Skip to content

Classes

spark_expectations.core.expectations.SparkExpectations dataclass

This class implements/supports running the data quality rules on a dataframe returned by a function

Parameters:

Name Type Description Default
product_id str

Name of the product

required
rules_df DataFrame

DataFrame which contains the rules. User is responsible for reading the rules_table in which ever system it is

required
stats_table str

Name of the table where the stats/audit-info need to be written

required
debugger bool

Mark it as "True" if the debugger mode need to be enabled, by default is False

False
stats_streaming_options Optional[Dict[str, Union[str, bool]]]

Provide options to override the defaults, while writing into the stats streaming table

None

Attributes

debugger: bool = False class-attribute instance-attribute

product_id: str instance-attribute

rules_df: DataFrame instance-attribute

stats_streaming_options: Optional[Dict[str, Union[str, bool]]] = None class-attribute instance-attribute

stats_table: str instance-attribute

stats_table_writer: WrappedDataFrameWriter instance-attribute

target_and_error_table_writer: WrappedDataFrameWriter instance-attribute

Functions

with_expectations(target_table: str, write_to_table: bool = False, write_to_temp_table: bool = False, user_conf: Optional[Dict[str, Union[str, int, bool]]] = None, target_table_view: Optional[str] = None, target_and_error_table_writer: Optional[WrappedDataFrameWriter] = None) -> Any

This decorator helps to wrap a function which returns dataframe and apply dataframe rules on it

Parameters:

Name Type Description Default
target_table str

Name of the table where the final dataframe need to be written

required
write_to_table bool

Mark it as "True" if the dataframe need to be written as table

False
write_to_temp_table bool

Mark it as "True" if the input dataframe need to be written to the temp table to break the spark plan

False
user_conf Optional[Dict[str, Union[str, int, bool]]]

Provide options to override the defaults, while writing into the stats streaming table

None
target_table_view Optional[str]

This view is created after the _row_dq process to run the target agg_dq and query_dq. If value is not provided, defaulted to {target_table}_view

None
target_and_error_table_writer Optional[WrappedDataFrameWriter]

Provide the writer to write the target and error table, this will take precedence over the class level writer

None

Returns:

Name Type Description
Any Any

Returns a function which applied the expectations on dataset

Source code in spark_expectations/core/expectations.py
 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
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
def with_expectations(
    self,
    target_table: str,
    write_to_table: bool = False,
    write_to_temp_table: bool = False,
    user_conf: Optional[Dict[str, Union[str, int, bool]]] = None,
    target_table_view: Optional[str] = None,
    target_and_error_table_writer: Optional["WrappedDataFrameWriter"] = None,
) -> Any:
    """
    This decorator helps to wrap a function which returns dataframe and apply dataframe rules on it

    Args:
        target_table: Name of the table where the final dataframe need to be written
        write_to_table: Mark it as "True" if the dataframe need to be written as table
        write_to_temp_table: Mark it as "True" if the input dataframe need to be written to the temp table to break
                            the spark plan
        user_conf: Provide options to override the defaults, while writing into the stats streaming table
        target_table_view: This view is created after the _row_dq process to run the target agg_dq and query_dq.
            If value is not provided, defaulted to {target_table}_view
        target_and_error_table_writer: Provide the writer to write the target and error table,
            this will take precedence over the class level writer

    Returns:
        Any: Returns a function which applied the expectations on dataset
    """

    def _except(func: Any) -> Any:
        # variable used for enabling notification at different level

        _default_notification_dict: Dict[
            str, Union[str, int, bool, Dict[str, str], None]
        ] = {
            user_config.se_notifications_on_start: False,
            user_config.se_notifications_on_completion: False,
            user_config.se_notifications_on_fail: True,
            user_config.se_notifications_on_error_drop_exceeds_threshold_breach: False,
            user_config.se_notifications_on_error_drop_threshold: 100,
            user_config.se_enable_agg_dq_detailed_result: False,
            user_config.se_enable_query_dq_detailed_result: False,
            user_config.se_job_metadata: None,
            user_config.querydq_output_custom_table_name: f"{self.stats_table}_querydq_output",
        }

