig_namespace=refs/original/
 116 force=
 117 while :
 118 do
 119         case "$1" in
 120         --)
 121                 shift
 122                 break
 123                 ;;
 124         --force|-f)
 125                 shift
 126                 force=t
 127                 continue
 128                 ;;
 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="$OPTARG"
 145                 ;;
 146         --env-filter)
 147                 filter_env="$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)
 162                 filter_commit="$functions; $OPTARG"
 163                 ;;
 164         --tag-name-filter)
 165                 filter_tag_name="$OPTARG"
 166                 ;;
 167         --subdirectory-filter)
 168                 filter_subdir="$OPTARG"
 169                 ;;
 170         --original)
 171                 orig_namespace=$(expr "$OPTARG/" : '\(.*[^/]\)/*$')/
 172                 ;;
 173         *)
 174                 usage
 175                 ;;
 176         esac
 177 done
 178
 179 case "$force" in
 180 t)
 181         rm -rf "$tempdir"
 182 ;;
 183 '')
 184         test -d "$tempdir" &&
 185                 die "$tempdir already exists, please remove it"
 186 esac
 187 mkdir -p "$tempdir/t" &&
 188 tempdir="$(cd "$tempdir"; pwd)" &&
 189 cd "$tempdir/t" &&
 190 workdir="$(pwd)" ||
 191 die ""
 192
 193 # Remove tempdir on exit
 194 trap 'cd ../..; rm -rf "$tempdir"' 0
 195
 196 # Make sure refs/original is empty
 197 git for-each-ref > "$tempdir"/backup-refs
 198 while read sha1 type name
 199 do
 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" $sha1
 206         ;;
 207         esac
 208 done < "$tempdir"/backup-refs
 209
 210 ORIG_GIT_DIR="$GIT_DIR"
 211 ORIG_GIT_WORK_TREE="$GIT_WORK_TREE"
 212 ORIG_GIT_INDEX_FILE="$GIT_INDEX_FILE"
 213 GIT_WORK_TREE=.
 214 export GIT_DIR GIT_WORK_TREE
 215
 216 # The refs should be updated if their heads were rewritten
 217 git rev-parse --no-flags --revs-only --symbolic-full-name --default HEAD "$@" |
 218 sed -e '/^^/d' >"$tempdir"/heads
 219
 220 test -s "$tempdir"/heads ||
 221         die "Which ref do you want to rewrite?"
 222
 223 GIT_INDEX_FILE="$(pwd)/../index"
 224 export GIT_INDEX_FILE
 225 git read-tree || die "Could not seed the index"
 226
 227 ret=0
 228
 229 # map old->new commit ids for rewriting parents
 230 mkdir ../map || die "Could not create map/ directory"
 231
 232 case "$filter_subdir" in
 233 "")
 234         git rev-list --reverse --topo-order --default HEAD \
 235                 --parents "$@"
 236         ;;
 237 *)
 238         git rev-list --reverse --topo-order --default HEAD \
 239                 --parents "$@" -- "$filter_subdir"
 240 esac > ../revs || die "Could not get the commits"
 241 commits=$(wc -l <../revs | tr -d " ")
 242
 243 test $commits -eq 0 && die "Found nothing to rewrite"
 244
 245 # Rewrite the commits
 246
 247 i=0
 248 while read commit parents; do
 249         i=$(($i+1))
 250         printf "\rRewrite $commit ($i/$commits)"
 251
 252         case "$filter_subdir" in
 253         "")
 254                 git read-tree -i -m $commit
 255                 ;;
 256         *)
 257                 # The commit may not have the subdirectory at all
 258                 err=$(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                 }
 267         esac || die "Could not initialize the index"
 268
 269         GIT_COMMIT=$commit
 270         export GIT_COMMIT
 271         git cat-file commit "$commit" >../commit ||
 272                 die "Cannot read commit $commit"
 273
 274         eval "$(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: $filter_env"
 280
 281         if [ "$filter_tree" ]; then
 282                 git checkout-index -f -u -a ||
 283                         die "Could not checkout the index"
