feat(sqlglot): pin to latest version of sqlglot (#372)

Pin minimum version of sqlglot and add `sqlglot[c]` group.

Author: cofin

.pre-commit-config.yaml

@@ -2,7 +2,7 @@ default_language_version:
   python: "3"
 repos:
   - repo: https://github.com/compilerla/conventional-pre-commit
-    rev: v4.3.0
+    rev: v4.4.0
     hooks:
       - id: conventional-pre-commit
         stages: [commit-msg]
@@ -17,7 +17,7 @@ repos:
       - id: mixed-line-ending
       - id: trailing-whitespace
   - repo: https://github.com/charliermarsh/ruff-pre-commit
-    rev: "v0.15.0"
+    rev: "v0.15.2"
     hooks:
       - id: ruff
         args: ["--fix"]

pyproject.toml

@@ -24,7 +24,7 @@ maintainers = [{ name = "Litestar Developers", email = "hello@litestar.dev" }]
 name = "sqlspec"
 readme = "README.md"
 requires-python = ">=3.10, <4.0"
-version = "0.39.0"
+version = "0.40.0"
 
 [project.urls]
 Discord = "https://discord.gg/litestar"
@@ -49,15 +49,15 @@ flask = ["flask"]
 fsspec = ["fsspec"]
 litestar = ["litestar"]
 msgspec = ["msgspec"]
-mypyc = []
+mypyc = ["sqlglot[c]"]
 mysql-connector = ["mysql-connector-python"]
 nanoid = ["fastnanoid>=0.4.1"]
 obstore = ["obstore"]
 opentelemetry = ["opentelemetry-instrumentation"]
 oracledb = ["oracledb"]
 orjson = ["orjson"]
 pandas = ["pandas", "pyarrow"]
-performance = ["sqlglot[rs]", "msgspec"]
+performance = ["msgspec"]
 polars = ["polars", "pyarrow"]
 prometheus = ["prometheus-client"]
 psqlpy = ["psqlpy"]
@@ -241,7 +241,7 @@ opt_level = "3"   # Maximum optimization (0-3)
 allow_dirty = true
 commit = false
 commit_args = "--no-verify"
-current_version = "0.39.0"
+current_version = "0.40.0"
 ignore_missing_files = false
 ignore_missing_version = false
 message = "chore(release): bump to v{new_version}"

sqlspec/adapters/spanner/dialect/_spangres.py

@@ -4,10 +4,11 @@
 
 from sqlglot import exp
 from sqlglot.dialects.postgres import Postgres
-from sqlglot.tokens import TokenType
+from sqlglot.tokenizer_core import TokenType
 
 __all__ = ("Spangres",)
 
+
 _ROW_DELETION_NAME = "ROW_DELETION_POLICY"
 _TTL_MIN_COMPONENTS = 2
 
@@ -16,16 +17,16 @@ class SpangresParser(Postgres.Parser):
     """Parse Spanner row deletion policies."""
 
     def _parse_property(self) -> exp.Expression:
-        if self._match_text_seq("ROW", "DELETION", "POLICY"):  # type: ignore[no-untyped-call]
-            self._match(TokenType.L_PAREN)  # type: ignore[no-untyped-call]
-            self._match_text_seq("OLDER_THAN")  # type: ignore[no-untyped-call]
-            self._match(TokenType.L_PAREN)  # type: ignore[no-untyped-call]
+        if self._match_text_seq("ROW", "DELETION", "POLICY"):
+            self._match(TokenType.L_PAREN)
+            self._match_text_seq("OLDER_THAN")
+            self._match(TokenType.L_PAREN)
             column = cast("exp.Expression", self._parse_id_var())
-            self._match(TokenType.COMMA)  # type: ignore[no-untyped-call]
-            self._match_text_seq("INTERVAL")  # type: ignore[no-untyped-call]
+            self._match(TokenType.COMMA)
+            self._match_text_seq("INTERVAL")
             interval = cast("exp.Expression", self._parse_expression())
-            self._match(TokenType.R_PAREN)  # type: ignore[no-untyped-call]
-            self._match(TokenType.R_PAREN)  # type: ignore[no-untyped-call]
+            self._match(TokenType.R_PAREN)
+            self._match(TokenType.R_PAREN)
 