        _notification_dict: Dict[
            str, Union[str, int, bool, Dict[str, str], None]
        ] = (
            {**_default_notification_dict, **user_conf}
            if user_conf
            else _default_notification_dict
        )
        _default_stats_streaming_dict: Dict[str, Union[bool, str]] = {
            user_config.se_enable_streaming: True,
            user_config.secret_type: "databricks",
            user_config.dbx_workspace_url: "https://workspace.cloud.databricks.com",
            user_config.dbx_secret_scope: "secret_scope",
            user_config.dbx_kafka_server_url: "se_streaming_server_url_secret_key",
            user_config.dbx_secret_token_url: "se_streaming_auth_secret_token_url_key",
            user_config.dbx_secret_app_name: "se_streaming_auth_secret_appid_key",
            user_config.dbx_secret_token: "se_streaming_auth_secret_token_key",
            user_config.dbx_topic_name: "se_streaming_topic_name",
        }
        _se_stats_streaming_dict: Dict[str, Any] = (
            {**self.stats_streaming_options}
            if self.stats_streaming_options
            else _default_stats_streaming_dict
        )

        enable_error_table = _notification_dict.get(
            user_config.se_enable_error_table, True
        )
        self._context.set_se_enable_error_table(
            enable_error_table if isinstance(enable_error_table, bool) else True
        )

        dq_rules_params = _notification_dict.get(user_config.se_dq_rules_params, {})
        self._context.set_dq_rules_params(
            dq_rules_params if isinstance(dq_rules_params, dict) else {}
        )

        # Overwrite the writers if provided by the user in the with_expectations explicitly
        if target_and_error_table_writer:
            self._context.set_target_and_error_table_writer_config(
                target_and_error_table_writer.build()
            )

        _agg_dq_detailed_stats: bool = (
            bool(_notification_dict[user_config.se_enable_agg_dq_detailed_result])
            if isinstance(
                _notification_dict[user_config.se_enable_agg_dq_detailed_result],
                bool,
            )
            else False
        )

        _query_dq_detailed_stats: bool = (
            bool(_notification_dict[user_config.se_enable_query_dq_detailed_result])
            if isinstance(
                _notification_dict[user_config.se_enable_query_dq_detailed_result],
                bool,
            )
            else False
        )

        if _agg_dq_detailed_stats or _query_dq_detailed_stats:
            if _agg_dq_detailed_stats:
                self._context.set_agg_dq_detailed_stats_status(
                    _agg_dq_detailed_stats
                )

            if _query_dq_detailed_stats:
                self._context.set_query_dq_detailed_stats_status(
                    _query_dq_detailed_stats
                )

            self._context.set_query_dq_output_custom_table_name(
                str(
                    _notification_dict[user_config.querydq_output_custom_table_name]
                )
            )

        # need to call the get_rules_frm_table function to get the rules from the table as expectations
        (
            dq_queries_dict,
            expectations,
            rules_execution_settings,
        ) = self.reader.get_rules_from_df(
            self.rules_df, target_table, params=self._context.get_dq_rules_params
        )

        _row_dq: bool = rules_execution_settings.get("row_dq", False)
        _source_agg_dq: bool = rules_execution_settings.get("source_agg_dq", False)
        _target_agg_dq: bool = rules_execution_settings.get("target_agg_dq", False)
        _source_query_dq: bool = rules_execution_settings.get(
            "source_query_dq", False
        )
        _target_query_dq: bool = rules_execution_settings.get(
            "target_query_dq", False
        )
        _target_table_view: str = (
            target_table_view
            if target_table_view
            else f"{target_table.split('.')[-1]}_view"
        )

        _notification_on_start: bool = (
            bool(_notification_dict[user_config.se_notifications_on_start])
            if isinstance(
                _notification_dict[user_config.se_notifications_on_start],
                bool,
            )
            else False
        )
        _notification_on_completion: bool = (
            bool(_notification_dict[user_config.se_notifications_on_completion])
            if isinstance(
                _notification_dict[user_config.se_notifications_on_completion],
                bool,
            )
            else False
        )
        _notification_on_fail: bool = (
            bool(_notification_dict[user_config.se_notifications_on_fail])
            if isinstance(
                _notification_dict[user_config.se_notifications_on_fail],
                bool,
            )
            else False
        )
        _notification_on_error_drop_exceeds_threshold_breach: bool = (
            bool(
                _notification_dict[
                    user_config.se_notifications_on_error_drop_exceeds_threshold_breach
                ]
            )
            if isinstance(
                _notification_dict[
                    user_config.se_notifications_on_error_drop_exceeds_threshold_breach
                ],
                bool,
            )
            else False
        )

        _job_metadata: str = user_config.se_job_metadata

        notifications_on_error_drop_threshold = _notification_dict.get(
            user_config.se_notifications_on_error_drop_threshold, 100
        )
        _error_drop_threshold: int = (
            notifications_on_error_drop_threshold
            if isinstance(notifications_on_error_drop_threshold, int)
            else 100
        )

        self.reader.set_notification_param(user_conf)
        self._context.set_notification_on_start(_notification_on_start)
        self._context.set_notification_on_completion(_notification_on_completion)
        self._context.set_notification_on_fail(_notification_on_fail)

        self._context.set_se_streaming_stats_dict(_se_stats_streaming_dict)
        self._context.set_dq_expectations(expectations)
        self._context.set_rules_execution_settings_config(rules_execution_settings)
        self._context.set_querydq_secondary_queries(dq_queries_dict)
        self._context.set_job_metadata(_job_metadata)

        @self._notification.send_notification_decorator
        @self._statistics_decorator.collect_stats_decorator
        @functools.wraps(func)
        def wrapper(*args: tuple, **kwargs: dict) -> DataFrame:
            try:
                _log.info("The function dataframe is getting created")
                # _df: DataFrame = func(*args, **kwargs)
                _df: DataFrame = func(*args, **kwargs)
                table_name: str = self._context.get_table_name

                _input_count = _df.count()
                _log.info("data frame input record count: %s", _input_count)
                _output_count: int = 0
                _error_count: int = 0
                _source_dq_df: Optional[DataFrame] = None
                _source_query_dq_df: Optional[DataFrame] = None
                _row_dq_df: DataFrame = _df
                _final_dq_df: Optional[DataFrame] = None
                _final_query_dq_df: Optional[DataFrame] = None

                # initialize variable with default values through set
                self._context.set_dq_run_status()
                self._context.set_source_agg_dq_status()
                self._context.set_source_query_dq_status()
                self._context.set_row_dq_status()
                self._context.set_final_agg_dq_status()
                self._context.set_final_query_dq_status()
                self._context.set_input_count()
                self._context.set_error_count()
                self._context.set_output_count()
                self._context.set_source_agg_dq_result()
                self._context.set_final_agg_dq_result()
                self._context.set_source_query_dq_result()
                self._context.set_final_query_dq_result()
                self._context.set_summarized_row_dq_res()
                self._context.set_dq_expectations(expectations)

                # initialize variables of start and end time with default values
                self._context._source_agg_dq_start_time = None
                self._context._final_agg_dq_start_time = None
                self._context._source_query_dq_start_time = None
                self._context._final_query_dq_start_time = None
                self._context._row_dq_start_time = None

                self._context._source_agg_dq_end_time = None
                self._context._final_agg_dq_end_time = None
                self._context._source_query_dq_end_time = None
                self._context._final_query_dq_end_time = None
                self._context._row_dq_end_time = None

                self._context.set_input_count(_input_count)
                self._context.set_error_drop_threshold(_error_drop_threshold)

