fast-export: do automatic reencoding of commit messages only if requested
Automatic re-encoding of commit messages (and dropping of the encoding header) hurts attempts to do reversible history rewrites (e.g. sha1sum <-> sha256sum transitions, some subtree rewrites), and seems inconsistent with the general principle followed elsewhere in fast-export of requiring explicit user requests to modify the output (e.g. --signed-tags=strip, --tag-of-filtered-object=rewrite). Add a --reencode flag that the user can use to specify, and like other fast-export flags, default it to 'abort'. Signed-off-by: Elijah Newren <newren@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Elijah Newren committed
May 13, 2019 at 21:31 UTC
e80001f8fd7f608c6d05ce0ab2ebfc90d724a307
3 files changed
+85
-6
Documentation/git-fast-export.txt
+7
@@ -129,6 +129,13 @@ marks the same across runs.
129
for intermediary filters (e.g. for rewriting commit messages
130
which refer to older commits, or for stripping blobs by id).
131
132
+--reencode=(yes|no|abort)::
133
+ Specify how to handle `encoding` header in commit objects. When
134
+ asking to 'abort' (which is the default), this program will die
135
+ when encountering such a commit object. With 'yes', the commit
136
+ message will be reencoded into UTF-8. With 'no', the original
137
+ encoding will be preserved.
138
+
139
--refspec::
140
Apply the specified refspec to each ref exported. Multiple of them can
141
be specified.
builtin/fast-export.c
+43
-3
@@ -33,6 +33,7 @@ static const char *fast_export_usage[] = {
33
static int progress;
34
static enum { SIGNED_TAG_ABORT, VERBATIM, WARN, WARN_STRIP, STRIP } signed_tag_mode = SIGNED_TAG_ABORT;
35
static enum { TAG_FILTERING_ABORT, DROP, REWRITE } tag_of_filtered_mode = TAG_FILTERING_ABORT;
36
+static enum { REENCODE_ABORT, REENCODE_YES, REENCODE_NO } reencode_mode = REENCODE_ABORT;
37
static int fake_missing_tagger;
38
static int use_done_feature;
39
static int no_data;
@@ -77,6 +78,31 @@ static int parse_opt_tag_of_filtered_mode(const struct option *opt,
78
return 0;
79
}
80
81
+static int parse_opt_reencode_mode(const struct option *opt,
82
+ const char *arg, int unset)
83
+{
84
+ if (unset) {
85
+ reencode_mode = REENCODE_ABORT;
86
+ return 0;
87
+ }
88
+
89
+ switch (git_parse_maybe_bool(arg)) {
90
+ case 0:
91
+ reencode_mode = REENCODE_NO;
92
+ break;
93
+ case 1:
94
+ reencode_mode = REENCODE_YES;
95
+ break;
96
+ default:
97
+ if (!strcasecmp(arg, "abort"))
98
+ reencode_mode = REENCODE_ABORT;
99
+ else
100
+ return error("Unknown reencoding mode: %s", arg);
101
+ }
102
+
103
+ return 0;
104
+}
105
+
106
static struct decoration idnums;
107
static uint32_t last_idnum;
108
@@ -633,10 +659,21 @@ static void handle_commit(struct commit *commit, struct rev_info *rev,
659
}
660
661
mark_next_object(&commit->object);
636
- if (anonymize)
662
+ if (anonymize) {
663
reencoded = anonymize_commit_message(message);
638
- else if (!is_encoding_utf8(encoding))
639
- reencoded = reencode_string(message, "UTF-8", encoding);
664
+ } else if (encoding) {
665
+ switch(reencode_mode) {
666
+ case REENCODE_YES:
667
+ reencoded = reencode_string(message, "UTF-8", encoding);
668
+ break;
669
+ case REENCODE_NO:
670
+ break;
671
+ case REENCODE_ABORT:
672
+ die("Encountered commit-specific encoding %s in commit "
673
+ "%s; use --reencode=[yes|no] to handle it",
674
+ encoding, oid_to_hex(&commit->object.oid));
675
+ }