 284                 # files that $commit removed are now still in the working tree;
 285                 # remove them, else they would be added again
 286                 git clean -d -q -f -x
 287                 eval "$filter_tree" < /dev/null ||
 288                         die "tree filter failed: $filter_tree"
 289
 290                 (
 291                         git diff-index -r --name-only $commit
 292                         git ls-files --others
 293                 ) |
 294                 git update-index --add --replace --remove --stdin
 295         fi
 296
 297         eval "$filter_index" < /dev/null ||
 298                 die "index filter failed: $filter_index"
 299
 300         parentstr=
 301         for parent in $parents; do
 302                 for reparent in $(map "$parent"); do
 303                         parentstr="$parentstr -p $reparent"
 304                 done
 305         done
 306         if [ "$filter_parent" ]; then
 307                 parentstr="$(echo "$parentstr" | eval "$filter_parent")" ||
 308                                 die "parent filter failed: $filter_parent"
 309         fi
 310
 311         sed -e '1,/^$/d' <../commit | \
 312                 eval "$filter_msg" > ../message ||
 313                         die "msg filter failed: $filter_msg"
 314         @SHELL_PATH@ -c "$filter_commit" "git commit-tree" \
 315                 $(git write-tree) $parentstr < ../message > ../map/$commit
 316 done <../revs
 317
 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
 320 # revision walker.  Fix it by mapping these heads to the next rewritten
 321 # ancestor(s), i.e. the boundaries in the set of rewritten commits.
 322
 323 # NEEDSWORK: we should sort the unmapped refs topologically first
 324 while read ref
 325 do
 326         sha1=$(git rev-parse "$ref"^0)
 327         test -f "$workdir"/../map/$sha1 && continue
 328         # Assign the boundarie(s) in the set of rewritten commits
 329         # as the replacement commit(s).
 330         # (This would look a bit nicer if --not --stdin worked.)
 331         for p in $( (cd "$workdir"/../map; ls | sed "s/^/^/") |
 332                 git rev-list $ref --boundary --stdin |
 333                 sed -n "s/^-//p")
 334         do
 335                 map $p >> "$workdir"/../map/$sha1
 336         done
 337 done < "$tempdir"/heads
 338
 339 # Finally update the refs
 340
 341 _x40='[0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f]'
 342 _x40="$_x40$_x40$_x40$_x40$_x40$_x40$_x40$_x40"
 343 echo
 344 while read ref
 345 do
 346         # avoid rewriting a ref twice
 347         test -f "$orig_namespace$ref" && continue
 348
 349         sha1=$(git rev-parse "$ref"^0)
 350         rewritten=$(map $sha1)
 351
 352         test $sha1 = "$rewritten" &&
 353                 warn "WARNING: Ref '$ref' is unchanged" &&
 354                 continue
 355
 356         case "$rewritten" in
 357         '')
 358                 echo "Ref '$ref' was deleted"
 359                 git update-ref -m "filter-branch: delete" -d "$ref" $sha1 ||
 360                         die "Could not delete $ref"
 361         ;;
 362         $_x40)
 363                 echo "Ref '$ref' was rewritten"
 364                 if ! git update-ref -m "filter-branch: rewrite" \
 365                                         "$ref" $rewritten $sha1 2>/dev/null; then
 366                         if test $(git cat-file -t "$ref") = tag; then
 367                                 if test -z "$filter_tag_name"; then
 368                                         warn "WARNING: You said to rewrite tagged commits, but not the corresponding tag."
 369                                         warn "WARNING: Perhaps use '--tag-name-filter cat' to rewrite the tag."
