Small fixes

[foxeng]

The majority is just about typos / formatting / rough edges.
Commit 0a71dbf undoes part of what 477d5cc introduced, where UmlDocTest would exit (successfully) before comparing the doclet output to the reference. This unfortunately masked some failing tests (e.g. testdata/umldoc-ref/gr/spinellis/invoice/gr.spinellis.invoice.dot), where a class relation multiplicity does not appear. These tests now fail.

[foxeng]

As it turns out, the multiplicity was indeed included in the generated diagrams, but there is an overlap between the labels and the arrows in the generated diagrams, which makes it very hard to discern (this seems to be a difficult issue to address, UmlGraph can only try to work around it as it is supposed to be taken care of by GraphViz). I manually checked the test output and everything seems fine, so I updated the reference files in the test suite.

[dspinellis]

Thanks! We still seem to have some failing tests. Can you please look into it?

BTW In the future we'd like to have the tests be less brittle.

[foxeng]

I 've spent some time trying to figure it out, but I haven't marked any progress yet because I can't seem to reproduce the failure: the tests succeed on all the setups I 've tried and unfortunately travis' log indicates a crash, but no relevant output (just an exit with a non-zero return code when running UmlGraphDoc).

I 'd be grateful if you could test it as well since I might be missing something here. Thank you for your time!

[dspinellis]

Sadly, I can't test it locally, until we resolve #57.