porndish fix

src/providers/porndish.rs

@@ -13,9 +13,9 @@ use futures::stream::{self, StreamExt};
 use htmlentity::entity::{ICodedDataTrait, decode};
 use regex::Regex;
 use scraper::{ElementRef, Html, Selector};
-use std::process::Command;
 use std::sync::{Arc, RwLock};
 use std::thread;
+use std::time::{SystemTime, UNIX_EPOCH};
 
 error_chain! {
     foreign_links {
@@ -263,52 +263,22 @@ impl PorndishProvider {
         );
     }
 
-    async fn fetch_with_curl_cffi(url: &str, referer: Option<&str>) -> Result<String> {
+    fn request_headers(referer: Option<&str>) -> Vec<(String, String)> {
-        let url = url.to_string();
+        let referer = referer
-        let referer = referer.unwrap_or("").to_string();
+            .filter(|referer| !referer.is_empty())
+            .unwrap_or("https://www.porndish.com/");
+        vec![("Referer".to_string(), referer.to_string())]
+    }
 
-        let output = tokio::task::spawn_blocking(move || {
+    async fn fetch_html(
-            Command::new("python3")
+        requester: &mut Requester,
-                .arg("-c")
+        url: &str,
-                .arg(
+        referer: Option<&str>,
-                    r#"
+    ) -> Result<String> {
-import sys
+        requester
-from curl_cffi import requests
+            .get_with_headers(url, Self::request_headers(referer), None)
-
-url = sys.argv[1]
-referer = sys.argv[2] if len(sys.argv) > 2 else ""
-headers = {"Referer": referer} if referer else {}
-response = requests.get(
-    url,
-    impersonate="chrome",
-    timeout=30,
-    allow_redirects=True,
-    headers=headers,
-)
-if response.status_code >= 400:
-    sys.stderr.write(f"status={response.status_code} url={response.url}\n")
-    sys.exit(1)
-sys.stdout.buffer.write(response.content)
-"#,
-                )
-                .arg(url)
-                .arg(referer)
-                .output()
-        })
             .await
-        .map_err(|error| Error::from(format!("spawn_blocking failed: {error}")))?
+            .map_err(|error| Error::from(format!("request failed: {error}")))
-        .map_err(|error| Error::from(format!("python3 execution failed: {error}")))?;
-
-        if !output.status.success() {
-            let stderr = String::from_utf8_lossy(&output.stderr).trim().to_string();
-            return Err(Error::from(format!("curl_cffi request failed: {stderr}")));
-        }
-
-        Ok(String::from_utf8_lossy(&output.stdout).to_string())
-    }
-
-    async fn fetch_html(url: &str) -> Result<String> {
-        Self::fetch_with_curl_cffi(url, None).await
     }
 
     async fn load_filters(
@@ -317,6 +287,7 @@ sys.stdout.buffer.write(response.content)
         tags: Arc<RwLock<Vec<FilterOption>>>,
         uploaders: Arc<RwLock<Vec<FilterOption>>>,
     ) -> Result<()> {
+        let mut requester = Requester::new();
         let link_selector = Self::selector("a[href]")?;
         let article_selector = Self::selector("article.entry-tpl-grid, article.post")?;
         let pages = vec![
@@ -328,7 +299,7 @@ sys.stdout.buffer.write(response.content)
         ];
 
         for url in pages {
-            let html = match Self::fetch_html(&url).await {
+            let html = match Self::fetch_html(&mut requester, &url, None).await {
                 Ok(html) => html,
                 Err(error) => {
                     report_provider_error_background(
@@ -648,59 +619,64 @@ sys.stdout.buffer.write(response.content)
         Ok(fragments)
     }
 
-    async fn resolve_myvidplay_stream(&self, iframe_url: &str) -> Result<String> {
+    async fn resolve_myvidplay_stream(
-        let iframe_url = iframe_url.to_string();
+        &self,
-        let output = tokio::task::spawn_blocking(move || {
+        requester: &mut Requester,
-            Command::new("python3")
+        iframe_url: &str,
-                .arg("-c")
+        referer: &str,
-                .arg(
+    ) -> Result<String> {
-                    r#"
+        let html = Self::fetch_html(requester, iframe_url, Some(referer)).await?;
-import re
+        let pass_regex = Self::regex(r#"\$\.get\(\s*['"](/pass_md5/[^'"]+)['"]"#)?;
-import sys
+        let path = pass_regex
-import time
+            .captures(&html)
-from curl_cffi import requests
+            .and_then(|captures| captures.get(1).map(|value| value.as_str().to_string()))
+            .ok_or_else(|| Error::from("myvidplay resolution failed: missing pass_md5 path"))?;
 
