I'm new to .net and C# but I have been able to use your library with very little effort and it gives me everything I need for my project. Thanks very much!
Now I am checking the code to see if it will be robust in a heavy-traffic server environment. And I just have a couple of questions.
My stress tests show that when parsing an HTML doc 100K times or more that no RAM is leaked, but average time to process the doc decreases for a while before levelling out. I am a bit puzzled as to why the avg. parsing time increases but I can live with that
for now. (It seems to be related to XPath usage, but I haven't spent much time looking into exactly where its slowing down.)
Anyway, I noticed in the source that the collections are not generics, and I understand that HAP was originally written before generics were available. However, now that we have them, I'm wondering is there any plan to convert the strongly-typed collections
over to generics?
I've also run FxCop program over the HAP and it raises a number of issues, some which may be important. Have you ever looked at conforming the code to the .net standards a la FxCop?