Refactor TensorNetwork internals to incidence matrix representation

[mofeing]

This PR refactors the TensorNetwork internals to be more like a incidence matrix.

[codecov]

Codecov Report

Attention: 11 lines in your changes are missing coverage. Please review.

Comparison is base (3424964) 89.73% compared to head (80181a2) 87.76%.

❗ Current head 80181a2 differs from pull request most recent head a2079e6. Consider uploading reports for the commit a2079e6 to get more accurate results

Additional details and impacted files

@@             Coverage Diff             @@
##           develop     #120      +/-   ##
===========================================
- Coverage    89.73%   87.76%   -1.97%     
===========================================
  Files           10       10              
  Lines          604      564      -40     
===========================================
- Hits           542      495      -47     
- Misses          62       69       +7     

Files	Coverage Δ
ext/TenetChainRulesTestUtilsExt.jl	100.00% <100.00%> (ø)
ext/TenetFiniteDifferencesExt.jl	100.00% <100.00%> (ø)
ext/TenetMakieExt.jl	100.00% <100.00%> (ø)
src/Transformations.jl	98.41% <100.00%> (+0.91%)	⬆️
ext/TenetChainRulesCoreExt.jl	75.00% <50.00%> (-4.32%)	⬇️
src/TensorNetwork.jl	74.52% <90.76%> (-5.35%)	⬇️

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

[mofeing]

@jofrevalles Makie tests are now failing because now I return tensors(tn) sorted by their inds (Look at the new definition of Tenet.tensors).

Would you mind looking at it?