Add hazenlib init tests coverage #144

Lucrezia-Cester · 2021-11-15T21:44:47Z

Hello,

I have been writing test functions for the hazenlib.init script.

There are a couple of things which seem odd to me:

1: the get_manufacturer function (in init ), which extracts the manufacturer tag https://dicom.innolitics.com/ciods/rt-plan/general-equipment/00080070), outputs not one but many manufacturers' names.

2: the rescale_to_byte function, which is then used in the spatial_resolution script rescales the data, but it is unclear to me why this is necessary. It also seems to decrease the image resolution.

I added the first test function for hazenlib __init__. It checks if a file is an enhanced dicom. In the test data I could only find files with the class: 1.2.840.10008.5.1.4.1.1.4 and no file with the class 1.2.840.10008.5.1.4.1.1.4.1 so I only tested part of the conditional statement. I will gradually add more tests functions. This pull request is not complete.

hshuaib90 · 2021-11-16T16:17:24Z

1: the get_manufacturer function (in init ), which extracts the manufacturer tag https://dicom.innolitics.com/ciods/rt-plan/general-equipment/00080070), outputs not one but many manufacturers' names.

def get_manufacturer(dcm: pydicom.Dataset) -> str:
    supported = ['ge', 'siemens', 'philips', 'toshiba']
    manufacturer = dcm.Manufacturer.lower()
    for item in supported:
        if item in manufacturer:
            return item

    raise Exception(f'{manufacturer} not recognised manufacturer')

I'm not sure what you mean. It just returns item once.

2: the rescale_to_byte function, which is then used in the spatial_resolution script rescales the data, but it is unclear to me why this is necessary. It also seems to decrease the image resolution.

This is in so that you can use OpenCV functions, which only work on 8 byte images. It doesn't directly change the image resolution but it does change the image contrast.

Lucrezia-Cester · 2021-11-17T21:03:39Z

Hello,

right, sorry, I hadn't noticed the manufacturer function was used in child classes.
Would it make sense to try to use another function to rescale, so that it only normalises the image? changing the contrast might alter the MTF for MRI images which have both magnitude and phase. However I am not entirely sure of this, maybe it would be worth testing it out?
should the main function be tested?

-all the other functions besides the main and get_field_of_view have a test function now. The get_field_of_view function returns an error when the manufacturer is ge. This is the error it outputs:

Traceback (most recent call last):
File "C:\Users\lucrezia\Documents\GitHub\hazen\tests\test_hazenlib.py", line 163, in setUp
print(self.dcm[0x19, 0x101e].value)
File "C:\Users\lucrezia\AppData\Local\Programs\Python\Python36\lib\site-packages\pydicom\dataset.py", line 852, in getitem
data_elem = self._dict[tag]
KeyError: (0019, 101e)

I am trying to work this out now.

github-actions · 2021-11-22T09:52:56Z

Coverage Report

File	Stmts	Miss	Cover	Missing
hazenlib
init.py	136	34	75%	137, 141, 151, 156, 193, 200–205, 216, 219–226, 246–248, 266–268, 287–289, 298, 303, 309, 339, 373–374
ghosting.py	151	55	64%	24, 89–90, 94, 104–105, 133–135, 151–153, 215–248, 255–275, 280
relaxometry.py	317	90	72%	238–256, 630, 689–691, 745, 793–815, 833–848, 1173–1176, 1185–1188, 1200–1213, 1216–1221, 1233–1263
slice_position.py	110	20	82%	73–74, 99, 183, 190–203, 211, 219–220, 231
slice_width.py	290	45	84%	24, 27, 31, 35–40, 106, 277, 282–283, 289, 294, 357–358, 593–632, 668–671, 674–677
snr.py	165	47	72%	42–47, 209–219, 262–266, 304–315, 321–329, 357–368, 392, 400–402, 405
snr_map.py	108	5	95%	403, 408–410, 440
spatial_resolution.py	293	91	69%	44–102, 116, 208, 248–255, 283, 416–449, 481–484, 488–491
tools.py	75	6	92%	43–50, 89
uniformity.py	79	22	72%	69–70, 77, 111–123, 135–137, 140, 144–147, 155
TOTAL	1760	415	76%

Tests	Skipped	Failures	Errors	Time
173	0 💤	0 ❌	0 🔥	131.523s ⏱️

laurencejackson · 2021-11-22T10:21:04Z

Hey Lu, is this ready for review?

