diff --git a/.github/workflows/ci.yml b/.forgejo/workflows/ci.yml similarity index 61% rename from .github/workflows/ci.yml rename to .forgejo/workflows/ci.yml index 7dabcc4..82ee799 100644 --- a/.github/workflows/ci.yml +++ b/.forgejo/workflows/ci.yml @@ -12,35 +12,26 @@ permissions: jobs: test: name: Test - runs-on: ubuntu-latest - strategy: - matrix: - go-version: ["1.25"] + runs-on: docker + container: + image: docker.io/library/golang:1.25 steps: - - uses: actions/checkout@v6 - - - name: Set up Go - uses: actions/setup-go@v6 - with: - go-version: ${{ matrix.go-version }} + - uses: actions/checkout@v4 - name: Run tests run: go test -v -race -count=1 ./... build: name: Build - runs-on: ubuntu-latest + runs-on: docker + container: + image: docker.io/library/golang:1.25 strategy: matrix: goos: [linux, darwin, windows] goarch: [amd64, arm64] steps: - - uses: actions/checkout@v6 - - - name: Set up Go - uses: actions/setup-go@v6 - with: - go-version: "1.25" + - uses: actions/checkout@v4 - name: Build env: @@ -50,30 +41,30 @@ jobs: lint: name: Lint - runs-on: ubuntu-latest + runs-on: docker + container: + image: docker.io/library/golang:1.25 steps: - - uses: actions/checkout@v6 + - uses: actions/checkout@v4 - - name: Set up Go - uses: actions/setup-go@v6 - with: - go-version: "1.25" + - name: Install golangci-lint + run: | + curl -sSfL https://raw.githubusercontent.com/golangci/golangci-lint/v2.11.4/install.sh \ + | sh -s -- -b /usr/local/bin v2.11.4 - name: Run golangci-lint - uses: golangci/golangci-lint-action@v9 - with: - version: v2.11.4 + run: golangci-lint run ./... coverage: name: Coverage - runs-on: ubuntu-latest + runs-on: docker + container: + image: docker.io/library/golang:1.25 steps: - - uses: actions/checkout@v6 + - uses: actions/checkout@v4 - - name: Set up Go - uses: actions/setup-go@v6 - with: - go-version: "1.25" + - name: Install python3 + run: apt-get update && apt-get install -y --no-install-recommends python3 - name: Run tests with coverage (all packages) run: | @@ -102,24 +93,13 @@ jobs: print('OK: Coverage meets minimum threshold') " - - name: Upload coverage to Codecov - uses: codecov/codecov-action@v6 - with: - files: ./coverage.out - fail_ci_if_error: false - env: - CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }} - vet: name: Vet - runs-on: ubuntu-latest + runs-on: docker + container: + image: docker.io/library/golang:1.25 steps: - - uses: actions/checkout@v6 - - - name: Set up Go - uses: actions/setup-go@v6 - with: - go-version: "1.25" + - uses: actions/checkout@v4 - name: Run go vet run: go vet ./... diff --git a/.forgejo/workflows/docker-rebuild.yml b/.forgejo/workflows/docker-rebuild.yml new file mode 100644 index 0000000..34cc3d6 --- /dev/null +++ b/.forgejo/workflows/docker-rebuild.yml @@ -0,0 +1,61 @@ +# Rebuilds and re-pushes the `latest` image without a version bump so newly +# *fixed* Alpine / ffmpeg / Go patches land between tagged releases. Versioned +# tags are immutable and never touched here. Runs weekly and on demand. +name: Docker rebuild + +on: + schedule: + # Mondays 04:17 UTC (off the hour to avoid the scheduler rush) + - cron: "17 4 * * 1" + workflow_dispatch: + +jobs: + rebuild: + runs-on: docker + container: + image: docker.io/library/docker:27-cli + steps: + - uses: actions/checkout@v4 + with: + fetch-depth: 0 + + - name: Install build deps + run: apk add --no-cache curl git bash + + - name: Install buildx + run: | + mkdir -p ~/.docker/cli-plugins + curl -sSL https://github.com/docker/buildx/releases/latest/download/buildx-linux-amd64 \ + -o ~/.docker/cli-plugins/docker-buildx + chmod +x ~/.docker/cli-plugins/docker-buildx + + - name: Set up qemu + run: docker run --rm --privileged tonistiigi/binfmt --install all + + # Stamp the binary with the most recent release tag (not "dev"). + - name: Resolve version + id: ver + run: | + v=$(git describe --tags --abbrev=0 2>/dev/null || echo dev) + echo "version=$v" >> "$GITHUB_OUTPUT" + + - name: Login to Docker Hub + env: + DH_USER: ${{ secrets.DOCKERHUB_USERNAME }} + DH_TOKEN: ${{ secrets.DOCKERHUB_TOKEN }} + run: echo "$DH_TOKEN" | docker login -u "$DH_USER" --password-stdin + + - name: Build + push (refresh latest) + env: + VERSION: ${{ steps.ver.outputs.version }} + run: | + docker buildx create --name builder --use --driver docker-container + # Refresh the floating tag only — never overwrite a versioned release. + # Force a fresh base pull so apk upgrade picks up new patches. + docker buildx build \ + --platform linux/amd64,linux/arm64 \ + --build-arg "VERSION=$VERSION" \ + --tag "torrentclaw/unarr:latest" \ + --no-cache \ + --push \ + . diff --git a/.forgejo/workflows/release.yml b/.forgejo/workflows/release.yml new file mode 100644 index 0000000..d757612 --- /dev/null +++ b/.forgejo/workflows/release.yml @@ -0,0 +1,118 @@ +name: Release + +on: + push: + tags: + - "v*" + workflow_dispatch: + +permissions: + contents: write + +jobs: + release: + runs-on: docker + container: + image: docker.io/library/golang:1.25 + steps: + - uses: actions/checkout@v4 + with: + fetch-depth: 0 + + - name: Install build deps (bash, curl, jq, ffmpeg fetch deps) + run: | + apt-get update + apt-get install -y --no-install-recommends bash curl ca-certificates jq xz-utils unzip + + - name: Install goreleaser + run: | + curl -sSfL https://github.com/goreleaser/goreleaser/releases/latest/download/goreleaser_Linux_x86_64.tar.gz \ + | tar -xz -C /usr/local/bin goreleaser + + - name: Run goreleaser + env: + # Forgejo runner auto-injects GITHUB_TOKEN (a per-job, instance-scoped + # token usable against the Forgejo REST API). goreleaser only accepts + # one token; with both GITHUB_TOKEN + GITEA_TOKEN set it errors out + # ("multiple tokens"). Unset GITHUB_TOKEN before invoking goreleaser so + # it picks the Gitea code path + the gitea_urls block in .goreleaser.yml. + GITEA_TOKEN: ${{ secrets.GITHUB_TOKEN }} + SENTRY_DSN: ${{ secrets.SENTRY_DSN }} + # Empty when RELEASE_SIGNING_PUBKEY variable is unset — goreleaser + # accepts it and the resulting binary disables signature checks + # (back-compat: pre-signing releases continue to update). Set + # RELEASE_SIGNING_PUBKEY (variable) + RELEASE_SIGNING_KEY (secret) + # to turn verification on. + RELEASE_SIGNING_PUBKEY: ${{ vars.RELEASE_SIGNING_PUBKEY }} + run: | + unset GITHUB_TOKEN + goreleaser release --clean + + - name: Sign checksums.txt with ed25519 + if: ${{ vars.RELEASE_SIGNING_PUBKEY != '' && secrets.RELEASE_SIGNING_KEY != '' }} + env: + RELEASE_SIGNING_KEY: ${{ secrets.RELEASE_SIGNING_KEY }} + RELEASE_TAG: ${{ github.ref_name }} + FORGEJO_TOKEN: ${{ secrets.GITHUB_TOKEN }} + # Tailscale IP — domain-agnostic; the runner shares the dokploy-network with + # forgejo (hostname `forgejo`), so the in-cluster hostname is fastest, but the + # Tailscale IP is the documented fallback. + FORGEJO_API: http://forgejo:3000/api/v1 + REPO: torrentclaw/unarr + run: | + set -euo pipefail + go run ./scripts/sign-checksums \ + -key "$RELEASE_SIGNING_KEY" \ + -in dist/checksums.txt \ + -out dist/checksums.txt.sig + + # Find the release ID for this tag, then upload the sig as an asset. + rel_id=$(curl -sSf "$FORGEJO_API/repos/$REPO/releases/tags/$RELEASE_TAG" \ + -H "Authorization: token $FORGEJO_TOKEN" | jq -r '.id') + curl -sSf -X POST \ + "$FORGEJO_API/repos/$REPO/releases/$rel_id/assets?name=checksums.txt.sig" \ + -H "Authorization: token $FORGEJO_TOKEN" \ + -F "attachment=@dist/checksums.txt.sig" + + docker: + needs: release + runs-on: docker + container: + # Docker-in-Docker capable image — buildx + qemu pre-installed. + image: docker.io/library/docker:27-cli + steps: + - uses: actions/checkout@v4 + + - name: Install buildx + run: | + apk add --no-cache curl + mkdir -p ~/.docker/cli-plugins + curl -sSL https://github.com/docker/buildx/releases/latest/download/buildx-linux-amd64 \ + -o ~/.docker/cli-plugins/docker-buildx + chmod +x ~/.docker/cli-plugins/docker-buildx + + - name: Login to Docker Hub + env: + DH_USER: ${{ secrets.DOCKERHUB_USERNAME }} + DH_TOKEN: ${{ secrets.DOCKERHUB_TOKEN }} + run: echo "$DH_TOKEN" | docker login -u "$DH_USER" --password-stdin + + - name: Set up qemu + run: docker run --rm --privileged tonistiigi/binfmt --install all + + - name: Build + push multi-arch image + env: + VERSION: ${{ github.ref_name }} + run: | + set -euo pipefail + VERSION_SEMVER="${VERSION#v}" + MAJOR_MINOR="${VERSION_SEMVER%.*}" + docker buildx create --name builder --use --driver docker-container + docker buildx build \ + --platform linux/amd64,linux/arm64 \ + --build-arg "VERSION=$VERSION" \ + --tag "torrentclaw/unarr:$VERSION_SEMVER" \ + --tag "torrentclaw/unarr:$MAJOR_MINOR" \ + --tag "torrentclaw/unarr:latest" \ + --push \ + . diff --git a/.github/workflows/docker-rebuild.yml b/.github/workflows/docker-rebuild.yml deleted file mode 100644 index c1634f1..0000000 --- a/.github/workflows/docker-rebuild.yml +++ /dev/null @@ -1,52 +0,0 @@ -# Rebuilds and re-pushes the `latest` image without a version bump so newly -# *fixed* Alpine / ffmpeg / Go patches land between tagged releases. Versioned -# tags are immutable and never touched here. Runs weekly and on demand. -name: Docker rebuild - -on: - schedule: - # Mondays 04:17 UTC (off the hour to avoid the scheduler rush) - - cron: "17 4 * * 1" - workflow_dispatch: - -jobs: - rebuild: - runs-on: ubuntu-latest - steps: - - uses: actions/checkout@v6 - with: - fetch-depth: 0 - - # Stamp the binary with the most recent release tag (not "dev"). - - name: Resolve version - id: ver - run: echo "version=$(git describe --tags --abbrev=0 2>/dev/null || echo dev)" >> "$GITHUB_OUTPUT" - - - uses: docker/setup-qemu-action@v4 - - uses: docker/setup-buildx-action@v4 - - - uses: docker/login-action@v4 - with: - username: ${{ secrets.DOCKERHUB_USERNAME }} - password: ${{ secrets.DOCKERHUB_TOKEN }} - - - uses: docker/build-push-action@v7 - with: - context: . - push: true - platforms: linux/amd64,linux/arm64 - # Refresh the floating tag only — never overwrite a versioned release. - tags: torrentclaw/unarr:latest - build-args: | - VERSION=${{ steps.ver.outputs.version }} - # Force a fresh base pull so apk upgrade picks up new patches. - no-cache: true - - - name: Scan image for fixable CVEs (gate) - uses: docker/scout-action@v1 - with: - command: cves - image: torrentclaw/unarr:latest - only-severities: critical,high - only-fixed: true - exit-code: true diff --git a/.github/workflows/pages.yml b/.github/workflows/pages.yml deleted file mode 100644 index d0c683d..0000000 --- a/.github/workflows/pages.yml +++ /dev/null @@ -1,52 +0,0 @@ -name: Deploy install scripts to Pages - -on: - push: - branches: [main] - paths: - - install.sh - - install.ps1 - - CNAME - - .nojekyll - - .github/workflows/pages.yml - workflow_dispatch: - -permissions: - contents: read - pages: write - id-token: write - -concurrency: - group: pages - cancel-in-progress: false - -jobs: - deploy: - runs-on: ubuntu-latest - environment: - name: github-pages - url: ${{ steps.deployment.outputs.page_url }} - steps: - - uses: actions/checkout@v4 - - uses: actions/configure-pages@v5 - - name: Stage install scripts - run: | - mkdir -p _site - cp install.sh install.ps1 _site/ - [ -f CNAME ] && cp CNAME _site/ - touch _site/.nojekyll - # Also index page (humans landing) - cat > _site/index.html <<'HTML' - - unarr installer -

