"""ParadeDB dialect extending PGVector with pg_search operators.
Adds support for ParadeDB pg_search operators (pg_search 0.18.0+):
- @@@ : Full-text search / complex query expressions (BM25)
- &&& : Match conjunction (all tokenized terms must match)
- ||| : Match disjunction (any tokenized term matches)
- === : Exact term match (no tokenization of the right-hand side)
- ### : Exact phrase match (token order and position enforced)
- ## : Proximity match, any order (``'a' ## n ## 'b'``)
- ##> : Ordered proximity match (left term must appear first)
Scoring and snippets are plain functions in ParadeDB (``pdb.score()``,
``pdb.snippet()``), not operators, so they need no dialect support.
Also supports pgvector distance operators independently. Registered with
sqlglot through the ``sqlglot.dialects`` entry-point group in
``pyproject.toml`` and by the ``Dialect`` metaclass on import.
"""
from sqlglot.dialects.postgres import Postgres
from sqlspec.dialects.postgres._generators import ParadeDBGenerator
from sqlspec.dialects.postgres._operators import (
PARADEDB_OPERATOR_TOKENS,
PGVECTOR_OPERATOR_TOKENS,
register_postgres_extension_operators,
)
__all__ = ("ParadeDB",)
register_postgres_extension_operators()
class ParadeDBTokenizer(Postgres.Tokenizer):
"""Tokenizer with ParadeDB search operators and pgvector distance operators."""
KEYWORDS = {**Postgres.Tokenizer.KEYWORDS, **PARADEDB_OPERATOR_TOKENS, **PGVECTOR_OPERATOR_TOKENS}
[docs]
class ParadeDB(Postgres):
"""ParadeDB dialect with pg_search and pgvector extension support."""
Tokenizer = ParadeDBTokenizer
Generator = ParadeDBGenerator
