Make parsing of translate tables in tree blocks more robust

[SimonGreenhill]

This PR fixes a bug where trees with a translation block ignored taxa with asterisks in their name. These taxa labels would be silently ignored, i.e.

#nexus
begin trees;
    translate
        1 Simon,
        2 *Russell,
        3 Robert,
        tree tree = (1,2,3);
end;

..would mean that the taxon "*Russell" would be lost, and the tree would look like (Simon,2,Robert).

I believe that asterisks are invalid in taxa names anyway, but they are found in the wild (e.g. Grollemund's Bantu dataset).

It might be better to fix the bigger issue -- the translate table should not be able to have a different number of taxa in it to those in the trees. But I could not see an easy way to enforce this.

[codecov-commenter]

Codecov Report

Merging #31 (2623c33) into master (296c521) will not change coverage. The diff coverage is 100.00%.

@@            Coverage Diff            @@
##            master       #31   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           59        59           
  Lines         2758      2774   +16     
=========================================
+ Hits          2758      2774   +16     

Impacted Files	Coverage Δ
src/nexus/exceptions.py	100.00% <100.00%> (ø)
src/nexus/handlers/tree.py	100.00% <100.00%> (ø)
tests/test_cli.py	100.00% <100.00%> (ø)
tests/test_handler_TreeHandler.py	100.00% <100.00%> (ø)
tests/test_regressions.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 296c521...2623c33. Read the comment docs.

[SimonGreenhill]

ahh, figured out a way to check for this. We now raise an exception in detranslate if the Translate table doesn't match the tree.

[SimonGreenhill]

...and this popped up another bug. If there's no translate section in the trees block, we parse the first tree to identify taxa names. However this can be messed up with comments in the tree like those found here, e.g.:

...(hax[&length_mean=0.1146902001157983,height_median=2.775557561562891E-17,prob(percent)="100"...

where "percent" inside the comment is identified as an extra taxa. The PR now strips comments from the tree string before doing this, which is a little brute force but the comment info isn't needed here.

[xrotwang]

...(hax[&length_mean=0.1146902001157983,height_median=2.775557561562891E-17,prob(percent)="100"...

where "percent" inside the comment is identified as an extra taxa. The PR now strips comments from the tree string before doing this, which is a little brute force but the comment info isn't needed here.

Wow. I would have expected that braces in comments would crash the newick parsing hard. But apparently it's only square brackets that can't appear in comments ... I guess, comments in newick in Nexus is just another PHP ... and we are writing the parser ...

[xrotwang]

Ah, it does https://github.com/phlorest/sicoli_and_holton2014/blob/cd41d90973b8838879b3bcbba79c07051081745c/cldfbench_sicoli_and_holton2014.py#L7-L9 Makes me sort-of happy.