summaryrefslogtreecommitdiffstats
path: root/sqlglot/dialects/teradata.py
blob: 3d43793db5637fbfb569b7527b5f874ccc02995d (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
from __future__ import annotations

from sqlglot import exp, generator, parser, tokens
from sqlglot.dialects.dialect import Dialect, max_or_greatest, min_or_least
from sqlglot.tokens import TokenType


class Teradata(Dialect):
    class Tokenizer(tokens.Tokenizer):
        # https://docs.teradata.com/r/Teradata-Database-SQL-Functions-Operators-Expressions-and-Predicates/March-2017/Comparison-Operators-and-Functions/Comparison-Operators/ANSI-Compliance
        KEYWORDS = {
            **tokens.Tokenizer.KEYWORDS,
            "BYTEINT": TokenType.SMALLINT,
            "SEL": TokenType.SELECT,
            "INS": TokenType.INSERT,
            "MOD": TokenType.MOD,
            "LT": TokenType.LT,
            "LE": TokenType.LTE,
            "GT": TokenType.GT,
            "GE": TokenType.GTE,
            "^=": TokenType.NEQ,
            "NE": TokenType.NEQ,
            "NOT=": TokenType.NEQ,
            "ST_GEOMETRY": TokenType.GEOMETRY,
        }

        # teradata does not support % for modulus
        SINGLE_TOKENS = {**tokens.Tokenizer.SINGLE_TOKENS}
        SINGLE_TOKENS.pop("%")

    class Parser(parser.Parser):
        CHARSET_TRANSLATORS = {
            "GRAPHIC_TO_KANJISJIS",
            "GRAPHIC_TO_LATIN",
            "GRAPHIC_TO_UNICODE",
            "GRAPHIC_TO_UNICODE_PadSpace",
            "KANJI1_KanjiEBCDIC_TO_UNICODE",
            "KANJI1_KanjiEUC_TO_UNICODE",
            "KANJI1_KANJISJIS_TO_UNICODE",
            "KANJI1_SBC_TO_UNICODE",
            "KANJISJIS_TO_GRAPHIC",
            "KANJISJIS_TO_LATIN",
            "KANJISJIS_TO_UNICODE",
            "LATIN_TO_GRAPHIC",
            "LATIN_TO_KANJISJIS",
            "LATIN_TO_UNICODE",
            "LOCALE_TO_UNICODE",
            "UNICODE_TO_GRAPHIC",
            "UNICODE_TO_GRAPHIC_PadGraphic",
            "UNICODE_TO_GRAPHIC_VarGraphic",
            "UNICODE_TO_KANJI1_KanjiEBCDIC",
            "UNICODE_TO_KANJI1_KanjiEUC",
            "UNICODE_TO_KANJI1_KANJISJIS",
            "UNICODE_TO_KANJI1_SBC",
            "UNICODE_TO_KANJISJIS",
            "UNICODE_TO_LATIN",
            "UNICODE_TO_LOCALE",
            "UNICODE_TO_UNICODE_FoldSpace",
            "UNICODE_TO_UNICODE_Fullwidth",
            "UNICODE_TO_UNICODE_Halfwidth",
            "UNICODE_TO_UNICODE_NFC",
            "UNICODE_TO_UNICODE_NFD",
            "UNICODE_TO_UNICODE_NFKC",
            "UNICODE_TO_UNICODE_NFKD",
        }

        FUNC_TOKENS = {*parser.Parser.FUNC_TOKENS}
        FUNC_TOKENS.remove(TokenType.REPLACE)

        STATEMENT_PARSERS = {
            **parser.Parser.STATEMENT_PARSERS,  # type: ignore
            TokenType.REPLACE: lambda self: self._parse_create(),
        }

        FUNCTION_PARSERS = {
            **parser.Parser.FUNCTION_PARSERS,  # type: ignore
            "RANGE_N": lambda self: self._parse_rangen(),
            "TRANSLATE": lambda self: self._parse_translate(self.STRICT_CAST),
        }

        def _parse_translate(self, strict: bool) -> exp.Expression:
            this = self._parse_conjunction()

            if not self._match(TokenType.USING):
                self.raise_error("Expected USING in TRANSLATE")

            if self._match_texts(self.CHARSET_TRANSLATORS):
                charset_split = self._prev.text.split("_TO_")
                to = self.expression(exp.CharacterSet, this=charset_split[1])
            else:
                self.raise_error("Expected a character set translator after USING in TRANSLATE")

            return self.expression(exp.Cast if strict else exp.TryCast, this=this, to=to)

        # FROM before SET in Teradata UPDATE syntax
        # https://docs.teradata.com/r/Enterprise_IntelliFlex_VMware/Teradata-VantageTM-SQL-Data-Manipulation-Language-17.20/Statement-Syntax/UPDATE/UPDATE-Syntax-Basic-Form-FROM-Clause
        def _parse_update(self) -> exp.Expression:
            return self.expression(
                exp.Update,
                **{  # type: ignore
                    "this": self._parse_table(alias_tokens=self.UPDATE_ALIAS_TOKENS),
                    "from": self._parse_from(),
                    "expressions": self._match(TokenType.SET)
                    and self._parse_csv(self._parse_equality),
                    "where": self._parse_where(),
                },
            )

        def _parse_rangen(self):
            this = self._parse_id_var()
            self._match(TokenType.BETWEEN)

            expressions = self._parse_csv(self._parse_conjunction)
            each = self._match_text_seq("EACH") and self._parse_conjunction()

            return self.expression(exp.RangeN, this=this, expressions=expressions, each=each)

    class Generator(generator.Generator):
        TYPE_MAPPING = {
            **generator.Generator.TYPE_MAPPING,  # type: ignore
            exp.DataType.Type.GEOMETRY: "ST_GEOMETRY",
        }

        PROPERTIES_LOCATION = {
            **generator.Generator.PROPERTIES_LOCATION,  # type: ignore
            exp.PartitionedByProperty: exp.Properties.Location.POST_INDEX,
        }

        TRANSFORMS = {
            **generator.Generator.TRANSFORMS,
            exp.Max: max_or_greatest,
            exp.Min: min_or_least,
            exp.ToChar: lambda self, e: self.function_fallback_sql(e),
        }

        def partitionedbyproperty_sql(self, expression: exp.PartitionedByProperty) -> str:
            return f"PARTITION BY {self.sql(expression, 'this')}"

        # FROM before SET in Teradata UPDATE syntax
        # https://docs.teradata.com/r/Enterprise_IntelliFlex_VMware/Teradata-VantageTM-SQL-Data-Manipulation-Language-17.20/Statement-Syntax/UPDATE/UPDATE-Syntax-Basic-Form-FROM-Clause
        def update_sql(self, expression: exp.Update) -> str:
            this = self.sql(expression, "this")
            from_sql = self.sql(expression, "from")
            set_sql = self.expressions(expression, flat=True)
            where_sql = self.sql(expression, "where")
            sql = f"UPDATE {this}{from_sql} SET {set_sql}{where_sql}"
            return self.prepend_ctes(expression, sql)

        def mod_sql(self, expression: exp.Mod) -> str:
            return self.binary(expression, "MOD")

        def datatype_sql(self, expression: exp.DataType) -> str:
            type_sql = super().datatype_sql(expression)
            prefix_sql = expression.args.get("prefix")
            return f"SYSUDTLIB.{type_sql}" if prefix_sql else type_sql

        def rangen_sql(self, expression: exp.RangeN) -> str:
            this = self.sql(expression, "this")
            expressions_sql = self.expressions(expression)
            each_sql = self.sql(expression, "each")
            each_sql = f" EACH {each_sql}" if each_sql else ""

            return f"RANGE_N({this} BETWEEN {expressions_sql}{each_sql})"