fix: parse multiple 16-bit and 32-bit Service UUIDs in BLE advertisement data

[bdraco]

fixes esphome/issues#7182

Fix parsing of multiple Service UUIDs in BLE advertisement data

Summary

This PR fixes a bug in the BLE GAP advertisement parser where only the first UUID was being parsed when multiple 16-bit or 32-bit Service UUIDs were present in a single advertisement data structure. The parser now correctly iterates through all UUIDs in the data.

Changes

Parser fixes (src/bluetooth_data_tools/gap.py)

1. 16-bit UUIDs: Now iterates through the data in 2-byte chunks to extract all UUIDs
2. 32-bit UUIDs: Added missing support for parsing 32-bit UUIDs (was completely missing), iterating in 4-byte chunks
3. 128-bit UUIDs: Kept as single UUID parsing (correct per BLE spec - only one 128-bit UUID fits per record)

Test updates (tests/test_gap.py)

1. Updated existing test: Fixed test_parse_adv_data to expect two separate 16-bit UUIDs instead of one combined UUID
2. Added new test: test_parse_multiple_16bit_uuids - Tests parsing of 3 16-bit UUIDs
3. Added new test: test_parse_multiple_32bit_uuids - Tests parsing of 2 32-bit UUIDs
4. Added new test: test_parse_mixed_16bit_32bit_uuids - Tests parsing mixed 16-bit and 32-bit UUIDs in one advertisement
5. Added new test: test_parse_mixed_16bit_128bit_uuids - Tests parsing mixed 16-bit and 128-bit UUIDs in one advertisement

Technical details

The BLE advertisement data format allows multiple UUIDs to be packed in a single record:

• For 16-bit UUIDs (types 0x02, 0x03): Each UUID occupies 2 bytes
• For 32-bit UUIDs (types 0x04, 0x05): Each UUID occupies 4 bytes
• For 128-bit UUIDs (types 0x06, 0x07): Each UUID occupies 16 bytes (only one fits per record)

The previous implementation incorrectly treated the entire data block as a single UUID, leading to malformed UUID strings.

Testing

All tests pass:

• 52 tests in test_gap.py (added 4 new tests)
• New tests verify correct parsing of multiple UUIDs and mixed UUID types
• No regressions in existing tests

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (45fe00b) to head (760c3a7).
⚠️ Report is 3 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff             @@
##             main      #155      +/-   ##
===========================================
+ Coverage   99.59%   100.00%   +0.40%     
===========================================
  Files           6         6              
  Lines         245       255      +10     
  Branches       32        38       +6     
===========================================
+ Hits          244       255      +11     
+ Partials        1         0       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[codspeed-hq]

CodSpeed Performance Report

Merging #155 will not alter performance

_{Comparing uuid_parsing (760c3a7) with main (45fe00b)}

Summary

✅ 9 untouched benchmarks