Escaped pipes are ignored within table cells

[robaxelsen]

Hi 👋 😄

I am running into this issue in Storybook, which uses markdown-to-jsx for markdown parsing: storybookjs/storybook#6869

Could not find any issues here regarding this, so reporting it. Would it be possible to fix this? Thanks in advance for any help!

[quantizor]

Yeah, open to PRs 👍

[robaxelsen]

Looking deeper into this, I see the same output when parsing a table with escaped pipe characters with react-native-simple-markdown as well. And both seem to be built on top of simple-markdown. Will close this for now, and report there.

[quantizor]

Reopened - this library has diverged a lot from simple-markdown so good to fix this separately.

[robaxelsen]

Reopened - this library has diverged a lot from simple-markdown so good to fix this separately.

@probablyup Ok, thanks. I tried to look into the code, but was not able to get the full picture. Would you be able to point me to any logic involved for this specific case, or hints on how you would go about making a fix for this?

Here is the new issue I created, which also mentions my crude test repo I've been playing with to solve this: https://github.com/robaxelsen/markdown-to-jsx-table-pipe-test

[quantizor]

I think probably here: https://github.com/probablyup/markdown-to-jsx/blob/master/index.js#L162

Checking and rejecting pipes that have an escape

[robaxelsen]

I think probably here: https://github.com/probablyup/markdown-to-jsx/blob/master/index.js#L162

Checking and rejecting pipes that have an escape

@probablyup Thanks! I did some testing, and seems that line is used to match any line that is part of a table markdown.

This one seems promising though, as it's used throughout the code to split the row into cells, if I am reading the code correctly: https://github.com/probablyup/markdown-to-jsx/blob/5d4470314de8d652435e581851e4e8e2ef78a6c0/index.js#L174

Tried replacing it with regex that should ignore the escaped pipe, but having issues getting it to match. Have any ideas?

Posted a relevant stackoverflow question here, which might help shed some more light on this: https://stackoverflow.com/questions/57306043/splitting-complex-string-with-while-ignoring-escaped-delimeter

[ariabuckles]

In simple-markdown I was able to fix this by actually parsing each row, rather than using String.prototype.split. The patch is here: ariabuckles/simple-markdown@f32a628 , and I might try to see how hard it would be to do the same thing here.

Edit: sent you a PR: #261