What HTML parsers have the following features:
- Reliable and bug-free
- Parses HTML and XML
- Handles erroneous HTML
- Has a DOM implementation
- Relatively simple, object-oriented API
What parser you think is better?
Check out Web Harvest. It’s both a library you can use and a data extraction tool, which sounds to me that’s exactly what you want to do. You create XML script files to instruct the scraper how to extract the information you need and from where. The provided GUI is very useful to quickly test the scripts.
Check out the project’s samples page to see if it’s a good fit for what you are trying to do.
JTidy is more intented into formatting your html code into something XML-valid, but is still very useful as an XML parser, producing a DOM tree if needed.
You could have a look at this list for other alternatives.
Another choice could be to use hpricot through jRuby.
Validator.nu’s HTML parser, definitely. It’s an implementation of the HTML5 parsing algorithm, and Gecko is in the process of replacing its own HTML parser with a C++ translation of this one.
Apache Tika is the best choice. Apache has recently extracted many sub-projects out of the existing projects and made them public. Tika is one of them that was previously a component of Apache Lucene. Because of Apache’s support and reputation and the widely-used parent project Lucene it must be a very good choice. Furthermore, it is open-source.
A brief introduction from Apache Tika web site:
The Apache Tika™ toolkit detects and extracts metadata and structured text content from various documents using existing parser libraries.
And the supported formats are:
HyperText Markup Language XML and derived formats Microsoft Office document formats OpenDocument Format Portable Document Format Electronic Publication Format Rich Text Format Compression and packaging formats Text formats Audio formats Image formats Video formats Java class files and archives The mbox format
there aren’t so many good HTML parsers in java as you need, but here are some alternatives:
you probably want to look at doing something like running Mozilla in headless mode. Here is a link to get you started, I am sure you can use Google to find out more information.