I successfully establish a WS as a part of the product. Thus ohter language-based site also can use the component, without strange feagure like link from .asp to .jsp. I have used c# to invoke the method and successfully got the rearch result.
And, i modified the default servlet and search page(.jsp), make it more powerful to suit all kinds of search type. But still need to be modified better.
My partner have found a way to make the search process interrupted and restart. Since we use lucene, ithe constructor of indexwrite can recieve a parameter to resume the index instead of overwrite. and i just need to save some other collect data. The function is possibly done soon.
Still, some problem remains:
like:
the encode problem. we parse the index page of SCDN. the charset is UTF-8, anyway, i just got mass code. but no problem in gb2312 and iso-8859-1 page,i may have some further research.Aonther problem is due to lucene. Since we use StandardAnalyze class to analyze the content and user's input query, but i found when i input such as C# , C++, it just search as C. The performance of collect still need to be improved
