| 12345678910111213141516171819202122232425262728293031323334353637383940 |
- This is a test system using OASIS XSLT test suite
- http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=xslt
- "make" will do what you need (downloading test archive, expanding,
- compiling test runner.
- "make run-test" will run only clean tests (check for regressions).
- "make run-failed-test" will run only failed tests and inverse the
- test results (check for progressions).
- "make run-all-test" will run all tests.
- knownFailures.lst and fixme.lst have the list of tests that are
- are considered known failures for run-test and run-failed-test.
- If you want to create reference output files by yourself,
- run "make create-reference-output" on Microsoft.NET (it requires cygwin).
- Both generation and test phases read the following environment variables:
- XSLTTEST_DOMXSL - use XmlDocument for xsl
- XSLTTEST_DOMINSTANCE - use XmlDocument for xml
- XSLTTEST_DOM - both of the above
- XSLTTEST_WSXSL - significant whitespace in xsl
- XSLTTEST_WSSRC - significant whitespace in xml
- XSLTTEST_WS - both of the above
- TODO:
- This testsuite should not generate dotnet results. Instead, one should compare
- to oasis reference output. For testcases that fail on dotnet, one can skip
- checks, as it's done in W3C testsuite. This will allow to run negative tests,
- too.
- But before that, it's necessary to fix the comparison, instead of simple text
- compare, it should make compare as specified in each testcase - text, xml or
- html. There are open source tools to make xml compare, and html can be done
- by first converting html to xhtml and then running xml compare.
- Atsushi Eno <[email protected]>
- Andrew Skiba <[email protected]>
|