From e252bf4fce70c56b33294fdd095b0dc08ce8b95f Mon Sep 17 00:00:00 2001
From: lolcat <will@lolcat.ca>
Date: Sat, 29 Jun 2024 12:55:28 -0400
Subject: soundcloud fix, for good this time

---
 scraper/sc.php | 103 +++++++++++++++++++++++++++++++++++++++++++++------------
 1 file changed, 81 insertions(+), 22 deletions(-)

(limited to 'scraper/sc.php')

diff --git a/scraper/sc.php b/scraper/sc.php
index 86ea979..c435c78 100644
--- a/scraper/sc.php
+++ b/scraper/sc.php
@@ -6,6 +6,9 @@ class sc{
 		
 		include "lib/backend.php";
 		$this->backend = new backend("sc");
+		
+		include "lib/fuckhtml.php";
+		$this->fuckhtml = new fuckhtml();
 	}
 	
 	public function getfilters($page){
@@ -25,7 +28,7 @@ class sc{
 		];
 	}
 	
-	private function get($proxy, $url, $get = []){
+	private function get($proxy, $url, $get = [], $web_req = false){
 		
 		$curlproc = curl_init();
 		
@@ -37,19 +40,42 @@ class sc{
 		curl_setopt($curlproc, CURLOPT_URL, $url);
 		
 		curl_setopt($curlproc, CURLOPT_ENCODING, ""); // default encoding
-		curl_setopt($curlproc, CURLOPT_HTTPHEADER,
-			["User-Agent: " . config::USER_AGENT,
-			"Accept: application/json, text/javascript, */*; q=0.01",
-			"Accept-Language: en-US,en;q=0.5",
-			"Accept-Encoding: gzip",
-			"Referer: https://soundcloud.com/",
-			"Origin: https://soundcloud.com",
-			"DNT: 1",
-			"Connection: keep-alive",
-			"Sec-Fetch-Dest: empty",
-			"Sec-Fetch-Mode: cors",
-			"Sec-Fetch-Site: same-site"]
-		);
+		
+		// use http2
+		curl_setopt($curlproc, CURLOPT_HTTP_VERSION, CURL_HTTP_VERSION_2_0);
+		
+		if($web_req === false){
+			
+			curl_setopt($curlproc, CURLOPT_HTTPHEADER,
+				["User-Agent: " . config::USER_AGENT,
+				"Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8",
+				"Accept-Language: en-US,en;q=0.5",
+				"Accept-Encoding: gzip",
+				"Referer: https://soundcloud.com/",
+				"Origin: https://soundcloud.com",
+				"DNT: 1",
+				"Connection: keep-alive",
+				"Sec-Fetch-Dest: empty",
+				"Sec-Fetch-Mode: cors",
+				"Sec-Fetch-Site: same-site"]
+			);
+		}else{
+			
+			curl_setopt($curlproc, CURLOPT_HTTPHEADER,
+				["User-Agent: " . config::USER_AGENT,
+				"Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8",
+				"Accept-Language: en-US,en;q=0.5",
+				"Accept-Encoding: gzip",
+				"DNT: 1",
+				"Connection: keep-alive",
+				"Upgrade-Insecure-Requests: 1",
+				"Sec-Fetch-Dest: document",
+				"Sec-Fetch-Mode: navigate",
+				"Sec-Fetch-Site: cross-site",
+				"Priority: u=1",
+				"TE: trailers"]
+			);
+		}
 		
 		curl_setopt($curlproc, CURLOPT_RETURNTRANSFER, true);
 		curl_setopt($curlproc, CURLOPT_SSL_VERIFYHOST, 2);
@@ -396,13 +422,47 @@ class sc{
 		
 		$token = apcu_fetch("sc_token");
 		
-		if($token === false){
+		if($token !== false){
+			
+			return $token;
+		}
+		
+		// search through all javascript components on the main page
+		try{
+			$html =
+				$this->get(
+					$proxy,
+					"https://soundcloud.com",
+					[]
+				);
+		}catch(Exception $error){
+			
+			throw new Exception("Failed to fetch front page");
+		}
+		
+		$this->fuckhtml->load($html);
+		
+		$scripts =
+			$this->fuckhtml
+			->getElementsByTagName(
+				"script"
+			);
+		
+		foreach($scripts as $script){
+			
+			if(
+				!isset($script["attributes"]["src"]) ||
+				strpos($script["attributes"]["src"], "sndcdn.com") === false
+			){
+				
+				continue;
+			}
 			
 			try{
 				$js =
 					$this->get(
 						$proxy,
-						"https://a-v2.sndcdn.com/assets/0-a901c1e0.js",
+						$script["attributes"]["src"],
 						[]
 					);
 			}catch(Exception $error){
@@ -416,16 +476,15 @@ class sc{
 				$token
 			);
 			
-			if(!isset($token[1])){
+			if(isset($token[1])){
 				
-				throw new Exception("Failed to get search token");
+				apcu_store("sc_token", $token[1]);
+				return $token[1];
+				break;
 			}
-			
-			apcu_store("sc_token", $token[1]);
-			return $token[1];
 		}
 		
-		return $token;
+		throw new Exception("Did not find a Soundcloud token in the Javascript blobs");
 	}
 	
 	private function limitstrlen($text){
-- 
cgit v1.2.3


From 8161f8e7b8b236892a2254b8876fe4363b33050b Mon Sep 17 00:00:00 2001
From: lolcat <will@lolcat.ca>
Date: Sat, 29 Jun 2024 17:58:05 -0400
Subject: forgor to set headers

---
 scraper/sc.php | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

(limited to 'scraper/sc.php')

diff --git a/scraper/sc.php b/scraper/sc.php
index c435c78..011293a 100644
--- a/scraper/sc.php
+++ b/scraper/sc.php
@@ -57,7 +57,8 @@ class sc{
 				"Connection: keep-alive",
 				"Sec-Fetch-Dest: empty",
 				"Sec-Fetch-Mode: cors",
-				"Sec-Fetch-Site: same-site"]
+				"Sec-Fetch-Site: same-site",
+				"Priority: u=1"]
 			);
 		}else{
 			
@@ -433,7 +434,8 @@ class sc{
 				$this->get(
 					$proxy,
 					"https://soundcloud.com",
-					[]
+					[],
+					true
 				);
 		}catch(Exception $error){
 			
-- 
cgit v1.2.3


From 6df9d17ada64663c8bfb80baafc42dc951951218 Mon Sep 17 00:00:00 2001
From: lolcat <will@lolcat.ca>
Date: Sat, 29 Jun 2024 18:03:10 -0400
Subject: fixed soundcloud crash

---
 scraper/sc.php | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

(limited to 'scraper/sc.php')

diff --git a/scraper/sc.php b/scraper/sc.php
index 011293a..7083c42 100644
--- a/scraper/sc.php
+++ b/scraper/sc.php
@@ -327,9 +327,12 @@ class sc{
 						$description[] = $song["title"];
 					}
 					
-					if(count($description) != 0){
+					if(count($description) !== 0){
 						
 						$description = trim($count . " songs. " . implode(", ", $description));
+					}else{
+						
+						$description = "";
 					}
 					
 					if(
-- 
cgit v1.2.3