I develop Japanese OpenType fonts using AFDKO (installed via pip), and I'd like to report what appears to be a bug in one of the CMap resource files bundled with the afdko package.
Issue
The file resources/Adobe-Japan1/UniJIS2004-UTF32-H bundled with the afdko pip package (confirmed with the latest release, v5.0.1, as of July 8, 2026) is still based on Adobe-Japan1-6 (2017, CMapVersion 1.018) and has not been updated. It does not contain the Unicode mapping for the Reiwa era ligature CID (23058, U+32FF SQUARE ERA NAME REIWA), which was added in Adobe-Japan1-7 Supplement 7 (2019).
Steps to reproduce
-
pip install --upgrade afdko
-
Open resources/Adobe-Japan1/UniJIS2004-UTF32-H under site-packages
-
The header shows: %%Title: (UniJIS2004-UTF32-H Adobe Japan1 6) / %%Version: 1.018 / Copyright 1990-2017
-
Searching the file for "23058" returns no match
Impact
When building a CID-keyed font that includes the Adobe-Japan1-7 (or later) CID set with makeotf using this bundled CMap, the Reiwa ligature glyph never receives a Unicode value (U+32FF), so the character cannot be entered or displayed via Unicode in the resulting font.
Workaround we used
I downloaded the current UniJIS2004-UTF32-H from your official cmap-resources repository (https://github.com/adobe-type-tools/cmap-resources, Adobe-Japan1-7/CMap/UniJIS2004-UTF32-H, %%Version: 1.022, confirmed to include the CID 23058 mapping) and substituted it for the bundled copy.
Request
Could you please update the CMap resource snapshot bundled inside the afdko package to match the latest version in your official cmap-resources repository? It would be very helpful for users like me to get CMaps that support the current Adobe-Japan1 character collection simply via pip install, without needing a manual workaround.
I develop Japanese OpenType fonts using AFDKO (installed via pip), and I'd like to report what appears to be a bug in one of the CMap resource files bundled with the afdko package.
Issue
The file resources/Adobe-Japan1/UniJIS2004-UTF32-H bundled with the afdko pip package (confirmed with the latest release, v5.0.1, as of July 8, 2026) is still based on Adobe-Japan1-6 (2017, CMapVersion 1.018) and has not been updated. It does not contain the Unicode mapping for the Reiwa era ligature CID (23058, U+32FF SQUARE ERA NAME REIWA), which was added in Adobe-Japan1-7 Supplement 7 (2019).
Steps to reproduce
pip install --upgrade afdko
Open resources/Adobe-Japan1/UniJIS2004-UTF32-H under site-packages
The header shows: %%Title: (UniJIS2004-UTF32-H Adobe Japan1 6) / %%Version: 1.018 / Copyright 1990-2017
Searching the file for "23058" returns no match
Impact
When building a CID-keyed font that includes the Adobe-Japan1-7 (or later) CID set with makeotf using this bundled CMap, the Reiwa ligature glyph never receives a Unicode value (U+32FF), so the character cannot be entered or displayed via Unicode in the resulting font.
Workaround we used
I downloaded the current UniJIS2004-UTF32-H from your official cmap-resources repository (https://github.com/adobe-type-tools/cmap-resources, Adobe-Japan1-7/CMap/UniJIS2004-UTF32-H, %%Version: 1.022, confirmed to include the CID 23058 mapping) and substituted it for the bundled copy.
Request
Could you please update the CMap resource snapshot bundled inside the afdko package to match the latest version in your official cmap-resources repository? It would be very helpful for users like me to get CMaps that support the current Adobe-Japan1 character collection simply via pip install, without needing a manual workaround.