Consolidation of Post-V2 Mapping Change Suggestions

[kenlunde]

The following table shows the mapping changes that were made for the Version 2.001 update, and unless otherwise noted, are new mapping overrides:

Character	Glyph Name	CMap Resource
U+2FBF ⾿	uni9B2F-HK	HK
U+2FCB ⿋	uni9EF9-CN	HK (see Issue #204)
U+4EFD 份	uni4EFD-CN	HK (see Issue #204)
U+4F2C 伬	uni4F2C-HK	TW (see Issue #203)
U+50BA 傺	uni50BA-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+50E0 僠	uni50E0-HK	TW (see Issue #214)
U+50F8 僸	uni50F8-JP	HK REMOVE MAPPING OVERRIDE (see Issue #206)
U+5191 冑	uni5191-JP	HK
U+544E 呎	uni544E-HK	TW (see Issue #203)
U+54AB 咫	uni54AB-HK	TW (see Issue #203)
U+55D7 嗗	uni55D7-HK	TW CHANGE MAPPING OVERRIDE (see Issue #204)
U+55D7 嗗	uni55D7-JP	HK (see Issue #204)
U+588F 墏	uni588F-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+58A6 墦	uni58A6-HK	TW (see Issue #214)
U+5912 夒	uni5912-CN	HK
U+5AFA 嫺	uni5AFA-HK	TW REMOVE MAPPING OVERRIDE (see Issue #206)
U+5B0F 嬏	uni5B0F-HK	TW (see Issue #214)
U+5B38 嬸	uni5B38-HK	TW (see Issue #214)
U+5BE9 審	uni5BE9-TW	HK CHANGE MAPPING OVERRIDE (see Issue #214)
U+5C07 將	uni5C07-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+5C3A 尺	uni5C3A-HK	TW (see Issue #203)
U+5D45 嵅	uni5D45-JP	TW (see Issue #213)
U+5D45 嵅	uni5D45-JP	HK REMOVE MAPPING OVERRIDE (see Issue #213)
U+5D88 嶈	uni5D88-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+5D93 嶓	uni5D93-HK	TW (see Issue #214)
U+5E61 幡	uni5E61-HK	TW (see Issue #214)
U+5EB9 庹	uni5EB9-HK	TW (see Issue #203)
U+611F 感	uni611F-TW	HK CHANGE MAPPING OVERRIDE (see Issue #213)
U+61BE 憾	uni61BE-HK	TW (see Issue #213)
U+6220 戠	uni6220-CN	HK (see Issue #204)
U+64AD 播	uni64AD-TW	HK CHANGE MAPPING OVERRIDE (see Issue #214)
U+64BC 撼	uni64BC-HK	TW (see Issue #213)
U+64DB 擛	uni64DB-JP	HK REMOVE MAPPING OVERRIDE (see Issue #215)
U+65DB 旛	uni65DB-TW	HK CHANGE MAPPING OVERRIDE (see Issue #214)
U+68D7 棗	uni68D7-JP	HK
U+6937 椷	uni6937-JP	TW (see Issue #213)
U+6937 椷	uni6937-JP	HK REMOVE MAPPING OVERRIDE (see Issue #213)
U+69F3 槳	uni69F3-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+69F9 槹	uni69F9-HK	CN (see Issue #204)
U+69F9 槹	uni69F9-HK	TW REMOVE MAPPING OVERRIDE (see Issue #204)
U+6A07 樇	uni6A07-CN	HK REMOVE MAPPING OVERRIDE (see Issue #215)
U+6A4E 橎	uni6A4E-TW	HK CHANGE MAPPING OVERRIDE (see Issue #214)
U+6BCC 毌	uni6BCC-CN	HK REMOVE MAPPING OVERRIDE (see Issue #215)
U+6C23 氣	uni6C23-JP	HK CHANGE MAPPING OVERRIDE
U+6C69 汩	uni6C69-CN	HK
U+6EEB 滫	uni6EEB-JP	HK CHANGE MAPPING OVERRIDE
U+6F08 漈	uni6F08-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+6F3F 漿	uni6F3F-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+6F58 潘	uni6F58-TW	HK CHANGE MAPPING OVERRIDE (see Issue #214)
U+6FB8 澸	uni6FB8-TW	HK CHANGE MAPPING OVERRIDE (see Issue #213)
U+700B 瀋	uni700B-TW	HK CHANGE MAPPING OVERRIDE (see Issue #214)
