Auto linking (w/ anchor) syntax issue

[naokazuterada]

Auto linking (w/ anchor) syntax issue

[naokazuterada]

Hi, thanks for your post.

Is problem you reporting appeared on Github ?
If so, it's same as #13 .
I will try to implement a new feature.

Thanks.

[naokazuterada]

@raybit

Hi,
I've misunderstood your post. (Sorry..., I'm not good at English)

Did you mean it should be like

<!-- MarkdownTOC depth=2 -->

- [Heading with anchor][with-anchor]

<!-- /MarkdownTOC -->

# Heading with anchor [with-anchor]

not like

<!-- MarkdownTOC depth=2 -->

- [Heading with anchor](#with-anchor)

<!-- /MarkdownTOC -->

# Heading with anchor [with-anchor]

?

It used to be like the former according to "Reference-style links", but I have been notified of that some markdown render (e.g. Github) cannot support it by users, and changed implementation.

You propose that this plugin should also support the former style ? Right?
I will consider that.
So please answer my understanding is correct or not?

[raybit]

Please copy and paste following text as markdown and then you will know what I'm talking about.

<!-- MarkdownTOC depth=3 autolink -->
- [random1](#random1)
    - [random2](#random2)
    - [random3](#random3)
        - [issue0](#ID-String)
        - [](#issue1)
        - [](#issue2)
        - [issue3      {#custom-ID}](#issue3------{custom-id})
        - [](#issue4)
        - [the chars in the mid ()+-*/ 日本 is Ignored](#the-chars-in-the-mid---日本-is-ignored)
        - [日本](#日本)

<!-- /MarkdownTOC -->

# random1 
## random2
## random3


### issue0 [ID-String]
This format/style is curretly allowed in the TOC's to define header's ID, but as I mentioned this syntax is wrong and it can't be explained in any Markdown style.  [According to this link syntax section](http://aringfireball.net/projects/markdown/syntax#link), the link text is always delimited by [square brackets]. However, your current supported style didn't use square bracket around link text. 

### [issue1][my-anchor]
If we use a normal reference link style on header, TOC will always treat the strings in the first square brackets as an anchor.
[my-anchor]: http://www.googl.com

### [issue2](www.google.com)
if we use a normal inline link style on header, it has same issue.

### issue3      {#custom-ID}
If we use this MarkDown-Extra style to define ID, TOC also can't recongise it, and treat ID as a part of header string. Well, this is not caused by wrong grammer but it is just annother helpful feature to implement, or say it could be a nice replacement of current wrong implement since they share similar syntax structure. 


Overall, due to the wrong usage of link syntax in header line, the detection and extraction of right link/ID is not implemented.  

-----
And here is an another relevant issue. 
Although headers could auto generate ID and that ID can be used as anchor and this auto extraction processing of header ID works fine on most of render, it is not true for non-english header as you may have noticed.
  > The render will ignore all of those non-english char and special sign in header line when it try to extract header's ID string. And that's why we still need to set ID for header for those non-english header.

### issue4: autolink + non-english char = wrong achnor
Since current autolink feature won't ignore non-english char. FE:
### the chars in the mid ()+-*/ 日本 is Ignored
### 日本 

You may have to also ignore non-english chars and find a way to generate the right ID String if header line only contains non-english and special chars.

[naokazuterada]

Hi, thanks for explanation.

Issue0, 1, 2, 3 has been fixed on 1.4.0. See more details on README.md( some feature has been added ).

About Issue4,
On github, Japanese chars are not ignored. I want this features to be optimized to github not to official markdown specifications.

[johnathanz]

@raybit, thanks for the comprehensive explanation.  I'm not sure if this is correct behaviour. When I use the following syntax:
### 1 Next Release [next-release]
OR
### 2 Next Release {#next-release}
MarkdownTOC generates:
[1 Next Release][next-release]
When I use MarkdownPreview (https://github.com/revolunet/sublimetext-markdown-preview), it displays the TOC as it is, there is no link.
Am I not using the right Markdown syntax?
Thanks!

[stlehmann]

@johnathanz got the same propblem.

If I use the square brackets for links there will be no Links when viewing it with MarkdownPreview. With the round brackets it works but only for Markdown style, not Github.

Also Links with non-ascii characters such as (Ü, Ä, Ö, ... ) don' t work.

[jonasbn]

is the last problem described duplicate of the issues: #79 and #56 ?

[naokazuterada]

This might be fixed in 2.4.0.