-
Notifications
You must be signed in to change notification settings - Fork 1.6k
/
generic_test_builders.py
389 lines (338 loc) · 13.8 KB
/
generic_test_builders.py
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
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
import re
from copy import deepcopy
from typing import Any, Dict, Generic, List, Optional, Tuple
from dbt.artifacts.resources import NodeVersion
from dbt.clients.jinja import GENERIC_TEST_KWARGS_NAME, get_rendered
from dbt.contracts.graph.nodes import UnpatchedSourceDefinition
from dbt.contracts.graph.unparsed import UnparsedModelUpdate, UnparsedNodeUpdate
from dbt.exceptions import (
CustomMacroPopulatingConfigValueError,
SameKeyNestedError,
TagNotStringError,
TagsNotListOfStringsError,
TestArgIncludesModelError,
TestArgsNotDictError,
TestDefinitionDictLengthError,
TestNameNotStringError,
TestTypeError,
UnexpectedTestNamePatternError,
)
from dbt.parser.common import Testable
from dbt.utils import md5
from dbt_common.exceptions.macros import UndefinedMacroError
def synthesize_generic_test_names(
test_type: str, test_name: str, args: Dict[str, Any]
) -> Tuple[str, str]:
# Using the type, name, and arguments to this generic test, synthesize a (hopefully) unique name
# Will not be unique if multiple tests have same name + arguments, and only configs differ
# Returns a shorter version (hashed/truncated, for the compiled file)
# as well as the full name (for the unique_id + FQN)
flat_args = []
for arg_name in sorted(args):
# the model is already embedded in the name, so skip it
if arg_name == "model":
continue
arg_val = args[arg_name]
if isinstance(arg_val, dict):
parts = list(arg_val.values())
elif isinstance(arg_val, (list, tuple)):
parts = list(arg_val)
else:
parts = [arg_val]
flat_args.extend([str(part) for part in parts])
clean_flat_args = [re.sub("[^0-9a-zA-Z_]+", "_", arg) for arg in flat_args]
unique = "__".join(clean_flat_args)
# for the file path + alias, the name must be <64 characters
# if the full name is too long, include the first 30 identifying chars plus
# a 32-character hash of the full contents
test_identifier = "{}_{}".format(test_type, test_name)
full_name = "{}_{}".format(test_identifier, unique)
if len(full_name) >= 64:
test_trunc_identifier = test_identifier[:30]
label = md5(full_name)
short_name = "{}_{}".format(test_trunc_identifier, label)
else:
short_name = full_name
return short_name, full_name
class TestBuilder(Generic[Testable]):
"""An object to hold assorted test settings and perform basic parsing
Test names have the following pattern:
- the test name itself may be namespaced (package.test)
- or it may not be namespaced (test)
"""
# The 'test_name' is used to find the 'macro' that implements the test
TEST_NAME_PATTERN = re.compile(
r"((?P<test_namespace>([a-zA-Z_][0-9a-zA-Z_]*))\.)?"
