Defect Report

With today's TOT 0be4b71; (NotoSansSyriac family), I see the Syriac Abbreviation Mark (SAM) character not correctly rendered for text that includes diacritics that are commonly used with this character.

One use of the SAM is to denote numerals when Syriac letters are used as numbers (see: r12a > Syriac > Numbers). A common representation for the number 2000 is ܒ݇ (U+0712, U+0747). Putting the SAM above the composed character renders correctly as ܏ܒ݇ (U+070F, U+0712, U+0747) - I believe the white space after the composed character breaks the SAM in this case.

If I were to write the current year (2025), I would write ܒ݇ܟܗ (U+0712, U+0747, U+071F, U+0717) preceded with a SAM character. When I write the fully composed year, while prepending the SAM, I get (܏ܒ݇ܟܗ):

For reference, if I were to omit the talqana character (U+0712) from the sequence, I'd have ܏ܒܟܗ which renders correctly:

The issue seems to be that the commonly used talqana character (U+0712) is breaking the SAM. This seems to occur with other commonly used diacritics used for writing larger numbers, for example:
10232: ܏ܐ̱ܪܠܒ‌

The SAM should cover the whole number, but only covers the ܐ̱ (here the line underneath ܐ is incorrectly breaking the SAM).

It looks like the SAM uses a stch (Stretching Glyph Decomposition) feature in its Glyphs file, where:

    sub uni070F by SAM4in SAMline SAMdot SAMline SAM4xout;

And this seems to get broken with diacritics. As is, writing many numbers (and relevant years), is broken.