Update dependency shapely to v2

[renovate]

This PR contains the following updates:

Package	Change	Age	Adoption	Passing	Confidence
shapely	==1.8.5.post1 -> ==2.0.0

---

Release Notes

shapely/shapely

v2.0.0

Compare Source

Shapely version 2.0.0 is a major release featuring a complete refactor of the
internals and new vectorized (element-wise) array operations providing
considerable performance improvements.

For a full changelog, see
https://shapely.readthedocs.io/en/latest/release/2.x.html#version-2-0-0

Relevant changes in behaviour compared to 2.0rc3:

• Added temporary support for unpickling shapely<2.0 geometries.

---

Configuration

📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined).

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR has been generated by Mend Renovate. View repository job log here.

[jwkaltz]

Current strategy:
First fix the Shapely deprecation warnings on current master

• Fix shapely deprecation warnings #1647
• Fix shapely deprecation warning for type attribute #1662

Then reevaluate the upgrade to Shapely 2.

[renovate]

Edited/Blocked Notification

Renovate will not automatically rebase this PR, because it does not recognize the last commit author and assumes somebody else may have edited the PR.
You can manually request rebase by checking the rebase/retry box above.

⚠ Warning: custom changes will be lost.

[jwkaltz]

It seems Shapely 2.0 raises a different exception type than Shapely 1.8:
shapely.errors.GEOSException: TopologyException: Input geom 1 is invalid: Self-intersection at 0.5 0.5

The type of exception in our test case is shapely.errors.GEOSException,
whereas the same case in Shapely 1.8 results in a shapely.errors.TopologicalError

The behavior seems equivalent, so I adapted the test case.

Further reviews appreciated.

[codecov]

Codecov Report

Merging #1642 (e19f912) into master (8f8b522) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #1642   +/-   ##
=======================================
  Coverage   77.50%   77.50%           
=======================================
  Files         127      127           
  Lines        5209     5209           
=======================================
  Hits         4037     4037           
  Misses       1172     1172           

Flag	Coverage Δ
unittests	77.50% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[jwkaltz]

@mki-c2c can you please see if the test case adaptation makes sense,
in light of what you summarized in #1111?

See also my previous comment.

[mki-c2c]

This looks good. Thanks.
I did the following review actions:

• look over the code changes
• read the migration doc for shapely 1.8/2.0 https://shapely.readthedocs.io/en/stable/migration.html
• checked use of exceptions in the shapely source code (tests)

Everything tells me the current use of shapely (and the caught exception) is fine and I could not find any deprecations.
We should continue checking for deprecation warnings in the test outputs.

[jwkaltz]

@voisardf @michmuel do you also want to review this upgrade?
To us it seems OK.
Tests are OK, extract generation with the dev data also.

[michmuel]

@jwkaltz I will test it in our environment.

[michmuel]

I made some tests in our environment and did not encounter any differences to version 1.8. I remember (or think to) that there was a discussion about shapely at the beginning of working on pyramid oereb v2. Was this related to shapely and python version 3.7 or was it related to the handling of not valid geometries in different versions of shapely?

[jwkaltz]

I made some tests in our environment and did not encounter any differences to version 1.8. I remember (or think to) that there was a discussion about shapely at the beginning of working on pyramid oereb v2. Was this related to shapely and python version 3.7 or was it related to the handling of not valid geometries in different versions of shapely?

Yes, there were some differences in behaviour for older versions, summarized here: #1111
But regarding invalid geometries, there seems to be no differences in behavior between 1.8.5 and 2.0.0.

[michmuel]

Then I have no objections. Thanks.