Google Drive connector - txt and markdown support (danswer-ai#1469)

UiPath · Oct 2, 2024 · 0b1db1c · 0b1db1c
1 parent 2648754
commit 0b1db1c
Showing 1 changed file with 9 additions and 0 deletions.
diff --git a/backend/danswer/connectors/google_drive/connector.py b/backend/danswer/connectors/google_drive/connector.py
@@ -62,6 +62,8 @@ class GDriveMimeType(str, Enum):
     POWERPOINT = (
         "application/vnd.openxmlformats-officedocument.presentationml.presentation"
     )
+    PLAIN_TEXT = "text/plain"
+    MARKDOWN = "text/markdown"
 
 
 GoogleDriveFileType = dict[str, Any]
@@ -315,12 +317,19 @@ def extract_text(file: dict[str, str], service: discovery.Resource) -> str:
         GDriveMimeType.DOC.value,
         GDriveMimeType.PPT.value,
         GDriveMimeType.SPREADSHEET.value,
+        GDriveMimeType.PLAIN_TEXT.value,
+        GDriveMimeType.MARKDOWN.value,
     ]:
         export_mime_type = "text/plain"
         if mime_type == GDriveMimeType.SPREADSHEET.value:
             export_mime_type = "text/csv"
         elif mime_type == GDriveMimeType.PPT.value:
             export_mime_type = "text/plain"
+        elif mime_type in [
+            GDriveMimeType.PLAIN_TEXT.value,
+            GDriveMimeType.MARKDOWN.value
+        ]:
+            export_mime_type = mime_type
 
         response = (
             service.files()