com.purplehillsbooks.testcase

Class TestJSONSchema

java.lang.Object

com.purplehillsbooks.testcase.TestAbstract

com.purplehillsbooks.testcase.TestJSONSchema

All Implemented Interfaces:

TestSet

public class TestJSONSchema
extends TestAbstract
implements TestSet

Constructor Summary

Constructors

Constructor and Description
TestJSONSchema()

Method Summary

Modifier and Type	Method and Description
static void	main(java.lang.String[] args)
void	runTests(TestRecorder newTr) The only method that must be implemented is "runTests".
void	schemaTests(java.lang.String testId, java.lang.String coreName)

Methods inherited from class com.purplehillsbooks.testcase.TestAbstract

compareGeneratedTextFile, compareStringIgnoringCR, initForTests, testNotNull, testNull, testOutput, testRawXML, testScalar, testVal, writeListToFile, writeLiteralValue, writeShortLiteralValue

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

TestJSONSchema

public TestJSONSchema()

Method Detail

runTests

public void runTests(TestRecorder newTr)
              throws java.lang.Exception

Description copied from interface: TestSet

The only method that must be implemented is "runTests". The tests must initialize, run, and cleanup within this single call.

This method can include any number of tests. It is expected that you will have to get set up in a particular configuration, and then run many individual test to assure the program is running correctly.

A test is allowed to do anything you can do in java. Clearly you can write your own subroutines and use them anyway you need. A single TestSet class can actually be any number of individual tests; you simple call the markPassed or markFailed as apropriate for each test. You should strive to design tests so that the number of tests remain constant. One run of the tests should not have 5 results, and a different run the next day have 31 results. If there are 31 tests, they all should produce a result every time. If some tests run only in certain configurations, then when skipping the test, you should record a result of "passed" and include in the id/description that the test was skipped. Try to avoid tests that read the database and report a success for each row in a table because that might vary in number. Instead, read and check every row, and then record a single result so that the number remains constant.

Specified by:

runTests in interface TestSet

Parameters:

newTr - Each test can record success or failure with the TestRecorder. The TestRecorder is the only way that the test should contact the outside world (except the software it is testing). All configuration information should be gotten from the config file support or the command line parameter support. All output should be to the TestRecorder which offers the ability to record that a test passed, that it failed, and all additional debug log output. Where this output goes, and how it is formatted depends upon the test framework that this is embedded in. For example, this test might be run as part of a JUnit suite, and then the output will go to the normal JUnit log. Or it might be part of anything else. But the test must only use the TestRecorder to get config info, and to write results.

Throws:

java.lang.Exception - if an error occurs for which you can not properly continue testing, the test set should throw an exception. This is the proper way to indicate that testing was stopped before all of the tests were run. The exception will be recorded as a fatal error by the test framework. An example of such a fatal error is unable to log into the server ... if this happens there is little point in trying further tests, so throw an exception.

schemaTests

public void schemaTests(java.lang.String testId,
                        java.lang.String coreName)
                 throws java.lang.Exception

Throws:

java.lang.Exception

main

public static void main(java.lang.String[] args)