Subjoined Bhaiksuki numbers include U+11C42 BHAIKSUKI DOUBLE DANDA

[dscorbett]

Font

NotoSansBhaiksuki-Regular.otf

Where the font came from, and when

Site: https://github.com/googlefonts/noto-fonts/blob/da636d8d8c6654fe3199aba049a1a278859a892b/phaseIII_only/unhinted/otf/NotoSansBhaiksuki/NotoSansBhaiksuki-Regular.otf
Date: 2020-04-21

Font version

Version 2.001

Issue

The glyphs for the subjoined Bhaiksuki numbers ten through ninety should not include double dandas. Are subjoined dandas attested? L2/14-091 doesn’t show any, but if they are, they should be encoded explicitly as U+11C42, not generated by the font. When they are included in the input string, though, they block subjoining.

Character data

𑱬𑱠𑱤 𑱬𑱠𑱂𑱤𑱂
U+11C6C BHAIKSUKI HUNDREDS UNIT MARK
U+11C60 BHAIKSUKI NUMBER SEVEN
U+11C64 BHAIKSUKI NUMBER TWENTY
U+0020 SPACE
U+11C6C BHAIKSUKI HUNDREDS UNIT MARK
U+11C60 BHAIKSUKI NUMBER SEVEN
U+11C42 BHAIKSUKI DOUBLE DANDA
U+11C64 BHAIKSUKI NUMBER TWENTY
U+11C42 BHAIKSUKI DOUBLE DANDA

Screenshot

[punchcutter]

This is really the same as https://github.com/googlefonts/noto-fonts/issues/1385, but just didn't get addressed in stacked sequences.

[nizarsq]