[Bugfix] Fix the backticks bug

[Kyle-Ye]

SR-15415

[Kyle-Ye]

I do not know this will break any other situation. But this did address the issue of wrongly parse backticks mentioned in the SR-15415.

When parse "`" and "``"
Before the PR: We'll get 1:2-1:2 and 1:3-1:3 for the Text Node.
After the PR: We'll get the intended 1:1-1:1 and 1:1-1:2 for the Text Node

[QuietMisdreavus]

Can you make sure to run swift run api_test locally to run cmark-gfm's built-in tests? You could also add something to the source_pos() test in api_test/main.c to make sure your fix works even outside swift-markdown. Otherwise, this looks great, thanks!

[akyrtzi]

Assuming this is code from upstream cmark, could we fix it in upstream and then sync-up the fork instead?

[Kyle-Ye]

Can you make sure to run swift run api_test locally to run cmark-gfm's built-in tests? You could also add something to the source_pos() test in api_test/main.c to make sure your fix works even outside swift-markdown. Otherwise, this looks great, thanks!

It can pass the test by running swift run api_test

And added test to source_pos_inlines() in api_test/main.c

[Kyle-Ye]

Assuming this is code from upstream cmark, could we fix it in upstream and then sync-up the fork instead?

Add a clean PR to the upstream. See commonmark#427

[QuietMisdreavus]

Thanks so much for adding the test, and for posting the patch upstream! It looks like they've already taken a look, which is nice. This looks good to go for me, now.

[akyrtzi]

Assuming this is code from upstream cmark, could we fix it in upstream and then sync-up the fork instead?

Add a clean PR to the upstream. See commonmark#427

Thank you! 🙇🏻‍♂️