Fix O(n²) offset tracking in scholarly_to_modern

[Copilot]

scholarly_to_modern called len("".join(result_chars)) four times per loop iteration to track transliteration offsets, making the method O(n²) for long inputs.

Changes

• transliterator.py: Replace all len("".join(result_chars)) offset computations with a running trans_pos integer counter, incremented by len(modern) or len(char) after each character is appended.

# Before — O(n) join on every iteration
trans_start = len("".join(result_chars))
result_chars.append(modern)
trans_end = len("".join(result_chars))

# After — O(1) counter update
trans_start = trans_pos
result_chars.append(modern)
trans_pos += len(modern)

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.