Remove tsk_diff_iter_t

[benjeffery]

Fixes #2797

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 89.61%. Comparing base (47572c1) to head (be8a0b2).
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3221      +/-   ##
==========================================
- Coverage   89.63%   89.61%   -0.03%     
==========================================
  Files          28       28              
  Lines       32004    31893     -111     
  Branches     5879     5870       -9     
==========================================
- Hits        28688    28581     -107     
+ Misses       1886     1884       -2     
+ Partials     1430     1428       -2     

Flag	Coverage Δ
c-tests	86.59% <ø> (-0.07%)	⬇️
lwt-tests	80.38% <ø> (ø)
python-c-tests	88.18% <ø> (ø)
python-tests	98.86% <ø> (ø)

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

Files with missing lines	Coverage Δ
c/tskit/tables.c	83.02% <ø> (-0.19%)	⬇️
c/tskit/trees.c	90.61% <ø> (-0.01%)	⬇️

🚀 New features to boost your workflow:

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

[jeromekelleher]

I don't think we ever documented this, so I'm happy with ditching it.

@molpopgen is the only person I can think of that this would potentially affect. Are you happy for us to ditch this? The functionality is replaced by the much more efficient and flexible tree_iter stuff.

[benjeffery]

@molpopgen Said it was ok at #2797 (comment)

[molpopgen]

It is fine for removal -- but even w/the tree iter stuff, isn't it often much more efficient to use edge differences for doing various calculations? (That has always been my experience.)

Also -- what tree iter stuff? (Sorry if that question seems odd, but I find a lot of the changes on the C side to be a bit opaque/undescribed.)

[jeromekelleher]

It's based on the "tree position" class that we've been working towards as a way of encapsulating the state of along the genome traversals in a better way. Some pointers:

• Ideas fleshed out and tested here in Python: https://github.com/tskit-dev/tskit/blob/main/python/tests/test_tree_positioning.py
• Implementation here:

  tskit/python/tests/tsutil.py

  Line 1739 in da094c4

  class TreePosition:

• C implementation here:

  tskit/c/tskit/trees.c

  Line 5150 in da094c4

  /* ======================================================== *

This is also the basis of the Numba interface we'll be adding (#3225 )