gmakstutis/LightRAG
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Fix delimiter processing and optimize case-sensitive handling

Browse source 
• Fix completion_delimiter reference bug
• Add case check before lowercase conversion
• Improve delimiter corruption handling
• Optimize redundant processing logic
This commit is contained in:
yangdx 2025-09-14 12:23:48 +08:00
parent ff705a2323
commit 619553021e
1 changed files with 7 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

11
lightrag/operate.py
View file

@ -879,7 +879,7 @@ async def _process_extraction_result(
# Split LLL output result to records by "\n" # Split LLL output result to records by "\n"
records = split_string_by_multi_markers( records = split_string_by_multi_markers(
result, result,
["\n", completion_delimiter], ["\n", completion_delimiter, completion_delimiter.lower()],
) )
# Fix LLM output format error which use tuple_delimiter to seperate record instead of "\n" # Fix LLM output format error which use tuple_delimiter to seperate record instead of "\n"
@ -926,9 +926,12 @@ async def _process_extraction_result(
# Fix various forms of tuple_delimiter corruption from the LLM output using the dedicated function # Fix various forms of tuple_delimiter corruption from the LLM output using the dedicated function
delimiter_core = tuple_delimiter[2:-2] # Extract "#" from "<|#|>" delimiter_core = tuple_delimiter[2:-2] # Extract "#" from "<|#|>"
record = fix_tuple_delimiter_corruption(record, delimiter_core, tuple_delimiter) record = fix_tuple_delimiter_corruption(record, delimiter_core, tuple_delimiter)
# change delimiter_core to lower case, and fix again if delimiter_core != delimiter_core.lower():
delimiter_core = delimiter_core.lower() # change delimiter_core to lower case, and fix again
record = fix_tuple_delimiter_corruption(record, delimiter_core, tuple_delimiter) delimiter_core = delimiter_core.lower()
record = fix_tuple_delimiter_corruption(
record, delimiter_core, tuple_delimiter
)
record_attributes = split_string_by_multi_markers(record, [tuple_delimiter]) record_attributes = split_string_by_multi_markers(record, [tuple_delimiter])