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

Refactor: Harden compute_args_hash against Unicode errors

Browse source
This commit is contained in:
yangdx 2025-08-19 17:19:39 +08:00
parent 2603e99005
commit 64058c771f
1 changed files with 10 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

13
lightrag/utils.py
View file

@ -265,7 +265,8 @@ class EmbeddingFunc:
def compute_args_hash(*args: Any) -> str: def compute_args_hash(*args: Any) -> str:
"""Compute a hash for the given arguments. """Compute a hash for the given arguments with safe Unicode handling.
Args: Args:
*args: Arguments to hash *args: Arguments to hash
Returns: Returns:
@ -276,8 +277,14 @@ def compute_args_hash(*args: Any) -> str:
# Convert all arguments to strings and join them # Convert all arguments to strings and join them
args_str = "".join([str(arg) for arg in args]) args_str = "".join([str(arg) for arg in args])
# Compute MD5 hash # Use 'replace' error handling to safely encode problematic Unicode characters
return hashlib.md5(args_str.encode()).hexdigest() # This replaces invalid characters with Unicode replacement character (U+FFFD)
try:
return hashlib.md5(args_str.encode("utf-8")).hexdigest()
except UnicodeEncodeError:
# Handle surrogate characters and other encoding issues
safe_bytes = args_str.encode("utf-8", errors="replace")
return hashlib.md5(safe_bytes).hexdigest()
def generate_cache_key(mode: str, cache_type: str, hash_value: str) -> str: def generate_cache_key(mode: str, cache_type: str, hash_value: str) -> str: