toc helper should not parse the text of children element(s)

[noraj]

Check List

Please check followings before submitting a new feature request.

• [x] I have already read Docs page
• [x] I have already searched existing issues

Feature Request

Fix the toc helper to select only text content and not other sub-elements like links.

In depth explanation here: https://github.com/hexojs/hexo-renderer-markdown-it/issues/102

[SukkaW]

@noraj

Fix the toc helper to select only text content and not other sub-elements like links.

Is this issue show up in Hexo 4.1.1? I am wondering if the problem is the different behavior between htmlparser2 & cheerio.

[noraj]

@SukkaW I updated https://github.com/hexojs/hexo-renderer-markdown-it/issues/102

[curbengh]

Can confirm the issue in hexo-util 1.8.1:

const { tocObj } = require('hexo-util')

const text = '<h3 id="create-a-new-post"><a class="header-anchor" href="#create-a-new-post">#</a>Create a new post</h3>'

console.log(tocObj(text))
/*
[
  {
    text: '#Create a new post',
    id: 'create-a-new-post',
    level: 3
  }
]
*/

[SukkaW]

@curbengh

<h1 id="title-3"><span>Title</span> 3</h3>

In this case Title 3 is what we needed for text.

[curbengh]

@SukkaW

Since the output is '\n#\nCreate a new post\n', what if tocObj trim the newlines and anything before and after \n? It may be more appropriate to trim in toc helper instead, since tocObj user might want to retain the original (rare, but possible).

Edit: no newline by default

[SukkaW]

@curbengh

Currently,

<h3 id="create-a-new-post">
<a class="header-anchor" href="#create-a-new-post">#</a>Create a new post
</h3>

will become \n#Create a new post\n. After trimming it will become Create a pos. Still not appropriate.

---

Update

cheerio has the same behavior.

https://runkit.com/sukkaw/5e186c9d286975001a1051c2

[curbengh]

Looks like a common question in cheerio/jquery,

• https://stackoverflow.com/questions/20832910/get-text-in-parent-without-children-using-cheerio
• https://www.designcise.com/web/tutorial/how-to-get-html-elements-text-excluding-children-using-jquery-and-javascript

Seems possible to fix, if htmlparser has the equivalent function of native DOM's document.childNodes or cheerio's $.first().

[Microbai]

When contents likes this

<h2><span id="title_1">title 1</span></h2>

tocObj won't parse the correct id, So I edit the getId function, then work.

const getId = ele => {
  const { id } = ele.attribs;
  const { children } = ele;
  return id || (!children[0] ? null : getId(children[0]));
};

[SukkaW]

@Microbai tocObj() is not designed to get id from child element but only from parent element.

[Microbai]

@SukkaW I Knew it.But when I use Hexo 4.2, the page can't render toc correctly.So I edit tocObj for now, and it works.