kafene/netscape-bookmark-parser

Generic Netscape bookmark parser

v0.0.1 2015-09-17 22:20 UTC

MIT

• parse
• bookmark
• netscape

README

About

This library provides a generic NetscapeBookmarkParser class that is able of parsing Netscape bookmark export files.

The motivations behind developing this parser are the following:

• the Netscape format has a very loose specification: no DTD nor XSL stylesheet to constrain how data is formatted
• software and web services export bookmarks using a wild variety of attribute names and values
• using standard SAX or DOM parsers is thus not straightforward.

How it works:

• the input bookmark file is trimmed and sanitized to improve parsing results
• the resulting data is then parsed using PCRE patterns to match attributes and values corresponding to the most likely:
  • attribute names: description vs. note, tags vs. labels, date vs. time, etc.
  • data formats: comma,separated,tags vs. space separated labels, UNIX epochs vs. human-readable dates, newlines & carriage returns, etc.
• an associative array containing all successfully parsed links with their attributes is returned

Example

Script:

<?php
require_once 'NetscapeBookmarkParser.php';

$parser = new NetscapeBookmarkParser();
$bookmarks = $parser->parseFile('./tests/input/netscape_basic.htm');
var_dump($bookmarks);

Output:

array(2) {
  [0] =>
  array(6) {
    'tags' =>
    string(14) "private secret"
    'uri' =>
    string(19) "https://private.tld"
    'title' =>
    string(12) "Secret stuff"
    'note' =>
    string(52) "Super-secret stuff you're not supposed to know about"
    'time' =>
    int(971175336)
    'pub' =>
    int(0)
  }
  [1] =>
  array(6) {
    'tags' =>
    string(18) "public hello world"
    'uri' =>
    string(17) "http://public.tld"
    'title' =>
    string(12) "Public stuff"
    'note' =>
    string(0) ""
    'time' =>
    int(1456433748)
    'pub' =>
    int(1)
  }
}