[ara-quran-la1] Two stray characters and 22 tripled consonants #72

Hello, and thank you for maintaining this API.

We use the ara-quran-la1 edition (quran411.com, transliteration by
M.A.H. Eliyasee) as a pronunciation aid in Deen Coach,
a free, ad-free, offline Quran app, always displayed under the Arabic text and
never replacing it. A quality check on all 6236 entries surfaced two categories
of findings. I am reporting them rather than silently forking, as you ask in
the README.

A. Two stray characters (certain)

Neither _ nor = belongs to any transliteration convention.

For 12:34 the same formula appears elsewhere in this very edition as
Huwas Samee'ul, with a space and no stray letter: see 2:137, 5:76, 6:13,
6:115, 8:61 and 10:65. So the expected reading is
innahoo Huwas Samee'ul 'Aleem. Note there are two defects in this token,
not one: the _, and an extra s in Huswas where the Arabic هُوَ has none.

For 12:97, zunoobanaa appears at 3:16, 3:147 and 3:193. The final tripling
in banaaa looks intentional and correct here: it marks the alif madda of
ذُنُوبَنَآ before the hamza of innaa, a notation this edition uses
consistently in that position (12:17, 21:14, 26:15, 28:53). So the only
defect is the =, and the expected reading is zunoobanaaa.

We have corrected these two entries locally, using the forms your own edition
attests, and this issue documents that divergence.

B. 22 tripled non-nasal consonants (please confirm)

The edition triples n and m in 154 entries, which we read as an
intentional ghunna notation, so we are not reporting those. It also triples
a non-nasal consonant in 22 entries, where that notation does not apply.

In almost every case the Arabic has a shadda on that consonant, so a doubled
letter appears written with three, which suggests a data-entry duplication
rather than a convention.

Two further matches are excluded from the list above because they look
intentional: 'Ayyyn in 19:1 and 42:2. Those verses are disjointed letters,
where the edition marks elongation by tripling the vowel throughout
(Kaaaf, Saaad, Meeem, Seeen, Qaaaf), so Ayyyn is consistent with
its own notation rather than an error.

If the section B tripling is a convention we have misread, please say so and
we will leave it untouched.

Thank you.