htmlparser-papandreou

Forgiving HTML/XML/RSS Parser in JS for <em>both</em> Node and Browsers

NodeHtmlParser

A forgiving HTML/XML/RSS parser written in JS for both the browser and NodeJS (yes, despite the name it works just fine in any modern browser). The parser can handle streams (chunked data) and supports custom handlers for writing custom DOMs/output.

npm install htmlparser
node runtests.js

View runtests.html in any browser

var htmlparser = require("htmlparser");
var rawHtml = "Xyz <script language= javascript>var foo = '<<bar>>';< /  script><!--<!-- Waah! -- -->";
var handler = new htmlparser.DefaultHandler(function (errordom) {
if (error)
[...do something for errors...]
else
[...parsing done, do something...]
});
var parser = new htmlparser.Parser(handler);
parser.parseComplete(rawHtml);
sys.puts(sys.inspect(handler.dom, false, null));
var handler = new Tautologistics.NodeHtmlParser.DefaultHandler(function (errordom) {
if (error)
[...do something for errors...]
else
[...parsing done, do something...]
});
var parser = new Tautologistics.NodeHtmlParser.Parser(handler);
parser.parseComplete(document.body.innerHTML);
alert(JSON.stringify(handler.dom, null, 2));
[ { raw: 'Xyz ', data: 'Xyz ', type: 'text' }
  , { raw: 'script language= javascript'
  , data: 'script language= javascript'
  , type: 'script'
  , name: 'script'
  , attribs: { language: 'javascript' }
  , children: 
     [ { raw: 'var foo = \'<bar>\';<'
       , data: 'var foo = \'<bar>\';<'
       , type: 'text'
       }
     ]
  }
, { raw: '<!-- Waah! -- '
  , data: '<!-- Waah! -- '
  , type: 'comment'
  }
]
while (...) {
...
parser.parseChunk(chunk);
}
parser.done();
new htmlparser.RssHandler(function (errordom) {
...
});
var handler = new htmlparser.DefaultHandler(
  function (error) { ... }
, { verbose: false, ignoreWhitespace: true }
);

Indicates whether the DOM should exclude text nodes that consists solely of whitespace. The default value is "false".

The following HTML:

<font>
<br>this is the text
<font>

becomes:

[ { raw: 'font'
  , data: 'font'
  , type: 'tag'
  , name: 'font'
  , children: 
     [ { raw: 'br', data: 'br', type: 'tag', name: 'br' }
     , { raw: 'this is the text\n'
       , data: 'this is the text\n'
       , type: 'text'
       }
     , { raw: 'font', data: 'font', type: 'tag', name: 'font' }
     ]
  }
]

The following HTML:

<font>
<br>this is the text
<font>

becomes:

[ { raw: 'font'
  , data: 'font'
  , type: 'tag'
  , name: 'font'
  , children: 
     [ { raw: '\n\t', data: '\n\t', type: 'text' }
     , { raw: 'br', data: 'br', type: 'tag', name: 'br' }
     , { raw: 'this is the text\n'
       , data: 'this is the text\n'
       , type: 'text'
       }
     , { raw: 'font', data: 'font', type: 'tag', name: 'font' }
     ]
  }
]

Indicates whether to include extra information on each node in the DOM. This information consists of the "raw" attribute (original, unparsed text found between "<" and ">") and the "data" attribute on "tag", "script", and "comment" nodes. The default value is "true".

The following HTML:

<a href="test.html">xxx</a>

becomes:

[ { raw: 'a href="test.html"'
  , data: 'a href="test.html"'
  , type: 'tag'
  , name: 'a'
  , attribs: { href: 'test.html' }
  , children: [ { raw: 'xxx', data: 'xxx', type: 'text' } ]
  }
]

The following HTML:

<a href="test.html">xxx</a>

becomes:

[ { type: 'tag'
  , name: 'a'
  , attribs: { href: 'test.html' }
  , children: [ { data: 'xxx', type: 'text' } ]
  }
]

Indicates whether the DOM should prevent children on tags marked as empty in the HTML spec. Typically this should be set to "true" HTML parsing and "false" for XML parsing. The default value is "true".

The following HTML:

<link>text</link>

becomes:

[ { raw: 'link', data: 'link', type: 'tag', name: 'link' }
, { raw: 'text', data: 'text', type: 'text' }
]

The following HTML:

<link>text</link>

becomes:

[ { raw: 'link'
  , data: 'link'
  , type: 'tag'
  , name: 'link'
  , children: [ { raw: 'text', data: 'text', type: 'text' } ]
  }
]

Looking for CSS selectors to search the DOM? Try Node-SoupSelect, a port of SoupSelect to NodeJS: http://github.com/harryf/node-soupselect

There's also a port of hpricot to NodeJS that uses HtmlParser for HTML parsing: http://github.com/silentrob/Apricot