We're making our UI framework ready to work in IE11+JAWS17 using the WAI-ARIA standard (not 508 compliant, which we supported for like a decade). As this is a huge task we'd like to write automated tests to be safe against regressions.
Therefore we search for a way to write assertions on the speech output of JAWS (after a specific action). E.g. tab to a button => assert that JAWS reads "Button, Text is 'My Button', ...".
We found no direct way to integrate JAWS into e.g. selenium tests. As this topic is rather important for us we'd be willing to develop something ourselves, but everything we found looked discouraging:
jaws_executable /z"logfile.log"writes the speech output tologfile.logbut locks the file while JAWS is open, so we'd have to close JAWS whenever we need new speech output (which would be like once every action our test executes)- JAWS Speech History Mode writes the speech output to a separate window, and Freedom Scientific advises to copy the output and email it to the web developers when testing.
- Testing With JAWS advises to test manually up-front.
Is there a way to get the JAWS speech output programmatically? For a single "sentence", not for the whole JAWS session?
I could think of
- "logfile.log" somehow not being locked while JAWS writes into it
jaws_executable /znot writing to a file but to e.g. stdout- JAWS Speech History Mode not writing to the separate window but to e.g. stdout
Any hint highly appreciated!