Lucrezia-Cester · 2021-11-22T10:25:21Z

@laurencejackson Hi Laurence, yes it is.

Just one thing to note: the get_field_of_view function wasn't working when the manufacturer was 'GE'.

laurencejackson · 2021-11-22T10:36:45Z

Great, and sorry, I've just seen your questions above!

I don't think we should be investigating a different rescaling function here, it's probably beyond the scope of this PR. I think @tomaroberts has been looking at an updating rescaling function as part of Porting hazen to python 3.9 #143, but that is still being investigated.
Ideally yes the main function should be tested - if it's not clear how to do this then I'm happy to help
looking at the error for GE scanners, could you push your most recent changes? when I look at the code on the PR branch I can't see a test for get_field_of_view? The GE data will almost definitely have that dicom field in the metadata so it's odd that it can't find it now. Could you double check that it is a GE dataset that raises this error and it's not getting a Philips/Siemens/etc dataset instead?

Lucrezia-Cester · 2021-11-22T10:56:22Z

Hi @laurencejackson, the get_field_of_view tests are at lines 84, 107, 128, 149. These tests work for all manufacturers but 'GE'.

For 'GE, If I just try to do

file = str(TEST_DATA_DIR / 'ge' / 'ge_eFilm.dcm')
dcm = pydicom.read_file(file)
fov = hazenlib.get_field_of_view(dcm)
print(fov)

I get this error (sorry, I just noticed I wrote the same thing in my previous comment)

Traceback (most recent call last):
File "C:\Users\lucrezia\Documents\GitHub\hazen\tests\test_hazenlib.py", line 170, in
fov = hazenlib.get_field_of_view(dcm)
File "c:\users\lucrezia\documents\github\hazen\hazenlib_init_.py", line 308, in get_field_of_view
fov = dcm[0x19, 0x101e].value
File "C:\Users\lucrezia\AppData\Local\Programs\Python\Python36\lib\site-packages\pydicom\dataset.py", line 852, in getitem
data_elem = self._dict[tag]
KeyError: (0019, 101e)

I am gonna give it a shot at making a test function for the main now, does that sound good?

laurencejackson · 2021-11-22T13:46:01Z

I am gonna give it a shot at making a test function for the main now, does that sound good?

yeah sounds great, let me know when ready for review.

I've had a look at it seems like none of the available GE test data in the hazen repo has the (0019, 101e) dicom tag filled, it is however in an old GE QA dataset I have left over from a few years ago. I think this is a limitation of the test data. @hshuaib90 could you advise on these options :

we can add some newer GE test data with the tag present for this test
we could add a try-except to adjust the get_field_of_view function to try an alternative method ? In the below, running lines 296 and 298 give the same answer on my GE dataset (fov = 256.0)- we could add a try-except clause here to get the FOV using the logic in line 298 when the dicom tag doesn't exist?
Or finally, is there a particular reason we choose to get the FOV from a GE specific dicom tag that isn't always present? Have you come across any examples where the dcm.Columns *dcm.PixelSpacing[0] doesn't work? If we could use this for all manufacturers first it seems like the most easily maintainable method.

slightly changed the main function to make it easier to test

Lucrezia-Cester · 2021-11-24T19:37:23Z

@laurencejackson this is now ready for review. Cheers

laurencejackson

Thanks Lu, this looks good and it's nearly there! Just requested a few things below.

One thing I'd like to suggest is that package up each test dicom with its true values in dictionaries so we can see where everything is easily. So at the top of the file we can define something like:

test_dicoms = {'philips': {'file': str(TEST_DATA_DIR / 'resolution' / 'philips' / 'IM-0004-0002.dcm'),
                          'MANUFACTURER': 'philips',
                          'ROWS': 512,
                          'COLUMNS': 512,
                          'TR_CHECK': 500,
                          'BW': 205.0,
                          'ENHANCED': False,
                          'PIX_ARRAY': 1},
                'siemens': {'file': str(TEST_DATA_DIR / 'resolution' / 'siemens' / 'IM-0004-0002.dcm'),
                            'MANUFACTURER': 'siemens', ...

