Update dependencies v0.61

[juhoinkinen]

This updates some outdated dependencies:

• Scikit-learn 1.2.2
• Spacy 3.5.*
• Schemathesis 3.19.*
• lmdb 1.4.1

Also pins (at least) minor versions of all non-dev dependencies to the versions currently being installed and removes some unnecessary dependencies (swagger_ui_bundle, codecov).

[codecov]

Codecov Report

Patch coverage has no change and project coverage change: +0.03 🎉

Comparison is base (3a5fc50) 99.57% compared to head (7d2e98d) 99.61%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #683      +/-   ##
==========================================
+ Coverage   99.57%   99.61%   +0.03%     
==========================================
  Files          88       89       +1     
  Lines        6138     6166      +28     
==========================================
+ Hits         6112     6142      +30     
+ Misses         26       24       -2     

see 33 files with indirect coverage changes

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[juhoinkinen]

I think swagger_ui_bundle = "*" is nowdays not needed as a direct dependency and could be removed.

The dependency was added in #215 due to a problem discussed in spec-first/connexion#406 (comment), which seems to be fixed now.

Done: Remove unnecessary swagger_ui_bundle dependency

[osma]

Would it make sense to pin major/minor versions, at least for all non-dev dependencies? At least nltk and gunicorn are not pinned currently and this could cause sudden API breakage.

[juhoinkinen]

TensorFlow 2.12.0 has some breaking changes:

tf.keras:

Moved all saving-related utilities to a new namespace, keras.saving, for example: keras.saving.load_model, keras.saving.save_model, keras.saving.custom_object_scope, keras.saving.get_custom_objects, keras.saving.register_keras_serializable,keras.saving.get_registered_name and keras.saving.get_registered_object. The previous API locations (in keras.utils and keras.models) will be available indefinitely, but we recommend you update your code to point to the new API locations.

Annif imports load_model, which could or should be updated:

Annif/annif/backend/nn_ensemble.py

Line 15 in 33bc561

from tensorflow.keras.models import Model, load_model

And also atomic_save apparently uses the old location.

But now I see there is a conflict in installing the new TF version and numpy:

Because tensorflow-cpu (2.12.0) depends on numpy (>=1.22,<1.24)
 and no versions of tensorflow-cpu match >2.12.0,<2.13.0, tensorflow-cpu (>=2.12.0,<2.13.0) requires numpy (>=1.22,<1.24).
So, because annif depends on both numpy (1.24.*) and tensorflow-cpu (2.12.*), version solving failed.

[juhoinkinen]

Force-pushed to drop upgrading to TensorFlow v2.12.0, which conflicted with Numpy.

[sonarcloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

No Coverage information
No Duplication information

[osma]

LGTM