test_gms_preprocessing.py 22 KB
Newer Older
1
2
3
#!/usr/bin/env python
# -*- coding: utf-8 -*-

4
###################################################################################
5

6
"""
7
test_gms_preprocessing
8
----------------------------------
9

10
The testcases contained in this testscript, are parametrized testcases. They test
11
12
the level-processing steps defined in the 'gms_preprocessing' module in the
"gms_preprocessing"-project with the help of the test datasets:
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
- Landsat-5, Pre-Collection Data,
- Landsat-5, Collection Data,
- Landsat-7, SLC on, Pre-Collection Data,
- Landsat-7, SLC off, Pre-Collection Data,
- Landsat-7, SLC off, Collection Data,
- Landsat-8, Pre-Collection Data,
- Landsat-8, Collection Data,
- Sentinel-2A, Pre-Collection Data and
- Sentinel-2A, Collection Data.
The test datasets can be found in the directory "tests/data/archive_data/...". The
respective SRTM-datasets needed in the data-processing can be found in the directory
"tests/data/archive_data/Endeavor".

The tests, defined in a base-testcase (not executed), are triggered by creating
jobs (based on given job-IDs) in individual testcases that inherit the tests
from the base-testcase. The exception: The job-ID used in the last testclass
contains 3 different test datasets of the above listed datasets.

Note that the testresults are outputted in the console as well as a log-textfile
that can be found in the directory "tests/logs".

Program edited in July 2017.
"""

37
# Import python standard libraries.
38
39
40
41
42
43
import itertools
import logging
import os
import pandas
import sys
import time
44
45
import unittest

46
47
# Imports regarding the 'gms_preprocessing' module.
from gms_preprocessing import process_controller, __file__
48
from gms_preprocessing.model.gms_object import GMS_object
49
50
51
52
53
54
55
from gms_preprocessing.algorithms.L1A_P import L1A_object
from gms_preprocessing.algorithms.L1B_P import L1B_object
from gms_preprocessing.algorithms.L1C_P import L1C_object
from gms_preprocessing.algorithms.L2A_P import L2A_object
from gms_preprocessing.algorithms.L2B_P import L2B_object
from gms_preprocessing.algorithms.L2C_P import L2C_object
from gms_preprocessing.misc.database_tools import get_info_from_postgreSQLdb
56

57
58
from . import db_host

59
__author__ = 'Daniel Scheffler'  # edited by Jessica Palka.
60

61
# Rootpath of the gms_preprocessing-repository.
62
63
64
65
gmsRepo_rootpath = os.path.abspath(os.path.join(os.path.dirname(__file__), '..'))

# Defining the configurations needed to start a job containing the different dataset scenes.
# TODO Change the job-configurations for selected datasets.
66
job_config_kwargs = dict(parallelization_level='scenes', db_host=db_host, delete_old_output=True, is_test=True)
67

Daniel Scheffler's avatar
Daniel Scheffler committed
68
##########################
69
# Test case: BaseTestCases
Daniel Scheffler's avatar
Daniel Scheffler committed
70
71
##########################

72
73
74
75

class BaseTestCases:
    """
    General testclass. The tests defined in this testclass test the processing steps Level-1A, Level-1B, Level-1C,
76
    Level-2A, Level-2B and Level-2C defined in the "gms_preprocessing"-repository.
77
78
79
80
    Note that the tests in this testclass are not executed directly. They are re-used in the other classes defined
    in this test-script.
    """
    class TestAll(unittest.TestCase):
81
        PC = None  # default
82
83
84

        @classmethod
        def tearDownClass(cls):
85
            cls.PC.config.DB_job_record.delete_procdata_of_entire_job(force=True)
86
87
88

        @classmethod
        def validate_db_entry(cls, filename):
89
90
            sceneID_res = get_info_from_postgreSQLdb(cls.PC.config.conn_database, 'scenes', ['id'],
                                                     {'filename': filename})
91
92
93
94
            assert sceneID_res and isinstance(sceneID_res[0][0], int), 'Invalid database entry.'

