diff --git a/page.php b/page.php
index b9ed3fc..86ba5ac 100644
--- a/page.php
+++ b/page.php
@@ -1,278 +1,301 @@
<?php
define('LONG_DATE_FORMAT', '%e %B %Y');
define('USER_AGENT', 'WikimediaTools/SourceTemplatesGenerator/0.1');
define('USER_AGENT_FALLBACK', 'Mozilla/5.0');
require_once('helpers/Encoding.php');
class Page {
public $url;
/**
* @var array Meta tags
*/
public $meta_tags;
/**
* @var string The page content
*/
public $data;
/**
* @var string The page title
*/
public $title;
/**
* @var string The page author
*/
public $author;
/**
* @var string The site ISSN
*/
public $issn;
//If we use the parameters yyyy mm dd, we describe CONTENT date:
/**
* @var int The page content's year
*/
public $yyyy;
/**
* @var int The page content's month
*/
public $mm;
/**
* @var int The page content's day
*/
public $dd;
//If not, we describe ONLINE RESOURCE PUBLISH date:
/**
* @var string The page publication date in relevant locale
*/
public $date;
/**
* @var int The page publication unixtime
*/
public $unixtime;
/**
* @var bool Indicates if we have to skip year/month/date template parameters
*/
public $skipYMD;
/**
* @var bool Indicates if we have to skip author template parameter
*/
public $skipAuthor;
/**
* @var string The last error occured while opening and parsing the page
*/
public $error;
/**
* Initializes a new Page instance. If an error occured, you can read it in $this->error.
*
* @param string $url the page URL
*/
function __construct ($url) {
$this->url = $url;
$this->get_data();
if ($this->data) {
$this->analyse();
}
}
function get_data () {
ini_set('user_agent', USER_AGENT);
$data = file_get_contents($this->url);
if (!$data) {
ini_set('user_agent', USER_AGENT_FALLBACK);
if (!$data = @file_get_contents($this->url)) {
$this->error = "Can't read URL";
return;
}
}
$encoding = mb_detect_encoding($data, "ISO-8859-15, ISO-8859-1, UTF-8, ASCII, auto");
if ($encoding && $encoding != 'UTF-8') {
$this->data = Encoding::toUTF8($data);
} else {
$this->data = $data;
}
}
/**
* Return a new Page instance, or if such class exists, an instance class specialized for your site.
*
* @param $url the page URL
*/
static function load ($url) {
//Classes list are stored in pages/index.dat file
//Each line contains the URL beginning, a tabulation, and the page analyser name
// * class is this name, appended by 'Page'
// * source file is the lowercase version of this name, appended by '.php'
$pages = file('pages/index.dat', true);
foreach ($pages as $line) {
$page = explode("\t", $line);
if (substr($url, 0, strlen($page[0])) == $page[0]) {
$file = strtolower(trim($page[1])) . '.php';
$class = trim($page[1]) . 'Page';
require("pages/$file");
return new $class($url);
}
}
return new Page($url);
}
/**
* Analyses metatags to process content
*/
function analyse () {
//Meta tags (including <meta property="" value=""> and <meta itemprop="" value="" syntax)
$this->meta_tags = $this->get_meta_tags();
$t = $this->meta_tags;
//Title
$this->title = $this->get_title();
//Date
if ($date = $this->getMetaTag($t, 'date')) {
$date = date_parse($date);
$this->yyyy = $date['year'];
$this->mm = $date['month'];
$this->dd = $date['day'];
}
//Site name
$this->site = $this->getMetaTag($t, 'og:site_name');
//Author
$this->author = $this->getMetaTag($t, 'author');
}
/**
* Gets page metatags
*
* @return array an array where the keys are the metatags' names and the values the metatags' values
*/
function get_meta_tags () {
return $this::get_all_meta_tags($this->url);
}
/**
* Gets all metatags, including those using meta property= and meta itemprop= syntax
*
* @return array an array where the keys are the metatags' names and the values the metatags' values
*/
function get_all_meta_tags () {
//Thank you to Michael Knapp and Mariano
//See http://php.net/manual/en/function.get-meta-tags.php comments
preg_match_all('/<[\s]*meta[\s]*\b(name|property|itemprop)\b="?' . '([^>"]*)"?[\s]*' . 'content="?([^>"]*)"?[\s]*[\/]?[\s]*>/si', $this->data, $match);
if (isset($match) && is_array($match) && count($match) == 4) {
$originals = $match[0];
$names = $match[2];
$values = $match[3];
if (count($originals) == count($names) && count($names) == count($values)) {
$metaTags = array();
for ($i=0, $limiti = count($names) ; $i < $limiti ; $i++) {
$metaTags[$names[$i]] = $values[$i];
}
}
}
return $metaTags;
}
/**
* Gets title
*
* @return string The page title
*/
function get_title () {
$title = $this->getMetaTag($this->meta_tags, 'title', 'og:title', 'DC.title', 'Title');
return $title ?: ((preg_match("#<title>(.+)<\/title>#iU", $this->data, $title)) ? trim($title[1]) : '');
}
/**
* Determines if the current page is an article published in a journal.
*
* @return bool true if the current page is an article ; otherwise, false
*/
function is_article () {
return
(array_key_exists('dc_type', $this->meta_tags) && $this->meta_tags['dc_type'] == 'journalArticle')
||
(array_key_exists('dcsext_pn-cat', $this->meta_tags) && $this->meta_tags['dcsext_pn-cat'] == 'Article')
||
array_key_exists('citation_journal_title', $this->meta_tags)
||
array_key_exists('prism_publicationname', $this->meta_tags);
}
/**
* Gets relevant metatag
*
* @param array the metatags
* @param string... the list of acceptable metatags
*
* @return string the first metatag value found
*/
static function getMetaTag () {
$tags = func_get_args();
$metatags = array_shift($tags);
foreach ($tags as $tag) {
$tag_lowercase = strtolower($tag);
foreach ($metatags as $key => $value) {
if ($tag_lowercase == strtolower($key)) return $value;
}
}
return '';
}
/**
* Finds a portion of text included between $before and $after strings on the current page
*
* @param string $before The string at the left of the text to be grabbed
* @param string $after The string at the right of the text to be grabbed
*
* @return string The text found between $before and $after
*/
function between ($before, $after) {
return self::grab($this->data, $before, $after);
}
/**
* Finds a portion of text included between $before and $after strings
*
* @param string $text The text where to find the substring
* @param string $before The string at the left of the text to be grabbed
* @param string $after The string at the right of the text to be grabbed
*
* @return string The text found between $before and $after
*/
static function grab ($text, $before, $after) {
$pos1 = strpos($text, $before);
if ($pos1 === false) { return false; } else { $pos1 += strlen($before); }
$pos2 = strpos($text, $after, $pos1 + 1);
if ($pos2 === false) { return false; }
return substr($text, $pos1, $pos2 - $pos1);
}
+
+ /**
+ * Downloads, through CURL library, accepting cookies.
+ *
+ * @param $url The URL to fetch
+ */
+ static function curl_download ($url) {
+ $ch = curl_init();
+ $timeout = 5;
+ $cookie_file = tmpfile();
+ $cookie_file = tempnam(sys_get_temp_dir(), "cookie-sourcesgen-");
+ curl_setopt($ch, CURLOPT_COOKIESESSION, true);
+ curl_setopt($ch, CURLOPT_COOKIEFILE, $cookie_file);
+ curl_setopt($ch, CURLOPT_COOKIEJAR, $cookie_file);
+ curl_setopt($ch, CURLOPT_URL, $url);
+ curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
+ curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
+ curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, $timeout);
+ $data = curl_exec($ch);
+ curl_close($ch);
+ unlink($cookie_file);
+ return $data;
+ }
}
diff --git a/pages/taylorandfrancis.php b/pages/taylorandfrancis.php
index 4913c70..9aa2e75 100644
--- a/pages/taylorandfrancis.php
+++ b/pages/taylorandfrancis.php
@@ -1,76 +1,57 @@
<?php
//Page analysis for www.tandfonline.com
class TaylorAndFrancisPage extends Page {
/**
* Initializes a new TaylorAndFrancisPage instance. If an error occured, you can read it in $this->error.
*
* @param string $url the page URL
*/
function __construct ($url) {
$this->url = $url;
$this->data = self::curl_download($url);
$this->analyse();
}
function analyse () {
parent::analyse();
$this->publisher = 'Taylor & Francis';
//DOI
$this->doi = self::between('meta name="dc.Identifier" scheme="doi" content="', '"');
//Gets the right dc.Identifier (coden scheme)
//Expected format: <Issue name>, Vol. <Issue volume>, No. <Issue number>, <Issue date>, pp. <article pages>
//e.g. Annals of Science, Vol. 68, No. 3, July 2011, pp. 325–350
$identifier = self::between('meta name="dc.Identifier" scheme="coden" content="', '"');
$identifier_data = explode(', ', $identifier);
$pos = strpos($identifier, ", Vol. ");
$this->journal = substr($identifier, 0, $pos);
$this->volume = self::grab($identifier, "Vol. ", ",");
$this->issue = self::grab($identifier, "No. ", ",");
$date = explode(' ', $identifier_data[3]);
$this->yyyy = array_pop($date);
$pos = strpos($identifier, "pp. ");
$this->pages = substr($identifier, $pos + 4);
//Author
//TODO: handle several authors
$author = trim(self::getMetaTag($this->meta_tags, 'dc.Creator'));
$names = explode(' ', $author);
if (count($names) == 2) {
$this->author = "$names[1], $names[0]";
} else {
$this->author = $author;
}
}
function is_article () {
return true;
}
-
- static function curl_download ($url) {
- $ch = curl_init();
- $timeout = 5;
- $cookie_file = tmpfile();
- $cookie_file = tempnam(sys_get_temp_dir(), "cookie-sourcesgen-");
- curl_setopt($ch, CURLOPT_COOKIESESSION, true);
- curl_setopt($ch, CURLOPT_COOKIEFILE, $cookie_file);
- curl_setopt($ch, CURLOPT_COOKIEJAR, $cookie_file);
- curl_setopt($ch, CURLOPT_URL, $url);
- curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
- curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
- curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, $timeout);
- $data = curl_exec($ch);
- curl_close($ch);
- unlink($cookie_file);
- return $data;
- }
-
}
?>