Merge pull request #7 from mattgotteiner/matt/update-prepskills

Matt/update prepskills

Author: pamelafox

app/backend/prepdocslib/blobmanager.py

@@ -424,7 +424,8 @@ async def upload_blob(self, file: File) -> str:
             await container_client.create_container()
 
         # Re-open and upload the original file
-        if file.url is None:
+        # URL may be a path to a local file or already set to a blob URL
+        if file.url is None or os.path.exists(file.url):
             with open(file.content.name, "rb") as reopened_file:
                 blob_name = self.blob_name_from_file_name(file.content.name)
                 logger.info("Uploading blob for document '%s'", blob_name)

app/backend/prepdocslib/cloudingestionstrategy.py

@@ -304,7 +304,7 @@ async def run(self) -> None:
                 configuration=IndexingParametersConfiguration(
                     query_timeout=None,
                     # markdown_parsing_submode=None,
-                    data_to_extract="contentAndMetadata",
+                    data_to_extract="storageMetadata",
                     # markdown_header_depth=None,
                     allow_skillset_to_read_file_data=True,
                 )