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
|
from __future__ import annotations
import typing as t
from sqlglot import exp, generator, parser, tokens, transforms
from sqlglot.dialects.dialect import (
Dialect,
create_with_partitions_sql,
datestrtodate_sql,
format_time_lambda,
no_trycast_sql,
rename_func,
str_position_sql,
timestrtotime_sql,
ts_or_ds_to_date_sql,
)
def _date_add_sql(kind: str) -> t.Callable[[Drill.Generator, exp.DateAdd | exp.DateSub], str]:
def func(self: Drill.Generator, expression: exp.DateAdd | exp.DateSub) -> str:
this = self.sql(expression, "this")
unit = exp.var(expression.text("unit").upper() or "DAY")
return f"DATE_{kind}({this}, {self.sql(exp.Interval(this=expression.expression.copy(), unit=unit))})"
return func
def _str_to_date(self: Drill.Generator, expression: exp.StrToDate) -> str:
this = self.sql(expression, "this")
time_format = self.format_time(expression)
if time_format == Drill.DATE_FORMAT:
return f"CAST({this} AS DATE)"
return f"TO_DATE({this}, {time_format})"
class Drill(Dialect):
NORMALIZE_FUNCTIONS: bool | str = False
NULL_ORDERING = "nulls_are_last"
DATE_FORMAT = "'yyyy-MM-dd'"
DATEINT_FORMAT = "'yyyyMMdd'"
TIME_FORMAT = "'yyyy-MM-dd HH:mm:ss'"
TIME_MAPPING = {
"y": "%Y",
"Y": "%Y",
"YYYY": "%Y",
"yyyy": "%Y",
"YY": "%y",
"yy": "%y",
"MMMM": "%B",
"MMM": "%b",
"MM": "%m",
"M": "%-m",
"dd": "%d",
"d": "%-d",
"HH": "%H",
"H": "%-H",
"hh": "%I",
"h": "%-I",
"mm": "%M",
"m": "%-M",
"ss": "%S",
"s": "%-S",
"SSSSSS": "%f",
"a": "%p",
"DD": "%j",
"D": "%-j",
"E": "%a",
"EE": "%a",
"EEE": "%a",
"EEEE": "%A",
"''T''": "T",
}
class Tokenizer(tokens.Tokenizer):
IDENTIFIERS = ["`"]
STRING_ESCAPES = ["\\"]
ENCODE = "utf-8"
class Parser(parser.Parser):
STRICT_CAST = False
CONCAT_NULL_OUTPUTS_STRING = True
SUPPORTS_USER_DEFINED_TYPES = False
FUNCTIONS = {
**parser.Parser.FUNCTIONS,
"DATE_FORMAT": format_time_lambda(exp.TimeToStr, "drill"),
"TO_TIMESTAMP": exp.TimeStrToTime.from_arg_list,
"TO_CHAR": format_time_lambda(exp.TimeToStr, "drill"),
}
LOG_DEFAULTS_TO_LN = True
class Generator(generator.Generator):
JOIN_HINTS = False
TABLE_HINTS = False
QUERY_HINTS = False
NVL2_SUPPORTED = False
TYPE_MAPPING = {
**generator.Generator.TYPE_MAPPING,
exp.DataType.Type.INT: "INTEGER",
exp.DataType.Type.SMALLINT: "INTEGER",
exp.DataType.Type.TINYINT: "INTEGER",
exp.DataType.Type.BINARY: "VARBINARY",
exp.DataType.Type.TEXT: "VARCHAR",
exp.DataType.Type.NCHAR: "VARCHAR",
exp.DataType.Type.TIMESTAMPLTZ: "TIMESTAMP",
exp.DataType.Type.TIMESTAMPTZ: "TIMESTAMP",
exp.DataType.Type.DATETIME: "TIMESTAMP",
}
PROPERTIES_LOCATION = {
**generator.Generator.PROPERTIES_LOCATION,
exp.PartitionedByProperty: exp.Properties.Location.POST_SCHEMA,
exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
}
TRANSFORMS = {
**generator.Generator.TRANSFORMS,
exp.CurrentTimestamp: lambda *_: "CURRENT_TIMESTAMP",
exp.ArrayContains: rename_func("REPEATED_CONTAINS"),
exp.ArraySize: rename_func("REPEATED_COUNT"),
exp.Create: create_with_partitions_sql,
exp.DateAdd: _date_add_sql("ADD"),
exp.DateStrToDate: datestrtodate_sql,
exp.DateSub: _date_add_sql("SUB"),
exp.DateToDi: lambda self, e: f"CAST(TO_DATE({self.sql(e, 'this')}, {Drill.DATEINT_FORMAT}) AS INT)",
exp.DiToDate: lambda self, e: f"TO_DATE(CAST({self.sql(e, 'this')} AS VARCHAR), {Drill.DATEINT_FORMAT})",
exp.If: lambda self, e: f"`IF`({self.format_args(e.this, e.args.get('true'), e.args.get('false'))})",
exp.ILike: lambda self, e: f" {self.sql(e, 'this')} `ILIKE` {self.sql(e, 'expression')}",
exp.Levenshtein: rename_func("LEVENSHTEIN_DISTANCE"),
exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'this')}",
exp.RegexpLike: rename_func("REGEXP_MATCHES"),
exp.StrPosition: str_position_sql,
exp.StrToDate: _str_to_date,
exp.Pow: rename_func("POW"),
exp.Select: transforms.preprocess([transforms.eliminate_distinct_on]),
exp.StrToTime: lambda self, e: f"TO_TIMESTAMP({self.sql(e, 'this')}, {self.format_time(e)})",
exp.TimeStrToDate: lambda self, e: f"CAST({self.sql(e, 'this')} AS DATE)",
exp.TimeStrToTime: timestrtotime_sql,
exp.TimeStrToUnix: rename_func("UNIX_TIMESTAMP"),
exp.TimeToStr: lambda self, e: f"TO_CHAR({self.sql(e, 'this')}, {self.format_time(e)})",
exp.TimeToUnix: rename_func("UNIX_TIMESTAMP"),
exp.ToChar: lambda self, e: self.function_fallback_sql(e),
exp.TryCast: no_trycast_sql,
exp.TsOrDsAdd: lambda self, e: f"DATE_ADD(CAST({self.sql(e, 'this')} AS DATE), {self.sql(exp.Interval(this=e.expression.copy(), unit=exp.var('DAY')))})",
exp.TsOrDsToDate: ts_or_ds_to_date_sql("drill"),
exp.TsOrDiToDi: lambda self, e: f"CAST(SUBSTR(REPLACE(CAST({self.sql(e, 'this')} AS VARCHAR), '-', ''), 1, 8) AS INT)",
}
def normalize_func(self, name: str) -> str:
return name if exp.SAFE_IDENTIFIER_RE.match(name) else f"`{name}`"
|