        @classmethod
        def create_job(cls, jobID, config):
95
            cls.PC = process_controller(jobID, **config)
96
97

            cls.PC.logger.info('Execution of entire GeoMultiSens pre-processing chain started for job ID %s...'
98
                               % cls.PC.config.ID)
99
100

            # update attributes of DB_job_record and related DB entry
101
            cls.PC.config.DB_job_record.reset_job_progress()
102
            GMS_object.proc_status_all_GMSobjs.clear()  # reset
103

Daniel Scheffler's avatar
Bugfix.    
Daniel Scheffler committed
104
            cls.PC.config.data_list = cls.PC.add_local_availability(cls.PC.config.data_list)
105

106
            [cls.validate_db_entry(ds['filename']) for ds in cls.PC.config.data_list]
107
108
109
110

        def test_L1A_processing(self):
            self.L1A_newObjects = self.PC.L1A_processing()
            self.assertIsInstance(self.L1A_newObjects, list)
111
            self.assertNotEqual(len(self.L1A_newObjects), 0, msg='L1A_processing did not output an L1A object.')
112
113
114
115
116
            self.assertIsInstance(self.L1A_newObjects[0], L1A_object)

        def test_L1B_processing(self):
            self.L1B_newObjects = self.PC.L1B_processing()
            self.assertIsInstance(self.L1B_newObjects, list)
117
            self.assertNotEqual(len(self.L1B_newObjects), 0, msg='L1B_processing did not output an L1B object.')
118
119
120
121
122
            self.assertIsInstance(self.L1B_newObjects[0], L1B_object)

        def test_L1C_processing(self):
            self.L1C_newObjects = self.PC.L1C_processing()
            self.assertIsInstance(self.L1C_newObjects, list)
123
            self.assertNotEqual(len(self.L1C_newObjects), 0, msg='L1C_processing did not output an L1C object.')
124
125
126
127
128
            self.assertIsInstance(self.L1C_newObjects[0], L1C_object)

        def test_L2A_processing(self):
            self.L2A_newObjects = self.PC.L2A_processing()
            self.assertIsInstance(self.L2A_newObjects, list)
129
            self.assertNotEqual(len(self.L2A_newObjects), 0, msg='L2A_processing did not output an L2A object.')
130
131
132
133
134
            self.assertIsInstance(self.L2A_newObjects[0], L2A_object)

        def test_L2B_processing(self):
            self.L2B_newObjects = self.PC.L2B_processing()
            self.assertIsInstance(self.L2B_newObjects, list)
135
            self.assertNotEqual(len(self.L2B_newObjects), 0, msg='L2B_processing did not output an L2B object.')
136
137
138
139
140
            self.assertIsInstance(self.L2B_newObjects[0], L2B_object)

        def test_L2C_processing(self):
            self.L2C_newObjects = self.PC.L2C_processing()
            self.assertIsInstance(self.L2C_newObjects, list)
141
            self.assertNotEqual(len(self.L2C_newObjects), 0, msg='L2C_processing did not output an L2C object.')
142
            self.assertIsInstance(self.L2C_newObjects[0], L2C_object)
Daniel Scheffler's avatar
Daniel Scheffler committed
143
            # Setting the config.status manually.
144
            # if self.L2C_newObjects:
145
            #     self.PC.config.status = "finished"
146
147
            # FIXME after updating the job.status-attribute for the level-processes, delete the code that is commented
            # FIXME out.
148

Daniel Scheffler's avatar
Daniel Scheffler committed
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
    class TestCompletePipeline(unittest.TestCase):
        PC = None  # default

        @classmethod
        def tearDownClass(cls):
            cls.PC.config.DB_job_record.delete_procdata_of_entire_job(force=True)

        @classmethod
        def validate_db_entry(cls, filename):
            sceneID_res = get_info_from_postgreSQLdb(cls.PC.config.conn_database, 'scenes', ['id'],
                                                     {'filename': filename})
            assert sceneID_res and isinstance(sceneID_res[0][0], int), 'Invalid database entry.'

        @classmethod
        def create_job(cls, jobID, config):
            cls.PC = process_controller(jobID, **config)

            cls.PC.logger.info('Execution of entire GeoMultiSens pre-processing chain started for job ID %s...'
                               % cls.PC.config.ID)
168
            cls.PC.config.CPUs = 1  # FIXME
Daniel Scheffler's avatar
Daniel Scheffler committed
169
170
171

            [cls.validate_db_entry(ds['filename']) for ds in cls.PC.config.data_list]

Daniel Scheffler's avatar
Bugfix.    
Daniel Scheffler committed
172
173
        def test_run_all_processors(self):
            self.PC.run_all_processors()
Daniel Scheffler's avatar
Daniel Scheffler committed
174
            self.assertIsInstance(self.PC.L2C_newObjects, list)
175
176
177
178
179
180
181
182
183
184

###################################################################################
# Test cases 1-9: Test_<Satelite-Dataset>_<PreCollection or Collection>Data
# Test case 10: Test_MultipleDatasetsInOneJob


# TESTDATA-CLASSES.
class Test_Landsat5_PreCollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-5 TM scene (pre-collection data).
185
    More information on the dataset will be output after the tests-classes are executed.
186
187
188
189
190
191
192
193
194
195
196
197
198
199
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186263, job_config_kwargs)

# class Test_Landsat5_CollectionData(BaseTestCases.TestAll):
#     """
#     Parametrized testclass. Tests the level-processes on a Landsat-5 TM scene (collection data).
#     More information on the dataset will be outputted after the tests-classes are executed.
#     """
#     @classmethod
#     def setUpClass(cls):
#         cls.create_job(26186263, job_config_kwargs) # FIXME job_ID!

Daniel Scheffler's avatar
Daniel Scheffler committed
200

201
202
203
class Test_Landsat7_SLC_on_PreCollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-7 ETM+_SLC_ON scene (pre-collection data).
204
    More information on the dataset will be output after after the tests-classes are executed.
205
206
207
208
209
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186262, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
210

211
212
213
class Test_Landsat7_SLC_off_PreCollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-7 ETM+_SLC_OFF scene (pre-collection data).
214
    More information on the dataset will be output after the tests-classes are executed.
215
216
217
218
219
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186267, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
220

221
222
223
224
225
226
227
228
229
# class Test_Landsat7_SLC_off_CollectionData(BaseTestCases.TestAll):
#     """
#     Parametrized testclass. Tests the level-processes on a Landsat-7 ETM+_SLC_OFF scene (collection data).
#     More information on the dataset will be outputted after the tests-classes are executed.
#     """
#     @classmethod
#     def setUpClass(cls):
#         cls.create_job(26186267, job_config_kwargs) # FIXME job_ID!

230
#
231
232
233
class Test_Landsat8_PreCollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-8 OLI_TIRS scene (pre-collection data).
234
    More information on the dataset will be output after the tests-classes are executed.
235
236
237
238
239
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186196, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
240

241
242
243
class Test_Landsat8_CollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-8 OLI_TIRS scene (collection data).
244
    More information on the dataset will be output after the tests-classes are executed.
245
246
247
248
249
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186261, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
250

251
class Test_Sentinel2A_SingleGranuleFormat(BaseTestCases.TestAll):
252
    """
253
254
    Parametrized testclass. Tests the level-processes on a Sentinel-2A MSI scene (1 granule in archive: > 2017).
    More information on the dataset will be output after the tests-classes are executed.
255
256
257
258
259
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186268, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
260

Daniel Scheffler's avatar
Daniel Scheffler committed
261
262
263
264
265
266
267
268
269
270
271
class Test_Sentinel2A_SingleGranuleFormat_CompletePipeline(BaseTestCases.TestCompletePipeline):
    """
    Parametrized testclass. Tests the level-processes on a Sentinel-2A MSI scene (1 granule in archive: > 2017).
    More information on the dataset will be output after the tests-classes are executed.
    """
    @classmethod
    def setUpClass(cls):
        # job_config_kwargs['CPUs'] = 1
        cls.create_job(26186268, job_config_kwargs)


272
class Test_Sentinel2A_MultiGranuleFormat(BaseTestCases.TestAll):
273
    """
274
275
    Parametrized testclass. Tests the level-processes on a Sentinel-2A MSI scene (multiple granules in archive: < 2017).
    More information on the dataset will be output after the tests-classes are executed.
276
277
278
279
280
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186272, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
281

282
283
284
285
286
287
288
289
290
291
class Test_Sentinel2B_SingleGranuleFormat(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Sentinel-2B MSI scene (1 granule in archive: > 2017).
    More information on the dataset will be output after the tests-classes are executed.
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186937, job_config_kwargs)


292
293
294
295
296
297
298
299
300
301
class Test_MultipleDatasetsInOneJob(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a job containing a Landsat-5 (pre-collection data),
    Landsat-7 SLC_off (pre-collection data) and a Sentinel-2A (collection data) scene.
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186273, job_config_kwargs)


Daniel Scheffler's avatar
Daniel Scheffler committed
302
303
304
305
306
307
308
309
310
311
class Test_MultipleDatasetsInOneJob_CompletePipeline(BaseTestCases.TestCompletePipeline):
    """
    Parametrized testclass. Tests the level-processes on a job containing a Landsat-5 (pre-collection data),
    Landsat-7 SLC_off (pre-collection data) and a Sentinel-2A (collection data) scene.
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186273, job_config_kwargs)


312
313
314
315
316
317
318
319
###################################################################################
# Summarizing the information regarding the test datasets.

