Refactor to move shared functionality to the library instead of the tests

CHANGELOG.md

@@ -17,8 +17,6 @@ and uses [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 * `compare.py` create new `assert_within_tolerance` function to use file-specific tolerances
 * `autorift_golden.json.j2` includes L5+5, L7+7, L7+8, and L8+7 pairs in different projections to test reprojection code
 * `--user-id` pytest CLI argument to allow finding products submitted by a different user than the authorized user
-
-### Added
 * `test_autorift.py` golden test for the autoRIFT plugin
 
 ### Changed
@@ -28,6 +26,7 @@ and uses [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 * InSAR Gamma tests so that they do not use per-image threshold and instead analyze metadata, coregistration, nodata coverage, and dataproduct quality
 * RTC and autoRIFT golden tests now sleep for 60 seconds between requests for job status
 * `conda-env.yml` has been renamed to `environment.yml` to follow standard naming conventions 
+* Refactored golden tests to move shared `golden_submission`, `golden_wait`, `golden_jobs_succeeds`, and `golden_tif_names` to `helpers.py`
 
 ## [0.1.0](https://github.com/ASFHyP3/hyp3-testing/compare/v0.0.0...v0.1.0)
 

hyp3_testing/helpers.py

@@ -1,3 +1,4 @@
+import json
 import os
 from contextlib import contextmanager
 from glob import glob
@@ -8,6 +9,8 @@
 from hyp3_sdk.util import extract_zipped_product
 from remotezip import RemoteZip
 
+from hyp3_testing import util
+
 
 def freeze_job_parameters(job: Job) -> tuple:
     job_parameters = job.job_parameters
@@ -107,3 +110,53 @@ def job_tifs(job_id, api, directory, keep=False):
             for ff in product_dir.rglob('*'):
                 ff.unlink()
             product_dir.rmdir()
+
+
+def golden_submission(comparison_environments: tuple[tuple[Path, str], tuple[Path, str]], job_template: str):
+    job_name = util.generate_job_name()
+    print(f'Job name: {job_name}')
+
+    testing_parameters = util.render_template(job_template, name=job_name)
+    submission_payload = [{k: item[k] for k in ['name', 'job_parameters', 'job_type']} for item in testing_parameters]
+
+    for dir_, api in comparison_environments:
+        dir_.mkdir(parents=True, exist_ok=True)
+
+        hyp3 = HyP3(api, os.environ.get('EARTHDATA_LOGIN_USER'), os.environ.get('EARTHDATA_LOGIN_PASSWORD'))
+        jobs = hyp3.submit_prepared_jobs(submission_payload)
+        request_time = jobs.jobs[0].request_time.isoformat(timespec='seconds')
+        print(f'{dir_.name} request time: {request_time}')
+
+        submission_details = {'name': job_name, 'request_time': request_time}
+        submission_report = dir_ / f'{dir_.name}_submission.json'
+        submission_report.write_text(json.dumps(submission_details))
+
+
+def golden_wait(comparison_environments: tuple[tuple[Path, str], tuple[Path, str]], job_name: str | None, user_id: str):
+    for dir_, api in comparison_environments:
+        if job_name is None:
+            submission_report = dir_ / f'{dir_.name}_submission.json'
+            submission_details = json.loads(submission_report.read_text())
+            job_name = submission_details['name']
+
+        hyp3 = HyP3(api, os.environ.get('EARTHDATA_LOGIN_USER'), os.environ.get('EARTHDATA_LOGIN_PASSWORD'))
+        jobs = hyp3.find_jobs(name=job_name, user_id=user_id)
+
+        assert len(jobs) > 0  # will throw if job_name not associated with user_id
+
+        _ = hyp3.watch(jobs)
+
+
+def golden_job_succeeds(jobs_info: dict):
+    main_succeeds = sum([value['main']['succeeded'] for value in jobs_info.values()])
+    develop_succeeds = sum([value['develop']['succeeded'] for value in jobs_info.values()])
+    assert main_succeeds != 0, 'Main jobs did not succeed'
+    assert develop_succeeds != 0, 'Develop jobs did not succeed'
+    assert main_succeeds == develop_succeeds, 'Main and develop job success counts do not match'
+
+
+def golden_tif_names(jobs_info: dict):
+    for pair_information in jobs_info.values():
+        main_normalized_files = pair_information['main']['normalized_files']
+        develop_normalized_files = pair_information['develop']['normalized_files']
+        assert main_normalized_files == develop_normalized_files

hyp3_testing/isce2_compare.py

@@ -0,0 +1,67 @@
+import rioxarray  # noqa: F401
+import xarray as xr
+from osgeo import gdal
+
+from hyp3_testing import compare
+from hyp3_testing.helpers import job_tifs
+
+
+gdal.UseExceptions()
+
+
+def _comparisons(main_ds, develop_ds, pixel_size):
+    compare.images_are_within_offset_threshold(main_ds, develop_ds, pixel_size=pixel_size, offset_threshold=5.0)
+    compare.maskes_are_within_similarity_threshold(main_ds, develop_ds, mask_rate=0.98)
+    compare.values_are_within_statistic(main_ds, develop_ds, confidence_level=0.99)
+
+
+def compare_isce2_derived_products(comparison_environments, jobs_info, keep):
+    (main_dir, main_api), (develop_dir, develop_api) = comparison_environments
+
+    failure_count = 0
+    messages = []
+    for pair, pair_information in jobs_info.items():
+        with (
+            job_tifs(pair_information['main']['job_id'], main_api, main_dir, keep) as main_tifs,
+            job_tifs(pair_information['develop']['job_id'], develop_api, develop_dir, keep) as develop_tifs,
+        ):
+            main_file_dir = main_dir / (main_product_name := pair_information['main']['dir'])
+            develop_file_dir = develop_dir / (develop_product_name := pair_information['develop']['dir'])
+
+            compare.compare_product_files(main_file_dir, develop_file_dir)
+
+            main_parameter_file = (main_file_dir / main_product_name).with_suffix('.txt')
+            develop_parameter_file = (develop_file_dir / develop_product_name).with_suffix('.txt')
+
+            compare.compare_parameter_files(str(main_parameter_file), str(develop_parameter_file))
+
+            for main_tif, develop_tif in zip(main_tifs, develop_tifs):
+                comparison_header = '\n'.join(['-' * 80, str(main_tif), str(develop_tif), '-' * 80])
+
+                main_ds = xr.open_dataset(main_tif, engine='rasterio').band_data.data[0]
+                develop_ds = xr.open_dataset(develop_tif, engine='rasterio').band_data.data[0]
+
+                try:
+                    compare.compare_raster_info(main_tif, develop_tif)
+
+                    pixel_size = gdal.Info(str(main_tif), format='json')['geoTransform'][1]
+                    # OpenCV does not support complex data, so we must compare each component as real values.
+                    if main_ds.dtype in ('complex32', 'complex64'):
+                        _comparisons(main_ds.real, develop_ds.real, pixel_size)
+                        _comparisons(main_ds.imag, develop_ds.imag, pixel_size)
+                    else:
+                        _comparisons(main_ds, develop_ds, pixel_size)
+
+                    if '_unw_phase.tif' in str(main_tif):
+                        compare.nodata_count_change_are_within_threshold(main_ds, develop_ds, threshold=0.01)
+
+                    if '_corr.tif' in str(main_tif):
+                        compare.corr_average_decrease_within_threshold(main_ds, develop_ds, threshold=0.05)
+
+                except compare.ComparisonFailure as e:
+                    messages.append(f'{comparison_header}\n{e}')
+                    failure_count += 1
+
+    if messages:
+        messages.insert(0, f'{failure_count} differences found!!')
+        raise compare.ComparisonFailure('\n\n'.join(messages))

tests/test_autorift.py

@@ -7,30 +7,15 @@
 import pytest
 import xarray as xr
 
-from hyp3_testing import compare, helpers, util
+from hyp3_testing import compare, helpers
 
 
 pytestmark = pytest.mark.golden
 
 
 @pytest.mark.nameskip
 def test_golden_submission(its_live_environments):
-    job_name = util.generate_job_name()
-    print(f'Job name: {job_name}')
-
-    submission_payload = util.render_template('autorift_golden.json.j2', name=job_name)
-
-    for dir_, api in its_live_environments:
-        dir_.mkdir(parents=True, exist_ok=True)
-
-        hyp3 = hyp3_sdk.HyP3(api, os.environ.get('EARTHDATA_LOGIN_USER'), os.environ.get('EARTHDATA_LOGIN_PASSWORD'))
-        jobs = hyp3.submit_prepared_jobs(submission_payload)
-        request_time = jobs.jobs[0].request_time.isoformat(timespec='seconds')
-        print(f'{dir_.name} request time: {request_time}')
-
-        submission_details = {'name': job_name, 'request_time': request_time}
-        submission_report = dir_ / f'{dir_.name}_submission.json'
-        submission_report.write_text(json.dumps(submission_details))
+    helpers.golden_submission(its_live_environments, 'autorift_golden.json.j2')
 
 
 @pytest.mark.timeout(10800)  # 3 hours

tests/test_burst_insar.py

@@ -1,129 +1,33 @@
-import json
-import os
-
-import hyp3_sdk.util
 import pytest
-import rioxarray  # noqa: F401
-import xarray as xr
-from osgeo import gdal
 
-from hyp3_testing import compare, util
-from hyp3_testing.helpers import job_tifs
+import hyp3_testing.helpers as helpers
+from hyp3_testing.isce2_compare import compare_isce2_derived_products
 
 
-gdal.UseExceptions()
 pytestmark = pytest.mark.golden
 
 
 @pytest.mark.nameskip
 def test_golden_submission(comparison_environments):
-    job_name = util.generate_job_name()
-    print(f'Job name: {job_name}')
-
-    testing_parameters = util.render_template('insar_isce_burst_golden.json.j2', name=job_name)
-    submission_payload = [{k: item[k] for k in ['name', 'job_parameters', 'job_type']} for item in testing_parameters]
-
-    for dir_, api in comparison_environments:
-        dir_.mkdir(parents=True, exist_ok=True)
-
-        hyp3 = hyp3_sdk.HyP3(api, os.environ.get('EARTHDATA_LOGIN_USER'), os.environ.get('EARTHDATA_LOGIN_PASSWORD'))
-        jobs = hyp3.submit_prepared_jobs(submission_payload)
-        request_time = jobs.jobs[0].request_time.isoformat(timespec='seconds')
-        print(f'{dir_.name} request time: {request_time}')
-
-        submission_details = {'name': job_name, 'request_time': request_time}
-        submission_report = dir_ / f'{dir_.name}_submission.json'
-        submission_report.write_text(json.dumps(submission_details))
+    helpers.golden_submission(comparison_environments, 'insar_isce_burst_golden.json.j2')
 
 
 @pytest.mark.timeout(10800)  # 180 minutes as InSAR jobs can take ~2.5 hrs
 @pytest.mark.dependency()
 def test_golden_wait(comparison_environments, job_name, user_id):
-    for dir_, api in comparison_environments:
-        if job_name is None:
-            submission_report = dir_ / f'{dir_.name}_submission.json'
-            submission_details = json.loads(submission_report.read_text())
-            job_name = submission_details['name']
-
-        hyp3 = hyp3_sdk.HyP3(api, os.environ.get('EARTHDATA_LOGIN_USER'), os.environ.get('EARTHDATA_LOGIN_PASSWORD'))
-        jobs = hyp3.find_jobs(name=job_name, user_id=user_id)
-
-        assert len(jobs) > 0  # will throw if job_name not associated with user_id
-
-        _ = hyp3.watch(jobs)
+    helpers.golden_wait(comparison_environments, job_name, user_id)
 
 
 @pytest.mark.dependency(depends=['test_golden_wait'])
 def test_golden_job_succeeds(jobs_info):
-    main_succeeds = sum([value['main']['succeeded'] for value in jobs_info.values()])
-    develop_succeeds = sum([value['develop']['succeeded'] for value in jobs_info.values()])
-    assert main_succeeds != 0
-    assert develop_succeeds != 0
-    assert main_succeeds == develop_succeeds
+    helpers.golden_job_succeeds(jobs_info)
 
 
 @pytest.mark.dependency(depends=['test_golden_wait'])
 def test_golden_tif_names(jobs_info):
-    for pair_information in jobs_info.values():
-        main_normalized_files = pair_information['main']['normalized_files']
-        develop_normalized_files = pair_information['develop']['normalized_files']
-        assert main_normalized_files == develop_normalized_files
-
-
-def _comparisons(main_ds, develop_ds, pixel_size):
-    compare.images_are_within_offset_threshold(main_ds, develop_ds, pixel_size=pixel_size, offset_threshold=5.0)
-    compare.maskes_are_within_similarity_threshold(main_ds, develop_ds, mask_rate=0.98)
-    compare.values_are_within_statistic(main_ds, develop_ds, confidence_level=0.99)
+    helpers.golden_tif_names(jobs_info)
 
