test_gms_preprocessing.py 19.9 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
48
49
50
51
52
53
54
# Imports regarding the 'gms_preprocessing' module.
from gms_preprocessing import process_controller, __file__
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
55

56
__author__ = 'Daniel Scheffler'  # edited by Jessica Palka.
57

58
# Rootpath of the gms_preprocessing-repository.
59
60
61
62
63
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.
64
job_config_kwargs = dict(parallelization_level='scenes', db_host='geoms', delete_old_output=True, is_test=True)
65
66


Daniel Scheffler's avatar
Daniel Scheffler committed
67
##########################
68
# Test case: BaseTestCases
Daniel Scheffler's avatar
Daniel Scheffler committed
69
70
##########################

71
72
73
74

class BaseTestCases:
    """
    General testclass. The tests defined in this testclass test the processing steps Level-1A, Level-1B, Level-1C,
75
    Level-2A, Level-2B and Level-2C defined in the "gms_preprocessing"-repository.
76
77
78
79
    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):
80
        PC = None  # default
81
82
83

        @classmethod
        def tearDownClass(cls):
84
            cls.PC.config.DB_job_record.delete_procdata_of_entire_job(force=True)
85
86
87

        @classmethod
        def validate_db_entry(cls, filename):
88
89
            sceneID_res = get_info_from_postgreSQLdb(cls.PC.config.conn_database, 'scenes', ['id'],
                                                     {'filename': filename})
90
91
92
93
            assert sceneID_res and isinstance(sceneID_res[0][0], int), 'Invalid database entry.'

        @classmethod
        def create_job(cls, jobID, config):
94
            cls.PC = process_controller(jobID, **config)
95
96

            cls.PC.logger.info('Execution of entire GeoMultiSens pre-processing chain started for job ID %s...'
97
                               % cls.PC.config.ID)
98
99

            # update attributes of DB_job_record and related DB entry
100
            cls.PC.config.DB_job_record.reset_job_progress()
101

102
            [cls.PC.add_local_availability(ds) for ds in cls.PC.config.data_list]
103

104
            [cls.validate_db_entry(ds['filename']) for ds in cls.PC.config.data_list]
105
106
107
108

        def test_L1A_processing(self):
            self.L1A_newObjects = self.PC.L1A_processing()
            self.assertIsInstance(self.L1A_newObjects, list)
109
            self.assertNotEqual(len(self.L1A_newObjects), 0, msg='L1A_processing did not output an L1A object.')
110
111
112
113
114
            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)
115
            self.assertNotEqual(len(self.L1B_newObjects), 0, msg='L1B_processing did not output an L1B object.')
116
117
118
119
120
            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)
121
            self.assertNotEqual(len(self.L1C_newObjects), 0, msg='L1C_processing did not output an L1C object.')
122
123
124
125
126
            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)
127
            self.assertNotEqual(len(self.L2A_newObjects), 0, msg='L2A_processing did not output an L2A object.')
128
129
130
131
132
            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)
133
            self.assertNotEqual(len(self.L2B_newObjects), 0, msg='L2B_processing did not output an L2B object.')
134
135
136
137
138
            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)
139
            self.assertNotEqual(len(self.L2C_newObjects), 0, msg='L2C_processing did not output an L2C object.')
140
            self.assertIsInstance(self.L2C_newObjects[0], L2C_object)
Daniel Scheffler's avatar
Daniel Scheffler committed
141
            # Setting the config.status manually.
142
            # if self.L2C_newObjects:
143
            #     self.PC.config.status = "finished"
144
145
            # FIXME after updating the job.status-attribute for the level-processes, delete the code that is commented
            # FIXME out.
146
147
148
149
150
151
152
153
154
155
156


###################################################################################
# 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).
157
    More information on the dataset will be output after the tests-classes are executed.
158
159
160
161
162
163
164
165
166
167
168
169
170
171
    """
    @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
172

173
174
175
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).
176
    More information on the dataset will be output after after the tests-classes are executed.
177
178
179
180
181
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186262, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
182

183
184
185
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).
186
    More information on the dataset will be output after the tests-classes are executed.
187
188
189
190
191
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186267, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
192

193
194
195
196
197
198
199
200
201
# 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!

202
#
203
204
205
class Test_Landsat8_PreCollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-8 OLI_TIRS scene (pre-collection data).
206
    More information on the dataset will be output after the tests-classes are executed.
207
208
209
210
211
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186196, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
212

213
214
215
class Test_Landsat8_CollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-8 OLI_TIRS scene (collection data).
216
    More information on the dataset will be output after the tests-classes are executed.
217
218
219
220
221
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186261, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
222

223
class Test_Sentinel2A_SingleGranuleFormat(BaseTestCases.TestAll):
224
    """
225
226
    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.
227
228
229
230
231
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186268, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
232

233
class Test_Sentinel2A_MultiGranuleFormat(BaseTestCases.TestAll):
234
    """
235
236
    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.
237
238
239
240
241
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186272, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
242

243
244
245
246
247
248
249
250
251
252
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)


253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
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)


###################################################################################
# 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.
271
272
273
274
275
276
277
278
279
280
281
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'])
     }
282
283
284
285
286
287
288
289
290
291
292
293

# 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)]

294
295
296
297

@unittest.SkipTest
class Test_in_normal_mode(unittest.TestCase):
    def setUp(self):
298
        # self.job_id = 26186740  # Testjob Landsat-8
299
300
        # self.job_id = 26186906  # Bug Input Validator
        self.job_id = 26186925  # 1 Sentinel-2A, Bug NoneType' object has no attribute 'find'
301
302

        self.PC = process_controller(self.job_id, **dict(is_test=False, parallelization_level='scenes', db_host='geoms',
303
                                                         delete_old_output=True, disable_exception_handler=True))
304
305
306
307
308

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


309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
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)

325
        # FIXME: If the job.status-issue is fixed, the commented out section can be nullified.
326
        # jobstatus.append(eval("Test_"+items).PC.status)
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342

    # 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.
343
344
    # 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.
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
    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.
369
    logger.info("\ntest_gms_preprocessing.py"
370
371
372
373
374
375
                "\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(),
376
377
                                            orient='index',
                                            columns=['Country', 'Characteristic', 'Clouds', 'Overlap_area']))
378
379
380
381
382
383
384
385
386
387
    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]
388
    logger.info(pandas.DataFrame(summary_testResults, columns=results, index=testdata_index))
389
390
391
392
393

    # 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 \
394
       list(itertools.chain(*summary_skipped)):
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
        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")
410

411
412
413
414
415
            logger.info("Test_" + test + ", SKIPS:")
            if summary_skipped[index]:
                logger.info(summary_skipped[index][0][1])
            else:
                logger.info("None.")
416

417
418
            if not index == (len(testdata) - 1):
                logger.info("\n---------------------------------------------------------------------------------------")
419

420
        logger.info("\n***************************************************************************************")
421

422
423
    else:
        pass
424

425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
    # 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
440

441
    logger.info("END.")  # OUTPUT, END.
442

443
444
445
    # 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 = []