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 USAGE="[--env-filter <command>] [--tree-filter <command>] \
  12 [--index-filter <command>] [--parent-filter <command>] \
  13 [--msg-filter <command>] [--commit-filter <command>] \
  14 [--tag-name-filter <command>] [--subdirectory-filter <directory>] \
  15 [--original <namespace>] [-d <directory>] [-f | --force] \
  16 [<rev-list options>...]"
  17
  18 . git-sh-setup
  19
  20 warn () {
  21         echo "$*" >&2
  22 }
  23
  24 map()
  25 {
  26         # if it was not rewritten, take the original
  27         if test -r "$workdir/../map/$1"
  28         then
  29                 cat "$workdir/../map/$1"
  30         else
  31                 echo "$1"
  32         fi
  33 }
  34
  35 # override die(): this version puts in an extra line break, so that
  36 # the progress is still visible
  37
  38 die()
  39 {
  40         echo >&2
  41         echo "$*" >&2
  42         exit 1
  43 }
  44
  45 # When piped a commit, output a script to set the ident of either
  46 # "author" or "committer
  47
  48 set_ident () {
  49         lid="$(echo "$1" | tr "A-Z" "a-z")"
  50         uid="$(echo "$1" | tr "a-z" "A-Z")"
  51         pick_id_script='
  52                 /^'$lid' /{
  53                         s/'\''/'\''\\'\'\''/g
  54                         h
  55                         s/^'$lid' \([^<]*\) <[^>]*> .*$/\1/
  56                         s/'\''/'\''\'\'\''/g
  57                         s/.*/export GIT_'$uid'_NAME='\''&'\''/p
  58
  59                         g
  60                         s/^'$lid' [^<]* <\([^>]*\)> .*$/\1/
  61                         s/'\''/'\''\'\'\''/g
  62                         s/.*/export GIT_'$uid'_EMAIL='\''&'\''/p
  63
  64                         g
  65                         s/^'$lid' [^<]* <[^>]*> \(.*\)$/\1/
  66                         s/'\''/'\''\'\'\''/g
  67                         s/.*/export GIT_'$uid'_DATE='\''&'\''/p
  68
  69                         q
  70                 }
  71         '
  72
  73         LANG=C LC_ALL=C sed -ne "$pick_id_script"
  74         # Ensure non-empty id name.
  75         echo "[ -n \"\$GIT_${uid}_NAME\" ] || export GIT_${uid}_NAME=\"\${GIT_${uid}_EMAIL%%@*}\""
  76 }
  77
  78 tempdir=.git-rewrite
  79 filter_env=
  80 filter_tree=
  81 filter_index=
  82 filter_parent=
  83 filter_msg=cat
  84 filter_commit='git commit-tree "$@"'
  85 filter_tag_name=
  86 filter_subdir=
  87 orig_namespace=refs/original/
  88 force=
  89 while case "$#" in 0) usage;; esac
  90 do
  91         case "$1" in
  92         --)
  93                 shift
  94                 break
  95                 ;;
  96         --force|-f)
  97                 shift
  98                 force=t
  99                 continue
 100                 ;;
 101         -*)
 102                 ;;
 103         *)
 104                 break;
 105         esac
 106
 107         # all switches take one argument
 108         ARG="$1"
 109         case "$#" in 1) usage ;; esac
 110         shift
 111         OPTARG="$1"
 112         shift
 113
 114         case "$ARG" in
 115         -d)
 116                 tempdir="$OPTARG"
 117                 ;;
 118         --env-filter)
 119                 filter_env="$OPTARG"
 120                 ;;
 121         --tree-filter)
 122                 filter_tree="$OPTARG"
 123                 ;;
 124         --index-filter)
 125                 filter_index="$OPTARG"
 126                 ;;
 127         --parent-filter)
 128                 filter_parent="$OPTARG"
 129                 ;;
 130         --msg-filter)
 131                 filter_msg="$OPTARG"
 132                 ;;
 133         --commit-filter)
 134                 filter_commit="$OPTARG"
 135                 ;;
 136         --tag-name-filter)
 137                 filter_tag_name="$OPTARG"
 138                 ;;
 139         --subdirectory-filter)
 140                 filter_subdir="$OPTARG"
 141                 ;;
 142         --original)
 143                 orig_namespace=$(expr "$OPTARG/" : '\(.*[^/]\)/*$')/
 144                 ;;
 145         *)
 146                 usage
 147                 ;;
 148         esac
 149 done
 150
 151 case "$force" in
 152 t)
 153         rm -rf "$tempdir"
 154 ;;
 155 '')
 156         test -d "$tempdir" &&
 157                 die "$tempdir already exists, please remove it"
 158 esac
 159 mkdir -p "$tempdir/t" &&
 160 tempdir="$(cd "$tempdir"; pwd)" &&
 161 cd "$tempdir/t" &&
 162 workdir="$(pwd)" ||
 163 die ""
 164
 165 # Make sure refs/original is empty
 166 git for-each-ref > "$tempdir"/backup-refs
 167 while read sha1 type name
 168 do
 169         case "$force,$name" in
 170         ,$orig_namespace*)
 171                 die "Namespace $orig_namespace not empty"
 172         ;;
 173         t,$orig_namespace*)
 174                 git update-ref -d "$name" $sha1
 175         ;;
 176         esac
 177 done < "$tempdir"/backup-refs
 178
 179 export GIT_DIR GIT_WORK_TREE=.
