updating page name doesn't update tags prefixed with special char

[Amoscey]

What happened?

笔记名字修改后对应的其他笔记提到的原名字不会一起更改
比如有一个名为“Note1”的笔记，但是我改成了“note01”，这个时候，其他笔记上只有对应的 “[[Note1]]” 会改成 “[[note01]]” ，但是 "#Note1" 就不会改成"#note01"

For example, there is a note named "Note1", but I changed it to "note01".
At this time, only the "[[Note1]]" in other notes will be changed to "[[note01]]", but "#Note1" will not be changed to "#note01"

Reproduce the Bug

Windows system

Expected Behavior

No response

Screenshots

No response

Desktop Platform Information

No response

Mobile Platform Information

No response

Additional Context

No response

[cnrpman]

It's an issue related to CJK character compatiblity in code. Actually the case you provide won't trigger the issue. It only happens on non-latin characters.

Fixed in 0.8.3 Already.

0.83已经修复。请更新最新版。

[Amoscey]

It's an issue related to CJK character compatiblity in code. Actually the case you provide won't trigger the issue. It only happens on non-latin characters.

Fixed in 0.8.3 Already.

0.83已经修复。请更新最新版。

我更新后测试了一下，其他情况下可以，不过又发现一个新情况，如果前面带等于号就不行了。

[cnrpman]

@Amoscey Thank you for your great dig-in!
The = in =#Hashtag Sounds like a magic character to me. I tested some other symbols and all of them can't trigger Hashtag link. I don't know if it belongs to the design and need some time to investigate the mldoc code.

[Bad3r]

This issue exists in 0.8.18

Steps to reproduce:
1. on a block add the following line [[tag0]] #tag0 -#tag0 =#tag0 !#tag0

2. navigate to [[tag0] and rename it to [[tag1]]
3. notice the results:

only [[tag]] and #tag were updated.

[cnrpman]

The entry in codebase of this part:

logseq/src/main/frontend/handler/page.cljs

Lines 257 to 270 in 60fbfdf

	(defn- replace-tag-ref!
	[content old-name new-name]
	(let [old-tag (util/format "#%s" old-name)
	new-tag (if (re-find #"[\s\t]+" new-name)
	(util/format "#[[%s]]" new-name)
	(str "#" new-name))]
	;; hash tag parsing rules https://github.com/logseq/mldoc/blob/701243eaf9b4157348f235670718f6ad19ebe7f8/test/test_markdown.ml#L631
	;; Safari doesn't support look behind, don't use
	;; TODO: parse via mldoc
	(string/replace content
	(re-pattern (str "(?i)(^|\\s)(" (util/escape-regex-chars old-tag) ")(?=[,\\.]*($|\\s))"))
	;; case_insense^ ^lhs ^_grp2 look_ahead^ ^_grp3
	(fn [[_match lhs _grp2 _grp3]]
	(str lhs new-tag)))))