unarr CLI installer

-
Linux/macOS:  curl -fsSL https://unarr.torrentclaw.com/install.sh | sh
-          Windows:      irm https://unarr.torrentclaw.com/install.ps1 | iex
-

Source: github.com/torrentclaw/unarr

- - HTML - - uses: actions/upload-pages-artifact@v3 - with: - path: _site - - id: deployment - uses: actions/deploy-pages@v4 diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml deleted file mode 100644 index dcb49ce..0000000 --- a/.github/workflows/release.yml +++ /dev/null @@ -1,210 +0,0 @@ -name: Release - -on: - push: - tags: - - "v*" - -permissions: - contents: write - -jobs: - release: - runs-on: ubuntu-latest - steps: - - uses: actions/checkout@v6 - with: - fetch-depth: 0 - - - uses: actions/setup-go@v6 - with: - go-version-file: go.mod - - - uses: goreleaser/goreleaser-action@v6 - with: - version: "~> v2" - args: release --clean - env: - GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }} - SENTRY_DSN: ${{ secrets.SENTRY_DSN }} - # Empty when RELEASE_SIGNING_PUBKEY variable is unset — goreleaser - # accepts it and the resulting binary disables signature checks - # (back-compat: pre-signing releases continue to update). Set - # RELEASE_SIGNING_PUBKEY (variable) + RELEASE_SIGNING_KEY (secret) - # to turn verification on. - RELEASE_SIGNING_PUBKEY: ${{ vars.RELEASE_SIGNING_PUBKEY }} - - - name: Sign checksums.txt with ed25519 - # Reference secrets.X directly — step-level env defined in this same - # step is unreliable to read from this step's own if: expression. - if: ${{ vars.RELEASE_SIGNING_PUBKEY != '' && secrets.RELEASE_SIGNING_KEY != '' }} - env: - RELEASE_SIGNING_KEY: ${{ secrets.RELEASE_SIGNING_KEY }} - RELEASE_TAG: ${{ github.ref_name }} - GH_TOKEN: ${{ secrets.GITHUB_TOKEN }} - run: | - set -euo pipefail - go run ./scripts/sign-checksums \ - -key "$RELEASE_SIGNING_KEY" \ - -in dist/checksums.txt \ - -out dist/checksums.txt.sig - gh release upload "$RELEASE_TAG" dist/checksums.txt.sig --clobber - - docker: - needs: release - runs-on: ubuntu-latest - steps: - - uses: actions/checkout@v6 - - - name: Docker meta - id: meta - uses: docker/metadata-action@v6 - with: - images: torrentclaw/unarr - tags: | - type=semver,pattern={{version}} - type=semver,pattern={{major}}.{{minor}} - type=raw,value=latest - - - uses: docker/setup-qemu-action@v4 - - uses: docker/setup-buildx-action@v4 - - - uses: docker/login-action@v4 - with: - username: ${{ secrets.DOCKERHUB_USERNAME }} - password: ${{ secrets.DOCKERHUB_TOKEN }} - - - uses: docker/build-push-action@v7 - with: - context: . - push: true - platforms: linux/amd64,linux/arm64 - tags: ${{ steps.meta.outputs.tags }} - labels: ${{ steps.meta.outputs.labels }} - build-args: | - VERSION=${{ github.ref_name }} - - # CVE gate. Fails the release on FIXABLE critical/high only — unfixed - # upstream ffmpeg codec CVEs are accepted (see SECURITY.md), so the - # codec noise does not block. Runs post-push (image already published); - # a failure here flags that a fixable CVE slipped through. - - name: Scan image for fixable CVEs (gate) - uses: docker/scout-action@v1 - with: - command: cves - image: torrentclaw/unarr:latest - only-severities: critical,high - only-fixed: true - exit-code: true - - # Sync the Docker Hub repo description from DOCKERHUB.md. Non-fatal: a - # description-API auth hiccup must not undo a successful image push. - - name: Update Docker Hub description - uses: peter-evans/dockerhub-description@v4 - continue-on-error: true - with: - username: ${{ secrets.DOCKERHUB_USERNAME }} - password: ${{ secrets.DOCKERHUB_TOKEN }} - repository: torrentclaw/unarr - readme-filepath: ./DOCKERHUB.md - short-description: "unarr — the single binary that replaces your *arr stack" - - - virustotal: - needs: release - runs-on: ubuntu-latest - if: vars.VT_ENABLED == 'true' - steps: - - name: Get release tag - id: tag - run: echo "tag=${GITHUB_REF#refs/tags/}" >> "$GITHUB_OUTPUT" - - - name: Download release assets - env: - GH_TOKEN: ${{ secrets.GITHUB_TOKEN }} - run: | - mkdir -p assets - gh release download "${{ steps.tag.outputs.tag }}" \ - --repo "${{ github.repository }}" \ - --dir assets \ - --pattern '*.tar.gz' \ - --pattern '*.zip' \ - --pattern 'checksums.txt' - - - name: Scan assets with VirusTotal - env: - VT_API_KEY: ${{ secrets.VT_API_KEY }} - run: | - mkdir -p results - for file in assets/*; do - filename=$(basename "$file") - echo "Uploading $filename to VirusTotal..." - - response=$(curl -s --request POST \ - --url https://www.virustotal.com/api/v3/files \ - --header "x-apikey: $VT_API_KEY" \ - --form "file=@$file") - - analysis_id=$(echo "$response" | jq -r '.data.id // empty') - if [ -z "$analysis_id" ]; then - echo "::warning::Failed to upload $filename: $response" - continue - fi - - echo "$filename=$analysis_id" >> results/scans.txt - echo " Analysis ID: $analysis_id" - - # Rate limit: VT free tier allows 4 req/min - sleep 16 - done - - - name: Wait for analysis completion - env: - VT_API_KEY: ${{ secrets.VT_API_KEY }} - run: | - echo "Waiting 60s for VirusTotal analysis to complete..." - sleep 60 - - vt_report="## 🛡️ VirusTotal Scan Results\n\n" - vt_report+="| File | Result | Link |\n" - vt_report+="|------|--------|------|\n" - - while IFS='=' read -r filename analysis_id; do - result=$(curl -s --request GET \ - --url "https://www.virustotal.com/api/v3/analyses/$analysis_id" \ - --header "x-apikey: $VT_API_KEY") - - malicious=$(echo "$result" | jq -r '.data.attributes.stats.malicious // 0') - undetected=$(echo "$result" | jq -r '.data.attributes.stats.undetected // 0') - sha256=$(echo "$result" | jq -r '.meta.file_info.sha256 // empty') - - if [ "$malicious" = "0" ]; then - status="✅ Clean ($undetected engines)" - else - status="⚠️ $malicious detections" - fi - - link="https://www.virustotal.com/gui/file/$sha256" - vt_report+="| \`$filename\` | $status | [View]($link) |\n" - - sleep 16 - done < results/scans.txt - - echo -e "$vt_report" > results/report.md - cat results/report.md - - - name: Append scan results to release notes - env: - GH_TOKEN: ${{ secrets.GITHUB_TOKEN }} - run: | - current_body=$(gh release view "${{ steps.tag.outputs.tag }}" \ - --repo "${{ github.repository }}" \ - --json body --jq '.body') - - new_body="${current_body} - - $(cat results/report.md)" - - gh release edit "${{ steps.tag.outputs.tag }}" \ - --repo "${{ github.repository }}" \ - --notes "$new_body" diff --git a/.gitignore b/.gitignore index 81f1284..8015bab 100644 --- a/.gitignore +++ b/.gitignore @@ -41,4 +41,7 @@ dist-ffbinaries/ # Docker tmp/ config/ -dist-ffbinaries/ \ No newline at end of file +dist-ffbinaries/ + +# Claude Code: keep entirely local, do not track +.claude/ \ No newline at end of file diff --git a/.goreleaser.yml b/.goreleaser.yml index 26ce802..6bc4a51 100644 --- a/.goreleaser.yml +++ b/.goreleaser.yml @@ -59,6 +59,22 @@ changelog: - "^test:" - "^chore:" +# Self-hosted Forgejo at git.torrentclaw.com. goreleaser detects GITEA_TOKEN + +# these URLs and publishes the release there instead of GitHub. Reachable via +# `forgejo` hostname inside the dokploy-network (the runner shares it); for +# local goreleaser runs outside the network, override via env GITEA_API_URL. +# +# In goreleaser v2 `gitea_urls` is a top-level key (was nested under `release` +# in v1). +gitea_urls: + api: http://forgejo:3000/api/v1 + download: https://git.torrentclaw.com + skip_tls_verify: false + +release: + draft: false + prerelease: auto + # Homebrew tap — requires PAT with repo scope (not GITHUB_TOKEN) # Enable when torrentclaw/homebrew-tap PAT is configured as HOMEBREW_TAP_TOKEN # brews: diff --git a/CHANGELOG.md b/CHANGELOG.md index ca49641..de1dd6e 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -5,104 +5,131 @@ All notable changes to this project will be documented in this file. The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). +## [0.9.15] - 2026-05-27 + + +### Added + +- **sentry**: enhance error handling by skipping user input errors in CaptureError + +### Changed + +- **ci**: point Forgejo URLs at torrentclaw org (post-transfer) +- **sentry**: decouple agent import via string-match, rename predicate + +### Documentation + +- **positioning**: reframe unarr around download/stream/transcode, drop misleading search-first wording + +### Fixed + +- **ci**: unset GITHUB_TOKEN so goreleaser uses GITEA_TOKEN +- **sentry**: skip "daemon not running" stop/reload errors + +### Other + +- **scripts**: harden release.sh against double-release and inline version bumps +- untrack .claude/ (private local config) +## [0.9.14] - 2026-05-27 + + +### Added + +- **vaapi**: hybrid CPU-scale + hwupload encode path (QW2, 0.9.14) + +### CI/CD + +- port workflows from .github/ to .forgejo/ (Forgejo Actions) + +### Fixed + +- **daemon**: defensive IsClosed check in watchSessionReady poll loop +- **daemon**: use parent ctx for MarkSessionReady so cancel propagates +- **release**: move gitea_urls to top-level (goreleaser v2 schema) +## [0.9.13] - 2026-05-27 + + +### Added + +- **agent**: session-ready webhook for SSE-driven player handshake (0.9.13) +- **agent**: send full transcoder diagnostic in register payload (0.9.12) + +### Fixed + +- **daemon**: defer probeCancel so a panic mid-diagnostic still releases ctx + +### Other + +- **release**: add ship.sh end-to-end pipeline as GH Actions backup +- **skills**: add /publish