This is a poor but better-than-nothing unit test for the XSL files. I use Python and its unittest module to do this.

A typical test is:

def test_id_numbers(self):
    self.confirm_content(["regno", "appno"], ["rn2178784", "rn2713476"])

This will check two fields ("regno" and "appno", i.e. the registration number and application number) against XML files for two trademarks ("rn2178784" and "rn2713476").

In order to support this, the testdata directory must contain one or more XML files for each of the listed trademarks, named in the form TM-FORMAT.xml, where TM corresponds to the same trademark number as in the list, and FORMAT is the XML format ("ST66" or "ST96"). In the above example, there are four XML files:

• rn2178784-ST66.xml
• rn2178784-ST96.xml
• rn2713476-ST66.xml
• rn2713476-ST96.xml

In addition, the trademarks forming the testbed are kept in the test_tms list in the unittest setUp method. (maybe someday I'll make this self-discovering, but it's probably not worth the effort.)

For each of the fields, the expected directory must contain a file, named in the form FIELD-TM.txt that contains the text expected to be found for that field. For example, for the above example, with two fields being checked against two trademarks, there are four (2 × 2) files, one for each field-trademark combination:

• appno-rn2178784.txt
• appno-rn2713476.txt
• regno-rn2178784.txt
• regno-rn2713476.txt

Each of the files contains the text expected to be found in the XSL-transformed CSV file. For example, regno-rn2178784.txt contains the string RegistrationNumber,"2178784".

A test is deemed passed when the string from the applicable examples file is found in the specified CSV file.

At the end of the test, the CSV files, as well as the standard output and standard error from each XSL transform, are kept in the working directory, which is destroyed and recreated each time the test is run.

This is, as noted above, a pretty simplistic test, and there's a limit to how much it can really do, but it is better than nothing.