U+7158 煘	uni7158-HK	JP/TW (see Issue #213)
U+71D4 燔	uni71D4-HK	TW (see Issue #214)
U+720B 爋	uni720B-CN	JP (see Issue #203)
U+7238 爸	uni7238-JP	HK
U+7247 片	uni7247-CN	HK
U+74A0 璠	uni74A0-TW	HK CHANGE MAPPING OVERRIDE (see Issue #214)
U+74A8 璨	uni74A8-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+752D 甭	uni752D-CN	HK REMOVE MAPPING OVERRIDE (see Issue #207)
U+752D 甭	uni752D-HK	JP (see Issue #207)
U+756A 番	uni756A-HK	TW (see Issue #214)
U+7690 皐	uni7690-HK	CN (see Issue #204)
U+7690 皐	uni7690-HK	TW REMOVE MAPPING OVERRIDE (see Issue #204)
U+76A4 皤	uni76A4-TW	HK CHANGE MAPPING OVERRIDE (see Issue #214)
U+78FB 磻	uni78FB-HK	TW (see Issue #214)
U+796D 祭	uni796D-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+7981 禁	uni7981-CN	HK REMOVE MAPPING OVERRIDE (see Issue #206)
U+7A44 穄	uni7A44-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+7BB4 箴	uni7BB4-HK	TW (see Issue #213)
U+7C53 籓	uni7C53-HK	TW (see Issue #214)
U+7DD8 緘	uni7DD8-CN	TW (see Issue #213)
U+7DD8 緘	uni7DD8-CN	HK REMOVE MAPPINGS OVERRIDE (see Issue #213)
U+7DFB 緻	uni7DFB-CN	HK REMOVE MAPPING OVERRIDE (see Issue #215)
U+7E59 繙	uni7E59-HK	TW (see Issue #214)
U+7FB3 羳	uni7FB3-HK	TW (see Issue #214)
U+808F 肏	uni808F-TW	JP
U+809E 肞	uni809EuE0101-JP	CN
U+819A 膚	uni819A-JP	HK REMOVE MAPPING OVERRIDE (see Issue #204)
U+81B0 膰	uni81B0-HK	TW (see Issue #214)
U+8543 蕃	uni8543-HK	TW (see Issue #214)
U+85E9 藩	uni85E9-HK	TW (see Issue #214)
U+8687 蚇	uni8687-HK	TW (see Issue #203)
U+87BF 螿	uni87BF-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+87E0 蟠	uni87E0-HK	TW (see Issue #214)
U+89BE 覾	uni89BE-TW	HK REMOVE MAPPING OVERRIDE (see Issue #214)
U+8B52 譒	uni8B52-HK	TW (see Issue #214)
U+8B85 讅	uni8B85-HK	TW (see Issue #214)
U+8C4B 豋	uni8C4B-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+8E41 蹁	uni8E41-JP	HK
U+8E61 蹡	uni8E61-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+8E6F 蹯	uni8E6F-HK	TW (see Issue #214)
U+8F53 轓	uni8F53-HK	TW (see Issue #214)
U+8F57 轗	uni8F57-HK	TW (see Issue #213)
U+8FBC 込	uni8FBCuE0101-JP	KR (see Issue #204)
U+91DC 釜	uni91DC-JP	HK
U+9407 鐇	uni9407-HK	TW (see Issue #214)
U+95B9 閹	uni95B9-JP	HK REMOVE MAPPING OVERRIDE (see Issue #215)
U+9699 隙	uni9699-JP90-JP	HK CHANGE MAPPING OVERRIDE
U+969B 際	uni969B-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+982F 頯	uni982F-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+9B0D 鬍	uni9B0D-CN	HK REMOVE MAPPING OVERRIDE (see Issue #204)
U+9B2F 鬯	uni9B2F-CN	HK REMOVE MAPPING OVERRIDE (see Issue #215)
U+9C36 鰶	uni9C36-JP	HK CHANGE MAPPING OVERRIDE (see Issue #210)
U+9C55 鱕	uni9C55-HK	TW (see Issue #214)
U+9EF9 黹	uni9EF9-CN	HK (see Issue #204)
U+9EFC 黼	uni9EFC-CN	HK (see Issue #204)
U+24A01 𤨁	uni7468-CN	JP (see Issue #203)

[Buernia]

CN glyph 甭 should use TW version. Below is the glyph in 通用规范汉字表 and GB 18030-2005.

[kenlunde]

@Buernia This one is a bit complex, mainly because the CN glyph is considered the primary Chinese glyph. In this case, the identities of the two glyphs will be changed, as now reflected in Issue #207, and there will also be mapping changes as reflected in the table above.

[tamcy]

Mappings of the following characters could be changed to better reflect the HK convention.

Character	Current	Proposed	Remarks
U+5191 冑	TW	JP	冃 component is different
U+5912 夒	TW	CN	Bottom component: TW uses 夊 while HK uses 夂
U+6C69 汩	TW	JP CN (corrected in 16 Feb)	Difference in 曰 component
U+6EEB 滫	CN	JP	Top-right component: CN uses 夂 while HK uses 攵

[lapomme]

For JP/KR, 肏 should use the TW (right) glyph instead of CN (left), since 入 is in the traditional print style.

[tamcy]

Some more remapping suggestions for HK region:

Character	Current	Proposed	Remarks
U+68D7 棗	CN	JP	The bottom-right of 冂 is different. Also applicable to TW.
U+8E41 蹁	TW	CN	The 扁 component is different.
U+6C23 氣	CN	JP	Just for better consistency. Ref: U+6DD7 淗, U+6ECA 滊.
U+7238 爸	CN	JP	Just for better consistency. Ref: U+7239 爹, U+723A 爺.
U+91DC 釜	CN	JP	Ditto.
U+7247 片	TW	CN	Not an error, CN glyph better matches the reference form. Also ref U+6C9C 沜.

[kenlunde]

@tamcy & @lapomme: The mapping change suggestions from 2018-12-04 and 2018-12-11 are now reflected in the table at the beginning of this issue. Please double-check.

@tamcy: Note that the HK glyph for U+6C69 汩 should map to the CN glyph, uni6C69-CN, not to a JP glyph that does not exist.

[tamcy]

The CN glyph of 隙 (U+9699) isn't suitable for HK as the top-right component is different. The KR glyph (I suppose it's uni9699-JP90-JP) could be used instead.

[Marcus98T]

I noticed that the TW form for U+595C 奜 is mapped to KR. It should have been mapped to JP (for Serif at least), but since that JP glyph was (EDIT: never included in Source Han Sans, so never mind), map it to CN.

[kenlunde]

@Marcus98T Ideographs with K2 sources are outside the scope of the Source Han projects. (I consider those with K1 sources to be barely within the scope given the limited extent to which ideographs are used in ROK.)

Also, I checked all prior versions of Source Han Sans, and there has never been a JP glyph for U+595C 奜. There is a JP glyph for U+595C 奜 in Source Han Serif, but it has been marked for removal in Version 2.000 for reasons explained in the previous paragraph.

[Marcus98T]

@Marcus98T Ideographs with K2 sources are outside the scope of the Source Han projects. (I consider those with K1 sources to be barely within the scope given the limited extent to which ideographs are used in ROK.)

Also, I checked all prior versions of Source Han Sans, and there has never been a JP glyph for U+595C 奜. There is a JP glyph for U+595C 奜 in Source Han Serif, but it has been marked for removal in Version 2.000 for reasons explained in the previous paragraph.

Oh. Got a bit of misunderstanding here. If that's the case, I suggest mapping U+595C 奜 to CN for KR. If time permits, try mapping it to JP in Serif, if not, map to CN in the future v2 Serif.

EDIT: Perhaps I wonder why was the TW mapping used for KR in the first place.

[Marcus98T]

I would like to ask, would any character with the 主 component (the drop part over 王) be shared amongst CN/TW/HK and JP/KR (never mind some JP glyphs with the vertical stroke on the top that was intended for hyogaiji, but maybe an angle compromise perhaps)? I will detail next time. If not, then forget it.

(JP left, KR in bracket for 註, CN right)

For 辶, could the lower left part of this component be shared with JP, then can apply to CN glyphs, unifying some (but not all) of the 辶 characters with the JP glyphs?

