Adjusting tiffile versioning #487
Conversation
Codecov ReportPatch coverage:
Additional details and impacted files@@ Coverage Diff @@
## main #487 +/- ##
==========================================
- Coverage 94.06% 93.99% -0.07%
==========================================
Files 48 48
Lines 4363 4363
==========================================
- Hits 4104 4101 -3
- Misses 259 262 +3
☔ View full report in Codecov by Sentry. |
BrianWhitneyAI
force-pushed
the
admin/tiffile_dependency
branch
from
0cfc766 to
01bf791
Compare
BrianWhitneyAI
force-pushed
the
admin/tiffile_dependency
branch
from
35b6898 to
8d52dd5
Compare
…llModeling/aicsimageio into admin/tiffile_dependency
|
During the span of this PR, Two additional dependency issues came up. First, the codecov package is no longer supported and was removed from setup.py. Additionally, the pillow plugin in imageio package broke tests with new version 2.28.0 and has been pinned to 2.27.0. |
|
Pretty sure the codecov thing was resolved in a prior PR and already taken care of in main branch. (it was noted here https://allencellscience.slack.com/archives/CK1L3Q2NP/p1681324180304249 ) |
Oh yep, after a merge from main this PR no longer covers that issue. |
| @@ -214,22 +214,10 @@ def test_ome_tiff_reader_large_files( | |||
| @pytest.mark.parametrize( | |||
| "filename, expected_reader", | |||
| [ | |||
| ( | |||
There was a problem hiding this comment.
These still seem like worthwhile tests, consider moving them to their respective test files (like the first one here, TiffReader, to the TiffReader test file)
There was a problem hiding this comment.
I believe they are already there! They're under test-selected_tiff_reader in test_tiff_reader and test_ome_tiff_reader.
There was a problem hiding this comment.
Oh interesting, I wonder why they were here then. Well if no one else has an issue removing them here I guess it is fine!
There was a problem hiding this comment.
It looks like they were intentionally added to every reader that handles tiff files, maybe just to ensure that the code that selects which reader to use doesn't get confused. This is kind of like an integration test that makes sense when all the readers are present. Maybe there's a better way to organize the tests so we don't have to have redundant ones like that.
| @@ -31,8 +31,8 @@ def run(self): | |||
| # "READER_TO_INSTALL" lookup table from aicsimageio/formats.py. | |||
| format_libs: Dict[str, List[str]] = { | |||
| "base-imageio": [ | |||
| "imageio[ffmpeg]>=2.11.0", | |||
| "Pillow>=9.3", | |||
| "imageio[ffmpeg]>=2.11.0,<2.28.0", | |||
There was a problem hiding this comment.
Ah I see what you meant by Pillow plug-in your in Slack message now. Bummer that this is now also an issue :( seems like it can be handled outside of this PR though
Tifffile 2023.3.15 is released;
2023.3.15
More type annotations.
“For mmstack images”
TiffFile Before:
TiffFile After:
This change results in our _read_delayed and _read_immediate functions in ome_tiff_reader and tiff_reader being unable to read single scenes.
Discussed Possible solutions: