Phase 5 — PoTokenProvider trait + stream_extractor wiring

Mirrors NPE PoTokenProvider.java + PoTokenResult.java; defines the
host-injection surface for BotGuard attestation. The Rust crate stays
out of the BotGuard business — embedders (Straw on Android, future
Sulkta CLI via Browserless, etc.) supply their own impl.

src/youtube/potoken/mod.rs
  * PoTokenResult { player_request_po_token, streaming_data_po_token,
                    visitor_data }  + ::new + ::single constructors
  * PoTokenError (Unavailable, MintFailed) — FIX vs NPE: split 'declined'
    (Ok(None)) from 'errored' (Err) so callers can react differently
  * trait PoTokenProvider with 4 client-scoped methods; default impl
    returns Ok(None) so embedders can override just what they support
  * set_po_token_provider / clear_po_token_provider / po_token_provider
    static registration via RwLock<Option<Arc<dyn PoTokenProvider>>>

src/youtube/potoken/noop.rs
  * NoopPoTokenProvider — safe default

src/youtube/stream_extractor.rs
  * resolve_po_token via options-first-then-provider helper
    (options_or_provider)
  * Android branch: pulls player_request_po_token + visitor_data into
    /player body, streams streaming_data_po_token through to URL &pot=
  * iOS branch: same shape, gated on fetch_ios_client AND non-empty
    provider result

Kotlin side (PoTokenWebView lift into Straw via UniFFI's foreign-trait
bridge) is separate work — strawcore just owns the contract.

Tests: 77 lib unit pass (+4 since Phase 4) + 7 Phase 2 offline + 7
Phase 4 offline = 91 green.
This commit is contained in:
Kayos 2026-05-24 17:10:13 -07:00
parent a47e142ab7
commit b4286b8236
4 changed files with 271 additions and 8 deletions

View file

@ -8,6 +8,7 @@ pub mod constants;
pub mod itag; pub mod itag;
pub mod js; pub mod js;
pub mod parsing; pub mod parsing;
pub mod potoken;
pub mod stream_extractor; pub mod stream_extractor;
pub mod stream_helper; pub mod stream_helper;

