Hiragana/katakana glyphs without Unicode

[mashabow]

In GlyphData.xml, I found there are some hiragana/katakana glyphs without Unicode; but where do they come from? What are they intended for?

$ cat GlyphData.xml | grep -v unicode | grep kana
        <glyph name="nga-hira" category="Letter" script="kana" description="HIRAGANA LETTER NGA" />
        <glyph name="ngi-hira" category="Letter" script="kana" description="HIRAGANA LETTER NGI" />
        <glyph name="ngu-hira" category="Letter" script="kana" description="HIRAGANA LETTER NGU" />
        <glyph name="kosmall-hira" category="Letter" script="kana" description="HIRAGANA LETTER SMALL KO" />
        <glyph name="tea-hira" category="Letter" script="kana" description="HIRAGANA LETTER TEA" />
        <glyph name="kosmall-kata" category="Letter" script="kana" description="KATAKANA LETTER SMALL KO" />
        <glyph name="tea-kata" category="Letter" script="kana" description="KATAKANA LETTER TEA" />
        <glyph name="tai-kata" category="Letter" script="kana" description="KATAKANA LETTER TAI" />
        <glyph name="ce-kata" category="Letter" script="kana" description="KATAKANA LETTER CE" />
        <glyph name="co-kata" category="Letter" script="kana" description="KATAKANA LETTER CO" />
        <glyph name="pusmall-kata" category="Letter" script="kana" description="KATAKANA LETTER SMALL PU" />

Although it seems that some of these glyphs (nga-hira, ngi-hira, ngu-hira, ce-kata, and pusumall-kata) come from JIS X 0213, they are not enough to support the standard. Please refer to NamedSequences.txt in the Unicode Character Database for this.

In addition, while tea-hira, tea-kata, tai-kata are for さ゚, サ゚, シ゚ respectively (according to @Tosche), I think these names should be tsa-hira, tsa-kara, tsi-kata or something, based on their pronunciation.

[brawer]

@schriftgestalt, would you accept a pull request for fixing these? For example:

1c1
< <glyph name="nga-hira" category="Letter" script="kana" description="HIRAGANA LETTER NGA" />
> <glyph name="nga-hira" category="Letter" script="kana" production="uni304B309A" description="HIRAGANA LETTER BIDAKUON NGA" />

[schriftgestalt]

I just found a response to the first issue (I did send it by mail and that didn't come through)

I don't remember exactly but I think I added them on one point as I needed them in the mapping file for Adobe-Japan1. In the mapping file they are called 'ka_semivoicedcomb-hira' now so I could remove them from the glyphData file.

I’ll update the mapping file from my internal file.