immich-automated-selfie-tim.../src/job/mod.rs

585 lines
20 KiB
Rust

//! Background job processing module.
//!
//! Handles the complete pipeline:
//! 1. Fetching assets from Immich for a specific person
//! 2. Downloading and processing images in parallel
//! 3. Processing images through the extensible pipeline
//! 4. Compiling processed images into a timelapse video
mod processing;
use crate::config::{TimeIntervalConfig, TimeRange};
use crate::error::{Error, Result};
use crate::immich_api::{Asset, FaceData, ImmichClient};
use crate::pipeline::Pipeline;
use crate::utils::sanitize_folder_name;
use crate::video::compile_timelapse;
use crate::web::{AppState, AtomicSkipStats, JobStatus, Progress, SkipStats};
use processing::{process_single_asset, AssetProcessResult, DebugDirs, OutputDirs};
use chrono::NaiveDate;
use std::collections::HashMap;
use std::path::PathBuf;
use std::sync::atomic::{AtomicU32, Ordering};
use std::sync::{Arc, RwLock};
use tokio::sync::Semaphore;
use tokio_util::sync::CancellationToken;
/// Tracks per-time-bucket photo counts for limiting.
///
/// Thread-safe: uses atomic counters with fetch_add + undo pattern
/// so concurrent workers can claim slots without races.
pub struct TimeIntervalTracker {
max_photos: u32,
time_range: TimeRange,
buckets: RwLock<HashMap<String, Arc<AtomicU32>>>,
}
impl TimeIntervalTracker {
/// Create a tracker if photo limiting is enabled, otherwise return None.
pub fn new(config: &TimeIntervalConfig) -> Option<Self> {
if !config.enabled {
return None;
}
Some(Self {
max_photos: config.max_photos,
time_range: config.time_range,
buckets: RwLock::new(HashMap::new()),
})
}
/// Try to claim a slot for the given timestamp.
/// Returns true if the slot was claimed, false if the bucket is full.
pub fn try_claim(&self, timestamp: &str) -> bool {
let key = self.bucket_key(timestamp);
// Get or create the atomic counter for this bucket
let counter = {
// Try read lock first
let buckets = self.buckets.read().expect("bucket lock poisoned");
if let Some(counter) = buckets.get(&key) {
counter.clone()
} else {
drop(buckets);
let mut buckets = self.buckets.write().expect("bucket lock poisoned");
buckets
.entry(key)
.or_insert_with(|| Arc::new(AtomicU32::new(0)))
.clone()
}
};
// Atomically try to claim: increment, check, undo if over limit
let prev = counter.fetch_add(1, Ordering::SeqCst);
if prev < self.max_photos {
true
} else {
counter.fetch_sub(1, Ordering::SeqCst);
false
}
}
/// Check if the bucket for the given timestamp is already full.
/// This is a non-mutating check used to skip processing early.
pub fn is_full(&self, timestamp: &str) -> bool {
let key = self.bucket_key(timestamp);
let buckets = self.buckets.read().expect("bucket lock poisoned");
if let Some(counter) = buckets.get(&key) {
counter.load(Ordering::SeqCst) >= self.max_photos
} else {
false
}
}
/// Compute the bucket key from a timestamp string.
fn bucket_key(&self, timestamp: &str) -> String {
// Parse date from ISO 8601 timestamp (e.g. "2024-01-15" or "2024-01-15T12:34:56Z")
let date_str = timestamp.get(..10).unwrap_or(timestamp);
let date = NaiveDate::parse_from_str(date_str, "%Y-%m-%d").unwrap_or_default();
match self.time_range {
TimeRange::Day => date.format("%Y-%m-%d").to_string(),
TimeRange::Week => {
use chrono::Datelike;
format!("{}-W{:02}", date.iso_week().year(), date.iso_week().week())
}
TimeRange::Month => date.format("%Y-%m").to_string(),
}
}
}
/// Parameters for starting a processing job.
#[derive(Debug, Clone)]
pub struct JobParams {
pub person_id: String,
pub person_name: Option<String>,
pub date_from: Option<String>,
pub date_to: Option<String>,
}
/// Run the complete processing pipeline.
///
/// This is the main entry point for background job processing.
/// It handles progress reporting and cancellation.
pub async fn run_job(state: AppState, params: JobParams, cancel_token: CancellationToken) {
let result = run_job_inner(state.clone(), params, cancel_token).await;
// Clear the cancellation token when done
state.clear_cancel_token().await;
// Get final state from progress (skip stats and person info)
let final_progress = state.progress.read().await.clone();
match result {
Ok(output_path) => {
state
.update_progress(Progress {
status: JobStatus::Completed,
completed: final_progress.completed,
total: final_progress.total,
message: Some(format!(
"Video saved to: {}\
<br><i>Please note that for the best \
looking results, you should do a manual pass in the gallery view \
(button available in the section below)</i>",
output_path.display()
)),
skip_stats: final_progress.skip_stats,
person_id: final_progress.person_id,
person_name: final_progress.person_name,
})
.await;
}
Err(Error::Cancelled) => {
state
.update_progress(Progress {
status: JobStatus::Cancelled,
completed: final_progress.completed,
total: final_progress.total,
message: Some("Job cancelled by user".to_string()),
skip_stats: final_progress.skip_stats,
person_id: final_progress.person_id,
person_name: final_progress.person_name,
})
.await;
}
Err(e) => {
tracing::error!("Job failed: {}", e);
let friendly = e.user_message();
state
.update_progress(Progress {
status: JobStatus::Error(friendly.clone()),
completed: final_progress.completed,
total: final_progress.total,
message: Some(friendly),
skip_stats: final_progress.skip_stats,
person_id: final_progress.person_id,
person_name: final_progress.person_name,
})
.await;
}
}
}
/// Inner implementation that returns Result for easier error handling.
async fn run_job_inner(
state: AppState,
params: JobParams,
cancel_token: CancellationToken,
) -> Result<PathBuf> {
let config = state.config.read().await.clone();
// Create person-specific output directory
let folder_name = sanitize_folder_name(params.person_name.as_deref(), &params.person_id);
let person_dir = config.output_dir.join(&folder_name);
// If the folder exists, delete all its contents to start fresh
if person_dir.exists() {
tracing::info!("Removing existing folder: {}", person_dir.display());
tokio::fs::remove_dir_all(&person_dir).await?;
}
// Create output directories
let images_dir = person_dir.join("images");
tokio::fs::create_dir_all(&images_dir)
.await
.map_err(|e| permission_aware_io_error(e, &images_dir))?;
// Create debug directories if enabled
let debug = if config.processing.output.keep_intermediates {
let debug_base = person_dir.join("debug");
tokio::fs::create_dir_all(&debug_base).await?;
Some(DebugDirs { base: debug_base })
} else {
None
};
let video_filename = format!("{}.mp4", folder_name);
let output_dirs = OutputDirs {
images: images_dir.clone(),
video: person_dir.join(&video_filename),
debug,
};
tracing::info!("Output directory: {}", person_dir.display());
// Create Immich client
let client = ImmichClient::new(&config.api)?;
// Update progress: fetching assets
state
.update_progress(Progress {
status: JobStatus::Running,
completed: 0,
total: 0,
message: Some("Fetching assets from Immich...".to_string()),
skip_stats: SkipStats::default(),
person_id: Some(params.person_id.clone()),
person_name: params.person_name.clone(),
})
.await;
// Check for cancellation
if cancel_token.is_cancelled() {
return Err(Error::Cancelled);
}
// Fetch all assets for this person
let assets = client
.get_assets_with_person(
&params.person_id,
params.date_from.as_deref(),
params.date_to.as_deref(),
)
.await?;
if assets.is_empty() {
return Err(Error::ImageProcessing(
"No assets found for this person".to_string(),
));
}
