Reported by project member bodfishj@oclc.org, May 2, 2012
Currently the dummy connector uses a "Dummy database" (which is actually a Java class) to contain the data it uses to respond to requests. If we instead used a mock database, for which there are several open-source choices (http://www.dbunit.org/, http://hsqldb.org/) where it behaves like a JDBC database but reloads every time the application starts and reads its initial data from XML files. This would serve several purposes:
1) Create test data for test-driven development.
2) Easier for developers to create test data for unit testing.
3) Simple mechanism to "refresh" the database to a known state.
4) Those who currently want to connect their NCIP clients to an implementation of the Toolkit can do so with the Dummy connector and operate independently of our group having to maintain a test instance of some ILS.
JIRA issue created by: rcook Originally opened: 2012-06-27 12:05 PM
Issue body:
GC issue http://code.google.com/p/xcncip2toolkit/issues/detail?id=50
Reported by project member bodfishj@oclc.org, May 2, 2012
Currently the dummy connector uses a "Dummy database" (which is actually a Java class) to contain the data it uses to respond to requests. If we instead used a mock database, for which there are several open-source choices (http://www.dbunit.org/, http://hsqldb.org/) where it behaves like a JDBC database but reloads every time the application starts and reads its initial data from XML files. This would serve several purposes: 1) Create test data for test-driven development. 2) Easier for developers to create test data for unit testing. 3) Simple mechanism to "refresh" the database to a known state. 4) Those who currently want to connect their NCIP clients to an implementation of the Toolkit can do so with the Dummy connector and operate independently of our group having to maintain a test instance of some ILS.