Repository navigation
Replies: 1 comment
|
We try to match GFM. If you have a fix we always appreciate PRs 😁👍 |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Is the paragraph split below intended, or would a fix be in scope?
Rechecked on current master
b81a00a(18.1.3), using default options:The input is:
Marked emits:
Both cmark-gfm and markdown-it 15.0.3 keep it as one paragraph:
The GFM table rule says that a table is not recognized when header and delimiter cell counts differ. I am not proposing that this input become a table, or changing the historical even-backslash/pipe behavior from #1262. The question is only whether the rejected candidate should end the preceding paragraph.
The paragraph regex appears to stop before
b|c, then the table tokenizer rejects the two-header/three-delimiter candidate, leaving a second paragraph. I have a private prototype and regression tests, but have not opened a PR because I wanted to check intended behavior first. Its current scope is the stock GFM paragraph tokenizer; custom paragraph wrappers and setext-heading variants would need separate consideration.I reproduced the outputs locally and checked the spec/reference comparisons.
All reactions