HTML in Markdown

[marian2js]

Subject of the issue

Strings starting with <p>...</p> cause everything else to be dropped.

remark()
  .use(strip)
  .process('<p>foo</p> bar', function(err, file) {
    if (err) throw err
    console.log(String(file))
  })

Actual: ""
Expected: "foo bar"

With some text before the tag, it works as expected:

remark()
  .use(strip)
  .process('foo <p>bar</p>', function(err, file) {
    if (err) throw err
    console.log(String(file))
  })

Actual: "foo bar"

Your environment

• OS: macOS 10.15.5
• Packages: strip-markdown "3.1.2"
• Env: node 14

Steps to reproduce

The issue can also be reproduced on the demo, just enter:

<p>foo</p>
bar

Expected behavior

<p> tag should be stripped and the inner text should be returned. This behavior works with other tags like <span>.

Actual behavior

<p>foo</p> bar returns an empty string, but foo <p>bar</p> returns "foo bar".

[wooorm]

hey, sorry for the wait, was busy with other stuff!

This is a duplicate of #14, the answer is still the same!

[marian2js]

Hi @wooorm thanks for the reply. I wasn't expecting this to be a hard fix, since it works for tags like span but not for p.

As a workaround I prefixed the text with a character and stripe it after words, it's not very elegant, but it seems to work for all cases. Maybe a similar workaround can be introduced in this library in order to prevent unexpected behaviours.

[wooorm]

The proper way to deal with HTML is through remark-rehype, rehype-raw, and then getting the text.
I understand this workaround works well for you, but I’d rather not introduce something fragile into the code here!

[marian2js]

Hi @wooorm, it's fine if the workaround is not added, but I think this should be documented. The readme says this library removes HTML, but now you say the proper way is to use extra plugins for it. I was replacing a different implementation with this library and I was lucky that a test was covering this case, otherwise I might not have detected it until production.

Anyway, thanks for maintaining this project, it was really useful for me.

[wooorm]

I think what’s confusing here is:

<p>foo</p> *bar*

^-- this whole thing is HTML to markdown. The emphasis doesn’t work. The docs refer to that aspect of HTML in markdown. But from what I gather, you’re reading html as just the XML-like parts of HTML, but not the text in HTML.

[wooorm]

Added a link to this issue next to where the readme mentions HTML!