Support IVS

[Transfusion]

https://github.com/w3c/csswg-drafts/issues/1710

A sample line from an IVS file would be:
4F96 E0101; Adobe-Japan1; CID+2729
To investigate:

• The current release's A file includes almost everything before the CJK Ext B block, in order to pass fc-validate on all of zh, ja, and kr. CJKATC already has 61673 CIDs assigned.
• Creating another font file specifically for IVSes and putting it at the front of a font stack will surely work, but depending on the fallback behavior, if locl is to work, do all 5 locales' glyphs have to be included too???
• Thus, is it possible to pass an IVS file to makeotf -ci without having the base unicode character assigned to a CID, so that it falls back to the A file if the character alone does not exist in the new IVS-only font file?

Original HanaMinA includes 5555 glyphs from MojiJoho alone https://glyphwiki.org/wiki/Group:IVD-MojiJoho

There are nearly 30,000 IVSes in Unicode Inc.'s IVD https://www.unicode.org/ivd/data/2020-11-06/

https://docs.microsoft.com/en-us/typography/opentype/spec/cff2#table-14-fdselect-format-4
Format 4 differs from Format 3 only in that it accommodates **more than 65535 glyphs** by using a uint32 type for the nRanges and sentinel fields, and a Range4 record array.

[Transfusion]

Perhaps implement this and #9 first as part of a CJK-related blocks only font (i.e. no Latin blocks, no Cyrillic blocks...) to take the CID limit out of the equation