test_gms_preprocessing.py 18.7 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(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
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101

        @classmethod
        def tearDownClass(cls):
            cls.PC.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.job.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, parallelization_level='scenes', db_host='geoms', delete_old_output=True,
                                        job_config_kwargs=config)
            [cls.PC.add_local_availability(ds) for ds in cls.PC.usecase.data_list]

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

        def test_L1A_processing(self):
            self.L1A_newObjects = self.PC.L1A_processing()
            self.assertIsInstance(self.L1A_newObjects, list)
102
            self.assertNotEqual(len(self.L1A_newObjects), 0, msg='L1A_processing did not output an L1A object.')
103
104
105
106
107
            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)
108
            self.assertNotEqual(len(self.L1B_newObjects), 0, msg='L1B_processing did not output an L1B object.')
109
110
111
112
113
            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)
114
            self.assertNotEqual(len(self.L1C_newObjects), 0, msg='L1C_processing did not output an L1C object.')
115
116
117
118
119
            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)
120
            self.assertNotEqual(len(self.L2A_newObjects), 0, msg='L2A_processing did not output an L2A object.')
121
122
123
124
125
            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)
126
            self.assertNotEqual(len(self.L2B_newObjects), 0, msg='L2B_processing did not output an L2B object.')
127
128
129
130
131
            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)
132
            self.assertNotEqual(len(self.L2C_newObjects), 0, msg='L2C_processing did not output an L2C object.')
133
134
135
            self.assertIsInstance(self.L2C_newObjects[0], L2C_object)
            # Setting the job.status manually.
            # if self.L2C_newObjects:
136
137
138
            #     self.PC.job.status = "finished"
            # FIXME after updating the job.status-attribute for the level-processes, delete the code that is commented
            # FIXME out.
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


###################################################################################
# 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).
    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)

# 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
165

166
167
168
169
170
171
172
173
174
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).
    More information on the dataset will be outputted after after the tests-classes are executed.
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186262, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
175

176
177
178
179
180
181
182
183
184
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).
    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)

Daniel Scheffler's avatar
Daniel Scheffler committed
185

186
187
188
189
190
191
192
193
194
# 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!

Daniel Scheffler's avatar
Daniel Scheffler committed
195

196
197
198
199
200
201
202
203
204
class Test_Landsat8_PreCollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-8 OLI_TIRS scene (pre-collection data).
    More information on the dataset will be outputted after the tests-classes are executed.
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186196, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
205

206
207
208
209
210
211
212
213
214
class Test_Landsat8_CollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Landsat-8 OLI_TIRS scene (collection data).
    More information on the dataset will be outputted after the tests-classes are executed.
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186261, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
215

216
217
218
219
220
221
222
223
224
class Test_Sentinel2A_CollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Sentinel-2A MSI scene (pre-collection data).
    More information on the dataset will be outputted after the tests-classes are executed.
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186268, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
225

226
227
228
229
230
231
232
233
234
class Test_Sentinel2A_PreCollectionData(BaseTestCases.TestAll):
    """
    Parametrized testclass. Tests the level-processes on a Sentinel-2A MSI scene (collection data).
    More information on the dataset will be outputted after the tests-classes are executed.
    """
    @classmethod
    def setUpClass(cls):
        cls.create_job(26186272, job_config_kwargs)

Daniel Scheffler's avatar
Daniel Scheffler committed
235

236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
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.
254
255
256
257
258
259
260
261
262
263
264
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'])
     }
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

# 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)]

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)

293
294
        # FIXME: If the job.status-issue is fixed, the commented out section can be nullified.
        # jobstatus.append(eval("Test_"+items).PC.job.status)
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310

    # 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.
311
312
    # 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.
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
    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.
337
    logger.info("\ntest_gms_preprocessing.py"
338
339
340
341
342
343
                "\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(),
344
345
                                            orient='index',
                                            columns=['Country', 'Characteristic', 'Clouds', 'Overlap_area']))
346
347
348
349
350
351
352
353
354
355
    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]
356
    logger.info(pandas.DataFrame(summary_testResults, columns=results, index=testdata_index))
357
358
359
360
361

    # 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 \
362
       list(itertools.chain(*summary_skipped)):
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
        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")
378

379
380
381
382
383
            logger.info("Test_" + test + ", SKIPS:")
            if summary_skipped[index]:
                logger.info(summary_skipped[index][0][1])
            else:
                logger.info("None.")
384

385
386
            if not index == (len(testdata) - 1):
                logger.info("\n---------------------------------------------------------------------------------------")
387

388
        logger.info("\n***************************************************************************************")
389

390
391
    else:
        pass
392

393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
    # 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
408

409
    logger.info("END.")  # OUTPUT, END.
410

411
412
413
    # 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 = []