Test Specification Reference

The test specification grammar provides a very flexible way to call out which tests to run. The basic structure is this:

specification ::= unary | binary | compound

unary ::= tags | suite | classMethods | properties

compound ::= '(' binary ')'

tags ::= '<' tagName (',' (all | success | fail | error | failError)? '>'

suite ::= '[' suiteName (',' suiteName)* ']'

properties ::= '{' propertyName ':' singleQuotedValue (',' propertyName ':' singleQuotedValue)* '}'

classMethods ::= (('%' packageName)? ('@' className) | qualifiedName) ('#' methodName)? ('$' operationName)?

Probably the best way to explain this is with some examples:

Specification	Result
@classname	Run every method in every class with 'classname' in the name.
%pack	Run every method in every class with 'pack' in the package name
@classname and %pack	Run every method which matches both of the previous conditions. This is a compound statement.
$operation	Run every class and method, but only execute operations named 'operation'
#method	Matches just the method name
%pack1@class1#method1$operation1	Selects packages containing 'pack1', classes named 'class1', methods named 'method1', and only executes operations named 'operation1'
<tag>	Run the tag named 'tag' using the 'all' set.
<tag, fail>	Run the tag named 'tag' using the 'fail' set.
[ aud ]	Run every method in every class with the @JoinSuite tag with the name aud.
{ property: 'value' }	Run every method in all classes which contain an operation with a parameter named 'property' and the value 'value'

All of the selectors have the effect of choosing any test to run that they match. If an operation is specified (using $), only those operations are executed. The properties selector will select the entire method if a single operation matches.