 
 @pytest.mark.dependency(depends=['test_golden_wait'])
-def test_golden_burst_insar(comparison_environments, jobs_info, keep):
-    (main_dir, main_api), (develop_dir, develop_api) = comparison_environments
-
-    failure_count = 0
-    messages = []
-    for pair, pair_information in jobs_info.items():
-        with (
-            job_tifs(pair_information['main']['job_id'], main_api, main_dir, keep) as main_tifs,
-            job_tifs(pair_information['develop']['job_id'], develop_api, develop_dir, keep) as develop_tifs,
-        ):
-            main_file_dir = main_dir / (main_product_name := pair_information['main']['dir'])
-            develop_file_dir = develop_dir / (develop_product_name := pair_information['develop']['dir'])
-
-            compare.compare_product_files(main_file_dir, develop_file_dir)
-
-            main_parameter_file = (main_file_dir / main_product_name).with_suffix('.txt')
-            develop_parameter_file = (develop_file_dir / develop_product_name).with_suffix('.txt')
-
-            compare.compare_parameter_files(str(main_parameter_file), str(develop_parameter_file))
-
-            for main_tif, develop_tif in zip(main_tifs, develop_tifs):
-                comparison_header = '\n'.join(['-' * 80, str(main_tif), str(develop_tif), '-' * 80])
-
-                main_ds = xr.open_dataset(main_tif, engine='rasterio').band_data.data[0]
-                develop_ds = xr.open_dataset(develop_tif, engine='rasterio').band_data.data[0]
-
-                try:
-                    compare.compare_raster_info(main_tif, develop_tif)
-
-                    pixel_size = gdal.Info(str(main_tif), format='json')['geoTransform'][1]
-                    # OpenCV does not support complex data, so we must compare each component as real values.
-                    if main_ds.dtype in ('complex32', 'complex64'):
-                        _comparisons(main_ds.real, develop_ds.real, pixel_size)
-                        _comparisons(main_ds.imag, develop_ds.imag, pixel_size)
-                    else:
-                        _comparisons(main_ds, develop_ds, pixel_size)
-
-                    if '_unw_phase.tif' in str(main_tif):
-                        compare.nodata_count_change_are_within_threshold(main_ds, develop_ds, threshold=0.01)
-
-                    if '_corr.tif' in str(main_tif):
-                        compare.corr_average_decrease_within_threshold(main_ds, develop_ds, threshold=0.05)
-
-                except compare.ComparisonFailure as e:
-                    messages.append(f'{comparison_header}\n{e}')
-                    failure_count += 1
-
-    if messages:
-        messages.insert(0, f'{failure_count} differences found!!')
-        raise compare.ComparisonFailure('\n\n'.join(messages))
+def test_golden_multi_burst_insar(comparison_environments, jobs_info, keep):
+    compare_isce2_derived_products(comparison_environments, jobs_info, keep)

tests/test_helpers.py

@@ -1,3 +1,5 @@
+import pytest
+
 from hyp3_testing import helpers
 
 
@@ -36,3 +38,38 @@ def test_find_files_in_products(tmp_path):
         (main_dir / 'b.tif', develop_dir / 'b.tif'),
         (main_dir / 'c.tif', develop_dir / 'c.tif'),
     ]
+
+
+def test_golden_job_succeeds():
+    jobs_info = {'pair1': {'main': {'succeeded': 1}, 'develop': {'succeeded': 1}}}
+    helpers.golden_job_succeeds(jobs_info)
+
+    jobs_info = {'pair1': {'main': {'succeeded': 0}, 'develop': {'succeeded': 1}}}
+    with pytest.raises(AssertionError, match='Main jobs did not succeed'):
+        helpers.golden_job_succeeds(jobs_info)
+
+    jobs_info = {'pair1': {'main': {'succeeded': 1}, 'develop': {'succeeded': 0}}}
+    with pytest.raises(AssertionError, match='Develop jobs did not succeed'):
+        helpers.golden_job_succeeds(jobs_info)
+
+    jobs_info = {'pair1': {'main': {'succeeded': 1}, 'develop': {'succeeded': 2}}}
+    with pytest.raises(AssertionError, match='Main and develop job success counts do not match'):
+        helpers.golden_job_succeeds(jobs_info)
+
+
+def test_golden_tif_names():
+    jobs_info = {
+        'pair1': {
+            'main': {'normalized_files': ['file1', 'file2']},
+            'develop': {'normalized_files': ['file1', 'file2']},
+        },
+        'pair2': {
+            'main': {'normalized_files': ['file3', 'file4']},
+            'develop': {'normalized_files': ['file3', 'file4']},
+        },
+    }
+    helpers.golden_tif_names(jobs_info)
+
+    jobs_info['pair2']['develop']['normalized_files'] = ['file3', 'file5']
+    with pytest.raises(AssertionError):
+        helpers.golden_tif_names(jobs_info)