tracing::info!("Found {} assets to process", assets.len());
// Filter assets that have face data for the target person
let assets_with_faces: Vec<(Asset, FaceData)> = assets
.into_iter()
.filter_map(|asset| {
let face = find_face_for_person(&asset, &params.person_id)?;
Some((asset, face))
})
.collect();
if assets_with_faces.is_empty() {
return Err(Error::ImageProcessing(
"No assets with face data found".to_string(),
));
}
let total = assets_with_faces.len() as u32;
tracing::info!("{} assets have face data", total);
// Update progress with total count
state
.update_progress(Progress {
status: JobStatus::Running,
completed: 0,
total,
message: Some(format!("Processing {} images...", total)),
skip_stats: SkipStats::default(),
person_id: Some(params.person_id.clone()),
person_name: params.person_name.clone(),
})
.await;
// Create the processing pipeline based on config
let pipeline = Arc::new(Pipeline::with_steps_from_config(&config));
tracing::debug!("Pipeline steps: {:?}", pipeline.step_ids());
// Create photo limit tracker if enabled
let time_interval_tracker =
TimeIntervalTracker::new(&config.processing.time_interval).map(Arc::new);
// Process images in parallel with concurrency limit
let completed = Arc::new(AtomicU32::new(0));
let semaphore = Arc::new(Semaphore::new(config.processing.max_workers));
let client = Arc::new(client);
let config = Arc::new(config);
let output_dirs = Arc::new(output_dirs);
// Atomic counters for real-time skip statistics (consolidated into single struct)
let skip_stats = Arc::new(AtomicSkipStats::new());
let mut handles = Vec::with_capacity(assets_with_faces.len());
// Clone person info for use in spawned tasks
let task_person_id = params.person_id.clone();
let task_person_name = params.person_name.clone();
for (asset, face_data) in assets_with_faces {
// Check for cancellation before spawning more tasks
if cancel_token.is_cancelled() {
break;
}
let permit = match semaphore.clone().acquire_owned().await {
Ok(permit) => permit,
Err(_) => continue, // Semaphore closed, skip remaining tasks
};
let client = client.clone();
let config = config.clone();
let output_dirs = output_dirs.clone();
let completed = completed.clone();
let state = state.clone();
let task_cancel_token = cancel_token.clone();
let pipeline = pipeline.clone();
// Clone skip stats for this task
let skip_stats = skip_stats.clone();
// Clone photo limit tracker for this task
let time_interval_tracker = time_interval_tracker.clone();
// Clone person info for this task
let person_id = task_person_id.clone();
let person_name = task_person_name.clone();
let handle = tokio::spawn(async move {
// Check cancellation at start of task
if task_cancel_token.is_cancelled() {
drop(permit);
return AssetProcessResult::Cancelled {
asset_id: asset.id.clone(),
};
}
let result = process_single_asset(
&client,
&config,
&asset,
&face_data,
&output_dirs,
&task_cancel_token,
&skip_stats,
&pipeline,
time_interval_tracker.as_ref(),
)
.await;
// Update progress with current skip stats (only if not cancelled)
if !task_cancel_token.is_cancelled() {
let done = completed.fetch_add(1, Ordering::SeqCst) + 1;
let current_skip_stats = skip_stats.snapshot();
state
.update_progress(Progress {
status: JobStatus::Running,
completed: done,
total,
message: Some(format!("Processing images... ({}/{})", done, total)),
skip_stats: current_skip_stats,
person_id: Some(person_id.clone()),
person_name: person_name.clone(),
})
.await;
}
drop(permit);
result
});
handles.push(handle);
}
// Wait for all tasks to complete
let mut results = Vec::with_capacity(handles.len());
for handle in handles {
match handle.await {
Ok(result) => results.push(result),
Err(e) => {
tracing::error!("Task panicked: {:?}", e);
}
}
}
// Check if we were cancelled
if cancel_token.is_cancelled() {
return Err(Error::Cancelled);
}
// Get final skip statistics from atomic counters
let final_skip_stats = skip_stats.snapshot();
// Count results and log errors (deduplicated)
