indieweb/original-post-discovery

Functions for doing original post discovery in PHP

v0.1.0 2014-01-20 19:10 UTC

Requires

• barnabywalters/mf-cleaner: *
• guzzle/guzzle: *
• mf2/mf2: ~0.2
• mf2/shim: ~0.2
• tantek/cassis: *

Requires (Dev)

• phpunit/phpunit: *

MIT

• Barnaby Walters

README

A set of PHP functions for determining the canonical URL for a post, given a POSSEd copy. A partial implementation of the algorithm outlined at original-post-discovery.

A demo can be found on waterpigs.co.uk/services/original-post.

Usage

Install using Composer ./composer.phar require indieweb/original-post-discovery:dev-master.

<?php

require __DIR__ . '/vendor/autoload.php';

list($url, $err) = IndieWeb\discoverOriginalPost('https://twitter.com/BarnabyWalters/status/423465842148671488');
if ($err !== null) {
	// handle HTTP errors here
}

// do stuff (e.g. auto-fill in-reply-to form controls) with $url

Other functions

• string $str = cleanString($str) cleans up a bunch of weird encoding and character issues which can occur, specifically converting non-breaking space codepoints into normal spaces to handle some Twitter.com bugs
• string|null $url = originalPostUrlFromTwitter($html) is a pure function for parsing HTML from Twitter.com and looking in it for trailing URLs
• string $str = stripHashtags($str) removes hashtags from a string
• string|null $url = getTrailingUrl($str) finds parenthesised (text text. (http://example.com)) or ellipsis (text text… http://example.com) trailing URLs in a string
• string|null $str = getUrlFromPermashortid($str) looks for a trailing permashortid ((cctld.me id)) and converts it into a URL (assumes HTTP)

Testing

A small PHPUnit test suite is provided — if making contributions please at least ensure that all the existing tests pass before/after your changes are made. If you could add new tests to cover the code you added that would be great too.

Version History

0.1.0 2014-01-20

• Initial extraction from Taproot, readme and basic test suite