slash command + allow .claude/ in git +## [0.9.11] - 2026-05-27 + + +### Added + +- **hls**: pre-segmentación delantada — 2 s segments + async session start (0.9.10) +- **hls**: faster first-start — probe cache + tighter encoder presets (0.9.9) + +### Changed + +- **hls**: critico-driven hardening of fase 3.2 + +### Fixed + +- **cors**: allow play from .to / staging / onion mirrors +- **library**: classify resolution by width + height, not height alone +- **transcode**: make preset libx264-only + restore quality opt-in + +### Other + +- **release**: 0.9.11 +## [0.9.8] - 2026-05-27 + + +### Fixed + +- **upgrade**: break auto-apply restart loop (0.9.8) +## [0.9.7] - 2026-05-26 + + +### Added + +- **hls**: persistent fMP4 segment cache + integrity + stats (0.9.7) ## [0.9.6] - 2026-05-26 + ### Added -- **auto-upgrade**: when the web flags the agent for upgrade - (`POST /api/internal/agent/upgrade` or the "Force update now" button), - the daemon now downloads and replaces the binary in-place, then exits so - the service supervisor (`systemd Restart=always` on Linux, the equivalent - on macOS/Windows) respawns on the new version. No `unarr update` step - required from the user. Still opt-in — only fires when the server sends - the upgrade signal. - -### Changed - -- The `OnUpgrade` daemon callback no longer just logs `run unarr self-update`; - it now triggers the actual upgrade in a background goroutine. - +- **daemon**: auto-apply upgrades when server signals (0.9.6) ## [0.9.5] - 2026-05-26 + ### Added -- **funnel**: optional CloudFlare Quick Tunnel subprocess. `unarr funnel on` - spawns `cloudflared` as a child process and registers an anonymous - `https://.trycloudflare.com` hostname tunnelled to the daemon's - HLS server. The hostname is reported back to the web on every sync so the - in-browser player picks it up automatically — cross-network playback now - works on torrentclaw.com without Tailscale or port forwarding. Bytes - proxy through CloudFlare; TorrentClaw still doesn't relay content. -- **funnel**: on by default for fresh installs (NAS/Docker get cross-network - HTTPS automatically); existing configs that pre-date the feature stay - off until the operator runs `unarr funnel on`. -- **funnel**: auto-downloads cloudflared to the unarr data dir when not on - PATH (Linux amd64/arm64/armhf/386). ELF magic + size sanity check on the - download; `O_EXCL` partial-write so concurrent daemons don't clobber - each other. -- **funnel**: subprocess supervisor keeps the tunnel up across cloudflared - crashes + CF's ~6h Quick Tunnel rotation. Exponential backoff (2 s → 5 min) - on persistent failures. The web's reported URL is cleared the moment - cloudflared exits so an outdated hostname doesn't keep handing out 502s. -- **funnel**: `unarr funnel status` shows the live URL once registered. - See README §`[downloads.funnel]` for the throughput / latency caveats of - CF's free Quick Tunnels. -- **docker**: the official `torrentclaw/unarr` image now bundles - `cloudflared` so the funnel works the moment the container starts — no - first-run download. - -### Fixed - -- **hls/libx264**: bump the H.264 level we hint to libx264 by one tier so - anamorphic (>16:9) sources stop emitting unplayable streams. 720p at - level 3.1 silently rejected 1728×720 cinemascope frames with - `frame MB size > level limit`; 720p now ships at level 4.0, 1080p at 4.1. - Decoder compatibility is unaffected — every device that handles 1080p - already handles ≥ 4.1. - +- **funnel**: cloudflare quick tunnel embedded subprocess (0.9.5) ## [0.9.4] - 2026-05-26 -### Removed - -- **streaming**: retire the custom WebRTC DataChannel pipeline. The daemon no - longer ships pion/webrtc, the WSS signaling client, or the wire framing - package — every in-browser session now uses HLS over HTTP from the daemon - (Tailscale / LAN / UPnP). Browser P2P (WebTorrent) bytes never re-enabled. -- **config**: `[downloads.webrtc]` block removed from the TOML schema; existing - config files with the section parse cleanly because go-toml ignores unknown - sections. -- **seed_file**: `mode=seed_file` task handler + `engine.SeedFile` helper - dropped — the last in-browser caller was retired with the WebRTC player. -- **wstracker-probe**: standalone probe binary removed. - -### Changed - -- **agent wire**: `SyncResponse.WebRTCSessions` (JSON: `webrtcSessions`) renamed - to `StreamSessions` (JSON: `streamSessions`). The Go type `agent.WebRTCSession` - is now `agent.StreamSession`. Wire-incompatible with web < 2026-05-26. -- **torrent**: `buildMagnet` no longer accepts an `extraTrackers` variadic — - the default tracker list is the only set used. - -### Fixed - -- **hls**: clamp the ffmpeg `-b:v` to the bitrate cap derived from the EFFECTIVE - output height instead of the requested quality. Previously asking for "2160p" - on a 1080p source overshot the H.264 level we resolved from the effective - height (4.0, max 20 Mbps) and made libx264 abort with - `VBV bitrate > level limit`. - -## [0.9.2] - 2026-05-21 ### Added -- **vpn**: `unarr vpn` command (`status`, `enable`, `disable`) to manage the managed - WireGuard split-tunnel, with `vpn status --check` to verify provisioning. -- **vpn**: report split-tunnel state (active, exit server) to the web on register - + every sync, so the dashboard shows which agent holds the single WireGuard slot. -- **vpn**: send the agent id when fetching the VPN config so the web can arbitrate - the single WireGuard slot — the first agent claims it; the rest are told to run - OpenVPN on their own host (1 agent on WireGuard + up to 9 on OpenVPN). +- **stream**: retire WebRTC, HLS-only, bump 0.9.4 (**BREAKING**) +## [0.9.3] - 2026-05-26 + +### Added + +- **usenet**: warn at startup when par2 or extractor is missing + +### Fixed + +- **engine**: truncate errorMessage before reporting status +- **hls**: clamp ffmpeg bitrate to the level we derive from outputHeight +## [0.9.2] - 2026-05-22 + + +### Added + +- **vpn**: unarr vpn command + report/arbitrate the WireGuard slot ## [0.9.1] - 2026-05-21 @@ -113,6 +140,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0 ### Fixed - **security**: bump golang.org/x deps and add container CVE scan gate + +### Other + +- **release**: 0.9.1 ## [0.9.0] - 2026-05-21 @@ -122,6 +153,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0 - **vpn**: local config_file for self-hosted/personal VPN testing - **vpn**: split-tunnel torrent traffic through managed WireGuard +### CI/CD + +- deploy install scripts to GitHub Pages + ### Documentation - **docker**: refresh Docker Hub README + sync description in CI @@ -135,6 +170,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0 ### Other +- **pages**: add .nojekyll to disable Jekyll processing +- **pages**: set custom domain unarr.torrentclaw.com - **release**: 0.9.0 ## [0.8.1] - 2026-05-08 @@ -508,6 +545,17 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0 ### Build - add -s -w -trimpath to Makefile, add build-small target with UPX +[0.9.15]: https://github.com/torrentclaw/unarr/compare/v0.9.14...v0.9.15 +[0.9.14]: https://github.com/torrentclaw/unarr/compare/v0.9.13...v0.9.14 +[0.9.13]: https://github.com/torrentclaw/unarr/compare/v0.9.11...v0.9.13 +[0.9.11]: https://github.com/torrentclaw/unarr/compare/v0.9.8...v0.9.11 +[0.9.8]: https://github.com/torrentclaw/unarr/compare/v0.9.7...v0.9.8 +[0.9.7]: https://github.com/torrentclaw/unarr/compare/v0.9.6...v0.9.7 +[0.9.6]: https://github.com/torrentclaw/unarr/compare/v0.9.5...v0.9.6 +[0.9.5]: https://github.com/torrentclaw/unarr/compare/v0.9.4...v0.9.5 +[0.9.4]: https://github.com/torrentclaw/unarr/compare/v0.9.3...v0.9.4 +[0.9.3]: https://github.com/torrentclaw/unarr/compare/v0.9.2...v0.9.3 +[0.9.2]: https://github.com/torrentclaw/unarr/compare/v0.9.1...v0.9.2 [0.9.1]: https://github.com/torrentclaw/unarr/compare/v0.9.0...v0.9.1 [0.9.0]: https://github.com/torrentclaw/unarr/compare/v0.8.1...v0.9.0 [0.8.1]: https://github.com/torrentclaw/unarr/compare/v0.8.0...v0.8.1 diff --git a/DOCKERHUB.md b/DOCKERHUB.md index 7a9bc0e..3df5b70 100644 --- a/DOCKERHUB.md +++ b/DOCKERHUB.md @@ -1,8 +1,9 @@ # unarr -**The single binary that replaces your whole *arr stack.** Search 30+ torrent -sources, inspect real quality before you download, grab subtitles, and manage -your media library — all from one terminal tool or a headless daemon. +**The single binary that replaces your whole *arr stack.** Built-in torrent, +debrid, and usenet engines. Stream, transcode, and organize your library from +one terminal — or run it as a headless daemon with a web dashboard, WireGuard +split-tunnel, and Cloudflare Funnel remote access. **[Website & docs](https://torrentclaw.com/unarr)** · **[Install guide](https://torrentclaw.com/cli)** · **[Get an API key](https://torrentclaw.com)** diff --git a/Makefile b/Makefile index 08462b6..b3325bc 100644 --- a/Makefile +++ b/Makefile @@ -1,4 +1,4 @@ -.PHONY: all build test lint coverage clean fmt vet check install-hooks changelog release release-patch release-minor release-major release-dry +.PHONY: all build test lint coverage clean fmt vet check install-hooks changelog release release-patch release-minor release-major release-dry ship ship-dry ship-push BINARY = unarr SENTRY_DSN ?= @@ -71,6 +71,19 @@ release-dry: @test -n "$(V)" || { echo "Usage: make release-dry V=patch|minor|major|0.5.0"; exit 1; } @./scripts/release.sh --dry-run $(V) +## Ship a release end-to-end (goreleaser + Hetzner + Docker Hub). Standalone backup for GH Actions. +## Reads version from internal/cmd/version.go unless V= is provided. +ship: + @./scripts/ship.sh $(V) + +## Ship + git push tag to GH afterwards +ship-push: + @./scripts/ship.sh --push $(V) + +## Preview ship steps without executing +ship-dry: + @./scripts/ship.sh --dry-run $(V) + ## Remove generated files clean: rm -f $(BINARY) coverage.out coverage.html diff --git a/README.md b/README.md index 4129ffb..75c9c62 100644 --- a/README.md +++ b/README.md @@ -11,9 +11,9 @@ [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE) [![Go Version](https://img.shields.io/github/go-mod/go-version/torrentclaw/unarr)](go.mod) -Powerful terminal tool for torrent search and management. **Free and open source.** +The single-binary terminal client for torrent, debrid, and usenet downloads. **Free and open source.