closes #399: on-the-fly calculation of checksum #400
Conversation
|
Why do we need the md5 checksum of a page content ? What use can be done with it ? |
If a crawler's visit algorithm performs expensive operations on pages and then stores only the extracted information it may be usefull to have a common checksum storage where they can check if an identical page has already been processed and ignore them in the future, for instance. The problem is that every page with a non-js clock, visit counter, a tiny little PHP dinamic banner or additional whitespace printed God knows why would break the equality. I'm currently using a more html-driven solution by checking a specific tag that contains a field that I know can be taken as a "primary key" for the website, but I've to parse the content into a jsoup Document first, so it is probably more expensive and requires me to exactly know the exact layout of crawled pages (which I know because I'm crawling information, not just html documents) |
No description provided.