195
src/youtube/potoken/mod.rs Normal file
View file

@ -0,0 +1,195 @@
// PoTokenProvider — the BotGuard / attestation surface NPE delegates to
// the embedder. Mirrors NPE PoTokenProvider.java + PoTokenResult.java.
//
// Architecturally: BotGuard runs in a real browser-y environment (WebView
// on Android). NPE doesn't ship a BotGuard impl — it asks the host app to
// produce po_tokens via this trait. Straw's Kotlin side will register an
// adapter that calls into PoTokenWebView (lifted verbatim from NewPipe
// app under GPL-3.0 compat).
//
// Two distinct token strings per call (audit Track E):
// * player_request_po_token — goes into JSON body
// serviceIntegrityDimensions.poToken
// * streaming_data_po_token — goes into URL &pot=<...>
// They CAN differ.
//
// visitor_data: the visitor token the po_token was minted against. The
// player request MUST send the same visitorData in context.client, or
// YT 403's the streaming URLs.
//
// FIX (audit Track E §2.2): NPE's Java API returns null both for
// "provider declined" and "provider errored." We split: Ok(None) =
// declined (no provider available for this client / video), Err = the
// provider tried and failed (the caller should still attempt extraction
// without po_token).
pub mod noop;
use std::sync::Arc;
use once_cell::sync::Lazy;
use parking_lot::RwLock;
use thiserror::Error;
pub use noop::NoopPoTokenProvider;
#[derive(Clone, Debug, Eq, PartialEq)]
pub struct PoTokenResult {
pub player_request_po_token: String,
pub streaming_data_po_token: String,
pub visitor_data: String,
}
impl PoTokenResult {
pub fn new(
player_request_po_token: impl Into<String>,
streaming_data_po_token: impl Into<String>,
visitor_data: impl Into<String>,
) -> Self {
Self {
player_request_po_token: player_request_po_token.into(),
streaming_data_po_token: streaming_data_po_token.into(),
visitor_data: visitor_data.into(),
}
}
/// Convenience for clients that mint a single token used for both
/// the JSON body and the URL `&pot=`.
pub fn single(token: impl Into<String>, visitor_data: impl Into<String>) -> Self {
let token = token.into();
Self {
streaming_data_po_token: token.clone(),
player_request_po_token: token,
visitor_data: visitor_data.into(),
}
}
}
#[derive(Debug, Error)]
pub enum PoTokenError {
#[error("po_token provider unavailable: {0}")]
Unavailable(String),
#[error("po_token mint failed: {0}")]
MintFailed(String),
}
pub trait PoTokenProvider: Send + Sync {
/// po_token for the WEB client. Returns Ok(None) when the provider
/// doesn't support this client (e.g. no browser available).
fn get_web_client_po_token(
&self,
_video_id: &str,
) -> Result<Option<PoTokenResult>, PoTokenError> {
Ok(None)
}
/// po_token for the WEB_EMBEDDED_PLAYER client.
fn get_web_embedded_client_po_token(
&self,
_video_id: &str,
) -> Result<Option<PoTokenResult>, PoTokenError> {
Ok(None)
}
/// po_token for the ANDROID client. NPE's primary use case — paired
/// with `getAndroidPlayerResponse()` to satisfy YT's anti-bot check.
fn get_android_client_po_token(
&self,
_video_id: &str,
) -> Result<Option<PoTokenResult>, PoTokenError> {
Ok(None)
}
/// po_token for the IOS client. Optional — iOS can extract without
/// a po_token but URL count + quality is reduced.
fn get_ios_client_po_token(
&self,
_video_id: &str,
) -> Result<Option<PoTokenResult>, PoTokenError> {
Ok(None)
}
}
static REGISTERED_PROVIDER: Lazy<RwLock<Option<Arc<dyn PoTokenProvider>>>> =
Lazy::new(|| RwLock::new(None));
pub fn set_po_token_provider(provider: Arc<dyn PoTokenProvider>) {
*REGISTERED_PROVIDER.write() = Some(provider);
}
pub fn clear_po_token_provider() {
*REGISTERED_PROVIDER.write() = None;
}
pub fn po_token_provider() -> Option<Arc<dyn PoTokenProvider>> {
REGISTERED_PROVIDER.read().clone()
}
#[cfg(test)]
mod tests {
use super::*;
struct CountingProvider {
android_calls: std::sync::atomic::AtomicU32,
}
impl PoTokenProvider for CountingProvider {
fn get_android_client_po_token(
&self,
_video_id: &str,
) -> Result<Option<PoTokenResult>, PoTokenError> {
self.android_calls
.fetch_add(1, std::sync::atomic::Ordering::SeqCst);
Ok(Some(PoTokenResult::single("mock-pot", "mock-visitor")))
}
}
#[test]
fn provider_registers_and_resolves() {
let p = Arc::new(CountingProvider {
android_calls: 0.into(),
});
set_po_token_provider(p.clone());
let got = po_token_provider().expect("registered");
let r = got.get_android_client_po_token("vid").unwrap().unwrap();
assert_eq!(r.streaming_data_po_token, "mock-pot");
assert_eq!(r.player_request_po_token, "mock-pot");
assert_eq!(r.visitor_data, "mock-visitor");
assert_eq!(
p.android_calls.load(std::sync::atomic::Ordering::SeqCst),
1
);
clear_po_token_provider();
}
#[test]
fn provider_default_decline() {
struct DeclineEverything;
impl PoTokenProvider for DeclineEverything {}
set_po_token_provider(Arc::new(DeclineEverything));
let p = po_token_provider().unwrap();
assert!(p.get_android_client_po_token("vid").unwrap().is_none());
assert!(p.get_ios_client_po_token("vid").unwrap().is_none());
assert!(p.get_web_client_po_token("vid").unwrap().is_none());
assert!(p.get_web_embedded_client_po_token("vid").unwrap().is_none());
clear_po_token_provider();
}
#[test]
fn po_token_result_constructors() {
let r = PoTokenResult::new("a", "b", "c");
assert_eq!(r.player_request_po_token, "a");
assert_eq!(r.streaming_data_po_token, "b");
let r = PoTokenResult::single("xx", "yy");
assert_eq!(r.player_request_po_token, "xx");
assert_eq!(r.streaming_data_po_token, "xx");
assert_eq!(r.visitor_data, "yy");
}
#[test]
fn unset_provider_returns_none() {
clear_po_token_provider();
assert!(po_token_provider().is_none());
}
}

View file

@ -0,0 +1,11 @@
// Default PoTokenProvider — always declines. The Rust crate isn't going
// to mint po_tokens by itself (BotGuard needs a WebView); embedders
// (Straw on Android, a future Sulkta CLI calling out to Browserless,
// etc.) provide their own impl. This noop is a safe default for tests
// and for code paths that don't require a po_token.
use crate::youtube::potoken::PoTokenProvider;
pub struct NoopPoTokenProvider;
impl PoTokenProvider for NoopPoTokenProvider {}