** -Search 30+ torrent sources, inspect torrent quality, discover popular content, find streaming providers, and manage your media collection — all from your terminal. +Built-in torrent engine, debrid (Real-Debrid / AllDebrid), and NZB support. Stream to mpv/vlc, transcode on the fly with hardware acceleration, and manage your library — one binary or a headless daemon with WireGuard split-tunnel and Cloudflare Funnel remote access. @@ -343,6 +343,58 @@ unarr self-update --force # reinstall even if up to date `unarr doctor` checks: config file, API key, server connectivity (with latency), agent registration, download directory, disk space, and version. +### Updating unarr + +unarr supports three update paths. Pick whichever fits your workflow. + +**1. Manual self-update (always available).** + +```bash +unarr self-update # interactive update to latest +unarr self-update --force # reinstall same version +unarr self-update --allow-unsigned # accept releases without checksum signature +``` + +The CLI downloads the new release archive over HTTPS (from +`torrentclaw.com/releases/download/v/`), verifies SHA-256, swaps the +binary in place (`.backup` kept next to it), and restarts the systemd +user unit if the daemon is running. + +**2. Auto-apply on server signal (default, since 0.9.6).** + +When you press **"Force update now"** on the web (Settings → Agent → Force +update), the server sets a flag your daemon polls every sync (~3 s). On +the next sync the daemon downloads the new binary, replaces itself, and +exits — `systemd Restart=always` respawns on the new version. No SSH, no +terminal access required. Works headless on NAS / Docker. + +The button shows an amber warning if your agent is below 0.9.6 (older +daemons see the signal but only log "run unarr update" — the operator +must run the command manually that one time). + +**Opt out of auto-apply.** Some users prefer reviewing CHANGELOG before +applying. Disable in `config.toml`: + +```toml +[daemon] +auto_upgrade = false +``` + +With `auto_upgrade = false`, pressing the web button still flags your +agent (so the daemon logs the new version on next sync), but the daemon +will not download / replace anything — you run `unarr self-update` when +you're ready. + +**3. Docker auto-restart with a new tag.** + +```bash +docker pull torrentclaw/unarr:latest +docker compose up -d +``` + +Tags published: `latest`, `0.9`, `0.9.7`, ... — pin to a minor (`0.9`) +for opt-in patch updates without surprises. + ## Clean Remove temporary files, logs, resume data, and other artifacts generated by unarr. Shows what will be removed and asks for confirmation before deleting. @@ -424,6 +476,7 @@ tv_shows_dir = "~/Media/TV Shows" [daemon] poll_interval = "30s" heartbeat_interval = "30s" +auto_upgrade = true # apply server-flagged upgrades in-place (since 0.9.6) [notifications] enabled = true @@ -466,6 +519,40 @@ If `transcode.enabled = true` but `ffmpeg` / `ffprobe` aren't on PATH, the daemon logs a warning at startup and HLS sessions are rejected at runtime with a clear error — install ffmpeg or set `enabled = false`. +#### `[downloads.hls_cache]` — persistent HLS segment cache + +```toml +[downloads.hls_cache] +enabled = true # on by default +size_gb = 5 # disk budget; LRU eviction once exceeded +dir = "" # custom path; empty = ~/.cache/unarr/hls-cache +``` + +| Key | Type | Default | Notes | +|-----|------|---------|-------| +| `enabled` | bool | `true` | Persists finished HLS encodes per `(source, quality, audio_index)`. A second play of the same file at the same quality reuses the segments — no ffmpeg, near-zero CPU, instant playback. Set to `false` to delete segments on session close (original behavior). | +| `size_gb` | int | `5` | Cache budget in gigabytes. When exceeded the LRU sweeper evicts the least-recently-used cached encodes hourly. Minimum 1 GB (smaller values are clamped up). | +| `dir` | string | `""` | Custom storage path. Empty defaults to `~/.cache/unarr/hls-cache` (Linux/macOS) or the user cache dir (Windows). | + +**What it does.** First play encodes normally (ffmpeg writes segments). +On session close, if every segment is on disk and ffmpeg exited cleanly, +the directory is sealed with a `.complete` marker and kept. Next time the +same source + quality combo is requested, the daemon serves segments +straight from disk — no transcode, no warm-up, no CPU cost. + +**Why per (source, quality, audio).** Renaming the file or switching +quality invalidates the entry: the segments are tied to the exact source +bytes and the exact ffmpeg parameters. Re-encoding generates a new key. + +**Eviction.** A background goroutine wakes every hour. If total cache size +exceeds `size_gb`, it deletes the oldest entries (by mtime) until under +budget. Active sessions are pinned — they never get evicted mid-play. + +**Disable.** Either edit the TOML to set `enabled = false`, or remove the +cache directory manually (it'll be recreated as needed). Disabling does +not delete existing cached segments — drop `dir` (or `~/.cache/unarr/hls-cache`) +to reclaim the space. + #### `[downloads.vpn]` | Key | Type | Default | Notes | diff --git a/internal/agent/client.go b/internal/agent/client.go index 9aa3c2a..e7f2c37 100644 --- a/internal/agent/client.go +++ b/internal/agent/client.go @@ -91,6 +91,45 @@ func (c *Client) Deregister(ctx context.Context, agentID string) error { return nil } +// ReportUpgradeResult tells the server the outcome of a previously requested +// upgrade so the server can clear `upgrade_requested`. Without this call the +// flag stays sticky and the daemon would re-trigger applyAutoUpgrade on every +// sync after upgrade — even for "already on target version" no-ops. +func (c *Client) ReportUpgradeResult(ctx context.Context, agentID string, success bool, version, errMsg string) error { + req := struct { + AgentID string `json:"agentId"` + Success bool `json:"success"` + Version string `json:"version,omitempty"` + Error string `json:"error,omitempty"` + }{AgentID: agentID, Success: success, Version: version, Error: errMsg} + var resp StatusResponse + if err := c.doPost(ctx, "/api/internal/agent/upgrade-result", req, &resp); err != nil { + return fmt.Errorf("report upgrade result: %w", err) + } + return nil +} + +// MarkSessionReady signals the server that the first HLS segment + init.mp4 +// landed on disk for the given session. The web side flips +// streaming_session.ready_at = NOW(), which its SSE endpoint emits to +// subscribed players so the "Preparando…" UI ends without polling HEAD +// on /hls//master.m3u8. +// +// Best-effort: the server is the source of truth for session state and +// will reach the same conclusion via HEAD probes anyway if this call +// fails. We log the error in the caller but don't retry — by the time +// a retry would land the user is likely already playing. +func (c *Client) MarkSessionReady(ctx context.Context, sessionID string) error { + req := struct { + SessionID string `json:"sessionId"` + }{SessionID: sessionID} + var resp StatusResponse + if err := c.doPost(ctx, "/api/internal/agent/session-ready", req, &resp); err != nil { + return fmt.Errorf("mark session ready: %w", err) + } + return nil +} + // ReportStatus reports download progress. Returns server-side flags the CLI must act on. func (c *Client) ReportStatus(ctx context.Context, update StatusUpdate) (*StatusResponse, error) { var resp StatusResponse diff --git a/internal/agent/daemon.go b/internal/agent/daemon.go index e5e4c60..f7994fb 100644 --- a/internal/agent/daemon.go +++ b/internal/agent/daemon.go @@ -28,6 +28,15 @@ type DaemonConfig struct { ScanPaths []string // configured scan paths for file deletion validation HWAccel string // detected encoder backend ("nvenc"/"qsv"/"vaapi"/"videotoolbox"/"none") MaxTranscodeHeight int // resolution cap the agent can transcode comfortably (px) + // Diagnostic data populated by engine.DetectHWAccelDiagnostic at daemon + // start. Surfaced in the web "Diagnose transcoder" modal — lets a user + // see which encoders the ffmpeg binary supports and which devices the + // host exposes without running `unarr probe-hwaccel`. + FFmpegVersion string // first line of `ffmpeg -version` + FFmpegPath string // resolved binary path + HWEncoders []string // HW-class encoder names found in `ffmpeg -encoders` + HWDevices []string // device files + driver bins detected at probe time + AutoUpgrade bool // honor server-flagged upgrades by downloading + restarting (default: true) } // Daemon manages agent registration and the sync loop. @@ -121,6 +130,10 @@ func (d *Daemon) Register(ctx context.Context) error { TailscaleIP: d.cfg.TailscaleIP, HWAccel: d.cfg.HWAccel, MaxTranscodeHeight: d.cfg.MaxTranscodeHeight, + FFmpegVersion: d.cfg.FFmpegVersion, + FFmpegPath: d.cfg.FFmpegPath, + HWEncoders: d.cfg.HWEncoders, + HWDevices: d.cfg.HWDevices, VPNActive: d.vpnActive, VPNMode: d.vpnMode, VPNServer: d.vpnServer, @@ -237,6 +250,10 @@ func (d *Daemon) Run(ctx context.Context) error { return } d.lastNotifiedVersion = version + if !d.cfg.AutoUpgrade { + log.Printf("[upgrade] new version available: %s — auto_upgrade=false, run `unarr update` to apply", version) + return + } log.Printf("[upgrade] new version available: %s — applying auto-upgrade", version) go d.applyAutoUpgrade(version) } @@ -289,8 +306,30 @@ func (d *Daemon) Deregister() { // supervisor (systemd Restart=always on Linux) respawns on the new binary. // Triggered by the server's upgrade signal — opt-in flag set by the user from // the web UI; the daemon never auto-upgrades on a passive version bump. +// +// Reports the outcome to /api/internal/agent/upgrade-result so the server +// clears `upgrade_requested`. Without this report the flag stays sticky and +// the daemon would loop on every sync — including the no-op case where it's +// already on the target version. func (d *Daemon) applyAutoUpgrade(targetVersion string) { currentClean := strings.TrimPrefix(d.cfg.Version, "v") + targetClean := strings.TrimPrefix(targetVersion, "v") + + // No-op: server signal arrived but we're already running the target. This + // happens when the daemon restarts after a previous auto-upgrade before + // reportUpgradeResult cleared the flag, or when the operator manually + // installed the same version off-band. Skip Execute (which would also + // no-op) AND