[jbosstools-issues] [JBoss JIRA] Created: (JBIDE-5675) Increase width and depth of Smooks automated testing

[Brian Fitzpatrick (JIRA)]

Increase width and depth of Smooks automated testing
----------------------------------------------------

                 Key: JBIDE-5675
                 URL: https://jira.jboss.org/jira/browse/JBIDE-5675
             Project: Tools (JBoss Tools)
          Issue Type: Task
          Components: smooks
            Reporter: Brian Fitzpatrick
             Fix For: LATER


Now that we have started putting together some automated tests, it's made me think again about what it is we should be testing...

At a high level, we have some basic functional areas... Setting the input data, mapping to some Java Classes, and then taking the data from the Java Classes and outputting to some other format (such as CSV or XML (XSD)).

We take input from:
* XML/XHTML files
* Incoming Java classes (meeting JavaBean-type interfaces)
* XSD/WSDL files
* CSV files
* EDI files
* JSON files
* Custom files (what requirements do custom Smooks inputs have?)

So we need to test to make sure that a variety of these input types can be successfully brought into the editor without errors. But since we only really support a limited set of inputs, this should remain pretty stable if we come up with a few solid examples of each. Tests should include edge cases such as Java Classes that don't match the Java Bean structure, invalid files, empty files, etc.

The next area is then the Java Mapping task, where we provide a facility for users to import Java classes to map their input to. These java classes need to meet the Java Bean-type structure. 

This Java Mapping should be tested against:

* Java Classes that don't match the Java Bean-type structure
* Simple classes (single class vs. collection)
* Collection of simple classes
* Collection of compound classes

And the final area at this point is applying a template - either CSV or XML (XSD). This mapping between the Java Mapping task and the output template should mix/match a number of things, including but not limited to:

* (Same things we tested the Java Mapping against)
* Java Classes that don't match the Java Bean-type structure
* Simple classes (single class vs. collection)
* Collection of simple classes
* Collection of compound classes

Plus:

* Single-field CSV (field1)
* Multi-field CSV (field1,field2,field3,...)
* Simple XML (XSD) - i.e. XSD with only simple types
* Complex XSD - multiple complex types
* Invalid XSD
* Blank XSD

Generally I just want us to try and automate as many tests as we can. Dart, Tom & I can help you with this process once we get through the release. 

Ultimately we need to think of as many weird or strange things as we can to start hitting fringe cases that our users may try. Users are talented in the ways they will try to make tools work. And I have to admit I've used the occasional screwdriver to pound in a nail (typically unsuccessfully) so we should anticipate the user will try and do the same thing with the Smooks editor. 

The better we build our test suite, the better off we'll be as we add additional template types, new functionality, and continue to bang on this thing.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira