sqlglot.optimizer.qualify

 1from __future__ import annotations
 2
 3import typing as t
 4
 5from sqlglot import exp
 6from sqlglot.dialects.dialect import Dialect, DialectType
 7from sqlglot.optimizer.isolate_table_selects import isolate_table_selects
 8from sqlglot.optimizer.normalize_identifiers import normalize_identifiers
 9from sqlglot.optimizer.qualify_columns import (
10    pushdown_cte_alias_columns as pushdown_cte_alias_columns_func,
11    qualify_columns as qualify_columns_func,
12    quote_identifiers as quote_identifiers_func,
13    validate_qualify_columns as validate_qualify_columns_func,
14)
15from sqlglot.optimizer.qualify_tables import qualify_tables
16from sqlglot.schema import Schema, ensure_schema
17
18
19def qualify(
20    expression: exp.Expression,
21    dialect: DialectType = None,
22    db: t.Optional[str] = None,
23    catalog: t.Optional[str] = None,
24    schema: t.Optional[dict | Schema] = None,
25    expand_alias_refs: bool = True,
26    expand_stars: bool = True,
27    infer_schema: t.Optional[bool] = None,
28    isolate_tables: bool = False,
29    qualify_columns: bool = True,
30    validate_qualify_columns: bool = True,
31    quote_identifiers: bool = True,
32    identify: bool = True,
33) -> exp.Expression:
34    """
35    Rewrite sqlglot AST to have normalized and qualified tables and columns.
36
37    This step is necessary for all further SQLGlot optimizations.
38
39    Example:
40        >>> import sqlglot
41        >>> schema = {"tbl": {"col": "INT"}}
42        >>> expression = sqlglot.parse_one("SELECT col FROM tbl")
43        >>> qualify(expression, schema=schema).sql()
44        'SELECT "tbl"."col" AS "col" FROM "tbl" AS "tbl"'
45
46    Args:
47        expression: Expression to qualify.
48        db: Default database name for tables.
49        catalog: Default catalog name for tables.
50        schema: Schema to infer column names and types.
51        expand_alias_refs: Whether to expand references to aliases.
52        expand_stars: Whether to expand star queries. This is a necessary step
53            for most of the optimizer's rules to work; do not set to False unless you
54            know what you're doing!
55        infer_schema: Whether to infer the schema if missing.
56        isolate_tables: Whether to isolate table selects.
57        qualify_columns: Whether to qualify columns.
58        validate_qualify_columns: Whether to validate columns.
59        quote_identifiers: Whether to run the quote_identifiers step.
60            This step is necessary to ensure correctness for case sensitive queries.
61            But this flag is provided in case this step is performed at a later time.
62        identify: If True, quote all identifiers, else only necessary ones.
63
64    Returns:
65        The qualified expression.
66    """
67    schema = ensure_schema(schema, dialect=dialect)
68    expression = normalize_identifiers(expression, dialect=dialect)
69    expression = qualify_tables(expression, db=db, catalog=catalog, schema=schema, dialect=dialect)
70
71    if isolate_tables:
72        expression = isolate_table_selects(expression, schema=schema)
73
74    if Dialect.get_or_raise(dialect).PREFER_CTE_ALIAS_COLUMN:
75        expression = pushdown_cte_alias_columns_func(expression)
76
77    if qualify_columns:
78        expression = qualify_columns_func(
79            expression,
80            schema,
81            expand_alias_refs=expand_alias_refs,
82            expand_stars=expand_stars,
83            infer_schema=infer_schema,
84        )
85
86    if quote_identifiers:
87        expression = quote_identifiers_func(expression, dialect=dialect, identify=identify)
88
89    if validate_qualify_columns:
90        validate_qualify_columns_func(expression)
91
92    return expression