first push

composer.json

@@ -0,0 +1,14 @@
+{
+	"name": "athlon1600/youtube-downloader",
+	"description": "PHP powered alternative for youtube-dl",
+	"keywords": ["youtube downloader", "download youtube", "download youtube videos"],
+	"license": "MIT",
+	"require": {
+		"ext-curl": "*"
+	},
+	"autoload": {
+		"files": [
+			"src/YouTubeDownloader.php"
+		]
+	}
+}

src/YouTubeDownloader.php

@@ -0,0 +1,379 @@
+<?php
+
+// utils.php
+function sig_js_decode($player_html){
+
+	// what javascript function is responsible for signature decryption?
+	// var l=f.sig||Xn(f.s)
+	// a.set("signature",Xn(c));return a
+	if(preg_match('/signature",([a-zA-Z0-9]+)\(/', $player_html, $matches)){
+
+		$func_name = $matches[1];
+
+		// extract code block from that function
+		// xm=function(a){a=a.split("");wm.zO(a,47);wm.vY(a,1);wm.z9(a,68);wm.zO(a,21);wm.z9(a,34);wm.zO(a,16);wm.z9(a,41);return a.join("")};
+		if(preg_match("/{$func_name}=function\([a-z]+\){(.*?)}/", $player_html, $matches)){
+
+			$js_code = $matches[1];
+
+			// extract all relevant statements within that block
+			// wm.vY(a,1);
+			if(preg_match_all('/([a-z0-9]{2})\.([a-z0-9]{2})\([^,]+,(\d+)\)/i', $js_code, $matches) != false){
+
+				// must be identical
+				$obj_list = $matches[1];
+
+				//
+				$func_list = $matches[2];
+
+				// extract javascript code for each one of those statement functions
+				preg_match_all('/('.implode('|', $func_list).'):function(.*?)\}/m', $player_html, $matches2,  PREG_SET_ORDER);
+
+				$functions = array();
+
+				// translate each function according to its use
+				foreach($matches2 as $m){
+
+					if(strpos($m[2], 'splice') !== false){
+						$functions[$m[1]] = 'splice';						
+					} else if(strpos($m[2], 'a.length') !== false){
+						$functions[$m[1]] = 'swap';
+					} else if(strpos($m[2], 'reverse') !== false){
+						$functions[$m[1]] = 'reverse';
+					}
+				}
+
+				// FINAL STEP! convert it all to instructions set
+				$instructions = array();
+
+				foreach($matches[2] as $index => $name){
+					$instructions[] = array($functions[$name], $matches[3][$index]);
+				}
+
+				return $instructions;
+			}
+		}
+	}
+
+	return false;
+}
+
+
+
+
+// YouTube is capitalized twice because that's how youtube itself does it:
+// https://developers.google.com/youtube/v3/code_samples/php
+class YouTubeDownloader {
+
+	private $storage_dir;
+	private $cookie_dir;
+
+	private $itag_info = array(
+
+		18 => "MP4 360p",
+		22 => "MP4 720p (HD)",
+		37 => "MP4 1080",
+		38 => "MP4 3072p",
+
+		// questionable MP4s
+		59 => "MP4 480p",
+		78 => "MP4 480p",
+
+		43 => "WebM 360p",
+
+		17 => "3GP 144p"
+	);
+
+	function __construct(){
+		$this->storage_dir = sys_get_temp_dir();
+		$this->cookie_dir = sys_get_temp_dir();
+	}
+
+	function setStorageDir($dir){
+		$this->storage_dir = $dir;
+	}
+
+	// what identifies each request? user agent, cookies...
