Punctuation not skipped by default for emphasis marks

[xfq]

https://www.w3.org/TR/jlreq/#n131

In practice, emphasis dots are not used for commas (cl-07), full stops (cl-06), opening brackets (cl-01) or closing brackets (cl-02).

慣行として，圏点は，句点類（cl-06），読点類（cl-07），始め括弧類（cl-01），終わり括弧類（cl-02）などには付けない．

Other punctuation marks are not mentioned here, so am I understanding correctly that hyphens, exclamation marks, question marks, and middle dots won't be skipped?

See also CLReq and CSS Text Decoration.

[himorin]

for other punctuation marks, text is written with 'or', and I suppose it indicates not limited (or lists as examples).
@kidayasuo I believe not, but do we have full set of targets as a replacement from current text? ('believe not' includes something wider like it depends on situation or target material, or defined by editing company or editors...)

[kidayasuo]

This describes the common practice of not applying emphasis dots to punctuation marks. However, the text does not explicitly prohibit doing so.

I agree that there is room for improvement, as the range of affected characters is unclear, and it is also ambiguous whether the text could be interpreted as prohibiting the use of emphasis dots on these characters. We'll discuss at the next JLReq TF (3/11) regarding this matter. Thank you for pointing this out.

---

p.s. I like the “NOTE” on 5.3.1 Emphasis Dots. As instances of referencing Japanese text within Chinese text, or vice versa, increase, establishing a common rule for their layout would be beneficial.

Regarding the rule that embedded text should follow, I think both cases exist. When the embedded text is part of a sentence and is expected to be read as such, it would be more natural to follow the rules of the surrounding text. However, when the embedded text serves as a reference, like an opaque embedded object within the text, there are cases where it would be preferable for it to follow the rules of its original language. I do not see a clear boundary between these cases.

Filed: #79

[kidayasuo]

Transferred to jlreq-d as jlreq document is in maintenance mode and we would like to improve the description in jlreq-d.

[kidayasuo]

We discussed this at the JLReq TF meeting on 2025-03-11 (meeting notes to be posted soon). The conclusion was as follows:

1. Although it is common practice not to attach emphasis marks to punctuation, this should be treated as a guideline rather than a strict rule. If an author or editor intentionally chooses to apply emphasis marks to any character, it should be allowed. Therefore, this rule should not be enforced.
2. As a convenience feature, it may still be helpful for systems to automatically skip certain character classes (such as punctuation). This makes it easier for users to select a full range of text without worrying about fine details, while still generally aligning with common practice.