 180
 181 # These refs should be updated if their heads were rewritten
 182
 183 git rev-parse --revs-only --symbolic "$@" |
 184 while read ref
 185 do
 186         # normalize ref
 187         case "$ref" in
 188         HEAD)
 189                 ref="$(git symbolic-ref "$ref")"
 190         ;;
 191         refs/*)
 192         ;;
 193         *)
 194                 ref="$(git for-each-ref --format='%(refname)' |
 195                         grep /"$ref")"
 196         esac
 197
 198         git check-ref-format "$ref" && echo "$ref"
 199 done > "$tempdir"/heads
 200
 201 test -s "$tempdir"/heads ||
 202         die "Which ref do you want to rewrite?"
 203
 204 export GIT_INDEX_FILE="$(pwd)/../index"
 205 git read-tree || die "Could not seed the index"
 206
 207 ret=0
 208
 209 # map old->new commit ids for rewriting parents
 210 mkdir ../map || die "Could not create map/ directory"
 211
 212 case "$filter_subdir" in
 213 "")
 214         git rev-list --reverse --topo-order --default HEAD \
 215                 --parents "$@"
 216         ;;
 217 *)
 218         git rev-list --reverse --topo-order --default HEAD \
 219                 --parents --full-history "$@" -- "$filter_subdir"
 220 esac > ../revs || die "Could not get the commits"
 221 commits=$(wc -l <../revs | tr -d " ")
 222
 223 test $commits -eq 0 && die "Found nothing to rewrite"
 224
 225 # Rewrite the commits
 226
 227 i=0
 228 while read commit parents; do
 229         i=$(($i+1))
 230         printf "\rRewrite $commit ($i/$commits)"
 231
 232         case "$filter_subdir" in
 233         "")
 234                 git read-tree -i -m $commit
 235                 ;;
 236         *)
 237                 git read-tree -i -m $commit:"$filter_subdir"
 238         esac || die "Could not initialize the index"
 239
 240         export GIT_COMMIT=$commit
 241         git cat-file commit "$commit" >../commit ||
 242                 die "Cannot read commit $commit"
 243
 244         eval "$(set_ident AUTHOR <../commit)" ||
 245                 die "setting author failed for commit $commit"
 246         eval "$(set_ident COMMITTER <../commit)" ||
 247                 die "setting committer failed for commit $commit"
 248         eval "$filter_env" < /dev/null ||
 249                 die "env filter failed: $filter_env"
 250
 251         if [ "$filter_tree" ]; then
 252                 git checkout-index -f -u -a ||
 253                         die "Could not checkout the index"
 254                 # files that $commit removed are now still in the working tree;
 255                 # remove them, else they would be added again
 256                 git ls-files -z --others | xargs -0 rm -f
 257                 eval "$filter_tree" < /dev/null ||
 258                         die "tree filter failed: $filter_tree"
 259
 260                 git diff-index -r $commit | cut -f 2- | tr '\n' '\0' | \
 261                         xargs -0 git update-index --add --replace --remove
 262                 git ls-files -z --others | \
 263                         xargs -0 git update-index --add --replace --remove
 264         fi
 265
 266         eval "$filter_index" < /dev/null ||
 267                 die "index filter failed: $filter_index"
 268
 269         parentstr=
 270         for parent in $parents; do
 271                 for reparent in $(map "$parent"); do
 272                         parentstr="$parentstr -p $reparent"
 273                 done
 274         done
 275         if [ "$filter_parent" ]; then
 276                 parentstr="$(echo "$parentstr" | eval "$filter_parent")" ||
 277                                 die "parent filter failed: $filter_parent"
 278         fi
 279
 280         sed -e '1,/^$/d' <../commit | \
 281                 eval "$filter_msg" > ../message ||
 282                         die "msg filter failed: $filter_msg"
 283         sh -c "$filter_commit" "git commit-tree" \
 284                 $(git write-tree) $parentstr < ../message > ../map/$commit
 285 done <../revs
 286
 287 # In case of a subdirectory filter, it is possible that a specified head
 288 # is not in the set of rewritten commits, because it was pruned by the
 289 # revision walker.  Fix it by mapping these heads to the next rewritten
 290 # ancestor(s), i.e. the boundaries in the set of rewritten commits.
