Strikethrough does not work with a space just inside of it

[thesunny]

Subject of the issue

Strikethrough works in this scenario:

This is ~~text that has strikethrough~~ in it

But is parsed as plain text with the literal ~~ when there is a space in it

This is ~~text that has strikethrough ~~ in it and does not work (space at end)
This is ~~ text that has strikethrough~~ in it and this also does not work (space at front)

These two lines also render fine in GitHub:

Notably, bold and italic work find with spaces just inside the tokens.

Your environment

• OS: Mac
• Packages: remark-parse 8.0.3, unified 9.0.0
• Env: Node 12.16.2

Steps to reproduce

Here is a minimal reproduction on GitHub.

Just clone, yarn and yarn start

git clone git@github.com:thesunny/remark-strikethrough.git
cd remark-strikethrough
yarn
yarn start

Expected behaviour

It should parse a strikethrough with a space in it as the formatting of a strikethrough

Actual behaviour

It parses it as literal text

[ChristianMurphy]

@thesunny this is to spec.
You can try it on Github

You example

This is ~~text that has strikethrough ~~ in it and does not work (space at end)
This is ~~ text that has strikethrough~~ in it and this also does not work (space at front)

renders as

---

This is text that has strikethrough ~~ in it and does not work (space at end)
This is ~~ text that has strikethrough in it and this also does not work (space at front)

---

which matches has remark parses it https://astexplorer.net/#/gist/9e65a4608b9e79c42eddbc5785d5c428/056450d5fa2badbd5f6051f266925a2b6c8975ef

You can find the spec here:
https://github.github.com/gfm/#emphasis-and-strong-emphasis
https://github.github.com/gfm/#strikethrough-extension-

[thesunny]

Hmm... I'm looking at the source code and I see that this behavior seems intentional.

I also notice that GFM parses this the same way which is IMO broken. It also parses bold and italic this way which remark does not.

According to the GFM docs here:

https://github.github.com/gfm/#strikethrough-extension-

It doesn't mention any special behavior with respect to spaces just inside the tokens for strikethrough and seems to suggest that strikethrough should be parsed the same as emphasis. From their docs:

As with regular emphasis delimiters, a new paragraph will cause strikethrough parsing to cease:

In other words, it seems strikethrough should work the same as bold and italic.

I don't want to make things worse for myself, but bold and italic are broken in GFM. For example, if I type in this code in GitHub (ie. right here).

**bold **_italic _~~strikethrough ~~

**bold **_italic _~~strikethrough ~~

None of these work. But bold and italic work in remark but strikethrough does not.

My suggestion is we fix strikethrough to be consistent with bold and italic.

[ChristianMurphy]

My suggestion is we fix strikethrough to be consistent with bold and italic.

See #306 and #504

#439 aims to resolve this

[thesunny]

Thanks for the update @ChristianMurphy and thanks for the quick response.