We can then use this identical structure to loop over a test example for each manufacturer in the test cases where the function should work on all manufacturers (everything in the base TestHazenLib class) e.g.

def test_get_pixel_size(self):
  for manufactuer in test_dicoms.keys():
      with pydicom.read_file(test_dicoms[manufactuer]['file']) as dcm:
          PIX_SIZE = dcm.PixelSpacing
          PIX_SIZE=tuple(PIX_SIZE)
          pix_size = hazenlib.get_pixel_size(dcm)
          assert pix_size == PIX_SIZE

Structuring the test dicoms this way should let us easily expand the scope of our tests!, which would be great!

laurencejackson · 2021-11-25T12:33:33Z

Hazen_logger.log

@@ -0,0 +1,3 @@
+2021-11-15 21:19:24 I [spatial_resolution.py:main:514] 'FileDataset' object has no attribute 'SeriesDescription'
+2021-11-15 21:20:56 I [spatial_resolution.py:main:514] 'FileDataset' object has no attribute 'SeriesDescription'
+2021-11-15 21:28:24 I [spatial_resolution.py:main:514] 'FileDataset' object has no attribute 'SeriesDescription'


Can you add all *.log files to the .gitignore so we don't track them?

Thanks, can you also delete this log from the repo?

laurencejackson · 2021-11-25T12:36:19Z

hazenlib/__init__.py

    task = importlib.import_module(f"hazenlib.{arguments['<task>']}")
    folder = arguments['<folder>']
    files = [os.path.join(folder, x) for x in os.listdir(folder) if x not in EXCLUDED_FILES]
-    dicom_objects = [pydicom.read_file(x, force=True) for x in files if is_dicom_file(x)]
+    dicom_objects = [pydicom.read_file(x, force=True) for x in files]


Can you re-add your if is_dicom_file condition here? It looks like this branch was forked before the is_dicom_file check was added. We need to add this back in here or it will overwrite this line.

I am giving thumbs up to comments I have acted upon

laurencejackson · 2021-11-25T12:38:31Z

hazenlib/__init__.py

    if not arguments['<task>'] == 'snr' and arguments['--measured_slice_width']:
        raise Exception("the (--measured_slice_width) option can only be used with snr")
    elif arguments['<task>'] == 'snr' and arguments['--measured_slice_width']:
        measured_slice_width = float(arguments['--measured_slice_width'])
-        return pp.pprint(task.main(dicom_objects, measured_slice_width, report_path=report))
+        return task.main(dicom_objects, measured_slice_width, report_path=report)


Why have you chosen to return the output outside of a pprint object? How does this effect the output the user sees?

this is not done to change the output the user sees, rather for me to be able to test the function because print returns NONE

Do you mean print returns none or pp.pprint returns none? Could you add some screenshots to example outputs to explain what you mean? I don't think we want to change this if we can.

@Lucrezia-Cester, would that mean that the user does not see the result printed to the command line when they run the SNR task? Could you try something like the below so we print the result and we also return a value?

if not arguments['<task>'] == 'snr' and arguments['--measured_slice_width']: raise Exception("the (--measured_slice_width) option can only be used with snr") elif arguments['<task>'] == 'snr' and arguments['--measured_slice_width']: measured_slice_width = float(arguments['--measured_slice_width']) result = task.main(dicom_objects, measured_slice_width, report_path=report) elif arguments['<task>'] == 'relaxometry': result = parse_relaxometry_data(task, arguments, dicom_objects, report) else: result = task.main(dicom_objects, report_path=report) pp.pprint(result) return result

laurencejackson · 2021-11-25T12:39:36Z

hazenlib/__init__.py

-
-        # Pass arguments with dictionary, stripping initial double dash ('--')
-        relaxometry_args = {}
+        return parse_relaxometry_data(task, arguments, dicom_objects, report)


As above we need to re-add this relaxometry code due to the earlier branching from develop

This is still there, I have put this relaxometry code in a function outside the main function.

laurencejackson · 2021-11-25T12:42:51Z

tests/test_hazenlib.py

@@ -37,9 +52,47 @@ def test_get_TR(self):
        TR = hazenlib.get_TR(self.dcm)
        assert TR == self.TR_CHECK

