unstructured-ingest 1.0.32__tar.gz → 1.0.34__tar.gz

{unstructured_ingest-1.0.32 → unstructured_ingest-1.0.34}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: unstructured_ingest
-Version: 1.0.32
+Version: 1.0.34
 Summary: Local ETL data pipeline to get data RAG ready
 Author-email: Unstructured Technologies <devops@unstructuredai.io>
 License-Expression: Apache-2.0

unstructured_ingest-1.0.34/unstructured_ingest/__version__.py

@@ -0,0 +1 @@
+__version__ = "1.0.34"  # pragma: no cover

{unstructured_ingest-1.0.32 → unstructured_ingest-1.0.34}/unstructured_ingest/processes/connectors/confluence.py

@@ -12,6 +12,7 @@ from unstructured_ingest.data_types.file_data import (
     SourceIdentifiers,
 )
 from unstructured_ingest.error import SourceConnectionError
+from unstructured_ingest.errors_v2 import UserAuthError, UserError
 from unstructured_ingest.interfaces import (
     AccessConfig,
     ConnectionConfig,
@@ -96,7 +97,7 @@ class ConfluenceConnectionConfig(ConnectionConfig):
 
     @requires_dependencies(["atlassian"], extras="confluence")
     @contextmanager
-    def get_client(self) -> "Confluence":
+    def get_client(self) -> Generator["Confluence", None, None]:
         from atlassian import Confluence
 
         access_configs = self.access_config.get_secret_value()
@@ -126,15 +127,36 @@ class ConfluenceIndexer(Indexer):
 
     def precheck(self) -> bool:
         try:
-            # Attempt to retrieve a list of spaces with limit=1.
-            # This should only succeed if all creds are valid
-            with self.connection_config.get_client() as client:
+            self.connection_config.get_client()
+        except Exception as e:
+            logger.exception(f"Failed to connect to Confluence: {e}")
+            raise UserAuthError(f"Failed to connect to Confluence: {e}")
+
+        with self.connection_config.get_client() as client:
+            # opportunistically check the first space in list of all spaces
+            try:
                 client.get_all_spaces(limit=1)
+            except Exception as e:
+                logger.exception(f"Failed to connect to find any Confluence space: {e}")
+                raise UserError(f"Failed to connect to find any Confluence space: {e}")
+
             logger.info("Connection to Confluence successful.")
-            return True
-        except Exception as e:
-            logger.error(f"Failed to connect to Confluence: {e}", exc_info=True)
-            raise SourceConnectionError(f"Failed to connect to Confluence: {e}")
+
+            # If specific spaces are provided, check if we can access them
+            errors = []
+
+            if self.index_config.spaces:
+                for space_key in self.index_config.spaces:
+                    try:
+                        client.get_space(space_key)
+                    except Exception as e:
+                        logger.exception(f"Failed to connect to Confluence: {e}")
+                        errors.append(f"Failed to connect to '{space_key}' space, cause: '{e}'")
+
+            if errors:
+                raise UserError("\n".join(errors))
+
+        return True
 
     def _get_space_ids_and_keys(self) -> List[Tuple[str, int]]:
         """
@@ -406,7 +428,7 @@ class ConfluenceDownloader(Downloader):
                     expand="history.lastUpdated,version,body.view",
                 )
         except Exception as e:
-            logger.error(f"Failed to retrieve page with ID {doc_id}: {e}", exc_info=True)
+            logger.exception(f"Failed to retrieve page with ID {doc_id}: {e}")
             raise SourceConnectionError(f"Failed to retrieve page with ID {doc_id}: {e}")
 
         if not page:

{unstructured_ingest-1.0.32 → unstructured_ingest-1.0.34}/unstructured_ingest/processes/connectors/google_drive.py

@@ -344,6 +344,7 @@ class GoogleDriveIndexer(Indexer):
                 },
             ),
             additional_metadata=root_info,
+            display_name=source_identifiers.fullpath,
         )
 
     def get_paginated_results(

unstructured_ingest-1.0.32/unstructured_ingest/__version__.py

@@ -1 +0,0 @@
-__version__ = "1.0.32"  # pragma: no cover