tests/test-hg-parseurl.py.out

tests: make a stab at approximating wall-clock times

Vaguely empirical observations:

* ".py" tests are about an order of magnitude faster than ".t" tests

* dividing size by 1000 gives an approximation to wall-clock
  run time (in seconds) that is not completely ridiculous.

http://example.com/no/anchor, branches: (None, [])
http://example.com/an/anchor, branches: ('foo', [])
http://example.com/no/anchor/branches, branches: (None, ['foo'])
http://example.com/an/anchor/branches, branches: ('bar', ['foo'])
http://example.com/an/anchor/branches-None, branches: ('foo', [])
http://example.com/, branches: (None, [])
http://example.com/, branches: (None, [])
http://example.com/, branches: ('foo', [])