MetadataFilters

pydantic model llama_index.vector_stores.MetadataFilters

Metadata filters for vector stores.

Currently only supports exact match filters. TODO: support more advanced expressions.

Show JSON schema

{
   "title": "MetadataFilters",
   "description": "Metadata filters for vector stores.\n\nCurrently only supports exact match filters.\nTODO: support more advanced expressions.",
   "type": "object",
   "properties": {
      "filters": {
         "title": "Filters",
         "type": "array",
         "items": {
            "$ref": "#/definitions/MetadataFilter"
         }
      },
      "condition": {
         "default": "and",
         "allOf": [
            {
               "$ref": "#/definitions/FilterCondition"
            }
         ]
      }
   },
   "required": [
      "filters"
   ],
   "definitions": {
      "FilterOperator": {
         "title": "FilterOperator",
         "description": "Vector store filter operator.",
         "enum": [
            "==",
            ">",
            "<",
            "!=",
            ">=",
            "<=",
            "in",
            "nin",
            "text_match"
         ],
         "type": "string"
      },
      "MetadataFilter": {
         "title": "MetadataFilter",
         "description": "Comprehensive metadata filter for vector stores to support more operators.\n\nValue uses Strict* types, as int, float and str are compatible types and were all\nconverted to string before.\n\nSee: https://docs.pydantic.dev/latest/usage/types/#strict-types",
         "type": "object",
         "properties": {
            "key": {
               "title": "Key",
               "type": "string"
            },
            "value": {
               "title": "Value",
               "anyOf": [
                  {
                     "type": "integer"
                  },
                  {
                     "type": "number"
                  },
                  {
                     "type": "string"
                  }
               ]
            },
            "operator": {
               "default": "==",
               "allOf": [
                  {
                     "$ref": "#/definitions/FilterOperator"
                  }
               ]
            }
         },
         "required": [
            "key",
            "value"
         ]
      },
      "FilterCondition": {
         "title": "FilterCondition",
         "description": "Vector store filter conditions to combine different filters.",
         "enum": [
            "and",
            "or"
         ],
         "type": "string"
      }
   }
}

Fields

• condition (Optional[llama_index.vector_stores.types.FilterCondition])

• filters (List[llama_index.vector_stores.types.MetadataFilter])

field condition: Optional[FilterCondition] = FilterCondition.AND

field filters: List[MetadataFilter] [Required]

classmethod from_dict(filter_dict: Dict) → MetadataFilters

Create MetadataFilters from json.

classmethod from_dicts(filter_dicts: List[Dict], condition: Optional[FilterCondition] = FilterCondition.AND) → MetadataFilters

Create MetadataFilters from dicts.

This takes in a list of individual MetadataFilter objects, along with the condition.

Parameters

• filter_dicts – List of dicts, each dict is a MetadataFilter.

• condition – FilterCondition to combine different filters.

legacy_filters() → List[MetadataFilter]

Convert MetadataFilters to legacy ExactMatchFilters.