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
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
|
from sqlglot import exp
from sqlglot.dialects.dialect import (
Dialect,
format_time_lambda,
inline_array_sql,
rename_func,
)
from sqlglot.expressions import Literal
from sqlglot.generator import Generator
from sqlglot.helper import list_get
from sqlglot.parser import Parser
from sqlglot.tokens import Tokenizer, TokenType
def _check_int(s):
if s[0] in ("-", "+"):
return s[1:].isdigit()
return s.isdigit()
# from https://docs.snowflake.com/en/sql-reference/functions/to_timestamp.html
def _snowflake_to_timestamp(args):
if len(args) == 2:
first_arg, second_arg = args
if second_arg.is_string:
# case: <string_expr> [ , <format> ]
return format_time_lambda(exp.StrToTime, "snowflake")(args)
# case: <numeric_expr> [ , <scale> ]
if second_arg.name not in ["0", "3", "9"]:
raise ValueError(f"Scale for snowflake numeric timestamp is {second_arg}, but should be 0, 3, or 9")
if second_arg.name == "0":
timescale = exp.UnixToTime.SECONDS
elif second_arg.name == "3":
timescale = exp.UnixToTime.MILLIS
elif second_arg.name == "9":
timescale = exp.UnixToTime.MICROS
return exp.UnixToTime(this=first_arg, scale=timescale)
first_arg = list_get(args, 0)
if not isinstance(first_arg, Literal):
# case: <variant_expr>
return format_time_lambda(exp.StrToTime, "snowflake", default=True)(args)
if first_arg.is_string:
if _check_int(first_arg.this):
# case: <integer>
return exp.UnixToTime.from_arg_list(args)
# case: <date_expr>
return format_time_lambda(exp.StrToTime, "snowflake", default=True)(args)
# case: <numeric_expr>
return exp.UnixToTime.from_arg_list(args)
def _unix_to_time(self, expression):
scale = expression.args.get("scale")
timestamp = self.sql(expression, "this")
if scale in [None, exp.UnixToTime.SECONDS]:
return f"TO_TIMESTAMP({timestamp})"
if scale == exp.UnixToTime.MILLIS:
return f"TO_TIMESTAMP({timestamp}, 3)"
if scale == exp.UnixToTime.MICROS:
return f"TO_TIMESTAMP({timestamp}, 9)"
raise ValueError("Improper scale for timestamp")
class Snowflake(Dialect):
null_ordering = "nulls_are_large"
time_format = "'yyyy-mm-dd hh24:mi:ss'"
time_mapping = {
"YYYY": "%Y",
"yyyy": "%Y",
"YY": "%y",
"yy": "%y",
"MMMM": "%B",
"mmmm": "%B",
"MON": "%b",
"mon": "%b",
"MM": "%m",
"mm": "%m",
"DD": "%d",
"dd": "%d",
"d": "%-d",
"DY": "%w",
"dy": "%w",
"HH24": "%H",
"hh24": "%H",
"HH12": "%I",
"hh12": "%I",
"MI": "%M",
"mi": "%M",
"SS": "%S",
"ss": "%S",
"FF": "%f",
"ff": "%f",
"FF6": "%f",
"ff6": "%f",
}
class Parser(Parser):
FUNCTIONS = {
**Parser.FUNCTIONS,
"ARRAYAGG": exp.ArrayAgg.from_arg_list,
"IFF": exp.If.from_arg_list,
"TO_TIMESTAMP": _snowflake_to_timestamp,
"ARRAY_CONSTRUCT": exp.Array.from_arg_list,
"RLIKE": exp.RegexpLike.from_arg_list,
}
FUNCTION_PARSERS = {
**Parser.FUNCTION_PARSERS,
"DATE_PART": lambda self: self._parse_extract(),
}
FUNC_TOKENS = {
*Parser.FUNC_TOKENS,
TokenType.RLIKE,
}
COLUMN_OPERATORS = {
**Parser.COLUMN_OPERATORS,
TokenType.COLON: lambda self, this, path: self.expression(
exp.Bracket,
this=this,
expressions=[path],
),
}
PROPERTY_PARSERS = {
**Parser.PROPERTY_PARSERS,
TokenType.PARTITION_BY: lambda self: self._parse_partitioned_by(),
}
class Tokenizer(Tokenizer):
QUOTES = ["'", "$$"]
ESCAPE = "\\"
SINGLE_TOKENS = {
**Tokenizer.SINGLE_TOKENS,
"$": TokenType.DOLLAR, # needed to break for quotes
}
KEYWORDS = {
**Tokenizer.KEYWORDS,
"QUALIFY": TokenType.QUALIFY,
"DOUBLE PRECISION": TokenType.DOUBLE,
"TIMESTAMP_LTZ": TokenType.TIMESTAMPLTZ,
"TIMESTAMP_NTZ": TokenType.TIMESTAMP,
"TIMESTAMP_TZ": TokenType.TIMESTAMPTZ,
"TIMESTAMPNTZ": TokenType.TIMESTAMP,
"SAMPLE": TokenType.TABLE_SAMPLE,
}
class Generator(Generator):
TRANSFORMS = {
**Generator.TRANSFORMS,
exp.If: rename_func("IFF"),
exp.StrToTime: lambda self, e: f"TO_TIMESTAMP({self.sql(e, 'this')}, {self.format_time(e)})",
exp.UnixToTime: _unix_to_time,
exp.Array: inline_array_sql,
exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'value')}",
}
TYPE_MAPPING = {
**Generator.TYPE_MAPPING,
exp.DataType.Type.TIMESTAMP: "TIMESTAMPNTZ",
}
ROOT_PROPERTIES = {
exp.PartitionedByProperty,
exp.ReturnsProperty,
exp.LanguageProperty,
exp.SchemaCommentProperty,
}
def except_op(self, expression):
if not expression.args.get("distinct", False):
self.unsupported("EXCEPT with All is not supported in Snowflake")
return super().except_op(expression)
def intersect_op(self, expression):
if not expression.args.get("distinct", False):
self.unsupported("INTERSECT with All is not supported in Snowflake")
return super().intersect_op(expression)
|
