github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-06-04 14:13:29 +00:00
Code Issues Packages Projects Releases Wiki Activity

community[patch]: Fix neo4j enhanced schema bugs (#21072)

Browse Source
This commit is contained in:
Tomaz Bratanic 2024-05-01 02:16:26 +02:00 committed by GitHub
parent 8d2909ee25
commit c9e96bb5e2
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 17 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
libs/community/langchain_community/graphs/neo4j_graph.py
View File

@ -13,7 +13,6 @@ EXHAUSTIVE_SEARCH_LIMIT = 10000
LIST_LIMIT = 128 LIST_LIMIT = 128
# Threshold for returning all available prop values in graph schema # Threshold for returning all available prop values in graph schema
DISTINCT_VALUE_LIMIT = 10 DISTINCT_VALUE_LIMIT = 10
NL = "\n"
node_properties_query = """ node_properties_query = """
CALL apoc.meta.data() CALL apoc.meta.data()
@ -52,6 +51,10 @@ include_docs_query = (
) )
def clean_string_values(text: str) -> str:
return text.replace("\n", " ").replace("\r", " ")
def value_sanitize(d: Any) -> Any: def value_sanitize(d: Any) -> Any:
"""Sanitize the input dictionary or list. """Sanitize the input dictionary or list.
@ -170,7 +173,7 @@ def _enhanced_schema_cypher(
f" distinct_count: size(`{prop_name}_values`)" f" distinct_count: size(`{prop_name}_values`)"
) )
) )
elif prop_type in ["INTEGER", "FLOAT", "DATE"]: elif prop_type in ["INTEGER", "FLOAT", "DATE", "DATE_TIME"]:
with_clauses.append(f"min(n.`{prop_name}`) AS `{prop_name}_min`") with_clauses.append(f"min(n.`{prop_name}`) AS `{prop_name}_min`")
with_clauses.append(f"max(n.`{prop_name}`) AS `{prop_name}_max`") with_clauses.append(f"max(n.`{prop_name}`) AS `{prop_name}_max`")
with_clauses.append( with_clauses.append(
@ -194,7 +197,8 @@ def _enhanced_schema_cypher(
f"min_size: `{prop_name}_size_min`, " f"min_size: `{prop_name}_size_min`, "
f"max_size: `{prop_name}_size_max`" f"max_size: `{prop_name}_size_max`"
) )
elif prop_type in ["BOOLEAN", "POINT", "DURATION"]:
continue
output_dict[prop_name] = "{" + return_clauses.pop() + "}" output_dict[prop_name] = "{" + return_clauses.pop() + "}"
else: else:
# Just sample 5 random nodes # Just sample 5 random nodes
@ -210,7 +214,7 @@ def _enhanced_schema_cypher(
) )
) )
return_clauses.append(f"values: `{prop_name}_values`") return_clauses.append(f"values: `{prop_name}_values`")
elif prop_type in ["INTEGER", "FLOAT", "DATE"]: elif prop_type in ["INTEGER", "FLOAT", "DATE", "DATE_TIME"]:
with_clauses.append( with_clauses.append(
f"collect(distinct toString(n.`{prop_name}`)) " f"collect(distinct toString(n.`{prop_name}`)) "
f"AS `{prop_name}_values`" f"AS `{prop_name}_values`"
@ -226,6 +230,8 @@ def _enhanced_schema_cypher(
return_clauses.append( return_clauses.append(
f"min_size: `{prop_name}_size_min`,max_size: `{prop_name}_size_max`" f"min_size: `{prop_name}_size_min`,max_size: `{prop_name}_size_max`"
) )
elif prop_type in ["BOOLEAN", "POINT", "DURATION"]:
continue
output_dict[prop_name] = "{" + return_clauses.pop() + "}" output_dict[prop_name] = "{" + return_clauses.pop() + "}"
@ -253,7 +259,7 @@ def _format_schema(schema: Dict, is_enhanced: bool) -> str:
if prop["type"] == "STRING": if prop["type"] == "STRING":
if prop.get("distinct_count", 11) > DISTINCT_VALUE_LIMIT: if prop.get("distinct_count", 11) > DISTINCT_VALUE_LIMIT:
example = ( example = (
f'Example: "{prop["values"][0].replace(NL, " ")}"' f'Example: "{clean_string_values(prop["values"][0])}"'
if prop["values"] if prop["values"]
else "" else ""
) )
@ -261,13 +267,13 @@ def _format_schema(schema: Dict, is_enhanced: bool) -> str:
example = ( example = (
( (
"Available options: " "Available options: "
f'{[el.replace(NL, " ") for el in prop["values"]]}' f'{[clean_string_values(el) for el in prop["values"]]}'
) )
if prop["values"] if prop["values"]
else "" else ""
) )
elif prop["type"] in ["INTEGER", "FLOAT", "DATE"]: elif prop["type"] in ["INTEGER", "FLOAT", "DATE", "DATE_TIME"]:
if prop.get("min") is not None: if prop.get("min") is not None:
example = f'Min: {prop["min"]}, Max: {prop["max"]}' example = f'Min: {prop["min"]}, Max: {prop["max"]}'
else: else:
@ -282,7 +288,7 @@ def _format_schema(schema: Dict, is_enhanced: bool) -> str:
f'Min Size: {prop["min_size"]}, Max Size: {prop["max_size"]}' f'Min Size: {prop["min_size"]}, Max Size: {prop["max_size"]}'
) )
formatted_node_props.append( formatted_node_props.append(
f" - `{prop['property']}: {prop['type']}` {example}" f" - `{prop['property']}`: {prop['type']}` {example}"
) )
# Enhanced formatting for relationships # Enhanced formatting for relationships
@ -293,7 +299,7 @@ def _format_schema(schema: Dict, is_enhanced: bool) -> str:
if prop["type"] == "STRING": if prop["type"] == "STRING":
if prop.get("distinct_count", 11) > DISTINCT_VALUE_LIMIT: if prop.get("distinct_count", 11) > DISTINCT_VALUE_LIMIT:
example = ( example = (
f'Example: "{prop["values"][0].replace(NL, " ")}"' f'Example: "{clean_string_values(prop["values"][0])}"'
if prop["values"] if prop["values"]
else "" else ""
) )
@ -301,12 +307,12 @@ def _format_schema(schema: Dict, is_enhanced: bool) -> str:
example = ( example = (
( (
"Available options: " "Available options: "
f'{[el.replace(NL, " ") for el in prop["values"]]}' f'{[clean_string_values(el) for el in prop["values"]]}'
) )
if prop["values"] if prop["values"]
else "" else ""
) )
elif prop["type"] in ["INTEGER", "FLOAT", "DATE"]: elif prop["type"] in ["INTEGER", "FLOAT", "DATE", "DATE_TIME"]:
if prop.get("min"): # If we have min/max if prop.get("min"): # If we have min/max
example = f'Min: {prop["min"]}, Max: {prop["max"]}' example = f'Min: {prop["min"]}, Max: {prop["max"]}'
else: # return a single value else: # return a single value