skip os.Exit, but DO clear the flag — otherwise we loop. + if currentClean == targetClean { + log.Printf("[upgrade] already on v%s — clearing server flag", currentClean) + ctxR, cancelR := context.WithTimeout(context.Background(), 10*time.Second) + defer cancelR() + if err := d.client.ReportUpgradeResult(ctxR, d.cfg.AgentID, true, currentClean, ""); err != nil { + log.Printf("[upgrade] report-result failed (will retry on next signal): %v", err) + } + return + } + upgrader := &upgrade.Upgrader{ CurrentVersion: currentClean, OnProgress: func(msg string) { @@ -302,10 +341,24 @@ func (d *Daemon) applyAutoUpgrade(targetVersion string) { result := upgrader.Execute(ctx, targetVersion) if !result.Success { log.Printf("[upgrade] auto-upgrade failed: %v", result.Error) + errMsg := "" + if result.Error != nil { + errMsg = result.Error.Error() + } + ctxR, cancelR := context.WithTimeout(context.Background(), 10*time.Second) + defer cancelR() + if err := d.client.ReportUpgradeResult(ctxR, d.cfg.AgentID, false, targetClean, errMsg); err != nil { + log.Printf("[upgrade] report-result failed: %v", err) + } return } - log.Printf("[upgrade] upgraded v%s → v%s; exiting so service supervisor restarts on new binary", + log.Printf("[upgrade] upgraded v%s → v%s; reporting result + exiting so service supervisor restarts on new binary", result.OldVersion, result.NewVersion) + ctxR, cancelR := context.WithTimeout(context.Background(), 10*time.Second) + if err := d.client.ReportUpgradeResult(ctxR, d.cfg.AgentID, true, result.NewVersion, ""); err != nil { + log.Printf("[upgrade] report-result failed: %v", err) + } + cancelR() time.Sleep(500 * time.Millisecond) os.Exit(0) } diff --git a/internal/agent/state.go b/internal/agent/state.go index 1f00033..cc08ae5 100644 --- a/internal/agent/state.go +++ b/internal/agent/state.go @@ -2,6 +2,8 @@ package agent import ( "encoding/json" + "errors" + "fmt" "os" "path/filepath" "time" @@ -9,6 +11,13 @@ import ( "github.com/torrentclaw/unarr/internal/config" ) +// ErrDaemonNotRunning is returned when no daemon state file exists on disk. +// Callers may wrap it with %w; downstream code uses errors.Is to detect it. +// NOTE: the message text is matched by the sentry package (string-match, to +// avoid an import cycle). Keep the prefix "daemon does not appear to be +// running" stable, or update sentry.daemonNotRunningMarker accordingly. +var ErrDaemonNotRunning = errors.New("daemon does not appear to be running (state file not found)") + // DaemonState is written to disk every heartbeat for external tools to read. type DaemonState struct { AgentID string `json:"agentId"` @@ -69,17 +78,31 @@ func WriteState(state *DaemonState) { os.Rename(tmp, path) } -// ReadState reads the daemon state from disk. Returns nil if not found. +// ReadState reads the daemon state from disk. Returns nil if not found or +// unreadable. Use LoadState when callers need to distinguish "not running" +// from "state file corrupted". func ReadState() *DaemonState { + state, _ := LoadState() + return state +} + +// LoadState reads the daemon state and returns explicit errors: +// - ErrDaemonNotRunning when the state file does not exist +// - a wrapped json error when the file exists but cannot be decoded +// (a real bug worth reporting to Sentry) +func LoadState() (*DaemonState, error) { data, err := os.ReadFile(StateFilePath()) if err != nil { - return nil + if errors.Is(err, os.ErrNotExist) { + return nil, ErrDaemonNotRunning + } + return nil, err } var state DaemonState - if json.Unmarshal(data, &state) != nil { - return nil + if err := json.Unmarshal(data, &state); err != nil { + return nil, fmt.Errorf("decode daemon state %s: %w", StateFilePath(), err) } - return &state + return &state, nil } // RemoveState deletes the state file (called on clean shutdown). diff --git a/internal/agent/state_test.go b/internal/agent/state_test.go index 6c9abdd..7e275be 100644 --- a/internal/agent/state_test.go +++ b/internal/agent/state_test.go @@ -1,6 +1,7 @@ package agent import ( + "errors" "os" "path/filepath" "testing" @@ -104,3 +105,39 @@ func TestReadStateCorruptedJSON(t *testing.T) { t.Errorf("ReadState() should return nil for corrupted JSON, got %+v", state) } } + +func TestLoadStateNotFound(t *testing.T) { + tmpDir := t.TempDir() + origFn := stateFilePathFn + stateFilePathFn = func() string { return filepath.Join(tmpDir, "nonexistent.json") } + defer func() { stateFilePathFn = origFn }() + + state, err := LoadState() + if state != nil { + t.Errorf("LoadState() state = %+v, want nil", state) + } + if !errors.Is(err, ErrDaemonNotRunning) { + t.Errorf("LoadState() err = %v, want ErrDaemonNotRunning", err) + } +} + +func TestLoadStateCorruptedJSON(t *testing.T) { + tmpDir := t.TempDir() + origFn := stateFilePathFn + path := filepath.Join(tmpDir, "daemon.state.json") + stateFilePathFn = func() string { return path } + defer func() { stateFilePathFn = origFn }() + + os.WriteFile(path, []byte("not valid json{{{"), 0o644) + + state, err := LoadState() + if state != nil { + t.Errorf("LoadState() state = %+v, want nil", state) + } + if err == nil { + t.Fatal("LoadState() err = nil, want decode error") + } + if errors.Is(err, ErrDaemonNotRunning) { + t.Error("corrupt state must not be reported as ErrDaemonNotRunning — it would be filtered from Sentry") + } +} diff --git a/internal/agent/types.go b/internal/agent/types.go index 00802bc..ae87bb6 100644 --- a/internal/agent/types.go +++ b/internal/agent/types.go @@ -26,6 +26,15 @@ type RegisterRequest struct { // up to 2160p. HWAccel string `json:"hwAccel,omitempty"` MaxTranscodeHeight int `json:"maxTranscodeHeight,omitempty"` + // Diagnostic surface filled by engine.DetectHWAccelDiagnostic at daemon + // start. Surfaced in the web "Diagnose transcoder" modal so users can + // see *why* their HWAccel landed on "none" without running + // `unarr probe-hwaccel` locally — most commonly the ffmpeg binary + // shipped without HW encoders (linuxbrew, brew's default formula). + FFmpegVersion string `json:"ffmpegVersion,omitempty"` + FFmpegPath string `json:"ffmpegPath,omitempty"` + HWEncoders []string `json:"hwEncoders,omitempty"` + HWDevices []string `json:"hwDevices,omitempty"` // Managed-VPN split-tunnel state. The web tracks which agent holds the single // WireGuard slot (1 VPNResellers account = 1 WG keypair = 1 concurrent // connection); other agents are told to use OpenVPN on their host instead. diff --git a/internal/cmd/daemon.go b/internal/cmd/daemon.go index 6c00e95..2e0c074 100644 --- a/internal/cmd/daemon.go +++ b/internal/cmd/daemon.go @@ -143,7 +143,19 @@ func runDaemonStart() error { // is what the web side uses to decide whether the user should pre-empt // transcoding by downloading a smaller version (4K source on a software // libx264-only host is the canonical case where pre-download wins). - hwAccelPick := engine.DetectHWAccel(context.Background(), cfg.Library.FFmpegPath) + // + // Use the full diagnostic (encoders + devices + ffmpeg version) instead + // of just the picked backend — the extra fields ride along in the + // register payload so the web "Diagnose transcoder" modal can show *why* + // libx264 was selected on a host with a GPU (e.g. brew's ffmpeg without + // --enable-nvenc). 10 s ceiling so a hung ffmpeg binary can't stall + // startup forever. + ffmpegResolved, _ := mediainfo.ResolveFFmpeg(cfg.Library.FFmpegPath) + probeCtx, probeCancel := context.WithTimeout(context.Background(), 10*time.Second) + defer probeCancel() // guard against a panic inside DetectHWAccelDiagnostic + hwDiag := engine.DetectHWAccelDiagnostic(probeCtx, ffmpegResolved) + log.Println(hwDiag.LogLine()) + hwAccelPick := hwDiag.Pick maxTranscodeHeight := 1080 if hwAccelPick != engine.HWAccelNone { maxTranscodeHeight = 2160 @@ -162,6 +174,11 @@ func runDaemonStart() error { ScanPaths: library.ResolveScanPaths(cfg.Download.Dir, cfg.Organize.MoviesDir, cfg.Organize.TVShowsDir, cfg.Library.ScanPath), HWAccel: string(hwAccelPick), MaxTranscodeHeight: maxTranscodeHeight, + FFmpegVersion: hwDiag.FFmpegVersion, + FFmpegPath: hwDiag.FFmpegPath, + HWEncoders: hwDiag.Encoders, + HWDevices: hwDiag.Devices, + AutoUpgrade: cfg.Daemon.AutoUpgradeEnabled(), } // Create HTTP client with mirror failover so a `.com` block-out rolls @@ -292,13 +309,47 @@ func runDaemonStart() error { // Create persistent stream server streamSrv := engine.NewStreamServer(cfg.Download.StreamPort) streamSrv.SetUPnPEnabled(cfg.Download.EnableUPnP) - streamSrv.SetCORSAllowedOrigins(cfg.Download.CORSExtraOrigins) + // CORS extras = operator config + dynamic mirror list from /api/mirrors. + // Without the mirror merge, a user playing from `torrentclaw.to` (or any + // future mirror) hits the daemon, gets 200 + body, but no + // `Access-Control-Allow-Origin` → browser drops the response → player + // reports "404 todos los canales". Fetching /api/mirrors at startup + // future-proofs against mirror additions without a CLI rebuild. + corsExtras := append([]string(nil), cfg.Download.CORSExtraOrigins...) + corsExtras = append(corsExtras, mirrorCORSOrigins(ctx, cfg, userAgent)...) + streamSrv.SetCORSAllowedOrigins(corsExtras) // Reap HLS tmpdirs left over from a previous daemon run before we start // accepting new sessions. The in-memory registry doesn't survive a // restart, so without this disk usage grows unbounded across restarts. if err := engine.CleanupHLSOrphanDirs(); err != nil { log.Printf("[hls] orphan tmpdir cleanup: %v", err) } + + // Persistent HLS segment cache — survives across sessions so re-plays + // of the same file at the same quality skip ffmpeg entirely. Off when + // hls_cache.enabled = false; size cap from hls_cache.size_gb; path from + // hls_cache.dir (defaults to ~/.cache/unarr/hls-cache). + var hlsCache *engine.HLSCache + if cfg.Download.HLSCache.Enabled { + cacheDir := cfg.Download.HLSCache.Dir + if cacheDir == "" { + if base, err := os.UserCacheDir(); err == nil { + cacheDir = filepath.Join(base, "unarr", "hls-cache") + } else { + cacheDir = filepath.Join(os.TempDir(), "unarr-hls-cache") + } + } + c, err := engine.NewHLSCache(cacheDir, cfg.Download.HLSCache.SizeGB) + if err != nil { + log.Printf("[hls_cache] init failed (%v) — falling back to per-session tmpdirs", err) + } else { + hlsCache = c + hlsCache.StartSweeper(ctx, time.Hour) + log.Printf("[hls_cache] enabled: dir=%s budget=%dGB", cacheDir, cfg.Download.HLSCache.SizeGB) + } + } else { + log.Printf("[hls_cache] disabled by config — every play re-encodes from scratch") + } if err := streamSrv.Listen(ctx); err != nil { return fmt.Errorf("start stream server: %w", err) } @@ -543,15 +594,30 @@ func runDaemonStart() error { Quality: sess.Quality, AudioIndex: sess.AudioIndex, Transcode: tcRuntime, + Cache: hlsCache, } - hsess, err := engine.StartHLSSession(hlsCtx, hlsCfg) - if err != nil { - playerSessionRegistry.remove(sess.SessionID) - hlsCancel() - log.Printf("[hls %s] start failed: %v", agent.ShortID(sess.SessionID), err) - return - } - streamSrv.HLS().Register(hsess) + // StartHLSSession runs ffprobe (15 s cap, typical 0.3–1 s) before + // returning. Doing this synchronously inside the sync handler holds + // the next sync HTTP cycle until ffprobe is done, so any other + // pending actions (new tasks, deletes) wait too. Hand it off so + // the sync loop returns immediately — browser HEAD probes already + // have a 30 s retry budget that absorbs the gap until + // `streamSrv.HLS().Register` lands. + go func() { + hsess, err := engine.StartHLSSession(hlsCtx, hlsCfg) + if err != nil { + playerSessionRegistry.remove(sess.SessionID) + hlsCancel() + log.Printf("[hls %s] start failed: %v", agent.ShortID(sess.SessionID), err) + return + } + streamSrv.HLS().Register(hsess) + // Tell the server seg-0 is on disk as soon as it lands so the + // player's SSE subscription flips its "Preparando…" UI without + // waiting for the browser HEAD-probe loop to discover it + // independently. Cache-HIT sessions are ready immediately. + go watchSessionReady(hlsCtx, agentClient, hsess, sess.SessionID) + }() } // Periodic DHT node persistence (every 5 min) @@ -834,3 +900,93 @@ func superviseFunnel(ctx context.Context, d *agent.Daemon, port int) { backoff = min(backoff*2, maxBackoff) } } + +// mirrorCORSOrigins fetches /api/mirrors from the configured primary (+ extra +// mirror candidates + static IPFS fallback) and returns the discovered URLs as +// Origin strings. Best-effort: any failure logs a warning and returns an empty +// slice; the static defaultCORSAllowedOrigins in validate.go covers the known +// mirrors (.com / .to / built-in onion) so the daemon still accepts the +// official surfaces when this call fails. +// +// Bounded to a short timeout so a slow /api/mirrors response can't delay +// daemon startup — every second here is a second the user can't play. +func mirrorCORSOrigins(parent context.Context, cfg config.Config, userAgent string) []string { + ctx, cancel := context.WithTimeout(parent, 10*time.Second) + defer cancel() + + candidates := append([]string{cfg.Auth.APIURL}, cfg.Auth.Mirrors...) + resp, err := agent.FetchMirrorsWithFallback(ctx, candidates, userAgent) + if err != nil { + log.Printf("[cors] mirror discovery failed (%v) — using static allowlist only", err) + return nil + } + + seen := make(map[string]struct{}) + out := make([]string, 0, len(resp.Mirrors)) + add := func(rawURL string) { + if rawURL == "" { + return + } + origin := strings.TrimRight(rawURL, "/") + if _, dup := seen[origin]; dup { + return + } + seen[origin] = struct{}{} + out = append(out, origin) + } + for _, m := range resp.Mirrors { + add(m.URL) + } + if resp.Tor != nil { + add(resp.Tor.URL) + } + if len(out) > 0 { + log.Printf("[cors] merged %d mirror origins from /api/mirrors", len(out)) + } + return out +} + +// watchSessionReady polls HLSSession.ReadyCount until the first segment + +// init.mp4 are on disk, then POSTs /api/internal/agent/session-ready so +// the web side flips streaming_session.ready_at — which its SSE endpoint +// pushes to subscribed players. Cache-HIT sessions are ready the moment +// StartHLSSession returns and POST immediately. +// +// Bounded by a 60 s deadline so a permanently stuck encoder doesn't keep +// a goroutine alive forever; if seg-0 never lands the player falls back +// to its existing HEAD-probe retry path anyway. +func watchSessionReady(ctx context.Context, client *agent.Client, hsess *engine.HLSSession, sessionID string) { + deadline := time.Now().Add(60 * time.Second) + ticker := time.NewTicker(200 * time.Millisecond) + defer ticker.Stop() + for { + // Session torn down through a path that didn't cancel ctx (registry + // replace, idle sweep, internal kill). Bail before polling further — + // without this check the watcher could keep alive for up to 60 s on + // a dead HLSSession that's never going to become ready. + if hsess.IsClosed() { + return + } + // Cache HIT or seg-0 ready → notify + done. + if hsess.FromCache() || hsess.ReadyCount() >= 1 { + // Parent ctx so a session cancel mid-POST (user closed tab, + // daemon shutdown) tears down the in-flight webhook instead of + // blocking the goroutine for up to 10 s on a now-orphan call. + rctx, cancel := context.WithTimeout(ctx, 10*time.Second) + if err := client.MarkSessionReady(rctx, sessionID); err != nil { + log.Printf("[hls %s] mark-ready failed: %v", agent.ShortID(sessionID), err) + } + cancel() + return + } + select { + case <-ctx.Done(): + return + case <-ticker.C: + } + if time.Now().After(deadline) { + log.Printf("[hls %s] mark-ready: timeout waiting for seg-0", agent.ShortID(sessionID)) + return + } + } +} diff --git a/internal/cmd/daemon_control.go b/internal/cmd/daemon_control.go index 558fb26..4ac4d10 100644 --- a/internal/cmd/daemon_control.go +++ b/internal/cmd/daemon_control.go @@ -1,6 +1,7 @@ package cmd import ( + "errors" "fmt" "os" "os/exec" @@ -262,9 +263,12 @@ func runDaemonReload() error { // stopDaemonByPID reads the state file and sends a graceful stop to the daemon PID. // Used as fallback on platforms without a service manager (and as Windows implementation). func stopDaemonByPID() error { - state := agent.ReadState() - if state == nil { - return fmt.Errorf("daemon does not appear to be running (state file not found)") + state, err := agent.LoadState() + if err != nil { + if errors.Is(err, agent.ErrDaemonNotRunning) { + return err + } + return fmt.Errorf("read daemon state: %w", err) } return killPID(state.PID) } diff --git a/internal/cmd/reload_unix.go b/internal/cmd/reload_unix.go index 056112f..34d8e4d 100644 --- a/internal/cmd/reload_unix.go +++ b/internal/cmd/reload_unix.go @@ -3,6 +3,7 @@ package cmd import ( + "errors" "fmt" "log" "os" @@ -43,9 +44,12 @@ func startReloadWatcher(rc *ReloadableConfig) { // sendReloadSignal sends SIGUSR1 to the running daemon process. func sendReloadSignal() error { - state := agent.ReadState() - if state == nil { - return fmt.Errorf("daemon does not appear to be running (state file not found)") + state, err := agent.LoadState() + if err != nil { + if errors.Is(err, agent.ErrDaemonNotRunning) { + return err + } + return fmt.Errorf("read daemon state: %w", err) } p, err := os.FindProcess(state.PID) if err != nil { diff --git a/internal/cmd/root.go b/internal/cmd/root.go index b28ec92..375d8e9 100644 --- a/internal/cmd/root.go +++ b/internal/cmd/root.go @@ -25,16 +25,20 @@ var ( func init() { rootCmd = &cobra.Command{ - Use: "unarr", - Short: "unarr — torrent search and management", - Long: `unarr is a powerful terminal tool for torrent search and management. - -Search 30+ torrent sources, inspect torrent quality, discover popular content, -find streaming providers, and manage your media collection — all from your terminal. + Use: "unarr", + Version: Version, + Short: "Terminal torrent + debrid + usenet client — download, stream, transcode", + Long: `unarr is a terminal-native client that downloads torrents, debrid links, +and usenet (NZB) — all from the same binary. It streams content straight +to mpv/vlc with sequential piece prioritization, transcodes on the fly via +ffmpeg with hardware acceleration (NVENC, QSV, VA-API, VideoToolbox), and +organizes your library into Movies/TV folders. Run it one-shot or as a +long-running daemon with a built-in WireGuard split-tunnel and remote +playback over Cloudflare Funnel. Get started: unarr init First-time configuration wizard - unarr search "breaking bad" Search for content + unarr download Grab a torrent one-shot unarr start Start the download daemon Documentation: https://torrentclaw.com/cli @@ -55,7 +59,7 @@ Source: https://github.com/torrentclaw/unarr`, // Command groups for organized help output rootCmd.AddGroup( &cobra.Group{ID: "start", Title: "Getting Started:"}, - &cobra.Group{ID: "search", Title: "Search & Discovery:"}, + &cobra.Group{ID: "search", Title: "Catalog & Discovery:"}, &cobra.Group{ID: "download", Title: "Downloads & Streaming:"}, &cobra.Group{ID: "daemon", Title: "Daemon Management:"}, &cobra.Group{ID: "system", Title: "System & Diagnostics:"}, diff --git a/internal/cmd/scan.go b/internal/cmd/scan.go index df66a18..d05ae29 100644 --- a/internal/cmd/scan.go +++ b/internal/cmd/scan.go @@ -241,7 +241,7 @@ func printScanSummary(cache *library.LibraryCache) { continue } - res := library.ResolveResolution(item.MediaInfo.Video.Height) + res := library.ResolveResolution(item.MediaInfo.Video.Width, item.MediaInfo.Video.Height) if res == "" { res = "other" } diff --git a/internal/cmd/version.go b/internal/cmd/version.go index 32f1ce3..194e3c0 100644 --- a/internal/cmd/version.go +++ b/internal/cmd/version.go @@ -1,4 +1,4 @@ package cmd // Version is the CLI version. Overridden by goreleaser ldflags at release time. -var Version = "0.9.6" +var Version = "0.9.15" diff --git a/internal/config/config.go b/internal/config/config.go index 12573b4..dd406a6 100644 --- a/internal/config/config.go +++ b/internal/config/config.go @@ -52,10 +52,22 @@ type DownloadConfig struct { EnableUPnP bool `toml:"enable_upnp"` // map StreamPort to the WAN via UPnP/NAT-PMP (default: false; opt-in because it exposes the unauthenticated /stream + /hls endpoints to the public internet) CORSExtraOrigins []string `toml:"cors_extra_origins"` // extra browser origins added on top of the baked-in allowlist (torrentclaw.com, app.torrentclaw.com, localhost:3030) Transcode TranscodeConfig `toml:"transcode"` + HLSCache HLSCacheConfig `toml:"hls_cache"` VPN VPNConfig `toml:"vpn"` Funnel FunnelConfig `toml:"funnel"` } +// HLSCacheConfig controls the persistent HLS segment cache. A completed encode +// is kept on disk so a second play of the same file at the same quality skips +// ffmpeg