update ingest test

Unstructured-IO · Jun 18, 2024 · a9d424d · a9d424d
1 parent 68e7296
commit a9d424d
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 1 deletion.
diff --git a/test_unstructured_ingest/src/google-drive.sh b/test_unstructured_ingest/src/google-drive.sh
@@ -48,6 +48,8 @@ PYTHONPATH=${PYTHONPATH:-.} "$RUN_SCRIPT" \
   --verbose \
   --drive-id 1OQZ66OHBE30rNsNa7dweGLfRmXvkT_jr \
   --service-account-key "$GCP_INGEST_SERVICE_KEY_FILE" \
+  --recursive \
+  --extensions "pdf,docx" \
   --work-dir "$WORK_DIR"
 
 set +e

diff --git a/unstructured/ingest/v2/processes/connectors/local.py b/unstructured/ingest/v2/processes/connectors/local.py
@@ -173,7 +173,12 @@ def run(self, contents: list[UploadContent], **kwargs: Any) -> None:
         for content in contents:
             if source_identifiers := content.file_data.source_identifiers:
                 identifiers = source_identifiers
-                new_path = self.upload_config.output_path / identifiers.relative_path
+                rel_path = (
+                    identifiers.relative_path[1:]
+                    if identifiers.relative_path.startswith("/")
+                    else identifiers.relative_path
+                )
+                new_path = self.upload_config.output_path / Path(rel_path)
                 final_path = str(new_path).replace(
                     identifiers.filename, f"{identifiers.filename}.json"
                 )