BF: Fix regression in continueRoutine behaviour

[m-macaskill]

This addresses a regression in version 3.2.0 in which setting continueRoutine = False no longer has an effect in Builder code components (because the Builder-generated code overrides the custom code by always resetting continueRoutine = True immediately before the routine begins). All this pull request does its shift continueRoutine = True to be first rather than last in the order of execution prior to a routine, so that custom code can override it.

For recent bug reports, see:
https://discourse.psychopy.org/t/continueroutine-false-in-begin-routine-tab-has-no-effect/9801

and for previous development history, see:
#1143

[coveralls]

Coverage decreased (-0.02%) to 47.974% when pulling 350c71a on m-macaskill:controutine into ac8d34e on psychopy:master.

[m-macaskill]

Unclear to me why this is failing on just one of the five Travis Build jobs, and on an unrelated test (of test_StaticPeriod()).

[codecov-io]

Codecov Report

Merging #2680 into master will decrease coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #2680      +/-   ##
==========================================
- Coverage   43.47%   43.47%   -0.01%     
==========================================
  Files         236      236              
  Lines       43691    43693       +2     
  Branches     7496     7496              
==========================================
  Hits        18994    18994              
- Misses      22691    22696       +5     
+ Partials     2006     2003       -3

Impacted Files	Coverage Δ
psychopy/experiment/routine.py	87.65% <100%> (+0.1%)	⬆️
psychopy/visual/textbox/fontmanager.py	71.08% <0%> (-0.7%)	⬇️
psychopy/visual/textbox/__init__.py	43.5% <0%> (ø)	⬆️
psychopy/app/builder/dialogs/dlgsConditions.py	7.51% <0%> (ø)	⬆️
...ychopy/localization/generateTranslationTemplate.py	0% <0%> (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 ac8d34e...350c71a. Read the comment docs.

[peircej]

Thanks Mike. I thnk the Travis error was a random dropout