Created
April 26, 2010 20:52
-
-
Save markjames/379906 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<?php | |
require( 'persistent.class.php' ); | |
class Episode extends Persistent { | |
public $series; | |
public $season; | |
public $number; | |
public $airdate; | |
public $title; | |
public $description; | |
protected $detailURL; | |
public function __construct() { | |
$args = func_get_args(); | |
$this->series = ''; | |
$this->season = ''; | |
$this->number = 0; | |
$this->airdate = 0; | |
$this->title = ''; | |
$this->description = ''; | |
switch( count($args) ) { | |
case 1: // String version | |
$this->split($args[0]); | |
break; | |
case 2: // Series/Number | |
list($this->series,$this->number) = $args; | |
break; | |
case 3: // Series/Season/Number | |
list($this->series,$this->season,$this->number) = $args; | |
break; | |
} | |
} | |
public function is_released() { | |
return mktime() > $this->airdate; | |
} | |
public function toString( $format = '%1$s - S%2$02dE%3$02d - %4$s', $dateformat='Ymd' ) { | |
return sprintf( $format, $this->series, $this->season, $this->number, $this->title, date( $dateformat, $this->airdate ), $this->description ); | |
} | |
public function __toString() { | |
return $this->toString(); | |
} | |
public function split( $in ) { | |
$matches = array(); | |
// Attempt format with dots/spaces and S##E## | |
$expression = '~^(?:\[\w+\][ .]*)?([-A-Za-z0-9. ]+?)[. \[](S\d\d)(E\d\d)[. \[]?~i'; | |
if( preg_match( $expression, $in, $matches ) ) { | |
$this->series = ucwords(strtolower(str_replace( '.', ' ', $matches[1] ))); | |
$this->season = intval(str_ireplace( 'S', '', $matches[2] )); | |
$this->number = intval(str_ireplace( 'E', '', $matches[3] )); | |
return true; | |
} | |
// Attempt format with dots/spaces and production code | |
$expression = '~^([-A-Za-z0-9. ]+?)[. \[]+(\d+?)[x.]?(\d+?)[. \]]~i'; | |
if( preg_match( $expression, $in, $matches ) ) { | |
$this->series = ucwords(strtolower(str_replace( '.', ' ', $matches[1] ))); | |
$this->season = intval($matches[2]); | |
$this->number = intval(str_ireplace( 'x', '', $matches[3] )); | |
return true; | |
} | |
// Simple text format | |
$expression = '~^([-A-Za-z0-9. ]+?)\s+(S|Season)?\s*?(\d+)\s*?(Ep|Episode)?\s*?(\d+)~i'; | |
if( preg_match( $expression, $in, $matches ) ) { | |
$this->series = ucwords(strtolower(str_replace( '.', ' ', $matches[1] ))); | |
$this->season = $matches[3]; | |
$this->number = $matches[5]; | |
return true; | |
} | |
return false; | |
} | |
public function save( $dir='./data/' ) { | |
return parent::save( $dir.preg_replace('~\W~', ' ', $this->toString('%1$s S%2$02dE%3$02d')).'.txt' ); | |
} | |
public function load( $dir='./data/' ) { | |
return parent::load( $dir.preg_replace('~\W~', ' ', $this->toString('%1$s S%2$02dE%3$02d')).'.txt' ); | |
} | |
public function cached( $dir='./data/' ) { | |
return file_exists( $dir.preg_replace('~\W~', ' ', $this->toString('%1$s S%2$02dE%3$02d')).'.txt' ); | |
} | |
public function load_from_epguides() { | |
if( empty($this->series) ) { | |
throw new Exception('Show not set'); | |
} | |
$contents = file_get_contents( 'http://epguides.com/' . preg_replace( '~\W~', '', $this->series ) ); | |
if( $contents === false ) { | |
$contents = file_get_contents( 'http://epguides.com/' . str_ireplace('The','',preg_replace( '~\W~', '', $this->series ))); | |
} | |
$matches = array(); | |
if( preg_match( sprintf("~^\s*\d+[.]\s+%d[-]\s*?%2d\s+(\S*?)\s+(.*?)[<].*?href=\"(.*?)\".*?[>](.*?)[<]~mi", $this->season, $this->number), $contents, $matches ) ) { | |
$this->productioncode = $matches[1]; | |
$this->airdate = strtotime(trim($matches[2])); | |
$this->detailURL = $matches[3]; | |
$this->title = $matches[4]; | |
} else if( preg_match( sprintf("~<li>\s*%d[-]\s*?%2d\s+(\S*?)\s+(.*?)[<].*?href=\"(.*?)\".*?[>](.*?)[<]~mi", $this->season, $this->number), $contents, $matches ) ) { | |
$this->productioncode = $matches[1]; | |
$this->airdate = strtotime(trim($matches[2])); | |
$this->detailURL = $matches[3]; | |
$this->title = $matches[4]; | |
} else { | |
return false; | |
} | |
$detailcontents = file_get_contents( $this->detailURL ); | |
//$detailcontents = ''; | |
$expression = '~<title>(.*?)</title>~si'; | |
$seriesmatches = array(); | |
if( preg_match( $expression, $detailcontents, $seriesmatches ) ) { | |
$this->series = trim(trim(trim(str_replace('- TV.com','',strip_tags($seriesmatches[1]))), ':')); | |
} | |
$expression = '~<p class="deck">(.*?)<a~si'; | |
$detailmatches = array(); | |
if( preg_match( $expression, $detailcontents, $detailmatches ) ) { | |
$this->description = trim(strip_tags($detailmatches[1])); | |
} else { | |
$expression = '~<div id="main-col">\s+<div>(.*?)(?:<div)~si'; | |
$detailmatches = array(); | |
if( preg_match( $expression, $detailcontents, $detailmatches ) ) { | |
$this->description = trim(strip_tags($detailmatches[1])); | |
} | |
} | |
return true; | |
} | |
public static function get( $title ) { | |
$episode = new Episode( $title ); | |
if( $episode->cached() ) { | |
$episode->load(); | |
} else { | |
$episode->load_from_epguides(); | |
$episode->save(); | |
} | |
return $episode; | |
} | |
} | |
?> |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment