mirror of
https://github.com/euzu/tuliprox.git
synced 2026-10-30 18:52:31 +01:00
euzuandGitHub
99e5044599
* **New Features** * Added HLS live and catch-up playback for archived streams, manifests, segments, and provider failover. * Improved playlist processing with filtering, renaming, mapping, EPG handling, favourites, watch groups, staged overlays, and disk-based processing. * **Documentation** * Added guidance for environment variables, credentials, password hashes, and authentication secrets. * **Changes** * Removed target editing from the target menu; refresh and delete remain available. * **Bug Fixes** * Improved recovery when cached playlist data is unreadable and strengthened archive stream URL handling.
71 lines
2.8 KiB
Rust
71 lines
2.8 KiB
Rust
//! A stream probe that did not come back with metadata.
|
|
//!
|
|
//! Failed `ffprobe` outcomes are persisted and published so notification
|
|
//! subscribers can alert operators when a provider stream becomes unavailable.
|
|
//!
|
|
//! # Why there is no success variant
|
|
//!
|
|
//! A metadata run probes every stream it does not already know about, so a
|
|
//! success event would fire thousands of times per refresh and carry news
|
|
//! nobody asked for. The failure is the occurrence; the success is the
|
|
//! expected case.
|
|
|
|
use serde::{Deserialize, Serialize};
|
|
use std::sync::Arc;
|
|
|
|
/// Why a probe produced no metadata.
|
|
///
|
|
/// Mirrors the actionable half of
|
|
/// `tuliprox_core::utils::ffmpeg::ProbeFailureKind`. Its `Cancelled` variant
|
|
/// has no counterpart here on purpose: a cancelled probe is a shutdown or a
|
|
/// preemption, not a statement about the stream, and reporting it as a
|
|
/// failure would teach subscribers to act on server lifecycle noise.
|
|
#[derive(Debug, Copy, Clone, PartialEq, Eq, Serialize, Deserialize)]
|
|
#[serde(rename_all = "snake_case")]
|
|
pub enum StreamProbeFailureReason {
|
|
/// The provider answered, with a 404.
|
|
NotFound,
|
|
/// The probe failed or timed out.
|
|
Unreachable,
|
|
}
|
|
|
|
/// One stream failed to probe.
|
|
///
|
|
/// # Redaction
|
|
///
|
|
/// `url` must be passed through
|
|
/// [`sanitize_sensitive_info`](crate::utils::sanitize_sensitive_info) by the
|
|
/// emitter: a provider stream URL carries the account username and password
|
|
/// in its path or query, and this record is rendered into notification
|
|
/// channels. The field is documented rather than typed as pre-sanitized
|
|
/// because the sanitizer is configuration-dependent, and a newtype implying
|
|
/// a guarantee the emitter may not have honoured would be worse than a
|
|
/// contract stated once.
|
|
#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
|
|
pub struct StreamProbeFailure {
|
|
/// The input whose stream this is.
|
|
pub input: Arc<str>,
|
|
/// The stream's id within that input, for correlating with the store.
|
|
pub unique_id: Arc<str>,
|
|
/// The probed URL, already sanitized. See the note above.
|
|
pub url: Arc<str>,
|
|
pub reason: StreamProbeFailureReason,
|
|
}
|
|
|
|
impl StreamProbeFailure {
|
|
#[must_use]
|
|
pub fn new(input: Arc<str>, unique_id: Arc<str>, url: Arc<str>, reason: StreamProbeFailureReason) -> Self {
|
|
Self { input, unique_id, url, reason }
|
|
}
|
|
|
|
/// Deliberately per *input*, not per stream.
|
|
///
|
|
/// When a provider goes down, every stream behind it fails in the same
|
|
/// run. An operator wants to hear "this input is failing" once; a
|
|
/// per-stream key would deliver one notification per channel and bury
|
|
/// the signal. Plugins still see every event on the bus — this only
|
|
/// bounds what reaches a phone.
|
|
#[must_use]
|
|
pub fn dedup_key(&self) -> String { format!("stream-probe:{}", self.input) }
|
|
}
|