Closed fnareoh closed 1 month ago
As discussed, I added back the display of patches hunk and unified the display of patch line "category".
The error on unit test on windows seems to have been due outdated cassettes with faulty indices I'm really surprised the other OS work (and it is a bit worrying).
Attention: Patch coverage is 98.00000%
with 2 lines
in your changes missing coverage. Please review.
Please upload report for BASE (
agateau/fix-extended-match@8b5d34a
). Learn more about missing BASE report.
Files | Patch % | Lines |
---|---|---|
ggshield/core/lines.py | 85.71% | 2 Missing :warning: |
: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.
Minor remarks with suggestions, but feel free to merge :smile:
Context
After having worked on factorizing
make_matches
the new goal is to start storing context lines in the Extended Match and have the output entirely depend those line and no longer theresult.file.content
. In particular this means a way to censor Extended Match based on the lines and the indices in MatchSpan and no longer relying on censor content.What has been done
This MR is divided into several commits and test should pass after each commit (best reviewed by commit?).
Several important change have been made:
Line
class has been changed from aNamedTuple
to aSimpleNamespace
to be able to modify the line.content for censoring.Validation
Test pass and output is more coherent now ? I did my best to make it reviewable but it was quite tricky to work with...
PR check list
skip-changelog
label has been added to the PR.