r"(?P<test_name>([a-zA-Z_][0-9a-zA-Z_]*))"
)
# args in the test entry representing test configs
CONFIG_ARGS = (
"severity",
"tags",
"enabled",
"where",
"limit",
"warn_if",
"error_if",
"fail_calc",
"store_failures",
"store_failures_as",
"meta",
"database",
"schema",
"alias",
)
def __init__(
self,
data_test: Dict[str, Any],
target: Testable,
package_name: str,
render_ctx: Dict[str, Any],
column_name: Optional[str] = None,
version: Optional[NodeVersion] = None,
) -> None:
test_name, test_args = self.extract_test_args(data_test, column_name)
self.args: Dict[str, Any] = test_args
if "model" in self.args:
raise TestArgIncludesModelError()
self.package_name: str = package_name
self.target: Testable = target
self.version: Optional[NodeVersion] = version
self.render_ctx: Dict[str, Any] = render_ctx
self.column_name: Optional[str] = column_name
self.args["model"] = self.build_model_str()
match = self.TEST_NAME_PATTERN.match(test_name)
if match is None:
raise UnexpectedTestNamePatternError(test_name)
groups = match.groupdict()
self.name: str = groups["test_name"]
self.namespace: str = groups["test_namespace"]
self.config: Dict[str, Any] = {}
# Process legacy args
self.config.update(self._process_legacy_args())
# Process config args if present
if "config" in self.args:
self.config.update(self._render_values(self.args.pop("config", {})))
if self.namespace is not None:
self.package_name = self.namespace
# If the user has provided a description for this generic test, use it
# Then delete the "description" argument to:
# 1. Avoid passing it into the test macro
# 2. Avoid passing it into the test name synthesis
# Otherwise, use an empty string
self.description: str = ""
if "description" in self.args:
self.description = self.args["description"]
del self.args["description"]
# If the user has provided a custom name for this generic test, use it
# Then delete the "name" argument to avoid passing it into the test macro
# Otherwise, use an auto-generated name synthesized from test inputs
self.compiled_name: str = ""
self.fqn_name: str = ""
if "name" in self.args:
# Assign the user-defined name here, which will be checked for uniqueness later
# we will raise an error if two tests have same name for same model + column combo
self.compiled_name = self.args["name"]
self.fqn_name = self.args["name"]
del self.args["name"]
else:
short_name, full_name = self.get_synthetic_test_names()
self.compiled_name = short_name
self.fqn_name = full_name
# use hashed name as alias if full name is too long
if short_name != full_name and "alias" not in self.config:
self.config["alias"] = short_name
def _process_legacy_args(self):
config = {}
for key in self.CONFIG_ARGS:
value = self.args.pop(key, None)
if value and "config" in self.args and key in self.args["config"]:
raise SameKeyNestedError()
if not value and "config" in self.args:
value = self.args["config"].pop(key, None)
config[key] = value
return self._render_values(config)
def _render_values(self, config: Dict[str, Any]) -> Dict[str, Any]:
rendered_config = {}
for key, value in config.items():
if isinstance(value, str):
try:
value = get_rendered(value, self.render_ctx, native=True)
except UndefinedMacroError as e:
raise CustomMacroPopulatingConfigValueError(
target_name=self.target.name,
column_name=self.column_name,
name=self.name,
key=key,
err_msg=e.msg,
)
if value is not None:
rendered_config[key] = value
return rendered_config
def _bad_type(self) -> TypeError:
return TypeError('invalid target type "{}"'.format(type(self.target)))
@staticmethod
def extract_test_args(data_test, name=None) -> Tuple[str, Dict[str, Any]]:
if not isinstance(data_test, dict):
raise TestTypeError(data_test)
# If the test is a dictionary with top-level keys, the test name is "test_name"
# and the rest are arguments
# {'name': 'my_favorite_test', 'test_name': 'unique', 'config': {'where': '1=1'}}
if "test_name" in data_test.keys():
test_name = data_test.pop("test_name")
test_args = data_test
# If the test is a nested dictionary with one top-level key, the test name
# is the dict name, and nested keys are arguments
# {'unique': {'name': 'my_favorite_test', 'config': {'where': '1=1'}}}
else:
data_test = list(data_test.items())
if len(data_test) != 1:
raise TestDefinitionDictLengthError(data_test)
test_name, test_args = data_test[0]
if not isinstance(test_args, dict):
