From 5ac2715f2eaacc7c76ac03680a0d7a16a30946f2 Mon Sep 17 00:00:00 2001 From: Junio C Hamano Date: Sat, 13 Jan 2007 13:33:07 -0800 Subject: [PATCH 1/2] Consistent message encoding while reusing log from an existing commit. The following commands can reuse log message from an existing commit while creating a new commit: git-cherry-pick git-rebase (both with and without --merge) git-commit (-c and -C) When the original commit was made in a different encoding from the current i18n.commitencoding, "cat-file commit" would give a string that is inconsistent with what the resulting commit will claim to be in. Replace them with "git show -s --encoding". "git-rebase" without --merge is "git format-patch" piped to "git am" in essence, and has been taken care of before this commit. Signed-off-by: Junio C Hamano --- git-commit.sh | 7 +++++-- git-revert.sh | 14 ++++++++++---- 2 files changed, 15 insertions(+), 6 deletions(-) diff --git a/git-commit.sh b/git-commit.sh index eddd863015..b6387239dd 100755 --- a/git-commit.sh +++ b/git-commit.sh @@ -435,7 +435,9 @@ then fi elif test "$use_commit" != "" then - git-cat-file commit "$use_commit" | sed -e '1,/^$/d' + encoding=$(git repo-config i18n.commitencoding || echo UTF-8) + git show -s --pretty=raw --encoding="$encoding" "$use_commit" | + sed -e '1,/^$/d' -e 's/^ //' elif test -f "$GIT_DIR/MERGE_MSG" then cat "$GIT_DIR/MERGE_MSG" @@ -497,7 +499,8 @@ then q } ' - set_author_env=`git-cat-file commit "$use_commit" | + encoding=$(git repo-config i18n.commitencoding || echo UTF-8) + set_author_env=`git show -s --pretty=raw --encoding="$encoding" "$use_commit" | LANG=C LC_ALL=C sed -ne "$pick_author_script"` eval "$set_author_env" export GIT_AUTHOR_NAME diff --git a/git-revert.sh b/git-revert.sh index fcca3ebb90..fcbefb4e68 100755 --- a/git-revert.sh +++ b/git-revert.sh @@ -78,6 +78,8 @@ prev=$(git-rev-parse --verify "$commit^1" 2>/dev/null) || git-rev-parse --verify "$commit^2" >/dev/null 2>&1 && die "Cannot run $me a multi-parent commit." +encoding=$(git repo-config i18n.commitencoding || echo UTF-8) + # "commit" is an existing commit. We would want to apply # the difference it introduces since its first parent "prev" # on top of the current HEAD if we are cherry-pick. Or the @@ -85,10 +87,11 @@ git-rev-parse --verify "$commit^2" >/dev/null 2>&1 && case "$me" in revert) - git-rev-list --pretty=oneline --max-count=1 $commit | + git show -s --pretty=oneline --encoding="$encoding" $commit | sed -e ' s/^[^ ]* /Revert "/ - s/$/"/' + s/$/"/ + ' echo echo "This reverts commit $commit." test "$rev" = "$commit" || @@ -117,14 +120,17 @@ cherry-pick) q }' - set_author_env=`git-cat-file commit "$commit" | + + logmsg=`git show -s --pretty=raw --encoding="$encoding" "$commit"` + set_author_env=`echo "$logmsg" | LANG=C LC_ALL=C sed -ne "$pick_author_script"` eval "$set_author_env" export GIT_AUTHOR_NAME export GIT_AUTHOR_EMAIL export GIT_AUTHOR_DATE - git-cat-file commit $commit | sed -e '1,/^$/d' + echo "$logmsg" | + sed -e '1,/^$/d' -e 's/^ //' case "$replay" in '') echo "(cherry picked from commit $commit)" From c34c6008bcf2c66e17a97acc89be1144a6216f3f Mon Sep 17 00:00:00 2001 From: Junio C Hamano Date: Sat, 13 Jan 2007 13:34:44 -0800 Subject: [PATCH 2/2] More tests in t3901. This adds tests for "cherry-pick" and "rebase --merge" (and indirectly "commit -C" since it is used in the latter) to make sure they create a new commit with correct encoding. Signed-off-by: Junio C Hamano --- t/t3901-i18n-patch.sh | 175 +++++++++++++++++++++++++++++++++--------- 1 file changed, 138 insertions(+), 37 deletions(-) diff --git a/t/t3901-i18n-patch.sh b/t/t3901-i18n-patch.sh index 7fecfe98d2..eda0e2d729 100755 --- a/t/t3901-i18n-patch.sh +++ b/t/t3901-i18n-patch.sh @@ -7,6 +7,29 @@ test_description='i18n settings and format-patch | am pipe' . ./test-lib.sh +check_encoding () { + # Make sure characters are not corrupted + cnt="$1" header="$2" i=1 j=0 bad=0 + while test "$i" -le $cnt + do + git format-patch --encoding=UTF-8 --stdout HEAD~$i..HEAD~$j | + grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" && + git-cat-file commit HEAD~$j | + case "$header" in + 8859) + grep "^encoding ISO-8859-1" ;; + *) + ! grep "^encoding ISO-8859-1" ;; + esac || { + bad=1 + break + } + j=$i + i=$(($i+1)) + done + (exit $bad) +} + test_expect_success setup ' git-repo-config i18n.commitencoding UTF-8 && @@ -66,7 +89,7 @@ test_expect_success 'format-patch output (UTF-8)' ' grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-u2 ' -test_expect_success 'rebase (UTF-8)' ' +test_expect_success 'rebase (U/U)' ' # We want the result of rebase in UTF-8 git-repo-config i18n.commitencoding UTF-8 && @@ -82,17 +105,10 @@ test_expect_success 'rebase (UTF-8)' ' git checkout -b test && git-rebase master && - # Check the results. - git format-patch --stdout HEAD~2..HEAD^ >out-r1 && - git format-patch --stdout HEAD^ >out-r2 && - grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-r1 && - grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-r2 - - ! git-cat-file commit HEAD | grep "^encoding ISO-8859-1" && - ! git-cat-file commit HEAD^ | grep "^encoding ISO-8859-1" + check_encoding 2 ' -test_expect_success 'rebase (ISO-8859-1)' ' +test_expect_success 'rebase (U/L)' ' git-repo-config i18n.commitencoding UTF-8 && git repo-config i18n.logoutputencoding ISO-8859-1 && . ../t3901-utf8.txt && @@ -100,17 +116,10 @@ test_expect_success 'rebase (ISO-8859-1)' ' git reset --hard side && git-rebase master && - git repo-config i18n.logoutputencoding UTF-8 && - git format-patch --stdout HEAD~2..HEAD^ >out-r1 && - git format-patch --stdout HEAD^ >out-r2 && - grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-r1 && - grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-r2 && - - ! git-cat-file commit HEAD | grep "^encoding ISO-8859-1" && - ! git-cat-file commit HEAD^ | grep "^encoding ISO-8859-1" + check_encoding 2 ' -test_expect_success 'rebase (ISO-8859-1)' ' +test_expect_success 'rebase (L/L)' ' # In this test we want ISO-8859-1 encoded commits as the result git-repo-config i18n.commitencoding ISO-8859-1 && git repo-config i18n.logoutputencoding ISO-8859-1 && @@ -119,18 +128,10 @@ test_expect_success 'rebase (ISO-8859-1)' ' git reset --hard side && git-rebase master && - # Make sure characters are not corrupted. - git repo-config i18n.logoutputencoding UTF-8 && - git format-patch --stdout HEAD~2..HEAD^ >out-r1 && - git format-patch --stdout HEAD^ >out-r2 && - grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-r1 && - grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-r2 && - - git-cat-file commit HEAD | grep "^encoding ISO-8859-1" && - git-cat-file commit HEAD^ | grep "^encoding ISO-8859-1" + check_encoding 2 8859 ' -test_expect_success 'rebase (UTF-8)' ' +test_expect_success 'rebase (L/U)' ' # This is pathological -- use UTF-8 as intermediate form # to get ISO-8859-1 results. git-repo-config i18n.commitencoding ISO-8859-1 && @@ -140,15 +141,115 @@ test_expect_success 'rebase (UTF-8)' ' git reset --hard side && git-rebase master && - # Make sure characters are not corrupted. - git repo-config i18n.logoutputencoding UTF-8 && - git format-patch --stdout HEAD~2..HEAD^ >out-r1 && - git format-patch --stdout HEAD^ >out-r2 && - grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-r1 && - grep "^From: =?UTF-8?q?