676
+ }
677
if (!commit->parents)
678
printf("reset %s\n", refname);
679
printf("commit %s\nmark :%"PRIu32"\n", refname, last_idnum);
@@ -1091,6 +1128,9 @@ int cmd_fast_export(int argc, const char **argv, const char *prefix)
1128
OPT_CALLBACK(0, "tag-of-filtered-object", &tag_of_filtered_mode, N_("mode"),
1129
N_("select handling of tags that tag filtered objects"),
1130
parse_opt_tag_of_filtered_mode),
1131
+ OPT_CALLBACK(0, "reencode", &reencode_mode, N_("mode"),
1132
+ N_("select handling of commit messages in an alternate encoding"),
1133
+ parse_opt_reencode_mode),
1134
OPT_STRING(0, "export-marks", &export_filename, N_("file"),
1135
N_("Dump marks to this file")),
1136
OPT_STRING(0, "import-marks", &import_filename, N_("file"),
t/t9350-fast-export.sh
+35
-3
@@ -94,14 +94,14 @@ test_expect_success 'fast-export --show-original-ids | git fast-import' '
94
test $MUSS = $(git rev-parse --verify refs/tags/muss)
95
'
96
97
-test_expect_success 'iso-8859-7' '
97
+test_expect_success 'reencoding iso-8859-7' '
98
99
test_when_finished "git reset --hard HEAD~1" &&
100
test_config i18n.commitencoding iso-8859-7 &&
101
test_tick &&
102
echo rosten >file &&
103
git commit -s -F "$TEST_DIRECTORY/t9350/simple-iso-8859-7-commit-message.txt" file &&
104
- git fast-export wer^..wer >iso-8859-7.fi &&
104
+ git fast-export --reencode=yes wer^..wer >iso-8859-7.fi &&
105
sed "s/wer/i18n/" iso-8859-7.fi |
106
(cd new &&
107
git fast-import &&
@@ -118,13 +118,45 @@ test_expect_success 'iso-8859-7' '
118
! grep ^encoding actual)
119
'
120
121
+test_expect_success 'aborting on iso-8859-7' '
122
+
123
+ test_when_finished "git reset --hard HEAD~1" &&
124
+ test_config i18n.commitencoding iso-8859-7 &&
125
+ echo rosten >file &&
126
+ git commit -s -F "$TEST_DIRECTORY/t9350/simple-iso-8859-7-commit-message.txt" file &&
127
+ test_must_fail git fast-export --reencode=abort wer^..wer >iso-8859-7.fi
128
+'
129
+
130
+test_expect_success 'preserving iso-8859-7' '
131
+
132
+ test_when_finished "git reset --hard HEAD~1" &&
133
+ test_config i18n.commitencoding iso-8859-7 &&
134
+ echo rosten >file &&
135
+ git commit -s -F "$TEST_DIRECTORY/t9350/simple-iso-8859-7-commit-message.txt" file &&
136
+ git fast-export --reencode=no wer^..wer >iso-8859-7.fi &&
137
+ sed "s/wer/i18n-no-recoding/" iso-8859-7.fi |
138
+ (cd new &&
139
+ git fast-import &&
140
+ # The commit object, if not re-encoded, is 240 bytes.
141
+ # Removing the "encoding iso-8859-7\n" header would drops 20
142
+ # bytes. Re-encoding the Pi character from \xF0 (\360) in
143
+ # iso-8859-7 to \xCF\x80 (\317\200) in UTF-8 adds a byte.
144
+ # Check for the expected size...
145
+ test 240 -eq "$(git cat-file -s i18n-no-recoding)" &&
146
+ # ...as well as the expected byte.
147
+ git cat-file commit i18n-no-recoding >actual &&
148
+ grep $(printf "\360") actual &&
149
+ # Also make sure the commit has the "encoding" header
150
+ grep ^encoding actual)
151
+'
152
+
153
test_expect_success 'encoding preserved if reencoding fails' '
154
155
test_when_finished "git reset --hard HEAD~1" &&
156
test_config i18n.commitencoding iso-8859-7 &&
157
echo rosten >file &&
158
git commit -s -F "$TEST_DIRECTORY/t9350/broken-iso-8859-7-commit-message.txt" file &&
127
- git fast-export wer^..wer >iso-8859-7.fi &&
159
+ git fast-export --reencode=yes wer^..wer >iso-8859-7.fi &&
160
sed "s/wer/i18n-invalid/" iso-8859-7.fi |
161
(cd new &&
162
git fast-import &&