feat: annotate indel hits with edit type and query position (Indel Positions column)

[Roman-Young]

Summary

Adds an Indel Positions output column for indel searches, annotating each hit with the edit type, residue, and 1-based query position as we discussed in our meeting.

Format

• <d/i>: <residue>[<pos or range>] — e.g. d: A[6] (deletion of A at query position 6), i: X[6] (insertion of X before query position 6). Exact matches report [].
• The deletion residue comes from the query; the insertion residue from the matched protein substring.
• In a repeat the exact position is ambiguous (indistinguishable residues), so the inclusive range of all valid positions is reported — e.g. d: A[2,4]. (We chose a range over listing all possible starting positions).

Implementation

• Computed purely in Python in _to_dataframe from (query, matched)
• Follows the existing one-column-per-mode pattern: Indel Positions in indel mode, Mutated Positions otherwise, in the same schema slot. Mismatch/exact output is unchanged.

Scope

1-indel only (single-residue edits), matching the currently supported max_indels=1. Will be naturally extended to 2-indels soon.

Testing

• Added hand-verified unit tests plus deletion/insertion end-to-end tests to ensure the output correctly matches the indel found.
• Full suite green (44 tests).