remove SLiM metadata; closes #2534

[petrelharp]

This just rips out all metadata stuff from the SLiMMutationModel. We should not merge this until SLiM v6 is released.

Potential breakage: if someone runs

msprime.SLiMMutationModel(type=1, next_id=1, slim_generation=10)

then they get a warning saying that type and slim_generation will be ignored.
If someone runs

msprime.SLiMMutationModel(1, 1, 10)

or

msprime.SLiMMutationModel(1, 10)

then they get a TypeError because the only valid positional argument is the first one.
The latter is necessary to exclude because previously if someone ran this they'd get a model with type 1 and next ID 10; now (if both arguments were positional) they'd get a model with next ID 1 and block_size 10.

One possible problem: previously, the only required argument was mutation type, and it was the first argument. So, if someone ran

msprime.SLiMMutationModel(k)

then they would get a model with mutation type k and next ID 0. Now, they would get a model with next ID k. This will cause no practical problems - both "mutation type" and "next ID" have to be a nonnegative integers, and so this just means msprime will not simulate mutations with SLiM ID 0, ..., k-1 when you might have thought it would.

The alternative to this PR would be to just leave everything alone: msprime puts in unneeded metadata, and we can just safely ignore it downstream. But, this is confusing.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 93.84%. Comparing base (79dfc96) to head (aad986c).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2535   +/-   ##
=======================================
  Coverage   93.83%   93.84%           
=======================================
  Files          20       20           
  Lines       12106    12060   -46     
  Branches     2243     2242    -1     
=======================================
- Hits        11360    11318   -42     
+ Misses        567      564    -3     
+ Partials      179      178    -1     

Flag	Coverage Δ
C	84.07% <100.00%> (-0.04%)	⬇️
c-python	73.11% <100.00%> (+0.21%)	⬆️
python-tests	98.58% <100.00%> (+<0.01%)	⬆️

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

Components	Coverage Δ
Python API	98.61% <100.00%> (+<0.01%)	⬆️
Python C interface	92.84% <100.00%> (-0.02%)	⬇️
C library	91.05% <100.00%> (-0.01%)	⬇️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[jeromekelleher]

LGTM, spotted on trivial thing.

Did we not set a metadata schema for the slim mutation metadata?

I don't think we need to wait until SLiM 6 is released to merge, but we shouldn't release msprime until it is, though.

Might be worth noting that this also means that these mutations won't work with SLiM < 6 now (this isn't obvious from the CHANGELOG).

[jeromekelleher]

This set_columns is redundant now, right?

[petrelharp]

Did we not set a metadata schema for the slim mutation metadata?

Huh, yeah, I guess not.

[petrelharp]

This should be ready now!