On Sun, Jul 5, 2009 at 4:04 PM, Bo Yang<[email protected]> wrote: > Hi, > > After three days testing, debugging, bug fixing, I think it is > better to report some status of the libDOM now. In the past three > days, I have spent all of my time on testing the libDOM implementation > with level 1 test cases. Now, the result is as: > > Total: 527 > Passed: 400 > Failed: 127 > Conversion Error: 2 > Compile Error: 19 > Run Error: 73 > Not Support: 33 > > Now, let me explain the above result. > > 1. Now, we did not support all the DOM features for three level. We > did not support feature such as Notation, Entity, EntityRef, DTD and > all other XML featues. So, there are 33 "Not Support" errors, and > because of this kind of not support we got about another 70 Run Error. > 2. Conversion Errors mean that we have some test case can't convert. I > have examine , the two conversion error is rooted at the element <try> > which I am not intending to deal with now. > 3. Compile errors means we did not provide the corresponding interface > now. Such as dom_notation, dom_entity, and element.normalize. > > And the test is done with the libxml2 binding, so there are many XML > related features get into the test cases. But we did not support all > of them now, that is the reason why there are almost 100 failed > testcase. If we use the hubbub bindings and test only HTML feature, I > think we can get nearlly 100% pass.
Sorry, I forget another important reason of failed test cases. That is the I did not deal with the invalid character exception, and this cause almost 20 Run Error. After this get done, I think we can reach neally 100% then. :) Regards! Bo
