filter-repo: implement --strip-blobs-with-ids

Add a flag allowing for specifying a file filled with blob-ids which
will be stripped from the repository.

Signed-off-by: Elijah Newren <[email protected]>

Author: newren

git-filter-repo

@@ -1639,6 +1639,8 @@ EXAMPLES
                           dest='max_blob_size', default=0,
         help=_("Strip blobs (files) bigger than specified size (e.g. '5M', "
                "'2G', etc)"))
+    contents.add_argument('--strip-blobs-with-ids', metavar='BLOB-ID-FILENAME',
+        help=_("Strip blob with the specified git object ids (hashes)"))
 
     refrename = parser.add_argument_group(title=_("Renaming of refs "
                                               "(see also --refname-callback)"))
@@ -1926,6 +1928,11 @@ EXAMPLES
       args.mailmap = MailmapInfo(args.mailmap)
     if args.replace_text:
       args.replace_text = FilteringOptions.get_replace_text(args.replace_text)
+    if args.strip_blobs_with_ids:
+      with open(args.strip_blobs_with_ids, 'br') as f:
+        args.strip_blobs_with_ids = set(f.read().split())
+    else:
+      args.strip_blobs_with_ids = set()
     return args
 
 class RepoAnalyze(object):
@@ -2919,6 +2926,9 @@ class RepoFilter(object):
     if self._args.max_blob_size and len(blob.data) > self._args.max_blob_size:
       blob.skip()
 
+    if blob.original_id in self._args.strip_blobs_with_ids:
+      blob.skip()
+
     if self._args.replace_text:
       for literal, replacement in self._args.replace_text['literals']:
         blob.data = blob.data.replace(literal, replacement)
@@ -3051,6 +3061,10 @@ class RepoFilter(object):
       if self._args.max_blob_size and \
          self._unpacked_size.get(change.blob_id, 0) > self._args.max_blob_size:
         continue
+      if self._args.strip_blobs_with_ids and \
+         change.blob_id in self._args.strip_blobs_with_ids:
+        continue
+      # Otherwise, record the change
       new_file_changes[change.filename] = change
     commit.file_changes = [v for k,v in sorted(new_file_changes.items())]
 

t/t9390-filter-repo.sh

@@ -710,6 +710,51 @@ test_expect_success '--strip-blobs-bigger-than' '
 	)
 '
 
+test_expect_success '--strip-blobs-with-ids' '
+	(
+		git clone file://"$(pwd)"/analyze_me strip_blobs_with_ids &&
+		cd strip_blobs_with_ids &&
+
+		# Verify certain files are present initially
+		git log --format=%n --name-only | sort | uniq >../filenames &&
+		test_line_count = 11 ../filenames &&
+		grep fake_submodule ../filenames &&
+
+		# Strip "a certain file" files
+		git filter-repo --strip-blobs-with-ids <(echo deadbeefdeadbeefdeadbeefdeadbeefdeadbeef) &&
+
+		git log --format=%n --name-only | sort | uniq >../filenames &&
+		test_line_count = 10 ../filenames &&
+		# Make sure fake_submodule was removed
+		! grep fake_submodule ../filenames &&
+
+		# Do it again, this time with --replace-text since that means
+		# we are operating without --no-data and have to go through
+		# a different codepath.  (The search/replace terms are bogus)
+		cat >../bad-ids <<-\EOF &&
+		34b6a0c9d02cb6ef7f409f248c0c1224ce9dd373
+		51b95456de9274c9a95f756742808dfd480b9b35
+		EOF
+		cat >../replace-rules <<-\EOF &&
+		not found==>was found
+		EOF
+		git filter-repo --strip-blobs-with-ids ../bad-ids --replace-text ../replace-rules &&
+
+		git log --format=%n --name-only | sort | uniq >../filenames &&
+		test_line_count = 5 ../filenames &&
+		! grep sequence/to ../filenames &&
+		! grep words/to ../filenames &&
+		! grep capricious ../filenames &&
+		! grep fickle ../filenames &&
+		! grep mercurial ../filenames
+
+		# Remove the temporary auxiliary files
+		rm ../bad-ids &&
+		rm ../replace-rules &&
+		rm ../filenames
+	)
+'
+
 test_expect_success 'setup commit message rewriting' '
 	test_create_repo commit_msg &&
 	(