transform stream to tokenize html
var fs = require('fs');
var tokenize = require('html-tokenize');
var through = require('through2');
fs.createReadStream(__dirname + '/table.html')
.pipe(tokenize())
.pipe(through.obj(function (row, enc, next) {
row[1] = row[1].toString();
console.log(row);
next();
}))
;this html:
<table>
<tbody>blah blah blah</tbody>
<tr><td>there</td></tr>
<tr><td>it</td></tr>
<tr><td>is</td></tr>
</table>generates this output:
[ 'open', '<table>' ]
[ 'text', '\n ' ]
[ 'open', '<tbody>' ]
[ 'text', 'blah blah blah' ]
[ 'close', '</tbody>' ]
[ 'text', '\n ' ]
[ 'open', '<tr>' ]
[ 'open', '<td>' ]
[ 'text', 'there' ]
[ 'close', '</td>' ]
[ 'close', '</tr>' ]
[ 'text', '\n ' ]
[ 'open', '<tr>' ]
[ 'open', '<td>' ]
[ 'text', 'it' ]
[ 'close', '</td>' ]
[ 'close', '</tr>' ]
[ 'text', '\n ' ]
[ 'open', '<tr>' ]
[ 'open', '<td>' ]
[ 'text', 'is' ]
[ 'close', '</td>' ]
[ 'close', '</tr>' ]
[ 'text', '\n' ]
[ 'close', '</table>' ]
[ 'text', '\n' ]
var tokenize = require('html-tokenize');Return a tokenize transform stream t that takes html input and produces rows
of output. The output rows are of the form:
[ name, buffer ]
The input stream maps completely onto the buffers from the object stream.
The types of names are:
- open
- close
- text
cdata, comments, and scripts all use 'open' with their contents appearing in
subsequent 'text' rows.
There is an html-tokenize command too.
usage: html-tokenize {FILE}
Tokenize FILE into newline-separated json arrays for each tag.
If FILE is not specified, use stdin.
With npm, to get the library do:
npm install html-tokenize
or to get the command do:
npm install -g html-tokenize
MIT