What is the best way to parse HTML?

Question

What is the best way to parse HTML?

You must Login before you can answer or comment on any questions.

Hi I'm trying to parse HTML. When I passed HTML as string into Titanium.XML.parseString(), it crashed. Tried something like this:

http.send(); // http is a synchronous http client
var result = http.responseText;
var dom = Titanium.XML.parseString(result);//crash!!

My error is like this:

[ERROR] Error Domain=com.google.GDataXML Code=-1 "The operation couldn’t be completed. (com.google.GDataXML error -1.)". in -[TiDOMDocumentProxy parseString:] (TiDOMDocumentProxy.m:48)

Am I doing something wrong? Titanium.XML.parseString just can't parse HTML? Then is there any way to parse HTML? I need something like getElementById, getElementsByClassName....

3 Answers

Your Answer

Think you can help? Login to answer this question!

Answer 1

node-htmlparser

node-soupselect

Modify these two to run under Titanium's system. These allow you to parse non XML correct HTML.

What I did was this:

Modified htmlparser to expose it's exports to a regular object and used Ti.include to "include the file as if it was written there".

I did the same for soupselect, and they worked well together and passed the unit tests :)

Essentially I added this to the top of the source files:

exports = {};

and this at the bottom:

htmlparser = exports;

with soupselect, I had to substitute the line:

var domUtils = require('htmlparser').DomUtils;

with

var domUtils = htmlparser.DomUtils;

Ti.include('htmlparser.js');
Ti.include('soupselect.js');
 
var select = soupselect.select;
 
var body = '<html><head><title>Test</title></head>'
+ '<body>'
+ '<img src="http://cdn.cad-comic.com/comics/2859286598c11964un2ya69354216.jpg" />'
+ '</body></html>';
 
var handler = new htmlparser.DefaultHandler(function(err, dom) {
  if (err) {
    alert('Error: ' + err);
  } else {
    var img = select(dom, 'img');
 
    img.forEach(function(img) {
      alert('src: ' + img.attribs.src);
    });
  }
});
 
var parser = new htmlparser.Parser(handler);
parser.parseComplete(body);

Answer 2

YQL it's the best way to parse html, as long as the webpage does not block it.

Answer 3

Finally, I implemented a parsing procedure based on string itself.