raise TestArgsNotDictError(test_args)
if not isinstance(test_name, str):
raise TestNameNotStringError(test_name)
test_args = deepcopy(test_args)
if name is not None:
test_args["column_name"] = name
return test_name, test_args
@property
def enabled(self) -> Optional[bool]:
return self.config.get("enabled")
@property
def alias(self) -> Optional[str]:
return self.config.get("alias")
@property
def severity(self) -> Optional[str]:
sev = self.config.get("severity")
if sev:
return sev.upper()
else:
return None
@property
def store_failures(self) -> Optional[bool]:
return self.config.get("store_failures")
@property
def store_failures_as(self) -> Optional[bool]:
return self.config.get("store_failures_as")
@property
def where(self) -> Optional[str]:
return self.config.get("where")
@property
def limit(self) -> Optional[int]:
return self.config.get("limit")
@property
def warn_if(self) -> Optional[str]:
return self.config.get("warn_if")
@property
def error_if(self) -> Optional[str]:
return self.config.get("error_if")
@property
def fail_calc(self) -> Optional[str]:
return self.config.get("fail_calc")
@property
def meta(self) -> Optional[dict]:
return self.config.get("meta")
@property
def database(self) -> Optional[str]:
return self.config.get("database")
@property
def schema(self) -> Optional[str]:
return self.config.get("schema")
def get_static_config(self):
config = {}
if self.alias is not None:
config["alias"] = self.alias
if self.severity is not None:
config["severity"] = self.severity
if self.enabled is not None:
config["enabled"] = self.enabled
if self.where is not None:
config["where"] = self.where
if self.limit is not None:
config["limit"] = self.limit
if self.warn_if is not None:
config["warn_if"] = self.warn_if
if self.error_if is not None:
config["error_if"] = self.error_if
if self.fail_calc is not None:
config["fail_calc"] = self.fail_calc
if self.store_failures is not None:
config["store_failures"] = self.store_failures
if self.store_failures_as is not None:
config["store_failures_as"] = self.store_failures_as
if self.meta is not None:
config["meta"] = self.meta
if self.database is not None:
config["database"] = self.database
if self.schema is not None:
config["schema"] = self.schema
return config
def tags(self) -> List[str]:
tags = self.config.get("tags", [])
if isinstance(tags, str):
tags = [tags]
if not isinstance(tags, list):
raise TagsNotListOfStringsError(tags)
for tag in tags:
if not isinstance(tag, str):
raise TagNotStringError(tag)
return tags[:]
def macro_name(self) -> str:
macro_name = "test_{}".format(self.name)
if self.namespace is not None:
macro_name = "{}.{}".format(self.namespace, macro_name)
return macro_name
def get_synthetic_test_names(self) -> Tuple[str, str]:
# Returns two names: shorter (for the compiled file), full (for the unique_id + FQN)
target_name = self.target.name
if isinstance(self.target, UnparsedModelUpdate):
name = self.name
if self.version:
target_name = f"{self.target.name}_v{self.version}"
elif isinstance(self.target, UnparsedNodeUpdate):
name = self.name
elif isinstance(self.target, UnpatchedSourceDefinition):
name = "source_" + self.name
else:
raise self._bad_type()
if self.namespace is not None:
name = "{}_{}".format(self.namespace, name)
return synthesize_generic_test_names(name, target_name, self.args)
def construct_config(self) -> str:
configs = ",".join(
[
f"{key}="
+ (
('"' + value.replace('"', '\\"') + '"')
if isinstance(value, str)
else str(value)
)
for key, value in self.config.items()
]
)
if configs:
return f"{{{{ config({configs}) }}}}"
else:
return ""
# this is the 'raw_code' that's used in 'render_update' and execution
# of the test macro
def build_raw_code(self) -> str:
return ("{{{{ {macro}(**{kwargs_name}) }}}}{config}").format(
macro=self.macro_name(),
config=self.construct_config(),
kwargs_name=GENERIC_TEST_KWARGS_NAME,
)
def build_model_str(self):
targ = self.target
if isinstance(self.target, UnparsedModelUpdate):
if self.version:
target_str = f"ref('{targ.name}', version='{self.version}')"
else:
target_str = f"ref('{targ.name}')"
elif isinstance(self.target, UnparsedNodeUpdate):
target_str = f"ref('{targ.name}')"
elif isinstance(self.target, UnpatchedSourceDefinition):
target_str = f"source('{targ.source.name}', '{targ.table.name}')"
return f"{{{{ get_where_subquery({target_str}) }}}}"