Potential bug when converting a "ਸ਼" with sihaari

[dsomel21]

ਵਾਹਿਗੁਰੂਜੀਕਾਖਾਲਸਾਵਾਹਿਗੁਰੂਜੀਕੀਫਤੇ

I noticed a couple of times when I was previewing words like: "ਪੋਸ਼ਿਸ਼" or "ਬਖਸ਼ਿਸ਼" the Gurmukhi script (the English conversion) would be a little off.

Take this example:

anv.unicode('ਸ਼', true); // expects sæ
anv.unicode('ਸ਼ਿ', true); // expects "isæ"
anv.unicode('ਸ਼ਿਸ਼', true); // expects "isæsæ"

But the results are different. Check it out:

Here is what "ਪੋਸ਼ਿਸ਼" looks like when incorrectly converted using anv.unicide('ਪੋਸ਼ਿਸ਼"', true);. The image below is being displayed with the OpenGurbaniAkhar-Regular.otf font.

I've only seen this happen with the Sihaari, other vowels are looking good.