@alvinwan said:
Is it maybe because of the following line? (maybe should be len(point)?
result = text.forward(len(point) + 1)
Yes indeed. After this change PunctuationCommandNames no longer "gobble" the following charaacter and seem to behave as expected.
I've added some tests (if tokens serialize back to string correctly, then they must have been tokenized together as a unit). By coincidence I happened to choose a delimiter \rfloor and test was failing... so we ended up with a bonus typo fix.
Last but not least I added three more verbatim-type environments (users can manually add those to skip_envs if needed, but seems like a good idea to have them "baked in").
Coverage remained the same at 100.0% when pulling 54a8c9ec89207eb53eeca46e5e2bb92c9666272a on minireference:punctuation_fix into 942bb572312c97b02881bb7486456a8bdab0c612 on alvinwan:master.
Coverage remained the same at 100.0% when pulling 54a8c9ec89207eb53eeca46e5e2bb92c9666272a on minireference:punctuation_fix into 942bb572312c97b02881bb7486456a8bdab0c612 on alvinwan:master.
Yes indeed. After this change
PunctuationCommandName
s no longer "gobble" the following charaacter and seem to behave as expected.I've added some tests (if tokens serialize back to string correctly, then they must have been tokenized together as a unit). By coincidence I happened to choose a delimiter
\rfloor
and test was failing... so we ended up with a bonus typo fix.Last but not least I added three more verbatim-type environments (users can manually add those to
skip_envs
if needed, but seems like a good idea to have them "baked in").