+	public function curl($url){
+
+		$ch = curl_init($url);
+
+		curl_setopt($ch, CURLOPT_USERAGENT, 'Mozilla/5.0 (Windows NT 6.3; WOW64; rv:49.0) Gecko/20100101 Firefox/49.0');
+		curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
+		curl_setopt($ch, CURLOPT_HEADER, 0);
+
+		//curl_setopt($ch, CURLOPT_COOKIEJAR, $tmpfname);
+		//curl_setopt($ch, CURLOPT_COOKIEFILE, $tmpfname);
+
+		//curl_setopt($ch, CURLOPT_IPRESOLVE, CURL_IPRESOLVE_V4);
+		curl_setopt($ch, CURLOPT_SSL_VERIFYPEER, 0);
+		curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 1);
+
+		$result = curl_exec($ch);
+		curl_close($ch);
+
+		return $result;
+	}
+
+	public static function head($url){
+
+		$ch = curl_init($url);
+
+		curl_setopt($ch, CURLOPT_HEADER, 1);
+		curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
+		curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 0);
+		curl_setopt($ch, CURLOPT_NOBODY, 1);
+		$result = curl_exec($ch);
+		curl_close($ch);
+
+		return http_parse_headers($result);
+	}
+
+	// html code of watch?v=aaa
+	private function getInstructions($html){
+
+		// check what player version that video is using
+		if(preg_match('@<script\s*src="([^"]+player-en_US[^"]+)@', $html, $matches)){
+
+			$player_url = str_replace('//', 'http://', $matches[1]);
+
+			// try to find instructions list already cached from previous requests...
+			$file_path = $this->storage_dir.'/'.md5($player_url);
+
+			if(file_exists($file_path)){
+
+				// unserialize could fail on empty file
+				$str = file_get_contents($file_path);
+				return unserialize($str);;
+
+			} else {
+
+				$js_code = $this->curl($player_url);
+				$instructions = sig_js_decode($js_code);
+
+				if($instructions){
+					file_put_contents($file_path, serialize($instructions));
+					return $instructions;
+				}
+			}
+		}
+
+		return false;
+	}
+
+	// this is in beta mode!!
+	public function stream($id){
+
+		$links = $this->getDownloadLinks($id, "mp4");
+
+		if(count($links) == 0){
+			die("no url found!");
+		}
+
+		$url = $links[0]['url'];
+
+		$ch = curl_init();
+
+		$headers = array();
+
+		if(isset($_SERVER['HTTP_RANGE'])){
+			$headers[] = 'Range: '.$_SERVER['HTTP_RANGE'];
+		}
+
+		$headers[] = 'User-Agent: Mozilla/5.0 (Windows NT 6.3; WOW64; rv:49.0) Gecko/20100101 Firefox/49.0';
+
+		curl_setopt($ch, CURLOPT_HTTPHEADER, $headers);
+
+
+		//curl_setopt($ch, CURLOPT_BUFFERSIZE, 4096);
+		curl_setopt($ch, CURLOPT_URL, $url);
+
+		//curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, $timeout);
+		curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 0);
+
+		// we deal with this ourselves
+		curl_setopt($ch, CURLOPT_RETURNTRANSFER, 0);
+		curl_setopt($ch, CURLOPT_HEADER, 0);
+
+		// whether request to video success
+		$headers = '';
+		$headers_sent = false;
+		$success = false;
+
+		curl_setopt($ch, CURLOPT_HEADERFUNCTION, function($ch, $data) use (&$headers, &$headers_sent){
+
+			$headers .= $data;
+
+			// this should be first line
+			if(preg_match('@HTTP\/\d\.\d\s(\d+)@', $data, $matches)){
+				$status_code = $matches[1];
+
+				if($status_code == 200 || $status_code == 206){
+					$headers_sent = true;
+					header(rtrim($data));
+				}
+
+			} else {
+
+				// only headers we wish to forward back to the client
+				$forward = array('content-type', 'content-length', 'accept-ranges', 'content-range');
+
+				$parts = explode(':', $data, 2);
+
+				if($headers_sent && count($parts) == 2 && in_array(trim(strtolower($parts[0])), $forward)){
+					header(rtrim($data));
+				}
+			}
+
+			return strlen($data);
+		});
+
+
+		// if response is empty - this never gets called
+		curl_setopt($ch, CURLOPT_WRITEFUNCTION, function($curl, $data) use (&$headers_sent){
+
+			if($headers_sent){
+				echo $data;
+				flush();
+			}
+
+			return strlen($data);
+		});
+
+		$ret = @curl_exec($ch);
+		curl_close($ch);
+
+		// if we are still here by now, return status_code
+		return true;
+	}
+
+	// extract youtube video_id from any piece of text
+	public function extractId($str){
+
+		if(preg_match('/[a-z0-9_-]{11}/i', $str, $matches)){
+			return $matches[0];
+		}
+
+		return false;
+	}
+
+	// selector by format: mp4 360, 
+	private function selectFirst($links, $selector){
+
+		$result = array();
+		$formats = preg_split('/\s*,\s*/', $selector);
+
+		// has to be in this order
+		foreach($formats as $f){
+
+			foreach($links as $l){
+
+				if(stripos($l['format'], $f) !== false || $f == 'any'){
+					$result[] = $l;
+				}
+			}
+		}
+
+		return $result;
+	}
+
+	// options | deep_links | append_redirector
+	public function getDownloadLinks($id, $selector = false){
+
+		$result = array();
+		$instructions = array();
+
+		// you can input HTML of /watch? page directory instead of id
+		if(strpos($id, '<div id="player') !== false){
+			$html = $id;
+		} else {
+			$video_id = $this->extractId($id);
+
+			if(!$video_id){
+				return false;
+			}
+
+			$html = $this->curl("https://www.youtube.com/watch?v={$video_id}");
+		}
+
+		// age-gate
+		if(strpos($html, 'player-age-gate-content') !== false){
+			// nothing you can do folks...
+			return false;
+		}
+
+		// http://stackoverflow.com/questions/35608686/how-can-i-get-the-actual-video-url-of-a-youtube-live-stream
+		if(preg_match('@url_encoded_fmt_stream_map["\']:\s*["\']([^"\'\s]*)@', $html, $matches)){
+
+			$parts = explode(",", $matches[1]);
+
+			foreach($parts as $p){
+				$query = str_replace('\u0026', '&', $p);
+				parse_str($query, $arr);
+
+				$url = $arr['url'];
+
+				if(isset($arr['sig'])){
+					$url = $url.'&signature='.$arr['sig'];
+
+				} else if(isset($arr['signature'])){
+					$url = $url.'&signature='.$arr['signature'];
+
+				} else if(isset($arr['s'])){
+
+					// this is probably a VEVO/ads video... signature must be decrypted first! We need instructions for doing that
+					if(count($instructions) == 0){
+						$instructions = (array)$this->getInstructions($html);
+					}
+
+					$dec = $this->sig_decipher($arr['s'], $instructions);
+					$url = $url.'&signature='.$dec;
+				}
+
+				// redirector.googlevideo.com
+				//$url = preg_replace('@(\/\/)[^\.]+(\.googlevideo\.com)@', '$1redirector$2', $url);
+
+				$itag = $arr['itag'];
+				$format = isset($this->itag_info[$itag]) ? $this->itag_info[$itag] : 'Unknown';
+
+				$result[$itag] = array(
+					'url' => $url,
+					'format' => $format
+				);
+			}
+		}
+
+		// do we want all links or just select few?
+		if($selector){
+			return $this->selectFirst($result, $selector);
+		}
+
+		return $result;
+	}
+
+	private function sig_decipher($signature, $instructions){
+
+		foreach($instructions as $opt){
+
+			$command = $opt[0];
+			$value = $opt[1];
+
+			if($command == 'swap'){
+
+				$temp = $signature[0];
+				$signature[0] = $signature[$value % strlen($signature)];
+				$signature[$value] = $temp;
+
+			} else if($command == 'splice'){
+				$signature = substr($signature, $value);
+			} else if($command == 'reverse'){
+				$signature = strrev($signature);
+			}
+		}
+
+		return trim($signature);
+	}
+}
+
+
+?>