=C3=81=C3=A9=C3=AD_=C3=B3=C3=BA?=" out-r2 && + check_encoding 2 8859 +' - git-cat-file commit HEAD | grep "^encoding ISO-8859-1" && - git-cat-file commit HEAD^ | grep "^encoding ISO-8859-1" +test_expect_success 'cherry-pick(U/U)' ' + # Both the commitencoding and logoutputencoding is set to UTF-8. + + git-repo-config i18n.commitencoding UTF-8 && + git repo-config i18n.logoutputencoding UTF-8 && + . ../t3901-utf8.txt && + + git reset --hard master && + git cherry-pick side^ && + git cherry-pick side && + EDITOR=: VISUAL=: git revert HEAD && + + check_encoding 3 +' + +test_expect_success 'cherry-pick(L/L)' ' + # Both the commitencoding and logoutputencoding is set to ISO-8859-1 + + git-repo-config i18n.commitencoding ISO-8859-1 && + git repo-config i18n.logoutputencoding ISO-8859-1 && + . ../t3901-8859-1.txt && + + git reset --hard master && + git cherry-pick side^ && + git cherry-pick side && + EDITOR=: VISUAL=: git revert HEAD && + + check_encoding 3 8859 +' + +test_expect_success 'cherry-pick(U/L)' ' + # Commitencoding is set to UTF-8 but logoutputencoding is ISO-8859-1 + + git-repo-config i18n.commitencoding UTF-8 && + git repo-config i18n.logoutputencoding ISO-8859-1 && + . ../t3901-utf8.txt && + + git reset --hard master && + git cherry-pick side^ && + git cherry-pick side && + EDITOR=: VISUAL=: git revert HEAD && + + check_encoding 3 +' + +test_expect_success 'cherry-pick(L/U)' ' + # Again, the commitencoding is set to ISO-8859-1 but + # logoutputencoding is set to UTF-8. + + git-repo-config i18n.commitencoding ISO-8859-1 && + git repo-config i18n.logoutputencoding UTF-8 && + . ../t3901-8859-1.txt && + + git reset --hard master && + git cherry-pick side^ && + git cherry-pick side && + EDITOR=: VISUAL=: git revert HEAD && + + check_encoding 3 8859 +' + +test_expect_success 'rebase --merge (U/U)' ' + git-repo-config i18n.commitencoding UTF-8 && + git repo-config i18n.logoutputencoding UTF-8 && + . ../t3901-utf8.txt && + + git reset --hard side && + git-rebase --merge master && + + check_encoding 2 +' + +test_expect_success 'rebase --merge (U/L)' ' + git-repo-config i18n.commitencoding UTF-8 && + git repo-config i18n.logoutputencoding ISO-8859-1 && + . ../t3901-utf8.txt && + + git reset --hard side && + git-rebase --merge master && + + check_encoding 2 +' + +test_expect_success 'rebase --merge (L/L)' ' + # In this test we want ISO-8859-1 encoded commits as the result + git-repo-config i18n.commitencoding ISO-8859-1 && + git repo-config i18n.logoutputencoding ISO-8859-1 && + . ../t3901-8859-1.txt && + + git reset --hard side && + git-rebase --merge master && + + check_encoding 2 8859 +' + +test_expect_success 'rebase --merge (L/U)' ' + # This is pathological -- use UTF-8 as intermediate form + # to get ISO-8859-1 results. + git-repo-config i18n.commitencoding ISO-8859-1 && + git repo-config i18n.logoutputencoding UTF-8 && + . ../t3901-8859-1.txt && + + git reset --hard side && + git-rebase --merge master && + + check_encoding 2 8859 ' test_done