 370                                 fi
 371                         else
 372                                 die "Could not rewrite $ref"
 373                         fi
 374                 fi
 375         ;;
 376         *)
 377                 # NEEDSWORK: possibly add -Werror, making this an error
 378                 warn "WARNING: '$ref' was rewritten into multiple commits:"
 379                 warn "$rewritten"
 380                 warn "WARNING: Ref '$ref' points to the first one now."
 381                 rewritten=$(echo "$rewritten" | head -n 1)
 382                 git update-ref -m "filter-branch: rewrite to first" \
 383                                 "$ref" $rewritten $sha1 ||
 384                         die "Could not rewrite $ref"
 385         ;;
 386         esac
 387         git update-ref -m "filter-branch: backup" "$orig_namespace$ref" $sha1
 388 done < "$tempdir"/heads
 389
 390 # TODO: This should possibly go, with the semantics that all positive given
 391 #       refs are updated, and their original heads stored in refs/original/
 392 # Filter tags
 393
 394 if [ "$filter_tag_name" ]; then
 395         git for-each-ref --format='%(objectname) %(objecttype) %(refname)' refs/tags |
 396         while read sha1 type ref; do
 397                 ref="${ref#refs/tags/}"
 398                 # XXX: Rewrite tagged trees as well?
 399                 if [ "$type" != "commit" -a "$type" != "tag" ]; then
 400                         continue;
 401                 fi
 402
 403                 if [ "$type" = "tag" ]; then
 404                         # Dereference to a commit
 405                         sha1t="$sha1"
 406                         sha1="$(git rev-parse "$sha1"^{commit} 2>/dev/null)" || continue
 407                 fi
 408
 409                 [ -f "../map/$sha1" ] || continue
 410                 new_sha1="$(cat "../map/$sha1")"
 411                 GIT_COMMIT="$sha1"
 412                 export GIT_COMMIT
 413                 new_ref="$(echo "$ref" | eval "$filter_tag_name")" ||
 414                         die "tag name filter failed: $filter_tag_name"
 415
 416                 echo "$ref -> $new_ref ($sha1 -> $new_sha1)"
 417
 418                 if [ "$type" = "tag" ]; then
 419                         new_sha1=$(git cat-file tag "$ref" |
 420                                 sed -n \
 421                                     -e "1,/^$/{
 422                                           s/^object .*/object $new_sha1/
 423                                           s/^type .*/type commit/
 424                                           s/^tag .*/tag $new_ref/
 425                                         }" \
 426                                     -e '/^-----BEGIN PGP SIGNATURE-----/q' \
 427                                     -e 'p' |
 428                                 git mktag) ||
 429                                 die "Could not create new tag object for $ref"
 430                         if git cat-file tag "$ref" | \
 431                            grep '^-----BEGIN PGP SIGNATURE-----' >/dev/null 2>&1
 432                         then
 433                                 warn "gpg signature stripped from tag object $sha1t"
 434                         fi
 435                 fi
 436
 437                 git update-ref "refs/tags/$new_ref" "$new_sha1" ||
 438                         die "Could not write tag $new_ref"
 439         done
 440 fi
 441
 442 cd ../..
 443 rm -rf "$tempdir"
 444
 445 trap - 0
 446
 447 if [ "$(is_bare_repository)" = false ]; then
 448         unset GIT_DIR GIT_WORK_TREE GIT_INDEX_FILE
 449         test -z "$ORIG_GIT_DIR" || {
 450                 GIT_DIR="$ORIG_GIT_DIR" && export GIT_DIR
 451         }
 452         test -z "$ORIG_GIT_WORK_TREE" || {
 453                 GIT_WORK_TREE="$ORIG_GIT_WORK_TREE" &&
 454                 export GIT_WORK_TREE
 455         }
 456         test -z "$ORIG_GIT_INDEX_FILE" || {
 457                 GIT_INDEX_FILE="$ORIG_GIT_INDEX_FILE" &&
 458                 export GIT_INDEX_FILE
 459         }
 460         git read-tree -u -m HEAD
 461 fi
 462
 463 exit $ret