-iframe_url = sys.argv[1]
+        let token = path
-session = requests.Session(impersonate="chrome")
+            .trim_end_matches('/')
-html = session.get(iframe_url, timeout=30).text
+            .rsplit('/')
-match = re.search(r"\$\.get\(\s*['\"](/pass_md5/[^'\"]+)['\"]", html)
+            .next()
-if not match:
+            .unwrap_or_default()
-    sys.stderr.write("missing pass_md5 path\n")
+            .to_string();
-    sys.exit(1)
+        if token.is_empty() {
-path = match.group(1)
+            return Err(Error::from(
-token = path.rstrip("/").split("/")[-1]
+                "myvidplay resolution failed: missing pass_md5 token".to_string(),
-if not token:
+            ));
-    sys.stderr.write("missing pass_md5 token\n")
+        }
-    sys.exit(1)
-if path.startswith("http://") or path.startswith("https://"):
-    pass_url = path
-else:
-    pass_url = "/".join(iframe_url.split("/")[:3]) + path
-base = session.get(pass_url, headers={"Referer": iframe_url}, timeout=30).text.strip()
-if not base or base == "RELOAD" or not base.startswith("http"):
-    sys.stderr.write(f"unusable pass_md5 response: {base[:120]}\n")
-    sys.exit(1)
-chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789"
-now = int(time.time() * 1000)
-suffix = "".join(chars[(now + i * 17) % len(chars)] for i in range(10))
-sys.stdout.write(f"{base}{suffix}?token={token}&expiry={now}")
-"#,
-                )
-                .arg(iframe_url)
-                .output()
-        })
-        .await
-        .map_err(|error| Error::from(format!("spawn_blocking failed: {error}")))?
-        .map_err(|error| Error::from(format!("python3 execution failed: {error}")))?;
 
-        if !output.status.success() {
+        let pass_url = if path.starts_with("http://") || path.starts_with("https://") {
-            let stderr = String::from_utf8_lossy(&output.stderr).trim().to_string();
+            path
+        } else {
+            let base = url::Url::parse(iframe_url)
+                .map_err(|error| Error::from(format!("invalid iframe url: {error}")))?;
+            base.join(&path)
+                .map_err(|error| Error::from(format!("invalid pass_md5 url: {error}")))?
+                .to_string()
+        };
+
+        let base = Self::fetch_html(requester, &pass_url, Some(iframe_url))
+            .await?
+            .trim()
+            .to_string();
+        if base.is_empty() || base == "RELOAD" || !base.starts_with("http") {
             return Err(Error::from(format!(
-                "myvidplay resolution failed: {stderr}"
+                "myvidplay resolution failed: unusable pass_md5 response: {}",
+                &base.chars().take(120).collect::<String>()
             )));
         }
 
-        let resolved = String::from_utf8_lossy(&output.stdout).trim().to_string();
+        let chars = b"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789";
+        let now = SystemTime::now()
+            .duration_since(UNIX_EPOCH)
+            .map_err(|error| Error::from(format!("time error: {error}")))?
+            .as_millis();
+        let suffix = (0..10)
+            .map(|index| {
+                let pos = ((now + (index as u128 * 17)) % chars.len() as u128) as usize;
+                chars[pos] as char
+            })
+            .collect::<String>();
+        let resolved = format!("{base}{suffix}?token={token}&expiry={now}");
         if resolved.is_empty() || !resolved.starts_with("http") {
             return Err(Error::from(
                 "myvidplay resolution returned empty url".to_string(),
@@ -725,7 +701,7 @@ sys.stdout.write(f"{base}{suffix}?token={token}&expiry={now}")
         html: &str,
         page_url: &str,
         options: &ServerOptions,
-        _requester: &mut Requester,
+        requester: &mut Requester,
     ) -> Result<VideoItem> {
         let (
             parsed_title,
@@ -864,7 +840,10 @@ sys.stdout.write(f"{base}{suffix}?token={token}&expiry={now}")
             });
 
             if iframe_url.contains("myvidplay.com") {
-                match self.resolve_myvidplay_stream(&iframe_url).await {
+                match self
+                    .resolve_myvidplay_stream(requester, &iframe_url, page_url)
+                    .await
+                {
                     Ok(stream_url) => {
                         item.url = stream_url.clone();
                         let mut format = VideoFormat::new(
@@ -919,7 +898,7 @@ sys.stdout.write(f"{base}{suffix}?token={token}&expiry={now}")
             None => Requester::new(),
         };
 