                _log.info(
                    "Spark Expectations run id for this run: %s",
                    self._context.get_run_id,
                )

                if isinstance(_df, DataFrame):
                    _log.info("The function dataframe is created")
                    self._context.set_table_name(table_name)
                    if write_to_temp_table:
                        _log.info("Dropping to temp table started")
                        self.spark.sql(f"drop table if exists {table_name}_temp")  # type: ignore
                        _log.info("Dropping to temp table completed")
                        _log.info("Writing to temp table started")
                        source_columns = _df.columns
                        self._writer.save_df_as_table(
                            _df,
                            f"{table_name}_temp",
                            self._context.get_target_and_error_table_writer_config,
                        )
                        _log.info("Read from temp table started")
                        _df = self.spark.sql(f"select * from {table_name}_temp")  # type: ignore
                        _df = _df.select(source_columns)
                        _log.info("Read from temp table completed")

                    func_process = self._process.execute_dq_process(
                        _context=self._context,
                        _actions=self.actions,
                        _writer=self._writer,
                        _notification=self._notification,
                        expectations=expectations,
                        table_name=table_name,
                        _input_count=_input_count,
                    )

                    if _source_agg_dq is True:
                        _log.info(
                            "started processing data quality rules for agg level expectations on soure dataframe"
                        )
                        self._context.set_source_agg_dq_status("Failed")
                        self._context.set_source_agg_dq_start_time()
                        # In this steps source agg data quality expectations runs on raw_data
                        # returns:
                        #        _source_dq_df: applied data quality dataframe,
                        #        _dq_source_agg_results: source aggregation result in dictionary
                        #        _: place holder for error data at row level
                        #        status: status of the execution

                        (
                            _source_dq_df,
                            _dq_source_agg_results,
                            _,
                            status,
                        ) = func_process(
                            _df,
                            self._context.get_agg_dq_rule_type_name,
                            source_agg_dq_flag=True,
                        )
                        self._context.set_source_agg_dq_status(status)
                        self._context.set_source_agg_dq_end_time()

                        _log.info(
                            "ended processing data quality rules for agg level expectations on source dataframe"
                        )

                    if _source_query_dq is True:
                        _log.info(
                            "started processing data quality rules for query level expectations on soure dataframe"
                        )
                        self._context.set_source_query_dq_status("Failed")
                        self._context.set_source_query_dq_start_time()
                        # In this steps source query data quality expectations runs on raw_data
                        # returns:
                        #        _source_query_dq_df: applied data quality dataframe,
                        #        _dq_source_query_results: source query dq results in dictionary
                        #        _: place holder for error data at row level
                        #        status: status of the execution

                        (
                            _source_query_dq_df,
                            _dq_source_query_results,
                            _,
                            status,
                        ) = func_process(
                            _df,
                            self._context.get_query_dq_rule_type_name,
                            source_query_dq_flag=True,
                        )
                        self._context.set_source_query_dq_status(status)
                        self._context.set_source_query_dq_end_time()
                        _log.info(
                            "ended processing data quality rules for query level expectations on source dataframe"
                        )

                    if _row_dq is True:
                        _log.info(
                            "started processing data quality rules for row level expectations"
                        )
                        self._context.set_row_dq_status("Failed")
                        self._context.set_row_dq_start_time()
                        # In this steps row level data quality expectations runs on raw_data
                        # returns:
                        #        _row_dq_df: applied data quality dataframe at row level on raw dataframe,
                        #        _: place holder for aggregation
                        #        _error_count: number of error records
                        #        status: status of the execution
                        (_row_dq_df, _, _error_count, status) = func_process(
                            _df,
                            self._context.get_row_dq_rule_type_name,
                            row_dq_flag=True,
                        )
                        self._context.set_error_count(_error_count)

