nv-ingest-api 2025.6.16.dev20250616__py3-none-any.whl → 2025.6.24.dev20250625__py3-none-any.whl

nv_ingest_api/internal/transform/split_text.py

@@ -31,9 +31,16 @@ def _build_split_documents(row, chunks: List[str]) -> List[dict[str, Any]]:
         metadata = row.metadata if hasattr(row, "metadata") and isinstance(row.metadata, dict) else {}
         metadata = copy.deepcopy(metadata)
 
-        metadata["content"] = text
-
-        documents.append({"document_type": ContentTypeEnum.TEXT.value, "metadata": metadata, "uuid": str(uuid.uuid4())})
+        if row.document_type == ContentTypeEnum.AUDIO:
+            metadata["audio_metadata"]["audio_transcript"] = text
+            documents.append(
+                {"document_type": ContentTypeEnum.AUDIO.value, "metadata": metadata, "uuid": str(uuid.uuid4())}
+            )
+        else:
+            metadata["content"] = text
+            documents.append(
+                {"document_type": ContentTypeEnum.TEXT.value, "metadata": metadata, "uuid": str(uuid.uuid4())}
+            )
 
     return documents
 
@@ -118,7 +125,7 @@ def transform_text_split_and_tokenize_internal(
     )
 
     # Filter to documents with text content.
-    text_type_condition = df_transform_ledger["document_type"] == ContentTypeEnum.TEXT
+    text_type_condition = df_transform_ledger["document_type"].isin([ContentTypeEnum.TEXT, ContentTypeEnum.AUDIO])
 
     normalized_meta_df = pd.json_normalize(df_transform_ledger["metadata"], errors="ignore")
     if "source_metadata.source_type" in normalized_meta_df.columns:
@@ -147,7 +154,14 @@ def transform_text_split_and_tokenize_internal(
 
     split_docs: List[Dict[str, Any]] = []
     for _, row in df_filtered.iterrows():
-        content: str = row["metadata"]["content"] if row["metadata"]["content"] is not None else ""
+        if row["document_type"] == ContentTypeEnum.AUDIO:
+            content: str = (
+                row["metadata"]["audio_metadata"]["audio_transcript"]
+                if row["metadata"]["audio_metadata"]["audio_transcript"] is not None
+                else ""
+            )
+        else:
+            content: str = row["metadata"]["content"] if row["metadata"]["content"] is not None else ""
         chunks: List[str] = _split_into_chunks(content, tokenizer_model, chunk_size, chunk_overlap)
         split_docs.extend(_build_split_documents(row, chunks))
 

{nv_ingest_api-2025.6.16.dev20250616.dist-info → nv_ingest_api-2025.6.24.dev20250625.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: nv-ingest-api
-Version: 2025.6.16.dev20250616
+Version: 2025.6.24.dev20250625
 Summary: Python module with core document ingestion functions.
 Author-email: Jeremy Dyer <jdyer@nvidia.com>
 License:                                  Apache License

{nv_ingest_api-2025.6.16.dev20250616.dist-info → nv_ingest_api-2025.6.24.dev20250625.dist-info}/RECORD

@@ -100,7 +100,7 @@ nv_ingest_api/internal/store/image_upload.py,sha256=GNlY4k3pfcHv3lzXxkbmGLeHFsf9
 nv_ingest_api/internal/transform/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
 nv_ingest_api/internal/transform/caption_image.py,sha256=0ILCG2F8ESqKtZiPUM-6F1BHUflFZ76Dzi2GNzkE-lU,8517
 nv_ingest_api/internal/transform/embed_text.py,sha256=F8kg-WXihtuUMwDQUUYjnfGDCdQp1Mkd-jeThOiJT0s,16507
-nv_ingest_api/internal/transform/split_text.py,sha256=DlVoyHLqZ-6_FiWwZmofPcq7TX8Ta23hIE0St9tw1IY,6822
+nv_ingest_api/internal/transform/split_text.py,sha256=-kwpRWSVZrPldm1hn3-tVz_TkzuKM-kPvNU3HTp9zOY,7476
 nv_ingest_api/util/__init__.py,sha256=wQSlVx3T14ZgQAt-EPzEczQusXVW0W8yynnUaFFGE3s,143
 nv_ingest_api/util/control_message/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nv_ingest_api/util/control_message/validators.py,sha256=KvvbyheJ5rbzvJbH9JKpMR9VfoI0b0uM6eTAZte1p44,1315
@@ -150,8 +150,8 @@ nv_ingest_api/util/service_clients/rest/rest_client.py,sha256=dZ-jrk7IK7oNtHoXFS
 nv_ingest_api/util/string_processing/__init__.py,sha256=mkwHthyS-IILcLcL1tJYeF6mpqX3pxEw5aUzDGjTSeU,1411
 nv_ingest_api/util/system/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nv_ingest_api/util/system/hardware_info.py,sha256=ORZeKpH9kSGU_vuPhyBwkIiMyCViKUX2CP__MCjrfbU,19463
-nv_ingest_api-2025.6.16.dev20250616.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-nv_ingest_api-2025.6.16.dev20250616.dist-info/METADATA,sha256=nrAshcKmKQTsbPtJTNzrdu37KttrTjLkAhUybCMLrV8,13919
-nv_ingest_api-2025.6.16.dev20250616.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-nv_ingest_api-2025.6.16.dev20250616.dist-info/top_level.txt,sha256=abjYMlTJGoG5tOdfIB-IWvLyKclw6HLaRSc8MxX4X6I,14
-nv_ingest_api-2025.6.16.dev20250616.dist-info/RECORD,,
+nv_ingest_api-2025.6.24.dev20250625.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+nv_ingest_api-2025.6.24.dev20250625.dist-info/METADATA,sha256=jnzs5A1QthJjwPAHRU2Z4PmVEAgddnZMiXNX5W-sMjI,13919
+nv_ingest_api-2025.6.24.dev20250625.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+nv_ingest_api-2025.6.24.dev20250625.dist-info/top_level.txt,sha256=abjYMlTJGoG5tOdfIB-IWvLyKclw6HLaRSc8MxX4X6I,14
+nv_ingest_api-2025.6.24.dev20250625.dist-info/RECORD,,