-        let html = match Self::fetch_with_curl_cffi(&page_url, None).await {
+        let html = match Self::fetch_html(&mut requester, &page_url, None).await {
             Ok(html) => html,
             Err(error) => {
                 report_provider_error_background(
@@ -959,10 +938,10 @@ sys.stdout.write(f"{base}{suffix}?token={token}&expiry={now}")
             }
         }
 
-        let _requester =
+        let mut requester =
             crate::providers::requester_or_default(options, "porndish", "missing_requester");
 
-        let html = match Self::fetch_with_curl_cffi(&url, None).await {
+        let html = match Self::fetch_html(&mut requester, &url, None).await {
             Ok(html) => html,
             Err(error) => {
                 report_provider_error(

src/proxies/porndishthumb.rs

@@ -1,15 +1,14 @@
-use ntex::http::header::CONTENT_TYPE;
+use ntex::http::header::{CONTENT_LENGTH, CONTENT_TYPE};
 use ntex::{
     http::Response,
     web::{self, HttpRequest, error},
 };
-use std::process::Command;
 
 use crate::util::requester::Requester;
 
 pub async fn get_image(
     req: HttpRequest,
-    _requester: web::types::State<Requester>,
+    requester: web::types::State<Requester>,
 ) -> Result<impl web::Responder, web::Error> {
     let endpoint = req.match_info().query("endpoint").to_string();
     let image_url = if endpoint.starts_with("http://") || endpoint.starts_with("https://") {
@@ -18,45 +17,41 @@ pub async fn get_image(
         format!("https://{}", endpoint.trim_start_matches('/'))
     };
 
-    let output = tokio::task::spawn_blocking(move || {
+    let upstream = match requester
-        Command::new("python3")
+        .get_ref()
-            .arg("-c")
+        .clone()
-            .arg(
+        .get_raw_with_headers(
-                r#"
+            image_url.as_str(),
-import sys
+            vec![(
-from curl_cffi import requests
+                "Referer".to_string(),
-
+                "https://www.porndish.com/".to_string(),
-url = sys.argv[1]
+            )],
-response = requests.get(
-    url,
-    impersonate="chrome",
-    timeout=30,
-    allow_redirects=True,
-    headers={"Referer": "https://www.porndish.com/"},
         )
-if response.status_code >= 400:
-    sys.stderr.write(f"status={response.status_code}\n")
-    sys.exit(1)
-sys.stderr.write(response.headers.get("content-type", "application/octet-stream"))
-sys.stdout.buffer.write(response.content)
-"#,
-            )
-            .arg(image_url)
-            .output()
-    })
         .await
-    .map_err(error::ErrorBadGateway)?
+    {
-    .map_err(error::ErrorBadGateway)?;
+        Ok(response) => response,
+        Err(_) => return Ok(web::HttpResponse::NotFound().finish()),
+    };
 
-    if !output.status.success() {
+    let status = upstream.status();
+    let headers = upstream.headers().clone();
+    let bytes = upstream.bytes().await.map_err(error::ErrorBadGateway)?;
+
+    if !status.is_success() {
         return Ok(web::HttpResponse::NotFound().finish());
     }
 
-    let content_type = String::from_utf8_lossy(&output.stderr).trim().to_string();
+    let mut resp = Response::build(status);
-    let mut resp = Response::build(ntex::http::StatusCode::OK);
+    if let Some(ct) = headers.get(CONTENT_TYPE) {
-    if !content_type.is_empty() {
+        if let Ok(ct_str) = ct.to_str() {
-        resp.set_header(CONTENT_TYPE, content_type);
+            resp.set_header(CONTENT_TYPE, ct_str);
+        }
+    }
+    if let Some(cl) = headers.get(CONTENT_LENGTH) {
+        if let Ok(cl_str) = cl.to_str() {
+            resp.set_header(CONTENT_LENGTH, cl_str);
+        }
     }
 
-    Ok(resp.body(output.stdout))
+    Ok(resp.body(bytes.to_vec()))
 }

supervisord/hottub.sh

@@ -1 +1,5 @@
-/app/target/release/hottub
+#!/usr/bin/env bash
+set -euo pipefail
+
+cd /app
+exec cargo run --release