fix: expand the source id regex item part

":" is required for the mail integration to work but the regex has been
expanded to allow alphanumeric text and hyphen too.

Signed-off-by: Anupam Kumar <[email protected]>

Author: kyteinsky

context_chat_backend/chain/ingest/injest.py

@@ -186,6 +186,16 @@ def embed_sources(
 			f'{source.filename} ({_decode_latin_1(source.headers["title"])})'
 			for source in sources_filtered
 		],
+		'invalid_source_ids': [
+			source.filename for source in sources
+			if not is_valid_source_id(source.filename)  # pyright: ignore[reportArgumentType]
+		],
+		'not_allowed_file_ids': [
+			source.filename for source in sources
+			if not _allowed_file(source)
+		],
+		'len(source_ids)': len(sources_filtered),
+		'len(total_source_ids)': len(sources),
 	})
 
 	vectordb = vectordb_loader.load()

context_chat_backend/utils.py

@@ -100,7 +100,8 @@ def exec_in_proc(group=None, target=None, name=None, args=(), kwargs={}, *, daem
 
 
 def is_valid_source_id(source_id: str) -> bool:
-	return re.match(r'^[a-zA-Z0-9_-]+__[a-zA-Z0-9_-]+: \d+$', source_id) is not None
+	# note the ":" in the item id part
+	return re.match(r'^[a-zA-Z0-9_-]+__[a-zA-Z0-9_-]+: [a-zA-Z0-9:-]+$', source_id) is not None
 
 
 def is_valid_provider_id(provider_id: str) -> bool: