Wikipedia.php

/* Not the nicest code we ever wrote... */

<?php

require_once 'HTTP/Request.php';
require_once 'Text/Wiki/Mediawiki.php';


class Wikipedia
{
    private function zapBetween ($string, $start, $end)
    {
        $nesting = 0;
        $result = "";

        $index = 0;
        $len = strlen ($string);

        while ($index < $len) {
            if (strpos ($string, $start, $index) === $index) {
                $nesting++;
                $index += strlen ($start);
            } else if (strpos ($string, $end, $index) === $index) {
                $nesting--;
                $index += strlen ($end);
            } else {
                if ($nesting == 0) {
                    $result = $result . substr ($string, $index, 1);
                }

                $index++;
            }
        }

        return $result;
    }


    private function extractImage ($wikitext)
    {
        $image = array ();
        preg_match ("/\n\s?\|\s?image\s?=\s? (.*?)\n/", $wikitext, $image);
        $image = $image[1];

        if ($image) {
            $imgurl = "http://en.wikipedia.org/w/api.php" .
                '?prop=imageinfo&action=query&iiprop=url&iiurlwidth=150&format=php' .
                '&titles=Image:' . str_replace (' ', '_', $image);

            $client = new HTTP_Request ();
            $client->setMethod (HTTP_REQUEST_METHOD_GET);
            $client->setURL ($imgurl);
            $response = $client->sendRequest ();

            if ($response = $client->getResponseBody ()) {
                if ($imageinfo = unserialize ($response)) {
                    // Hack for wikipedia api
                    preg_match ('/\"http:\/\/(.*)\"/', $response, $matches);
                    $image = 'http://' . substr ($matches[1],
                                                 0,
                                                 strpos ($matches[1], '"'));
                    return $image;
                }
            }
        }

        return false;
    }


    private function cleanWikiText ($wikitext)
    {
        $wikitext = $this->zapBetween ($wikitext, "{{", "}}");
        $wikitext = preg_replace ("/\[\[Image:.*?\]\]\n/", "", $wikitext);

        if (!preg_match ('/disambiguation/', $result['name'])) {
            $wikitext = preg_replace ('/\=\=.*?$/s', '', $wikitext);
        }

        return $wikitext;
    }


    private function wikipediaBio ($lookfor, $author)
    {
        $wiki =& new Text_Wiki_Mediawiki ();
        $wiki->setRenderConf ('xhtml', 'wikilink', 'pages', false);

        $wiki->setFormatConf ('Xhtml', 'charset', 'utf-8');
        $wiki->setFormatConf ('Xhtml', 'translate', false);
        $wiki->setRenderConf ('xhtml', 'image', 'base',
                              'http://en.wikipedia.org/wiki/Image:');
        $wiki->setRenderConf ('xhtml', 'wikilink', 'view_url',
                              'http://en.wikipedia.org/wiki/%s');

        $wiki->disableRule ('image');

        $url = 'http://en.wikipedia.org/w/api.php?action=query&prop=revisions&rvprop=content&format=php&redirects=1&titles=' . $lookfor;
        $requestParams = array ('timeout' => 2, 'readTimeout' => array ('4','0'));
        $client = new HTTP_Request ('', $requestParams);

        $client->setMethod (HTTP_REQUEST_METHOD_GET);
        $client->setURL ($url);

        PEAR::setErrorHandling (PEAR_ERROR_RETURN);

        $response = $client->sendRequest ();

        if (!PEAR::isError ($response)) {

            $result = array();
            $body = @unserialize ($client->getResponseBody ());

            //Check if data exists or not
            if ($body &&
                !isset($body['error']) &&
                !isset($body['query']['pages']['-1'])) {

                $body = array_shift ($body['query']['pages']);

                $result['name'] = $body['title'];

                $body = array_shift ($body['revisions']);

                $wikitext = $body['*'];

                $image_url = $this->extractImage ($wikitext);

                if ($image_url) {
                    $result['image'] = $image_url;
                }

                $wikitext = $this->cleanWikiText ($wikitext);

                $result['description'] = $wiki->transform ($wikitext, 'xhtml');

                /* If we have life dates, extract them now... */
                $year = false;
                if (preg_match ("/[0-9]{4}/", array_pop ($author),
                                $matches)) {
                    if (count ($matches) > 0) {
                        $year = $matches[0];
                    }
                }

                $result['provider'] = "Wikipedia";
                $result['provider_link'] = "http://www.wikipedia.org/wiki/" . $result['name'];

                if (preg_match ('/disambiguation/', $result['name']) ||
                    preg_match ('/^[^\.]+(author|writer|journalist|novelist)[^\.]*\./',
                                $wikitext) ||
                    ($year && preg_match ("/$year/", $wikitext))) {
                    return $result;
                }

                return false;
            }
        }

        return false;
    }


    private function findBiography ($name)
    {
        global $configArray;

        $author = preg_replace ("/,$/", "", $name);

        /* Remove any parenthetical remarks from the author's name */
        $author = $this->zapBetween ($author, "(", ")");
        $author = explode (',', $author);

        // Look for a disambiguation page
        $wikiInfo = $this->wikipediaBio (urlencode ("$author[1] $author[0] (disambiguation)"),
                                          $author);

        if ($wikiInfo) {
            return false;
        }

        // Look for an author page
        $wikiInfo = $this->wikipediaBio (urlencode ("$author[1] $author[0]"),
                                     $author);

        if ($wikiInfo) {
            return $wikiInfo;
        }

        // Give up!
        return false;
    }


    function biographyFor ($name)
    {
        $result = $this->findBiography ($name);

        if ($result) {
            $result['snippet'] = (preg_replace ("/<\/p>.*/s", "</p>",
                                                $result['description']));

            $result['hasmore'] =
                !preg_match ('/^\s+$/',
                             substr ($result['description'],
                                     strlen ($result['snippet'])));
        }

        return $result;
    }
}

?>