fix(types): make AutohighlightResult.rank Optional to tolerate missing field

assemblyai/types.py

@@ -2140,7 +2140,12 @@ class Timestamp(BaseModel):
 
 class AutohighlightResult(BaseModel):
     count: int
-    rank: float
+    # The Auto Highlights API has been observed in production to return
+    # individual `results` entries with no `rank` field set (see #148).
+    # Treat it as optional so a TranscriptResponse parses successfully when
+    # any single highlight is missing the rank; the absence is then
+    # observable to callers via `rank is None`.
+    rank: Optional[float] = None
     text: str
     timestamps: List[Timestamp]
 

tests/unit/test_auto_highlights.py

@@ -93,3 +93,39 @@ def test_auto_highlights_enabled(httpx_mock: HTTPXMock):
         ):
             assert transcript_timestamp.start == response_timestamp["start"]
             assert transcript_timestamp.end == response_timestamp["end"]
+
+
+def test_auto_highlights_parses_result_without_rank(httpx_mock: HTTPXMock):
+    """
+    Regression for #148. The Auto Highlights API has been observed in
+    production to return individual `results` entries with no `rank` field
+    set, which previously raised a Pydantic ValidationError and made the
+    entire TranscriptResponse unparseable. The field is now Optional, so a
+    missing rank parses as None and the rest of the response is preserved.
+    """
+    mock_response = factories.generate_dict_factory(
+        AutohighlightTranscriptResponseFactory
+    )()
+    # Strip the `rank` from the first highlight to simulate the API's
+    # observed behavior. Other fields (count, text, timestamps) are left
+    # in place.
+    assert mock_response["auto_highlights_result"]["results"], (
+        "factory should produce at least one highlight"
+    )
+    del mock_response["auto_highlights_result"]["results"][0]["rank"]
+
+    _, transcript = unit_test_utils.submit_mock_transcription_request(
+        httpx_mock,
+        mock_response=mock_response,
+        config=aai.TranscriptionConfig(auto_highlights=True),
+    )
+
+    assert transcript.error is None
+    assert transcript.auto_highlights is not None
+    assert transcript.auto_highlights.results is not None
+    assert transcript.auto_highlights.results[0].rank is None
+    # The rest of the first result is still present, and other entries
+    # are unaffected.
+    assert transcript.auto_highlights.results[0].text is not None
+    if len(transcript.auto_highlights.results) > 1:
+        assert transcript.auto_highlights.results[1].rank is not None