+    def test_is_enhanced_dicom(self):


Can we have a second test for an ENHANCED = True dicom too? just need to copy/paste this one but with self.dcm overwritten with an enhanced dicom and assert that returns True

laurencejackson · 2021-11-25T12:51:45Z

tests/test_hazenlib.py

@@ -49,41 +102,178 @@ def setUp(self):
        self.file = str(TEST_DATA_DIR / 'resolution' / 'philips' / 'IM-0004-0002.dcm')
        self.dcm = pydicom.read_file(self.file)

+file = str(TEST_DATA_DIR / 'resolution' / 'philips' / 'IM-0004-0002.dcm')


as above, this bit looks like it needs to be ina function

laurencejackson · 2021-11-25T12:52:15Z

tests/test_hazenlib.py


    def setUp(self):
        self.file = str(TEST_DATA_DIR / 'resolution' / 'eastkent' / '256_sag.IMA')
        self.dcm = pydicom.read_file(self.file)
+        FOV = self.dcm.Columns * self.dcm.PixelSpacing[0]
+
+file = str(TEST_DATA_DIR / 'resolution' / 'eastkent' / '256_sag.IMA')


laurencejackson · 2021-11-25T12:52:21Z

tests/test_hazenlib.py

+        FOV = self.dcm.Columns * self.dcm.PixelSpacing[0]
+
+
+file = str(TEST_DATA_DIR / 'toshiba' / 'TOSHIBA_TM_MR_DCM_V3_0.dcm')


laurencejackson · 2021-11-25T12:55:24Z

tests/test_hazenlib.py


    def setUp(self):
        self.file = str(TEST_DATA_DIR / 'ge' / 'ge_eFilm.dcm')
        self.dcm = pydicom.read_file(self.file)


+file = str(TEST_DATA_DIR / 'ge' / 'ge_eFilm.dcm')


laurencejackson · 2021-11-25T12:56:32Z

tests/test_hazenlib.py

+
+
+
+from docopt import docopt


Can you move all import statements so they're at the top of the file?

Hi @laurencejackson, thank you for all the feedback. I'll get to it in the following days

@laurencejackson

Hi @laurencejackson, I have added the changes you suggested (If my interpretation was correct). I could not add the test_field_of_view with the other tests because the GE manufacturer was not working with the hazenlib function. Also, I could not add the enhanced_dicom = True bit because I do not have a test image which returns true for that condition.

Lucrezia-Cester · 2021-12-15T18:23:26Z

Hi @laurencejackson, this is now ready for review

laurencejackson

Sorry that this took so long to review @Lucrezia-Cester, but it looks fantastic! thank you! MY My one main suggestion is that we split out the tests again so they are individual functions, this helps a lot with debugging. thanks!

laurencejackson · 2021-12-16T10:24:51Z

Hazen_logger.log

@@ -0,0 +1,3 @@
+2021-11-15 21:19:24 I [spatial_resolution.py:main:514] 'FileDataset' object has no attribute 'SeriesDescription'
+2021-11-15 21:20:56 I [spatial_resolution.py:main:514] 'FileDataset' object has no attribute 'SeriesDescription'
+2021-11-15 21:28:24 I [spatial_resolution.py:main:514] 'FileDataset' object has no attribute 'SeriesDescription'


Thanks, can you also delete this log from the repo?

laurencejackson · 2021-12-16T10:25:26Z

.gitignore

@@ -5,6 +5,7 @@ coverage.xml
 **/__pycache__
 .idea
 logs
+*.logs


I think you meant this to be *.log

laurencejackson · 2021-12-16T10:28:13Z

hazenlib/__init__.py

    if not arguments['<task>'] == 'snr' and arguments['--measured_slice_width']:
        raise Exception("the (--measured_slice_width) option can only be used with snr")
    elif arguments['<task>'] == 'snr' and arguments['--measured_slice_width']:
        measured_slice_width = float(arguments['--measured_slice_width'])
-        return pp.pprint(task.main(dicom_objects, measured_slice_width, report_path=report))
+        return task.main(dicom_objects, measured_slice_width, report_path=report)


Do you mean print returns none or pp.pprint returns none? Could you add some screenshots to example outputs to explain what you mean? I don't think we want to change this if we can.

tests/test_hazenlib.py

Lucrezia-Cester · 2021-12-22T07:30:48Z

Hi @laurencejackson

regarding the pprint issue, here is what happens if I don't just use return rather than using pprint. I get None as output rather than a dictionary.

laurencejackson · 2021-12-22T14:17:38Z

hazenlib/__init__.py

    if not arguments['<task>'] == 'snr' and arguments['--measured_slice_width']:
        raise Exception("the (--measured_slice_width) option can only be used with snr")
    elif arguments['<task>'] == 'snr' and arguments['--measured_slice_width']:
        measured_slice_width = float(arguments['--measured_slice_width'])
-        return pp.pprint(task.main(dicom_objects, measured_slice_width, report_path=report))
+        return task.main(dicom_objects, measured_slice_width, report_path=report)


@Lucrezia-Cester, would that mean that the user does not see the result printed to the command line when they run the SNR task? Could you try something like the below so we print the result and we also return a value?

if not arguments['<task>'] == 'snr' and arguments['--measured_slice_width']: raise Exception("the (--measured_slice_width) option can only be used with snr") elif arguments['<task>'] == 'snr' and arguments['--measured_slice_width']: measured_slice_width = float(arguments['--measured_slice_width']) result = task.main(dicom_objects, measured_slice_width, report_path=report) elif arguments['<task>'] == 'relaxometry': result = parse_relaxometry_data(task, arguments, dicom_objects, report) else: result = task.main(dicom_objects, report_path=report) pp.pprint(result) return result

Lucrezia-Cester added 2 commits November 9, 2021 21:32

added test functions to __hazenlib.init

fd1bea4

Lucrezia-Cester requested review from hshuaib90 and laurencejackson November 15, 2021 21:44

Lucrezia-Cester linked an issue Nov 15, 2021 that may be closed by this pull request

[HZN-90] Hazenlib: improve __init__.py coverage #90

Closed

finished adding tests for functions in hazenlib

98241ce

Lucrezia-Cester added 4 commits November 18, 2021 15:55

Update __init__.py

d3a3217

added test for get_field_of_view

bf4f53e

Merge branch 'develop' into add-hazelib.__init__-tests-coverage

507526c

Update __init__.py

0ee77f6

Lucrezia-Cester added 5 commits November 24, 2021 18:04

added tests for hazenlib.main()

6b5491b

updated hazenlib.__main__

bc42917

slightly changed the main function to make it easier to test

fixed failing tests

d6137d8

updated hazenlib_tests

d2d224e

updated hazenlib_tests

008df9e

laurencejackson requested changes Nov 25, 2021

View reviewed changes

Lucrezia-Cester added 4 commits November 29, 2021 17:42

re added is_dicom_file

39353b7

put import statements at the beginning of the file

4afff59

added *.logs to gitignore

1ee8679

Lucrezia-Cester changed the base branch from develop to release/0.5.0 December 9, 2021 13:27

Merge branch 'release/0.5.0' into add-hazelib.__init__-tests-coverage

858c2f2

Lucrezia-Cester changed the base branch from release/0.5.0 to develop December 9, 2021 14:50

Lucrezia-Cester changed the base branch from develop to release/0.5.0 December 9, 2021 15:12

Update __init__.py

238be87

tomaroberts changed the title ~~Add hazelib. init tests coverage~~ Add hazenlib init tests coverage Dec 10, 2021

tomaroberts changed the base branch from release/0.5.0 to release/0.5.1 December 10, 2021 16:40

laurencejackson changed the base branch from release/0.5.1 to release/0.5.2 December 16, 2021 10:25

laurencejackson requested changes Dec 16, 2021

View reviewed changes

tomaroberts added this to the 0.5.2 milestone Dec 21, 2021

tomaroberts changed the base branch from release/0.5.2 to release/0.5.1 December 21, 2021 16:39