                        _row_dq_df.createOrReplaceTempView(_target_table_view)

                        _output_count = _row_dq_df.count() if _row_dq_df else 0
                        self._context.set_output_count(_output_count)

                        self._context.set_row_dq_status(status)
                        self._context.set_row_dq_end_time()

                        if (
                            _notification_on_error_drop_exceeds_threshold_breach
                            is True
                            and (100 - self._context.get_output_percentage)
                            >= _error_drop_threshold
                        ):
                            self._notification.notify_on_exceeds_of_error_threshold()
                            # raise SparkExpectationsErrorThresholdExceedsException(
                            #     "An error has taken place because"
                            #     " the set limit for acceptable"
                            #     " errors, known as the error"
                            #     " threshold, has been surpassed"
                            # )
                        _log.info(
                            "ended processing data quality rules for row level expectations"
                        )

                    if _row_dq is True and _target_agg_dq is True:
                        _log.info(
                            "started processing data quality rules for agg level expectations on final dataframe"
                        )
                        self._context.set_final_agg_dq_status("Failed")
                        self._context.set_final_agg_dq_start_time()
                        # In this steps final agg data quality expectations run on final dataframe
                        # returns:
                        #        _final_dq_df: applied data quality dataframe at row level on raw dataframe,
                        #        _dq_final_agg_results: final agg dq result in dictionary
                        #        _: number of error records
                        #        status: status of the execution

                        (
                            _final_dq_df,
                            _dq_final_agg_results,
                            _,
                            status,
                        ) = func_process(
                            _row_dq_df,
                            self._context.get_agg_dq_rule_type_name,
                            final_agg_dq_flag=True,
                            error_count=_error_count,
                            output_count=_output_count,
                        )
                        self._context.set_final_agg_dq_status(status)
                        self._context.set_final_agg_dq_end_time()
                        _log.info(
                            "ended processing data quality rules for agg level expectations on final dataframe"
                        )

                    if _row_dq is True and _target_query_dq is True:
                        _log.info(
                            "started processing data quality rules for query level expectations on final dataframe"
                        )
                        self._context.set_final_query_dq_status("Failed")
                        self._context.set_final_query_dq_start_time()
                        # In this steps final query dq data quality expectations run on final dataframe
                        # returns:
                        #        _final_query_dq_df: applied data quality dataframe at row level on raw dataframe,
                        #        _dq_final_query_results: final query dq result in dictionary
                        #        _: number of error records
                        #        status: status of the execution

                        _row_dq_df.createOrReplaceTempView(_target_table_view)

                        (
                            _final_query_dq_df,
                            _dq_final_query_results,
                            _,
                            status,
                        ) = func_process(
                            _row_dq_df,
                            self._context.get_query_dq_rule_type_name,
                            final_query_dq_flag=True,
                            error_count=_error_count,
                            output_count=_output_count,
                        )
                        self._context.set_final_query_dq_status(status)
                        self._context.set_final_query_dq_end_time()

                        _log.info(
                            "ended processing data quality rules for query level expectations on final dataframe"
                        )

                    # TODO if row_dq is False and source_agg/source_query is True then we need to write the
                    #  dataframe into the target table
                    if write_to_table:
                        _log.info("Writing into the final table started")
                        self._writer.save_df_as_table(
                            _row_dq_df,
                            f"{table_name}",
                            self._context.get_target_and_error_table_writer_config,
                        )
                        _log.info("Writing into the final table ended")

                else:
                    raise SparkExpectationsDataframeNotReturnedException(
                        "error occurred while processing spark "
                        "expectations due to given dataframe is not type of dataframe"
                    )
                # self.spark.catalog.clearCache()

                return _row_dq_df

            except Exception as e:
                raise SparkExpectationsMiscException(
                    f"error occurred while processing spark expectations {e}"
                )

        return wrapper

    return _except