             return exp.Property(
                 this=exp.Literal.string(_ROW_DELETION_NAME), value=exp.Tuple(expressions=[column, interval])

sqlspec/adapters/spanner/dialect/_spanner.py

@@ -9,7 +9,7 @@
 
 from sqlglot import exp
 from sqlglot.dialects.bigquery import BigQuery
-from sqlglot.tokens import TokenType
+from sqlglot.tokenizer_core import TokenType
 
 __all__ = ("Spanner",)
 
@@ -41,14 +41,14 @@ def _parse_table_parts(
         """Parse Spanner table options including interleaving metadata."""
         table = super()._parse_table_parts(schema=schema, is_db_reference=is_db_reference, wildcard=wildcard)
 
-        if self._match_text_seq("INTERLEAVE", "IN", "PARENT"):  # type: ignore[no-untyped-call]
+        if self._match_text_seq("INTERLEAVE", "IN", "PARENT"):
             parent = cast("exp.Expression", self._parse_table(schema=True, is_db_reference=True))
             on_delete: str | None = None
 
-            if self._match_text_seq("ON", "DELETE"):  # type: ignore[no-untyped-call]
-                if self._match_text_seq("CASCADE"):  # type: ignore[no-untyped-call]
+            if self._match_text_seq("ON", "DELETE"):
+                if self._match_text_seq("CASCADE"):
                     on_delete = "CASCADE"
-                elif self._match_text_seq("NO", "ACTION"):  # type: ignore[no-untyped-call]
+                elif self._match_text_seq("NO", "ACTION"):
                     on_delete = "NO ACTION"
 
             table.set("interleave_parent", parent)
@@ -59,25 +59,25 @@ def _parse_table_parts(
 
     def _parse_property(self) -> exp.Expression:
         """Parse Spanner row deletion policy or PostgreSQL-style TTL."""
-        if self._match_text_seq("ROW", "DELETION", "POLICY"):  # type: ignore[no-untyped-call]
-            self._match(TokenType.L_PAREN)  # type: ignore[no-untyped-call]
-            self._match_text_seq("OLDER_THAN")  # type: ignore[no-untyped-call]
-            self._match(TokenType.L_PAREN)  # type: ignore[no-untyped-call]
+        if self._match_text_seq("ROW", "DELETION", "POLICY"):
+            self._match(TokenType.L_PAREN)
+            self._match_text_seq("OLDER_THAN")
+            self._match(TokenType.L_PAREN)
             column = cast("exp.Expression", self._parse_id_var())
-            self._match(TokenType.COMMA)  # type: ignore[no-untyped-call]
-            self._match_text_seq("INTERVAL")  # type: ignore[no-untyped-call]
+            self._match(TokenType.COMMA)
+            self._match_text_seq("INTERVAL")
             interval = cast("exp.Expression", self._parse_expression())
-            self._match(TokenType.R_PAREN)  # type: ignore[no-untyped-call]
-            self._match(TokenType.R_PAREN)  # type: ignore[no-untyped-call]
+            self._match(TokenType.R_PAREN)
+            self._match(TokenType.R_PAREN)
 
             return exp.Property(
                 this=exp.Literal.string(_ROW_DELETION_NAME), value=exp.Tuple(expressions=[column, interval])
             )
 
