Extract text and some metainfo from HTML, coping with malformed pages as well as possible.
This package was written for a search engine, to allow it to extract the textual content and metadata from HTML pages. It tries to cope with invalid markup and incorrectly specified character sets, and strips out HTML tags (splitting words at tags appropriately). It also discards the contents of script tags and style tags.
As well as text from the body of the page, it extracts the page title, and the content of meta description and keyword tags. It also parses meta robots tags to determine whether the page should be indexed.
The HTML parser used by this module was extracted from the Xapian search engine library (and specifically, from the omindex indexing utility in that library).