Lucrezia-Cester added 2 commits December 22, 2021 05:58

changed *.logs to *.log

f9cc8c4

deleted logs file

44c2ddc

Lucrezia-Cester changed the base branch from release/0.5.1 to release/0.5.2 December 22, 2021 06:12

split big function into single functions

2510d13

Merge branch 'release/0.5.2' into add-hazelib.__init__-tests-coverage

aef3f7a

laurencejackson requested changes Dec 22, 2021

View reviewed changes

Lucrezia-Cester and others added 6 commits December 22, 2021 15:08

inserted pprint before return

51fa679

Merge branch 'release/0.5.2' into add-hazelib.__init__-tests-coverage

658eab3

updated init.py

ada0244

updates tests to work with pprint strings

0c05753

almsot equal relaxometry tests and set -Eeuxo pipefail

fdbecc9

adds entry point to allow us to return a string without exit code 0

28f4690

laurencejackson merged commit c292347 into release/0.5.2 Dec 22, 2021

laurencejackson deleted the add-hazelib.__init__-tests-coverage branch December 22, 2021 16:20

tomaroberts mentioned this pull request Feb 16, 2022

[HZN-90] Hazenlib: improve __init__.py coverage #90

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add hazenlib init tests coverage #144

Add hazenlib init tests coverage #144

Lucrezia-Cester commented Nov 15, 2021 •

edited

Loading

hshuaib90 commented Nov 16, 2021

Lucrezia-Cester commented Nov 17, 2021 •

edited

Loading

github-actions bot commented Nov 22, 2021 •

edited

Loading

laurencejackson commented Nov 22, 2021

Lucrezia-Cester commented Nov 22, 2021

laurencejackson commented Nov 22, 2021

Lucrezia-Cester commented Nov 22, 2021 •

edited

Loading

laurencejackson commented Nov 22, 2021 •

edited

Loading

Lucrezia-Cester commented Nov 24, 2021

laurencejackson left a comment •

edited

Loading

laurencejackson Nov 25, 2021

laurencejackson Dec 16, 2021

laurencejackson Nov 25, 2021

Lucrezia-Cester Nov 30, 2021

laurencejackson Nov 25, 2021

Lucrezia-Cester Nov 29, 2021

laurencejackson Dec 16, 2021

laurencejackson Dec 22, 2021

laurencejackson Nov 25, 2021

Lucrezia-Cester Nov 30, 2021

laurencejackson Nov 25, 2021

laurencejackson Nov 25, 2021

laurencejackson Nov 25, 2021

laurencejackson Nov 25, 2021

laurencejackson Nov 25, 2021

laurencejackson Nov 25, 2021

Lucrezia-Cester Nov 29, 2021

Lucrezia-Cester commented Dec 15, 2021 •

edited

Loading

laurencejackson left a comment

laurencejackson Dec 16, 2021

laurencejackson Dec 16, 2021

laurencejackson Dec 16, 2021

Lucrezia-Cester commented Dec 22, 2021

laurencejackson Dec 22, 2021

		FOV = self.dcm.Columns * self.dcm.PixelSpacing[0]


		file = str(TEST_DATA_DIR / 'toshiba' / 'TOSHIBA_TM_MR_DCM_V3_0.dcm')

Add hazenlib init tests coverage #144

Add hazenlib init tests coverage #144

Conversation

Lucrezia-Cester commented Nov 15, 2021 • edited Loading

hshuaib90 commented Nov 16, 2021

Lucrezia-Cester commented Nov 17, 2021 • edited Loading

github-actions bot commented Nov 22, 2021 • edited Loading

laurencejackson commented Nov 22, 2021

Lucrezia-Cester commented Nov 22, 2021

laurencejackson commented Nov 22, 2021

Lucrezia-Cester commented Nov 22, 2021 • edited Loading

laurencejackson commented Nov 22, 2021 • edited Loading

Lucrezia-Cester commented Nov 24, 2021

laurencejackson left a comment • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Lucrezia-Cester commented Dec 15, 2021 • edited Loading

laurencejackson left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Lucrezia-Cester commented Dec 22, 2021

Choose a reason for hiding this comment

Lucrezia-Cester commented Nov 15, 2021 •

edited

Loading

Lucrezia-Cester commented Nov 17, 2021 •

edited

Loading

github-actions bot commented Nov 22, 2021 •

edited

Loading

Lucrezia-Cester commented Nov 22, 2021 •

edited

Loading

laurencejackson commented Nov 22, 2021 •

edited

Loading

laurencejackson left a comment •

edited

Loading

Lucrezia-Cester commented Dec 15, 2021 •

edited

Loading