Closed ddeclerck closed 2 months ago
the test looks good, but it seems something is indeed broken in the parser ?!?
the test looks good, but it seems something is indeed broken in the parser ?!?
Yeah but this is unrelated to the conflicts mentionned in the GC3-GC4 merge PR.
It seems USAGE PROGRAM-POINTER
with a program-prototype argument has never been parsed correctly.
Are you working on this?
Are you working on this?
I'd rather postpone this issue for now for I fear it would take me too far away from the GC3-GC4 merge (plus I'm already being taken away by PR 109).
Make it an expected failure then and commit upstream, but don't merge it yet - just verify manually that the result with your GC4merge is not worse.
Make it an expected failure then and commit upstream, but don't merge it yet - just verify manually that the result with your GC4merge is not worse.
Maybe I can split the 3 cases POINTER/FUNCTION-POINTER/PROGRAM-POINTER so that we can at least "protect" what already works ?
Good idea. Possibly with a comment to combine again later. Can directly go upstream.
Also includes fixups for recently introduced CI failures (there were a C++-style comment in scanner.l
and missing stdio.h
includes both in scanner.l
and pplex.l
).
:warning: Please install the to ensure uploads and comments are reliably processed by Codecov.
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 77.55%. Comparing base (
8b7350a
) to head (8d1ba62
). Report is 30 commits behind head on gcos4gnucobol-3.x.
:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
As requested here and here.
Also it seems at some point recently some changes broke the CI - I'll investigate and update this PR.