Thread (40 messages) 40 messages, 5 authors, 5d ago

Re: [PATCH v2 2/2] ci: point test failures and fixed known breakages at their file and line

flat view

From: Junio C Hamano <hidden>
Date: 2026-09-28 20:53:30

"Harald Nordgren via GitGitGadget" [off-list ref] writes:
quoted hunk ↗ jump to hunk
 t/test-lib-github-workflow-markup.sh | 41 ++++++++++++++++++++++------
 1 file changed, 33 insertions(+), 8 deletions(-)
diff --git a/t/test-lib-github-workflow-markup.sh b/t/test-lib-github-workflow-markup.sh
index 0d54496358..67c5c3461c 100644
--- a/t/test-lib-github-workflow-markup.sh
+++ b/t/test-lib-github-workflow-markup.sh
@@ -31,6 +31,21 @@ start_test_output () {
 	github_markup_script_name=${0##*/}
 }
 
+github_escape_message_ () {
+	# A test description is always one line, so only % and CR need
+	# escaping here. Escape % first, or CR's own %-encoding gets mangled.
+	sed -e 's/%/%25/g' -e 's/\r/%0D/g'
+}
Is it portable to feed a two-letter sequence "\r" to "sed" and
expect it to be interpreted as Carriage Return?  Implementations of
BSD lineage "sed" don't grok it if I recall correctly.
quoted hunk ↗ jump to hunk
+find_test_case_line_ () {
+	# A description can contain characters like [ or * that would
+	# corrupt a regex search, so match it literally and take the first
+	# hit; -- keeps a description starting with "-" from being read as
+	# an option.
+	grep -n -F -- "$1" "$TEST_DIRECTORY/$github_markup_script_name" |
+	head -n 1 | cut -d: -f1
+}
+
 github_annotation_ () {
 	echo >>$github_markup_output "::$1 file=$2,line=$3::$4"
 }
@@ -40,21 +55,31 @@ github_annotation_ () {
 finalize_test_case_output () {
 	test_case_result=$1
 	shift
+
+	case "$test_case_result" in
+	ok|broken)
+		# Exit without printing the "ok" or "broken" tests
+		return
+		;;
+	esac
+
+	test_case_line=$(find_test_case_line_ "$1")
+	test_case_description=$(printf '%s' "$1" | github_escape_message_)
+
 	case "$test_case_result" in
 	failure)
-		echo >>$github_markup_output "::error::failed: $this_test.$test_count $1"
+		github_annotation_ error "t/$github_markup_script_name" "${test_case_line:-1}" \
+			"failed: $this_test.$test_count $test_case_description"
 		;;
 	fixed)
-		echo >>$github_markup_output "::notice::fixed: $this_test.$test_count $1"
-		;;
-	ok|broken)
-		# Exit without printing the "ok" or ""broken" tests
-		return
+		github_annotation_ notice "t/$github_markup_script_name" "${test_case_line:-1}" \
+			"fixed: $this_test.$test_count $test_case_description"
 		;;
 	esac
+
All of the above may make sense, but ...
quoted hunk ↗ jump to hunk
 	echo >>$github_markup_output "::group::$test_case_result: $this_test.$test_count $*"
-	test-tool >>$github_markup_output path-utils skip-n-bytes \
-		"$GIT_TEST_TEE_OUTPUT_FILE" $GIT_TEST_TEE_OFFSET
+	test-tool path-utils skip-n-bytes \
+		"$GIT_TEST_TEE_OUTPUT_FILE" $GIT_TEST_TEE_OFFSET >>$github_markup_output
 	echo >>$github_markup_output "::endgroup::"
What is this change about?  In the original, all surrounding code
has redirection early on the command line, and breaking that pattern
is the only difference between the removed and added lines here as
far as I can see.
 }
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help