 291
 292 # NEEDSWORK: we should sort the unmapped refs topologically first
 293 while read ref
 294 do
 295         sha1=$(git rev-parse "$ref"^0)
 296         test -f "$workdir"/../map/$sha1 && continue
 297         # Assign the boundarie(s) in the set of rewritten commits
 298         # as the replacement commit(s).
 299         # (This would look a bit nicer if --not --stdin worked.)
 300         for p in $( (cd "$workdir"/../map; ls | sed "s/^/^/") |
 301                 git rev-list $ref --boundary --stdin |
 302                 sed -n "s/^-//p")
 303         do
 304                 map $p >> "$workdir"/../map/$sha1
 305         done
 306 done < "$tempdir"/heads
 307
 308 # Finally update the refs
 309
 310 _x40='[0-9a-f][0-9a-f][0-9a-f][0-9a-f][0-9a-f]'
 311 _x40="$_x40$_x40$_x40$_x40$_x40$_x40$_x40$_x40"
 312 count=0
 313 echo
 314 while read ref
 315 do
 316         # avoid rewriting a ref twice
 317         test -f "$orig_namespace$ref" && continue
 318
 319         sha1=$(git rev-parse "$ref"^0)
 320         rewritten=$(map $sha1)
 321
 322         test $sha1 = "$rewritten" &&
 323                 warn "WARNING: Ref '$ref' is unchanged" &&
 324                 continue
 325
 326         case "$rewritten" in
 327         '')
 328                 echo "Ref '$ref' was deleted"
 329                 git update-ref -m "filter-branch: delete" -d "$ref" $sha1 ||
 330                         die "Could not delete $ref"
 331         ;;
 332         $_x40)
 333                 echo "Ref '$ref' was rewritten"
 334                 git update-ref -m "filter-branch: rewrite" \
 335                                 "$ref" $rewritten $sha1 ||
 336                         die "Could not rewrite $ref"
 337         ;;
 338         *)
 339                 # NEEDSWORK: possibly add -Werror, making this an error
 340                 warn "WARNING: '$ref' was rewritten into multiple commits:"
 341                 warn "$rewritten"
 342                 warn "WARNING: Ref '$ref' points to the first one now."
 343                 rewritten=$(echo "$rewritten" | head -n 1)
 344                 git update-ref -m "filter-branch: rewrite to first" \
 345                                 "$ref" $rewritten $sha1 ||
 346                         die "Could not rewrite $ref"
 347         ;;
 348         esac
 349         git update-ref -m "filter-branch: backup" "$orig_namespace$ref" $sha1
 350         count=$(($count+1))
 351 done < "$tempdir"/heads
 352
 353 # TODO: This should possibly go, with the semantics that all positive given
 354 #       refs are updated, and their original heads stored in refs/original/
 355 # Filter tags
 356
 357 if [ "$filter_tag_name" ]; then
 358         git for-each-ref --format='%(objectname) %(objecttype) %(refname)' refs/tags |
 359         while read sha1 type ref; do
 360                 ref="${ref#refs/tags/}"
 361                 # XXX: Rewrite tagged trees as well?
 362                 if [ "$type" != "commit" -a "$type" != "tag" ]; then
 363                         continue;
 364                 fi
 365
 366                 if [ "$type" = "tag" ]; then
 367                         # Dereference to a commit
 368                         sha1t="$sha1"
 369                         sha1="$(git rev-parse "$sha1"^{commit} 2>/dev/null)" || continue
 370                 fi
 371
 372                 [ -f "../map/$sha1" ] || continue
 373                 new_sha1="$(cat "../map/$sha1")"
 374                 export GIT_COMMIT="$sha1"
 375                 new_ref="$(echo "$ref" | eval "$filter_tag_name")" ||
 376                         die "tag name filter failed: $filter_tag_name"
 377
 378                 echo "$ref -> $new_ref ($sha1 -> $new_sha1)"
 379
 380                 if [ "$type" = "tag" ]; then
 381                         # Warn that we are not rewriting the tag object itself.
 382                         warn "unreferencing tag object $sha1t"
 383                 fi
 384
 385                 git update-ref "refs/tags/$new_ref" "$new_sha1" ||
 386                         die "Could not write tag $new_ref"
 387         done
 388 fi
 389
 390 cd ../..
 391 rm -rf "$tempdir"
 392 echo
 393 test $count -gt 0 && echo "These refs were rewritten:"
 394 git show-ref | grep ^"$orig_namespace"
 395
 396 exit $ret