1. Index on book item's attributes: is category or book, [belong to which category(name and id), id, used to identified which? ]name, file extension, size(will not be implemented)
2. Search book by category, by extension, by name,

1. exact each book item from 2000 *.xml files
parse the first tag to get the belonged category for its children(search the category object from categories map), parse the following tag, if is category add to categories map, if created related book tree,
binary traverse the tree and filter those limitted resource, and binary traverse the tree again the add book items.
2. add the book item as document to lucene
3. search, and result hashed result links(use the redirect?*&*?), belonged category, name, extension