WebWalker keeps on walking
I coded up a little automated web testing tool, WebWalker, just under 2 years ago and released it on Sourceforge. It’s the only tool I’ve actually completed well enough for others to use with documentation. Most of my stuff I code for myself and I know all the little spots to kick when something isn’t working right.
I’d forgotten all about it until I just now found a bookmark to it. Surprisingly, it’s had over 3000 visits to the page, with over 300 downloads. I’m impressed. I’ve actually received some emails for feature additions in the past, but never had the time to add to it, nor had the need to continue using it.
My code is really just a wrapper around HTTPUnit allowing non-programmers to write web tests to walk a website without having to program. They could write simple xml to do the dirty work. WebWalker interprets the XML and then changes that into code HTTPUnit can understand. I’d tried using HTTPUnit at a previous job, but it didn’t make sense to write a Java class for a test. If a webpage changed, you’d have to recompile your code to rerun the tests. And it limited the test writers to coders. Many QA departments don’t have full coders, so WebWalker would abstract the test with a layer of XML that was much easier to update. No recompiling required. It’s command line scriptable as well, so it makes a great tool for nightly tests.
The kinds of “walking” it can do right now is limited to opening URLS, clicking links, submitting forms & some verification (word checking, etc). If there’s interest in this, maybe I’ll start it up again. Might be a good pet project. Leave some comments if you’ve got feature requests.