View file

@ -32,6 +32,7 @@ use crate::stream::{
}; };
use crate::youtube::itag::{lookup as itag_lookup, ItagType, MediaFormat}; use crate::youtube::itag::{lookup as itag_lookup, ItagType, MediaFormat};
use crate::youtube::js::PlayerManager; use crate::youtube::js::PlayerManager;
use crate::youtube::potoken::{po_token_provider, PoTokenResult};
use crate::youtube::stream_helper::{self, generate_content_playback_nonce}; use crate::youtube::stream_helper::{self, generate_content_playback_nonce};
#[derive(Clone, Copy, Debug, Eq, PartialEq)] #[derive(Clone, Copy, Debug, Eq, PartialEq)]
@ -64,14 +65,32 @@ pub fn stream_info_with(
let localization = NewPipe::preferred_localization(); let localization = NewPipe::preferred_localization();
let content_country = NewPipe::preferred_content_country(); let content_country = NewPipe::preferred_content_country();
// Resolve po_token via the registered provider if present, falling
// back to caller-supplied options. The trait split (Ok(None) vs
// Err) lets us treat "provider declined" as "go anonymous" and
// "provider errored" as "still try anonymous but log the failure."
let provider = po_token_provider();
let android_token: Option<PoTokenResult> = options_or_provider(
options.android_player_request_pot.as_deref(),
options.android_streaming_pot.as_deref(),
options.android_visitor_data.as_deref(),
|| {
provider
.as_ref()
.and_then(|p| p.get_android_client_po_token(video_id).ok().flatten())
},
);
let android_cpn = generate_content_playback_nonce(); let android_cpn = generate_content_playback_nonce();
let player_response = fetch_android( let player_response = fetch_android(
video_id, video_id,
&localization, &localization,
&content_country, &content_country,
&android_cpn, &android_cpn,
options.android_player_request_pot.as_deref(), android_token
options.android_visitor_data.as_deref(), .as_ref()
.map(|t| t.player_request_po_token.as_str()),
android_token.as_ref().map(|t| t.visitor_data.as_str()),
)?; )?;
check_playability_status(&player_response)?; check_playability_status(&player_response)?;
@ -87,6 +106,21 @@ pub fn stream_info_with(
.unwrap_or(Value::Null); .unwrap_or(Value::Null);
// Optional iOS — best-effort. // Optional iOS — best-effort.
let ios_token: Option<PoTokenResult> = if options.fetch_ios_client {
options_or_provider(
options.ios_player_request_pot.as_deref(),
options.ios_streaming_pot.as_deref(),
options.ios_visitor_data.as_deref(),
|| {
provider
.as_ref()
.and_then(|p| p.get_ios_client_po_token(video_id).ok().flatten())
},
)
} else {
None
};
let (ios_streaming_data, ios_cpn) = if options.fetch_ios_client { let (ios_streaming_data, ios_cpn) = if options.fetch_ios_client {
let ios_cpn = generate_content_playback_nonce(); let ios_cpn = generate_content_playback_nonce();
match stream_helper::get_ios_player_response( match stream_helper::get_ios_player_response(
@ -94,8 +128,8 @@ pub fn stream_info_with(
&localization, &localization,
&content_country, &content_country,
&ios_cpn, &ios_cpn,
options.ios_player_request_pot.as_deref(), ios_token.as_ref().map(|t| t.player_request_po_token.as_str()),
options.ios_visitor_data.as_deref(), ios_token.as_ref().map(|t| t.visitor_data.as_str()),
) { ) {
Ok(r) if !is_player_response_not_valid(&r, video_id) => ( Ok(r) if !is_player_response_not_valid(&r, video_id) => (
r.get("streamingData").cloned().unwrap_or(Value::Null), r.get("streamingData").cloned().unwrap_or(Value::Null),
@ -107,6 +141,15 @@ pub fn stream_info_with(
(Value::Null, None) (Value::Null, None)
}; };
let android_streaming_pot = android_token
.as_ref()
.map(|t| t.streaming_data_po_token.clone())
.or_else(|| options.android_streaming_pot.clone());
let ios_streaming_pot = ios_token
.as_ref()
.map(|t| t.streaming_data_po_token.clone())
.or_else(|| options.ios_streaming_pot.clone());
let signature_timestamp = PlayerManager::instance() let signature_timestamp = PlayerManager::instance()
.signature_timestamp(video_id) .signature_timestamp(video_id)
.unwrap_or(0); .unwrap_or(0);
@ -129,21 +172,34 @@ pub fn stream_info_with(
video_id, video_id,
&android_cpn, &android_cpn,
ios_cpn.as_deref(), ios_cpn.as_deref(),
options.android_streaming_pot.as_deref(), android_streaming_pot.as_deref(),
options.ios_streaming_pot.as_deref(), ios_streaming_pot.as_deref(),
)?; )?;
populate_manifests( populate_manifests(
&mut info, &mut info,
&android_streaming_data, &android_streaming_data,
&ios_streaming_data, &ios_streaming_data,
options.android_streaming_pot.as_deref(), android_streaming_pot.as_deref(),
options.ios_streaming_pot.as_deref(), ios_streaming_pot.as_deref(),
); );
populate_captions(&mut info, &player_response); populate_captions(&mut info, &player_response);
Ok(info) Ok(info)
} }
fn options_or_provider(
opt_player_token: Option<&str>,
opt_streaming_token: Option<&str>,
opt_visitor: Option<&str>,
provider_fn: impl FnOnce() -> Option<PoTokenResult>,
) -> Option<PoTokenResult> {
// Caller-supplied wins when ALL three are present.
if let (Some(p), Some(s), Some(v)) = (opt_player_token, opt_streaming_token, opt_visitor) {
return Some(PoTokenResult::new(p, s, v));
}
provider_fn()
}
fn fetch_android( fn fetch_android(
video_id: &str, video_id: &str,
localization: &Localization, localization: &Localization,