-        if self._match_text_seq("TTL"):  # type: ignore[no-untyped-call]  # PostgreSQL-dialect style, keep for compatibility
-            self._match_text_seq("INTERVAL")  # type: ignore[no-untyped-call]
+        if self._match_text_seq("TTL"):
+            self._match_text_seq("INTERVAL")
             interval = cast("exp.Expression", self._parse_expression())
-            self._match_text_seq("ON")  # type: ignore[no-untyped-call]
+            self._match_text_seq("ON")
             column = cast("exp.Expression", self._parse_id_var())
 
             return exp.Property(this=exp.Literal.string("TTL"), value=exp.Tuple(expressions=[interval, column]))

sqlspec/builder/_base.py

@@ -394,7 +394,7 @@ def _parameterize_expression(self, expression: exp.Expression) -> exp.Expression
             A new expression with literals replaced by parameter placeholders
         """
 
-        return expression.transform(_ExpressionParameterizer(self), copy=False)
+        return cast("exp.Expression", expression.transform(_ExpressionParameterizer(self), copy=False))
 
     def add_parameter(self: Self, value: Any, name: str | None = None) -> tuple[Self, str]:
         """Explicitly adds a parameter to the query.
@@ -426,9 +426,6 @@ def load_parameters(self, parameters: "Mapping[str, Any]") -> None:
 
         Args:
             parameters: Mapping of parameter names to values.
-
-        Raises:
-            SQLBuilderError: If a parameter name already exists on the builder.
         """
         if not parameters:
             return
@@ -444,9 +441,6 @@ def load_ctes(self, ctes: "Iterable[exp.CTE]") -> None:
 
         Args:
             ctes: Iterable of CTE expressions to register.
-
-        Raises:
-            SQLBuilderError: If a CTE alias is missing or duplicated.
         """
         for cte in ctes:
             alias = self._resolve_cte_alias(cte)
@@ -543,7 +537,7 @@ def _update_placeholders_in_expression(
             Updated expression with new placeholder names
         """
 
-        return expression.transform(_PlaceholderReplacer(param_mapping), copy=False)
+        return cast("exp.Expression", expression.transform(_PlaceholderReplacer(param_mapping), copy=False))
 
     def _generate_builder_cache_key(self, config: "StatementConfig | None" = None) -> str:
         """Generate cache key based on builder state and configuration.
@@ -893,7 +887,7 @@ def _is_oracle_dialect(self, dialect: "DialectType | str | None") -> bool:
     def _unquote_identifiers_for_oracle(self, expression: exp.Expression) -> exp.Expression:
         """Remove identifier quoting to avoid Oracle case-sensitive lookup issues."""
 
-        return expression.copy().transform(_unquote_identifier, copy=False)
+        return cast("exp.Expression", expression.copy().transform(_unquote_identifier, copy=False))
 
     def _strip_lock_identifier_quotes(self, sql_string: str) -> str:
         for keyword in ("FOR UPDATE OF ", "FOR SHARE OF "):

tests/unit/adapters/test_spanner/test_dialect.py

@@ -703,7 +703,7 @@ def test_dialect_inheritance() -> None:
 
 def test_dialect_keywords() -> None:
     """Test that Spanner-specific keywords are registered."""
-    from sqlglot.tokens import TokenType
+    from sqlglot.tokenizer_core import TokenType
 
     # Keywords are only registered if sqlglot supports them
     keywords = Spanner.Tokenizer.KEYWORDS

tests/unit/builder/test_temporal.py

@@ -96,8 +96,9 @@ def test_join_as_of_dialect() -> None:
     sql_str = query.build(dialect="bigquery").sql
     normalized = normalize_sql(sql_str)
     # Should use FOR SYSTEM_TIME AS OF because dialect is passed at build time
-    assert "LEFT JOIN audit_log FOR SYSTEM_TIME AS OF '-1h'" in normalized
-    assert "log ON orders.id = log.order_id" in normalized
+    assert "LEFT JOIN audit_log" in normalized
+    assert "FOR SYSTEM_TIME AS OF '-1h'" in normalized
+    assert "ON orders.id = log.order_id" in normalized
 
 
 def test_join_as_of_dialect_override() -> None: