Repeated names

[mcabbott]

Maybe it's time to define & test the behaviour with repeated indices. This fixes:

• sum(A, dims=:i) to use the first dimension named :i, as it already did, and
• A[i=1] is an error if more than one dimension is named :i. Previously it fixed all such dimensions.

The one exception I can think of is that, for a Diagonal matrix, it may make sense to index both dimensions together. But maybe I have overlooked other uses?

The indexing change did not break any tests. But may still count as a breaking change, not sure.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 93.76%. Comparing base (1ef0b73) to head (1d3f4c7).
⚠️ Report is 145 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #148      +/-   ##
==========================================
+ Coverage   93.71%   93.76%   +0.04%     
==========================================
  Files          10       10              
  Lines         382      385       +3     
==========================================
+ Hits          358      361       +3     
  Misses         24       24              

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[oxinabox]

I know @nickrobinson251 likes for covarience matrixes that indexing with just the keyword gives you on the diagonal.
Not enough to use undefined behavour, but he thinks it is cute