Commonmark-compliance: Reader should NOT parse intra word underscore (CJK and Cyrillic)

[ob-user]

Steps to reproduce:

Copy and paste 字体_斜体_字体*斜体*字体

Actual result:

Expected:

[WhiteNoise]

this is not a bug

https://spec.commonmark.org/0.30/#emphasis-and-strong-emphasis

[ob-user]

Hi,

The “actual result” was the result from live preview, and the “expected” was the result from reading mode and the output file. The two results are inconsistent.

The two results are inconsistent. I guess the two modes use different parsers. You can say this is not a bug for its specific implementation, but I consider the inconsistency a bug.

[ob-user]

And btw do you consider this a bug?

Copy and paste 啦。**啦。**啦。

The result in live preview:

From the website you cited, “Many implementations have also restricted intraword emphasis to the * forms”, so the asterisks here should work, right?

[WhiteNoise]

@ob-user

The “actual result” was the result from live preview, and the “expected” was the result from reading mode and the output file. The two results are inconsistent.

According to commonmark spec, The bug is in the reader, non LP.

[WhiteNoise]

@ob-user

Copy and paste 啦。**啦。**啦。

The result in live preview:

From the website you cited, “Many implementations have also restricted intraword emphasis to the * forms”, so the asterisks here should work, right?

This is a whole other can of worms related to how markdown should handle Chinese special punctuation characters.

Markdown: Treat CJK punctuation symbols as letters

Some aspects of the commonmark spec do not play well with CJK languages. For example, according to the spec punctuation should be handled as whitespaces hence a**.test.**b should NOT be bold, whereas a**test**b has test in bold. This is problematic for Chinese users. This FR is used to track a proposed deviation from the commonmark spec: handle Chinese punctuation symbols (that are different from regular punctuation symbols) as regular letters.

[eightning]

How about not making the full stop bold?
啦。啦。啦。
啦。**啦**。啦。

Or using a zero-width space after the full stop (not sure if that will copy and paste here correctly)? Both parse.
啦。啦。​啦。
啦。**啦。​**啦。

[ob-user]

Hi,

Thank you. These are good workarounds!

[ob-user]

btw, how did you type that zero-width space?

[eightning]

On macOS: Edit → Emoji & Symbols → insert the zero-width space.

[Arachnid_O]

underscore do not work correctly with cyrilic symbols and broke links in preview mode.
in line-preview it works correctly

Steps to reproduce

enter text and link with underscore
for example:

текст_подчёркивание [[текст_ссылка]]

see what happens in preview mode

Expected result

working link to note текст_ссылка in preview

Actual result

preview do not work correctly

Environment

• Debug info:

SYSTEM INFO:
	Obsidian version: v1.1.9
	Installer version: v1.1.9
	Operating system: #1 SMP PREEMPT_DYNAMIC Mon Feb 6 11:26:13 UTC 2023 6.0.19-4-MANJARO
	Login status: not logged in
	Insider build toggle: off
	Live preview: on
	Legacy editor: off
	Base theme: dark
	Community theme: none
	Snippets enabled: 0
	Restricted mode: on

---

Additional information

example on sandbox vault

this problem works on desktop client
on mobile this not happens.

[WhiteNoise]

@Arachnid_O

on mobile this not happens.

I am sure it happens too.

[Arachnid_O]

i was test that
links are not broken on android client. they only formatting to be italic
live-preview

preview:

[WhiteNoise]

You are not separating with a blank line on mobile.