Fix: _batch_fetch_headers UID parsing for aioimaplib response format

[jbkjr]

Summary

• Fixes regression in _batch_fetch_headers introduced in PR Perf: Batch fetch emails to reduce IMAP round trips #107 where aioimaplib response format was incorrectly assumed
• aioimaplib returns FETCH responses in 3 separate parts, but code expected UID on same line as BODY[HEADER]
• Updates test mocks to use correct response format

Problem

PR #107 ("Batch fetch emails to reduce IMAP round trips") introduced a regression in _batch_fetch_headers. The function fails to parse email metadata because it assumes the UID appears on the same line as BODY[HEADER], but aioimaplib returns them separately.

Actual aioimaplib response format:

[
    b'2 FETCH (BODY[HEADER] {6149}',  # Item i:   has BODY[HEADER], NO UID
    bytearray(b'From: ...headers...'),  # Item i+1: header content
    b' UID 47)',                         # Item i+2: UID here!
]

What the code expected (and tests mocked):

[
    b'1 FETCH (UID 100 BODY[HEADER] {50}',  # UID on same line - WRONG
    bytearray(b'From: ...'),
    b')',
]

Fix

Updated _batch_fetch_headers to look for UID in data[i + 2] instead of data[i], matching the actual aioimaplib response format.

Test plan

• Run existing tests: uv run pytest tests/test_email_client.py tests/test_classic_handler.py -v (43 passed)
• Test mocks updated to use correct response format

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!