TextMesh Pro Tamil Font Support

Tamil like many other languages relies on OpenType features such as Diacritical Marks, Glyph Substitution and Ligatures which is not currently supported in TextMesh Pro. Support for these features is planned.

Looking at the particular example you provided, there are 3 different issues. The first is the position of ் over the ற. This we can address by creating a glyph adjustment pair in the Glyph Adjustment Table as such (values will be different depending on the font). Note the ID is the index of the glyph and not its unicode.



The வ and ே might be a chaining context substitution but I think (not sure since I don't understand the language) we might be able to also define an adjustment pair to reverse them as seen below.



Then the last combination of டand ு appears to be a ligature that results in டு which is not currently supported.

With the exception of the last substitution this is the result with the adjustments



Support for these features is coming. I don't have a firm ETA yet.

 

Right now this manual process is pretty difficult / horrible.

Once support for OpenType Font Features is available (no firm ETA yet but this is something I am working on) this process will be much easier as the glyph positional adjustment and substitution data will be extracted from the font file itself.

In terms of potentially, copying the data from one font asset to the other, you might be able to create a script where you grab the relevant data from the fontFeatureTable of one font asset and then re-assign the "internal" m_GlyphPairAdjustmentRecords to the other. Note these are marked as internal but since the package includes source code, you should be able to modify these for your needs. This is certainly note ideal but it might save you time.

P.S. Hopefully the glyph index remain consistent between the Light and Medium typefaces as copying the data would then be pointless as it would reference the incorrect glyphs.

 

I suggest testing just a few pairs but this should work between pairs within each font asset but not across them since the glyph indexes are not the same.