EDIT: In hindsight, the 之 component should have been under redesign, not mapping changes. It’s unlikely any 之 component will be shared, and I will retract my 之 comment.

And finally, for some glyphs, I think there are imperceptible differences with the duplicate glyphs that can be shared without a problem:

For SC/TC/HK, 嘉 needs to map to JP/KR, then the SC version removed. 朔 and 溯 needs to map to CN, then remove the TW glyphs. 閣 should map the CN glyph to JP/KR, then remove the JP glyph, because the right feet in 各 needs to be removed when inside 門 (the right hook could be interfering with the right feet), and then the CN glyph already had the right feet removed.

[tamcy]

For 辶, could the lower left part of this component be shared with JP, then can apply to CN glyphs, unifying some (but not all) of the 辶 characters with the JP glyphs?

This was raised (incidentally, by me) and rejected before. See https://github.com/adobe-fonts/source-han-sans/issues/98#issuecomment-94805891, https://github.com/adobe-fonts/source-han-sans/issues/98#issuecomment-122543563

[Marcus98T]

I must have been an idiot here. I didn’t have time to read through the whole stack of glyph sharing issues (that were already closed) so I simply posted without reading through them first, and perhaps wasted everyone’s time.

I’m hoping the preliminary list of v2.002 changes can come soon enough.

[kenlunde]

@Marcus98T & others: There's a remarkable feature provided by GitHub that allows one to search issues. By default, it searches only open issues, but if you remove that portion of the default search parameter, you can search all issues, open and closed:

But, to answer @tamcy's suggestion, we have no plans to unify the CN/JP ⻌ (aka Radical #162) component.

[Marcus98T]

I already know that unifying 辶 was rejected due to complying with national standards. I am still wondering if v3 can have the unified 𠆢 roof component (except for 人 and 从?) following the JP form for CN/TW/HK, like @tamcy did in one of the SHS forks. I’m sure the Pingfang font have this kind of roof design.

[kenlunde]

@Marcus98T Comparing to Ping Fang isn't helpful, as this is a typeface design difference. At this point, we have no plans to unify the JP/KR and CN/TW/HK forms of the 𠆢 component.

While it does not apply here, I feel compelled to point out that it can sometimes be dangerous to suggest changes to a typeface that make it more similar to another typeface.

[Marcus98T]

Thank you for the clarification. Certainly while it would have been pretty useful to remove even more unnecessary glyphs to make way for new glyphs (as I saw in the list of preliminary additions for v3), I conclude conforming to Chinese national standards and avoiding plagiarism from other commercial fonts are more important than that (even for extremely subtle differences in design), plus the typeface designer’s preference to keep certain aesthetics in the JP/KR glyphs, which would not make unifying shapes possible in the near future.

With that said, I think it’s better to wait until the time when v3 is underway, then discuss again, especially if there is really no more room to add new glyphs and something has to give way.

By the way, even when I remove “is:issue” in the issue search bar, Github’s search will only get results from titles, not contents from within. This I might try to raise with support.

[Marcus98T]

Should the KR version of 隙 (U+9699) be mapped to the current JP glyph? Most commercial Korean fonts do not show the JIS90 form, but rather the JIS2004 form.

However, Malgun Gothic shows the JIS90 form as well as the Unicode reference. I'll let Dr Lunde decide. I think given the glyph inconsistency between Korean Hanja forms (for other characters) in different fonts, this is unlikely to be done, given that I should not be suggesting edits that should follow the glyphs of commercial typefaces. But I just want to bring this up for a little discussion.

However, the next one is more of a mapping error. Please remap the KR version of 冤 (U+51A4) to the JIS2004 form. While most Korean fonts do not cover that character, the Microsoft Batang and Malgun Gothic fonts, plus the Unicode reference show the JIS2004 form, not the JIS90 form that Source Han Sans KR is currently having.

[Marcus98T]

Please remap the TW/HK version of U+5067 偧 and U+86E5 蛥 from the current CN form to the JP form since the JP form will be more consistent with the other 多 components in the TW/HK forms, even if they're outside the TW/HK scope.

Edit: Added U+5067 偧

[zhenlige]

Please map the CN version of U+20E6D 𠹭 from CID+59698 (with a hook in the top part of the 哥 component) to CID+59699 (without the hook).