|
I should have also noted that the Java client driver version needs to be modified to parse the file before executing, rather than executing as it parses the file as it does now. There are probably other problems with it too, but it should provide a decent start. In terms of my thinking on where this could go from here:
Ok, here's an update. Not all tests pass now, but I'm ok with the reasons for the failures, as some don't pass in Selenium IDE anyway, and in general they're things that won't pass via selenium RC/Grid Changes:
Not done:
I'm also interested, I'd like to run the html tests along with the java tests and display aggregated results in the form of junit test reports Daiji, could you upload the files with a license header to clarify that your work is open source before I make any modifications? I'm also interested in running html tests on Selenium Grid. Thanks, Marcelo I'm also interested in running html tests on Selenium Grid. Thanks, Rich This is nice-to-have for 1.0. But it will not ship with 1.0. |
|||||||||||||||||||||||||||||||||||
This isn't structured properly (several classes in a single file, no proper property encapsulation, no Javadoc, no unit tests) but was written this way to simplify the dev environment requirements and followed on my thinking to producing a fully-functional version that people can actually use right away. I've also attached the scripts I used to build and run. Here is the command I used to run the full test suite against Opera:
./selenium-java.sh --browser "*opera" --testsuite ../../selenium-core-0.8.3/tests/TestSuite.html --baseurl http://mymachine/selenium-core/tests/
--out results.html
The NekoHTML parser is required in order to properly parse HTML4 syntax (used by the bulk of the Selenium Core testsuite), but on the other hand, the standard XML parser is required to properly parse XHTML syntax (produced by the Selenium IDE).
I also added support for a number of functions that exist in Core but not in the IDE.