let mut successful = 0usize;
let mut errors = 0usize;
let mut first_error: Option<String> = None;
let mut all_same_error = true;
for result in &results {
match result {
AssetProcessResult::Success { .. } => successful += 1,
AssetProcessResult::Error { error, .. } => {
errors += 1;
match &first_error {
None => first_error = Some(error.clone()),
Some(first) if first != error => all_same_error = false,
_ => {}
}
}
_ => {}
}
}
if let Some(ref err) = first_error {
if all_same_error {
tracing::error!("{} assets failed with: {}", errors, err);
} else {
// Log distinct errors individually (up to 3)
let mut logged = 0;
for result in &results {
if let AssetProcessResult::Error { asset_id, error } = result {
if logged < 3 {
tracing::error!("Asset {} failed: {}", asset_id, error);
logged += 1;
}
}
}
if errors > 3 {
tracing::error!("... and {} more errors", errors - 3);
}
}
}
let skipped = final_skip_stats.total();
tracing::info!(
"Processing complete: {} successful, {} skipped, {} errors",
successful,
skipped,
errors
);
// Update progress with final skip stats
state
.update_progress(Progress {
status: JobStatus::Running,
completed: total,
total,
message: Some("Processing complete, preparing video...".to_string()),
skip_stats: final_skip_stats.clone(),
person_id: Some(params.person_id.clone()),
person_name: params.person_name.clone(),
})
.await;
if successful == 0 {
let detail = first_error
.map(|e| format!(" First error: {}", e))
.unwrap_or_default();
return Err(Error::ImageProcessing(format!(
"No images were successfully processed ({} errors).{}",
errors, detail
)));
}
// Check for cancellation before video compilation
if cancel_token.is_cancelled() {
return Err(Error::Cancelled);
}
// Compile video if enabled
if config.video.enabled {
state
.update_progress(Progress {
status: JobStatus::CompilingVideo,
completed: 0,
total: successful as u32,
message: Some("Compiling video...".to_string()),
skip_stats: final_skip_stats.clone(),
person_id: Some(params.person_id.clone()),
person_name: params.person_name.clone(),
})
.await;
compile_timelapse(
&output_dirs.images,
&output_dirs.video,
&config.video,
Some(&cancel_token),
|frame, total| {
// Sync callback - just log progress. State updates would require
// async context or channels, which adds complexity for minimal benefit
// since video compilation is typically fast.
tracing::debug!("Video progress: {}/{}", frame, total);
},
)
.await?;
Ok(output_dirs.video.clone())
} else {
Ok(output_dirs.images.clone())
}
}
/// Convert a permission-denied I/O error into a Config error with Docker fix instructions.
fn permission_aware_io_error(e: std::io::Error, path: &std::path::Path) -> Error {
if e.kind() == std::io::ErrorKind::PermissionDenied {
Error::Config(format!(
"Permission denied for '{}'. {}",
path.display(),
crate::error::PERMISSION_HINT
))
} else {
Error::Io(e)
}
}
/// Find the face data for a specific person in an asset.
fn find_face_for_person(asset: &Asset, person_id: &str) -> Option<FaceData> {
let people = asset.people.as_ref()?;
for person in people {
if person.id == person_id {
if let Some(faces) = &person.faces {
// Return the first face (Immich shouldn't detect the same person more than once per image)
return faces.first().cloned();
}
}
}
None
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_find_face_for_person() {
use crate::immich_api::PersonWithFaces;
let asset = Asset {
id: "asset1".to_string(),
device_asset_id: None,
original_file_name: None,
file_created_at: Some("2024-01-15T10:30:00Z".to_string()),
local_date_time: None,
people: Some(vec![PersonWithFaces {
id: "person1".to_string(),
name: Some("Test Person".to_string()),
faces: Some(vec![FaceData {
bounding_box_x1: 0.2,
bounding_box_y1: 0.1,
bounding_box_x2: 0.5,
bounding_box_y2: 0.6,
image_width: 1920,
image_height: 1080,
}]),
}]),
};
let face = find_face_for_person(&asset, "person1");
assert!(face.is_some());
let face = find_face_for_person(&asset, "person2");
assert!(face.is_none());
}
}