# The information: 'country' (3-letter country code, UN), 'characteristic features of the shown scene', 'cloud cover
# present' and 'overlap area present' of each dataset are summarized in the dictionary "testdata_scenes". The
# information are sorted according to the testdata.
# 3-letter code:
# UKR-Ukraine, KGZ-Kyrgyztan, POL-Poland, AUT-Austria, JPN-Japan, BOL-Bolivia, TUR-Turkey, DEU-Germany, CHE-Switzerland.
320
321
322
323
324
325
326
327
328
329
330
testdata_scenes = \
    {'Landsat5_PreCollectionData': list(['UKR', 'City region, forest', 'Sparsely', 'Zone 34/35']),
     # 'Landsat5_CollectionData': list(['KGZ', 'Snowy Mountains', 'Yes', 'None']),
     'Landsat7_SLC_on_PreCollectionData': list(['POL', 'City region, lakes', 'Yes', 'None']),
     'Landsat7_SLC_off_PreCollectionData': list(['AUT', 'Stripes (partly), Mountains', 'None', 'None']),
     # 'Landsat7_SLC_off_CollectionData': list(['JPN', 'Stripes (completly), Mountains', 'Yes', 'Zone 53/54']),
     'Landsat8_PreCollectionData': list(['BOL', 'Forest', 'Yes', 'None']),
     'Landsat8_CollectionData': list(['TUR', 'Snowy Mountains', 'Yes', 'None']),
     'Sentinel2A_PreCollectionData': list(['DEU', 'Potsdam', 'Sparsely', 'None']),
     'Sentinel2A_CollectionData': list(['CHE', 'City region, on the Rhine', 'Yes', 'None'])
     }
331
332
333
334
335
336
337
338
339
340
341
342

# The key of the dictionary is the key-value to parametrize the testclasses so that each testclass is executed
# automatically.
testdata = list(testdata_scenes.keys())
testdata.append('MultipleDatasetsInOneJob')


###################################################################################
# Parametrizing the test cases and creating a summary of the testresults.

summary_testResults, summary_errors, summary_failures, summary_skipped, jobstatus = [[] for _ in range(5)]

343
344
345
346

@unittest.SkipTest
class Test_in_normal_mode(unittest.TestCase):
    def setUp(self):
347
        # self.job_id = 26186740  # Testjob Landsat-8
348
349
        # self.job_id = 26186906  # Bug Input Validator
        self.job_id = 26186925  # 1 Sentinel-2A, Bug NoneType' object has no attribute 'find'
350

351
        self.PC = process_controller(self.job_id, **dict(is_test=False, parallelization_level='scenes', db_host=db_host,
352
                                                         delete_old_output=True, disable_exception_handler=True))
353
354
355
356
357

    def test(self):
        self.PC.run_all_processors()


358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
if __name__ == '__main__':
    # Part 1: Creating and running a testsuite for each dataset-testcase, and querying the job.status of the job.
    for items in testdata:
        suite = unittest.TestLoader().loadTestsFromTestCase(eval("Test_"+items))
        alltests = unittest.TestSuite(suite)

        # Part 2: Saving the results of each testsuite and the query for the job.status in individual variables.
        testResult = unittest.TextTestRunner(verbosity=2).run(alltests)

        summary_testResults.append([testResult.testsRun, testResult.wasSuccessful(),
                                    len(testResult.errors), len(testResult.failures),
                                    len(testResult.skipped)])
        summary_errors.append(testResult.errors)
        summary_failures.append(testResult.failures)
        summary_skipped.append(testResult.skipped)

374
        # FIXME: If the job.status-issue is fixed, the commented out section can be nullified.
375
        # jobstatus.append(eval("Test_"+items).PC.status)
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391

    # Part 3: Summarizing the testresults of each testsuite and outputting the results in an orderly fashion on the
    # console and in a textfile.
    # Note that the testresults are outputted as usual after each test is executed. Since the output of each
    # level-process is rather long, the output of the testresults become lost. Therefore, the purpose to output the
    # testresults again is simply to summarize the testresults in one place and to give an overview over the results.

    # Output: a) Information on the test datasets (table), b) testresults summarized in a table, c)if existing,
    # a list of errors, failures and skips in the testcases and d) the job.status that is not set to "finished".

    time.sleep(0.5)

    # Path of the textfile the results will be logged to.
    test_log_path = os.path.join(gmsRepo_rootpath, 'tests', 'data', 'logs', time.strftime('%Y%m%d_%H%M%S_log.txt'))

    # Creating a logging system for the testresults.
392
393
    # Source: The "GMS_logger"-function in the "gms_preprocessing" --> "misc" --> "logging.py"-script was used and
    # slightly altered to meet the needs of the current problem.
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
    logger = logging.getLogger("log_Test")
    logger.setLevel(logging.INFO)

    # Defining the format of the console and the file-output.
    formatter_fileH = logging.Formatter('')
    formatter_ConsoleH = logging.Formatter('')

    # Creating a handler for the file for the logging level "INFO".
    fileHandler = logging.FileHandler(test_log_path)
    fileHandler.setFormatter(formatter_fileH)
    fileHandler.setLevel(logging.INFO)

    # Creating a handler for the console for the logging level "INFO". "sys.stdout" is used for the logging output.
    consoleHandler_out = logging.StreamHandler(stream=sys.stdout)
    consoleHandler_out.setFormatter(formatter_ConsoleH)
    consoleHandler_out.set_name('console handler stdout')
    consoleHandler_out.setLevel(logging.INFO)

    # Adding the defined handlers to the instantiated logger.
    logger.addHandler(fileHandler)
    logger.addHandler(consoleHandler_out)

    # OUPUT, START.
    # Header of the file.
418
    logger.info("\ntest_gms_preprocessing.py"
419
420
421
422
423
424
                "\nREVIEW OF ALL TEST RESULTS, SUMMARY:"
                "\n***************************************************************************************"
                "\n--> SPECIFIC FEATURES OF DATA:")

    # Adding a table displaying the characteristic features of each dataset.
    logger.info(pandas.DataFrame.from_items(testdata_scenes.items(),
425
426
                                            orient='index',
                                            columns=['Country', 'Characteristic', 'Clouds', 'Overlap_area']))
427
428
429
430
431
432
433
434
435
436
    logger.info("\nThe jobID used in Test_" + testdata[-1] + " contains the datasets: "
                "\n-Landsat5_PreCollectionData,\n-Landsat7_SLC_off_PreCollectionData and "
                "\n-Sentinel2A_CollectionData.")

    # Adding a table displaying the testresults.
    logger.info("\n***************************************************************************************"
                "\n--> TESTRESULTS:")

    results = ["Run", "Success", "Errors", "Failures", "Skips"]
    testdata_index = ["Test_" + item for item in testdata]
437
    logger.info(pandas.DataFrame(summary_testResults, columns=results, index=testdata_index))
438
439
440
441
442

    # If errors, failures or skips (there is yet nothing to skip in the code) occurres, the respective message will
    # be printed.
    logger.info("\n***************************************************************************************")
    if list(itertools.chain(*summary_errors)) or list(itertools.chain(*summary_failures)) or \
443
       list(itertools.chain(*summary_skipped)):
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
        logger.info("--> ERRORS/FAILURES/SKIPS:")
        logger.info("\n---------------------------------------------------------------------------------------")

        for index, test in enumerate(testdata):
            logger.info("Test_" + test + ", ERRORS:")
            if summary_errors[index]:
                logger.info(summary_errors[index][0][1])
            else:
                logger.info("None. \n")

            logger.info("Test_" + test + ", FAILURES:")
            if summary_failures[index]:
                logger.info(summary_failures[index][0][1])
            else:
                logger.info("None. \n")
459

460
461
462
463
464
            logger.info("Test_" + test + ", SKIPS:")
            if summary_skipped[index]:
                logger.info(summary_skipped[index][0][1])
            else:
                logger.info("None.")
465

466
467
            if not index == (len(testdata) - 1):
                logger.info("\n---------------------------------------------------------------------------------------")
468

469
        logger.info("\n***************************************************************************************")
470

471
472
    else:
        pass
473

474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
    # Checking, if the job.status of each job is set to "finished". Is it not set to "finished", a dataframe is created
    # containing the test-name with and the different job.status itself.
    # FIXME: If the job.status-issue is fixed, the commented out section can be nullified.
    # jobstatus_table, index_table = [[] for _ in range(2)]
    # for index, test in enumerate(testdata):
    #     if jobstatus[index] != "finished":
    #         jobstatus_table.append(jobstatus[index])
    #         index_table.append("Test_" + test)
    #
    # if jobstatus_table:
    #     logger.info("--> WARNING!!! JOBSTATUS of the following testcase(s) is not set to 'finished': \n")
    #     logger.info(pandas.DataFrame(jobstatus_table, columns=["jobstatus"], index=index_table))
    #     logger.info("\n***************************************************************************************")
    # else:
    #     pass
489

490
    logger.info("END.")  # OUTPUT, END.
491

492
493
494
    # Delete the handlers added to the "log_Test"-logger to ensure that no message is outputted twice in a row, when
    # the logger is used again.
    logger.handlers = []