Parsing html with SAX parser

16,349

Solution 1

As flash says, you need an HTML parser, not an XML parser. HTML is not XML.

Good parsers i've used are Neko and TagSoup. Neko is a good all-round parser; TagSoup specifically aims to be able to parse anything, no matter how ill-formed.

Solution 2

Generally speaking, you cannot parse HTML with an XML parser:

  • HTML's element tags are not required to match in all cases. (For example a <p> tag does not require a matching </p> tag.) This will cause terminal indigestion for an XML parser.

  • Real-world HTML is notorious for not being conformant to the HTML spec, let alone an XML compatible subset of HTML.

However, if your input document is XHTML, you should in theory be able to use an XML parser such as SAX. You should even be able to validate the document against the XHTML schema.

Solution 3

Please have a look at HtmlParser. Normally SAX is not a good solution to parse html.

Share:
16,349
user972590
Author by

user972590

Updated on June 14, 2022

Comments

  • user972590
    user972590 over 1 year

    I am trying to parse the normal html file using SAX parser.

    SAXBuilder builder2 = new SAXBuilder();
             try {
                Document sdoc = (Document)builder2.build(readFile);
                NodeList nl=sdoc.getElementsByTagName("body");
                System.out.println("nodelist>>>>>>>>>>>"+nl.getLength());
    
            } catch (JDOMException e1) {
                e1.printStackTrace();
            }
    

    but i am getting the exception

    Open quote is expected for attribute "{1}" associated with an  element type  "class".
    

    can anyone please tell me why i am getting this exception, the html document is well formed and it has all the open and close tags properly.

    Thanks in advance.

  • slott
    slott over 8 years
    SAX is always a good option for parsing huge amounts of data - such as HTML. Try looking at TagSoup which is quite awesome for doing just that!
  • Russ Bateman
    Russ Bateman almost 8 years
    The thing about TagSoup is that, being based on SAX, it's lightning fast and it solves all the stuff basic SAX chokes on including &lt; and &gt;. It's as easy to set up for as SAX; the handler methods are just the same ones--no learning curve beyond the SAX you already know.