Fix #2

ums/agent/agent.py

@@ -192,22 +192,45 @@ class ExtractAgent(BasicAgent):
 		"""
 
 	@validate_call
-	def store_extracted(self, data:RiddleData, extracted:ExtractedData) -> str:
+	def store_extracted(self, data:RiddleData, extracted:ExtractedData, allow_overwrite:bool=True) -> str:
 		"""
 			Stores the newly extracted data (in `extracted`) from `data` (i.e., `data.file_plain`) 
 			and returns the filename to use in `data.file_extracted`.
+
+			If there already exists an extracted file for this `data`, the file will be overwritten if `allow_overwrite=True`.
+			Generally the system will check, if the contents of the current file are equal to the contents to write. 
+			File with equal content will not be written again.
 		"""
 
+		# get path and name
 		path_name = data.file_plain[:data.file_plain.rfind('.')]
-		
 		candidate = "{}.json".format(path_name)
-		cnt = 0
-		while os.path.isfile(candidate):
-			cnt += 1
-			candidate = "{}-{}.json".format(path_name, cnt)
 
+		# data to write
+		data = extracted.model_dump_json()
+
+		# check for file
+		if os.path.isfile(candidate):
+
+			# get current content
+			with open(candidate, 'r') as f:
+				content = f.read()
+
+			# files equal -> no need to rewrite
+			if content == data:
+				return candidate
+			
+			# not equal and overwrite not allowed
+			elif not allow_overwrite:
+				# get non-existent file name
+				cnt = 0
+				while os.path.isfile(candidate):
+					cnt += 1
+					candidate = "{}-{}.json".format(path_name, cnt)
+
+		# write file
 		with open(candidate, 'w+') as f:
-			f.write(extracted.model_dump_json())
+			f.write(data)
 
 		return candidate