search

apache / lucene-jira-archive

Jira archive for Apache Lucene
https://lucene.apache.org/
2 stars 6 forks source link

Numbered list item is not correctly converted if it follows TAB character #112

Closed mocobeta closed 2 years ago

mocobeta commented 2 years ago

Seems to occur only in commitbot's comments.

https://github.com/mocobeta/forks-migration-test-2/issues/10514#issuecomment-1205989333

スクリーンショット 2022-08-08 18 30 36

Looks like it's not easy to fix.

mocobeta commented 2 years ago

I tried this, and I'd conclude it's difficult to investigate (while it should be possible).

mocobeta commented 2 years ago

Or, we could just convert \t to space characters (2 or 4 whitespaces maybe)?

mocobeta commented 2 years ago

Or, we could just convert \t to space characters (2 or 4 whitespaces maybe)?

But I can't estimate whether there are no side effects.

msokolov commented 2 years ago

Maybe a no-break-space   would be safer (less likely to have some special meaning to github markdown parser) than a regular space?

mocobeta commented 2 years ago

I think in most cases it's fine to replace \t to whitespace or  , but noformat or code block can also contain \t; modifying any characters can break a code block (I once learned it).

mocobeta commented 2 years ago

I finally figured out a solution in #140.