Added survey feedback template and method for returning the templated feedback for a given response
Used this to clean up the default survey template, and updated the i18n and help
refactored out repetitive checking of takenCount
added i18n template variables to Survey template so that user is told what happened for timeouts etc..
Change the Ship, Cart to do exception handling when calling a shipping driver,
as that's the message passing interface between the modules.
Add a noFormProcess field property that is recognized by processPropertiesFromFormPost.
Refactored www_submitQuestions to use utility submitQuestions sub (for ease of testing)
Replaced numeric flags, completedCode etc..
Added restart() and exitUrl() to the Expression Engine
Added more Survey instance tests
Does nothing if GraphViz module not installed (the default until wre 0.9.3)
Shows sections and questions as nodes, answers as edges.
Branch targets also show up as edges
Branch expression edges are currently disabled
Survey Expression Engine now detects invalid variable names passed to value(), score(), etc..
Also now detects situations where you have jump targets/expressions defined at multiple levels - thus causing precedence rules to kick in (normally this indicates a mistake)
Fixed Survey edit page bug where TextEditor would not move after gotoExpression textarea resize caused items to move
Now stores all supplied answers to multi-answer question as arrayref
Variable [[var]] text replacement converts to comma-separated list
Expression Engine's value() returns appropriate value depending on context
Expression Engine's score() does sum of multi-answers, as you would expect
Updated related i18n & hover-help, and tests
Cleaned up survey js somewhat
it in the response data structure along with the actual user input. Tag data
can be used in subsequent expressions, in [[tag]] templated text replacement,
and to classify responses in an arbitrary way.
Refactored Survey expression utility subs that lookup values/scores/tags in
external assets.
Fixed bug whereby only highest precedence Survey expression was being evaluated
rather than letting them all run and do their own short-circuiting.