entirely. Old entries are evicted (LRU) once the cache exceeds the +// size budget. Enabled by default — disable to save disk space at the cost of +// re-encoding every play. +type HLSCacheConfig struct { + Enabled bool `toml:"enabled"` // default: true + SizeGB int `toml:"size_gb"` // size budget in gigabytes; default: 5; minimum: 1 + Dir string `toml:"dir"` // override storage path; default: ~/.cache/unarr/hls-cache +} + // FunnelConfig gates the optional CloudFlare Quick Tunnel that exposes the // daemon's HLS server over a public HTTPS hostname (https://.try // cloudflare.com). Enabling it lets the web player on torrentclaw.com play @@ -84,9 +96,27 @@ type VPNConfig struct { // Disabled by default; enabling requires ffmpeg + ffprobe on PATH (or // explicit paths via the library config). type TranscodeConfig struct { - Enabled bool `toml:"enabled"` // master switch - HWAccel string `toml:"hw_accel"` // "auto" | "none" | "nvenc" | "qsv" | "vaapi" | "videotoolbox" - Preset string `toml:"preset"` // libx264 preset; "veryfast" by default + Enabled bool `toml:"enabled"` // master switch + HWAccel string `toml:"hw_accel"` // "auto" | "none" | "nvenc" | "qsv" | "vaapi" | "videotoolbox" + // Preset is the encoder speed/quality dial. Only used on software encode + // (libx264) — HW backends (NVENC/QSV/VAAPI/VideoToolbox) use vendor + // presets that don't share libx264's vocabulary and would be rejected + // by ffmpeg if passed here. + // + // Empty (default) → engine picks "superfast" — latency-biased, ~3 s + // first-play on 1080p source on a modern x86 CPU. Marginal quality loss + // at 5-25 Mbps target bitrates. + // + // For better quality at slower first-play (1-2 s slower per seg): + // "veryfast" — previous default; balanced + // "faster" — slight quality bump + // "fast" — meaningful quality bump + // "medium" — libx264 stock default; CPU-bound on 4K + // "slow" / "slower" / "veryslow" — only for batch encodes, not real-time HLS + // + // Or faster: + // "ultrafast" — lowest quality, fastest encode + Preset string `toml:"preset"` VideoBitrate string `toml:"video_bitrate"` // e.g. "5M" AudioBitrate string `toml:"audio_bitrate"` // e.g. "192k" MaxHeight int `toml:"max_height"` // optional downscale cap (e.g. 720) @@ -101,8 +131,27 @@ type OrganizeConfig struct { type DaemonConfig struct { StatusInterval string `toml:"status_interval"` + // AutoUpgrade gates the daemon's response to a server-flagged upgrade + // (set via the "Force update" button on the web). When true the daemon + // downloads + replaces the binary in-place and exits so the service + // supervisor respawns on the new version. When false the daemon only + // logs "new version available" and the operator must run `unarr update` + // manually. Default: true. Available since unarr 0.9.6. + AutoUpgrade *bool `toml:"auto_upgrade"` } +// AutoUpgradeEnabled returns the resolved AutoUpgrade flag — defaults to true +// when the user has not set it explicitly. Pointer-vs-bool because Go's +// zero-value bool would collapse "unset" and "false" together. +func (d DaemonConfig) AutoUpgradeEnabled() bool { + if d.AutoUpgrade == nil { + return true + } + return *d.AutoUpgrade +} + +func boolPtr(v bool) *bool { return &v } + type NotificationsConfig struct { Enabled bool `toml:"enabled"` } @@ -145,7 +194,10 @@ func Default() Config { Transcode: TranscodeConfig{ Enabled: true, HWAccel: "auto", - Preset: "veryfast", + // Empty preset → engine.ResolveEncoderProfile picks the + // latency-biased default ("superfast" on libx264). Override + // in config.toml when quality > first-start latency matters. + Preset: "", AudioBitrate: "192k", MaxConcurrent: 2, }, @@ -156,11 +208,24 @@ func Default() Config { // `unarr funnel off` (sets enabled=false in the TOML). Enabled: true, }, + HLSCache: HLSCacheConfig{ + // On by default — second play of a recently watched file at the + // same quality skips ffmpeg (instant start, near-zero CPU). + // Users can opt out (hls_cache.enabled=false) or shrink the + // budget (hls_cache.size_gb) when disk is tight. + Enabled: true, + SizeGB: 5, + }, + }, + Daemon: DaemonConfig{ + // Pointer-to-true so Default() round-trips through TOML marshal + // as `auto_upgrade = true` instead of an omitted key — keeps the + // freshly-written config aligned with what README documents. + AutoUpgrade: boolPtr(true), }, Organize: OrganizeConfig{ Enabled: true, }, - Daemon: DaemonConfig{}, Notifications: NotificationsConfig{ Enabled: true, }, @@ -236,7 +301,12 @@ func applyDefaults(cfg *Config, meta toml.MetaData) { cfg.Download.Transcode.HWAccel = "auto" } if !meta.IsDefined("downloads", "transcode", "preset") { - cfg.Download.Transcode.Preset = "veryfast" + // Empty = let engine.ResolveEncoderProfile pick the latency-biased + // default ("superfast" on libx264). Users wanting better quality at + // slower first-play can override to "veryfast" / "fast" / "medium" in + // config.toml. Ignored when hw_accel picks NVENC/QSV/VAAPI/VideoToolbox + // (those have built-in vendor presets). + cfg.Download.Transcode.Preset = "" } if !meta.IsDefined("downloads", "transcode", "audio_bitrate") { cfg.Download.Transcode.AudioBitrate = "192k" diff --git a/internal/config/config_test.go b/internal/config/config_test.go index 8097395..c43599f 100644 --- a/internal/config/config_test.go +++ b/internal/config/config_test.go @@ -215,8 +215,11 @@ name = "Test" if cfg.Download.Transcode.HWAccel != "auto" { t.Errorf("Transcode.HWAccel = %q, want auto", cfg.Download.Transcode.HWAccel) } - if cfg.Download.Transcode.Preset != "veryfast" { - t.Errorf("Transcode.Preset = %q, want veryfast", cfg.Download.Transcode.Preset) + if cfg.Download.Transcode.Preset != "" { + // Default is now empty — engine.ResolveEncoderProfile picks + // "superfast" on libx264 for first-start latency. Users + // wanting better quality override in config.toml. + t.Errorf("Transcode.Preset = %q, want empty", cfg.Download.Transcode.Preset) } if cfg.Download.Transcode.MaxConcurrent != 2 { t.Errorf("Transcode.MaxConcurrent = %d, want 2", cfg.Download.Transcode.MaxConcurrent) diff --git a/internal/engine/hls.go b/internal/engine/hls.go index 9524627..86219d5 100644 --- a/internal/engine/hls.go +++ b/internal/engine/hls.go @@ -32,10 +32,46 @@ import ( "time" ) -// hlsSegmentDuration is the target seconds per HLS fragment. Four seconds is -// the Plex/Apple default — short enough that seek granularity is acceptable, -// long enough that GOP overhead doesn't dominate. -const hlsSegmentDuration = 4 +// hlsSegmentDuration is the target seconds per HLS fragment. +// +// We use 2 seconds (not the more common 4-6 s). Trade-off: 2× more segments +// per source (a 2 h movie produces 3600 segments instead of 1800), but the +// player's first-frame wait drops to ~half — ffmpeg only needs to encode +// 2 s before seg-0 lands. For software encodes on 4K this is ~1 s instead +// of ~3 s of cold-cache wait. Well within HLS spec (Apple recommends 6 s, +// but 2-6 s is acceptable; Low-Latency HLS uses 1-2 s segments). +// +// Caveat for existing cached encodes: cache entries from 0.9.9 used 4 s +// segments. After this bump, VerifyComplete (which checks the highest +// expected segment index) returns false for those entries — they're +// invalidated + re-encoded with 2 s segments on next play. Self-healing. +const hlsSegmentDuration = 2 + +// segmentDurationFor returns the target duration (in whole seconds) for the +// segment at index idx. With uniform-duration segments this is always +// hlsSegmentDuration; the helper exists so a future short-first-segment +// variant can be slotted in here without touching every call site. +func segmentDurationFor(idx int) int { + return hlsSegmentDuration +} + +// segmentStartSec returns the wall-clock start time of segment idx. Used +// to compute the `-ss` flag when ffmpeg restarts at a mid-file segment. +func segmentStartSec(idx int) float64 { + if idx <= 0 { + return 0 + } + return float64(idx * hlsSegmentDuration) +} + +// segmentCountForDuration returns how many segments cover a source of the +// given duration. Always returns at least 1. +func segmentCountForDuration(dur float64) int { + if dur <= 0 { + return 1 + } + return int((dur + float64(hlsSegmentDuration) - 1) / float64(hlsSegmentDuration)) +} // hlsSessionTTL is how long a session can sit idle (no segment requests) // before the manager kills ffmpeg + cleans the tmpdir. @@ -100,6 +136,11 @@ type HLSSessionConfig struct { Quality string // "2160p"|"1080p"|"720p"|"480p"|"original"|"" AudioIndex int // 0-based ffmpeg audio stream selection (-map 0:a:N). -1 = default. Transcode TranscodeRuntime + // Cache is an optional persistent segment cache keyed by (source, quality, + // audio). When set, completed encodes are kept across sessions so re-plays + // of the same file at the same quality skip ffmpeg entirely. nil disables + // caching (per-session tmpdir, deleted on Close — original behavior). + Cache *HLSCache } // HLSSession owns a tmpdir + ffmpeg subprocess producing HLS fragments. @@ -131,14 +172,29 @@ type HLSSession struct { restartCount int // bounded auto-restart counter (resets on Close) lastRestartAt time.Time - // readyCond + readyMax track which segments ffmpeg has finished writing. - // Handlers waiting on a future segment block on readyCond until the - // poller advances readyMax past their index (or ffmpeg exits). + // readyCh + readyMax track how many segments ffmpeg has finished writing. + // readyMax is a COUNT (not an index): readyMax=N means seg-0 … seg-(N-1) + // are fully on disk. A handler waiting on `idx` blocks until + // `idx < readyMax` (segment idx is present). The pollSegments goroutine + // advances readyMax and re-creates readyCh on every step. readyMu sync.Mutex - readyMax int // highest segment index whose .m4s file is fully written + readyMax int exitErr error exited bool readyCh chan struct{} // closed + replaced each time readyMax advances + + // Persistent cache state. cache==nil means caching disabled for this session. + // fromCache=true means the session is replaying a completed encode and no + // ffmpeg subprocess was spawned. writerLockHeld=true means this session + // owns the per-key TryAcquireWriter claim — Close must ReleaseWriter. + // subsDone closes when the subtitle extractor goroutine returns (or is + // nil when the source had no subtitle tracks); MarkComplete waits on it + // so a HIT replay never serves partial .vtt files. + cache *HLSCache + cacheKey string + fromCache bool + writerLockHeld bool + subsDone chan struct{} } // hlsSeekAhead is how many segments past the writer's current position the @@ -263,18 +319,78 @@ func StartHLSSession(ctx context.Context, cfg HLSSessionConfig) (*HLSSession, er return nil, errors.New("hls: source has no duration") } - tmpDir := filepath.Join(hlsTmpDirRoot(), cfg.SessionID) + // Resolve tmpDir + cache placement. Three states: + // 1. cache disabled → per-session tmpdir, deleted on Close. + // 2. cache HIT (.complete found) → read from cache dir, no ffmpeg, Pin. + // 3. cache MISS, writer-lock OK → ffmpeg writes to cache dir, Pin + writer-lock. + // 4. cache MISS, writer-lock NO → another session already writing this + // key; fall back to private per-session tmpdir + // (no caching for this session — second-writer + // would corrupt the first one's segments). + var ( + tmpDir string + cacheKey string + fromCache bool + writerLockHeld bool + ) + if cfg.Cache != nil { + cacheKey = cfg.Cache.KeyFor(cfg.SourcePath, cfg.Quality, cfg.AudioIndex) + // Integrity gate: HasComplete just stats the marker. If init.mp4 or + // the last segment vanished (external rm, partial-disk failure), we + // can't actually serve a HIT — drop the dir and re-encode. + segCountForVerify := segmentCountForDuration(probe.DurationSec) + if cfg.Cache.HasComplete(cacheKey) && !cfg.Cache.VerifyComplete(cacheKey, segCountForVerify) { + log.Printf("[hls %s] cache %s sealed but failed integrity check — re-encoding", + shortHLSID(cfg.SessionID), cacheKey) + _ = cfg.Cache.Invalidate(cacheKey) + } + if cfg.Cache.HasComplete(cacheKey) { + // HIT: read-only replay — many concurrent HITs are fine. + tmpDir = cfg.Cache.DirFor(cacheKey) + cfg.Cache.Pin(cacheKey) + fromCache = true + cfg.Cache.RecordHit() + _ = cfg.Cache.Touch(cacheKey) + } else if cfg.Cache.TryAcquireWriter(cacheKey) { + tmpDir = cfg.Cache.DirFor(cacheKey) + cfg.Cache.Pin(cacheKey) + writerLockHeld = true + cfg.Cache.RecordMiss() + } else { + // Another session is writing this key — fall back to private + // dir so we don't trample its segments. + log.Printf("[hls %s] cache key %s busy, falling back to per-session tmpdir", + shortHLSID(cfg.SessionID), cacheKey) + tmpDir = filepath.Join(hlsTmpDirRoot(), cfg.SessionID) + cacheKey = "" // disable caching for this session + cfg.Cache.RecordMiss() + } + } else { + tmpDir = filepath.Join(hlsTmpDirRoot(), cfg.SessionID) + } + + cleanupOnError := func() { + if cfg.Cache != nil && cacheKey != "" { + cfg.Cache.Unpin(cacheKey) + if writerLockHeld { + cfg.Cache.ReleaseWriter(cacheKey) + _ = cfg.Cache.Invalidate(cacheKey) + } + } else { + _ = os.RemoveAll(tmpDir) + } + } + if err := os.MkdirAll(filepath.Join(tmpDir, "video"), 0o755); err != nil { + cleanupOnError() return nil, fmt.Errorf("hls: mkdir video: %w", err) } if err := os.MkdirAll(filepath.Join(tmpDir, "subs"), 0o755); err != nil { + cleanupOnError() return nil, fmt.Errorf("hls: mkdir subs: %w", err) } - segCount := int((probe.DurationSec + float64(hlsSegmentDuration) - 1) / float64(hlsSegmentDuration)) - if segCount < 1 { - segCount = 1 - } + segCount := segmentCountForDuration(probe.DurationSec) s := &HLSSession{ cfg: cfg, @@ -285,10 +401,30 @@ func StartHLSSession(ctx context.Context, cfg HLSSessionConfig) (*HLSSession, er startedAt: time.Now(), lastTouch: time.Now(), readyCh: make(chan struct{}), + cache: cfg.Cache, + cacheKey: cacheKey, + fromCache: fromCache, + writerLockHeld: writerLockHeld, } s.manifestVideo = renderVideoPlaylist(probe.DurationSec, segCount) s.manifestRoot = renderMasterPlaylist(probe, cfg.Quality) + // Cache HIT: every segment + init.mp4 is already on disk. Skip ffmpeg + // entirely and mark readyMax so handlers don't wait. Background subtitle + // extraction is also unnecessary — subs were extracted on the original run. + if fromCache { + s.readyMu.Lock() + s.readyMax = segCount - 1 + s.exited = true + close(s.readyCh) + s.readyCh = nil + s.readyMu.Unlock() + log.Printf("[hls %s] cache HIT %s: %s, %.1fs, %d segs (quality=%s)", + shortHLSID(cfg.SessionID), cacheKey, filepath.Base(cfg.SourcePath), + probe.DurationSec, segCount, coalesce(cfg.Quality, "auto")) + return s, nil + } + // Spawn ffmpeg under a dedicated context so Close() can kill it without // touching the parent ctx. ffCtx, cancel := context.WithCancel(context.Background()) @@ -298,7 +434,7 @@ func StartHLSSession(ctx context.Context, cfg HLSSessionConfig) (*HLSSession, er cmd.Stderr = &hlsStderrCapture{owner: s} if err := cmd.Start(); err != nil { cancel() - _ = os.RemoveAll(tmpDir) + cleanupOnError() return nil, fmt.Errorf("hls: start ffmpeg: %w", err) } s.cmd = cmd @@ -307,12 +443,30 @@ func StartHLSSession(ctx context.Context, cfg HLSSessionConfig) (*HLSSession, er go s.pollSegments(ffCtx) if len(probe.SubtitleTracks) > 0 { - go s.extractSubtitles(ffCtx) + s.subsDone = make(chan struct{}) + go func() { + defer close(s.subsDone) + s.extractSubtitles(ffCtx) + }() } - log.Printf("[hls %s] started: %s, %.1fs, %d segs (quality=%s)", + cachedNote := "" + if cfg.Cache != nil { + cachedNote = fmt.Sprintf(" (cache-miss %s)", cacheKey) + } + // Surface the encoder profile so a "first-start was slow" report can be + // triaged from the agent log alone — `encoder=libx264 accel=none` means + // the user's ffmpeg has no HW encoders compiled in, which is the most + // common root cause (linuxbrew, default brew formula on macOS). + profile := ResolveEncoderProfile(cfg.Transcode.HWAccel, cfg.Transcode.Preset) + presetNote := "" + if profile.Preset != "" { + presetNote = " preset=" + profile.Preset + } + log.Printf("[hls %s] started: %s, %.1fs, %d segs (quality=%s, encoder=%s accel=%s%s)%s", shortHLSID(cfg.SessionID), filepath.Base(cfg.SourcePath), - probe.DurationSec, segCount, coalesce(cfg.Quality, "auto")) + probe.DurationSec, segCount, coalesce(cfg.Quality, "auto"), + profile.Codec, string(cfg.Transcode.HWAccel), presetNote, cachedNote) return s, nil } @@ -365,6 +519,28 @@ func (s *HLSSession) ProbeInfo() map[string]any { } } +// ReadyCount returns how many segments are currently fully on disk. +// Caller can `>= 1` it to check whether seg-0 has landed (and so the +// player can be told to attach). For cache-HIT sessions this is always +// `segmentCount` from the moment StartHLSSession returns. +func (s *HLSSession) ReadyCount() int { + s.readyMu.Lock() + defer s.readyMu.Unlock() + return s.readyMax +} + +// FromCache reports whether this session was served from the HLS cache +// (no ffmpeg subprocess spawned). Used by ready-watcher logic to short- +// circuit polling — a cache HIT is ready the moment we return. +func (s *HLSSession) FromCache() bool { return s.fromCache } + +// IsClosed reports whether Close() has been invoked. Exposed (vs the +// internal isClosed) so external watchers — the ready-webhook +// goroutine in cmd/daemon.go — can short-circuit polling on a session +// that was torn down through a different code path (registry replace, +// idle sweep) without racing on the unexported helper. +func (s *HLSSession) IsClosed() bool { return s.isClosed() } + // MasterPlaylist returns the rendered master.m3u8 contents. func (s *HLSSession) MasterPlaylist() string { return s.manifestRoot } @@ -385,8 +561,15 @@ func (s *HLSSession) Touch() { s.mu.Unlock() } -// Close stops ffmpeg, deletes the tmpdir, and prevents further requests from -// blocking on segment readiness. Idempotent. +// Close stops ffmpeg and prevents further requests from blocking on segment +// readiness. Idempotent. +// +// Disk lifecycle: +// - cache disabled → delete tmpDir (original behavior). +// - cache enabled + this session was a HIT → keep dir, just unpin. +// - cache enabled + this was a write session → if ffmpeg exited cleanly and +// every segment is on disk, persist with .complete and keep dir. Otherwise +// drop the dir so a half-written cache doesn't survive into the next play. func (s *HLSSession) Close() error { s.mu.Lock() if s.closed { @@ -407,7 +590,47 @@ func (s *HLSSession) Close() error { s.readyCh = nil } s.exited = true + exitErr := s.exitErr s.readyMu.Unlock() + + if s.cache != nil && s.cacheKey != "" { + defer s.cache.Unpin(s.cacheKey) + if s.writerLockHeld { + defer s.cache.ReleaseWriter(s.cacheKey) + } + if s.fromCache { + log.Printf("[hls %s] closed (cache reuse)", shortHLSID(s.cfg.SessionID)) + return nil + } + // Wait briefly for the subtitle extractor to finish so a cached + // replay never serves half-written .vtt files. Bounded so a stuck + // extractor can't block Close indefinitely; on timeout we treat + // the cache as incomplete and drop it. + subsOK := true + if s.subsDone != nil { + select { + case <-s.subsDone: + case <-time.After(15 * time.Second): + log.Printf("[hls %s] subtitle extractor timeout — not caching", shortHLSID(s.cfg.SessionID)) + subsOK = false + } + } + if subsOK && exitErr == nil && s.allSegmentsPresent() { + if err := s.cache.MarkComplete(s.cacheKey); err == nil { + log.Printf("[hls %s] cache persisted %s", shortHLSID(s.cfg.SessionID), s.cacheKey) + return nil + } else { + log.Printf("[hls %s] cache persist failed: %v", shortHLSID(s.cfg.SessionID), err) + } + } + // Partial / failed → drop so we re-encode next time. + if err := s.cache.Invalidate(s.cacheKey); err != nil { + log.Printf("[hls %s] cache invalidate failed: %v", shortHLSID(s.cfg.SessionID), err) + } + log.Printf("[hls %s] closed (cache discarded)", shortHLSID(s.cfg.SessionID)) + return nil + } + if tmpDir != "" { _ = os.RemoveAll(tmpDir) } @@ -415,6 +638,31 @@ func (s *HLSSession) Close() error { return nil } +// allSegmentsPresent reports whether every expected segment (and init.mp4) is +// on disk AND validated by the segment poller. Used to decide whether a +// finished session is cacheable. We trust readyMax (advanced by pollSegments +// only after the next segment exists, proving the predecessor is fully closed) +// over a naive Size>0 stat that could accept truncated mid-write files. +func (s *HLSSession) allSegmentsPresent() bool { + if fi, err := os.Stat(filepath.Join(s.tmpDir, "video", "init.mp4")); err != nil || fi.Size() == 0 { + return false + } + s.readyMu.Lock() + readyMax := s.readyMax + s.readyMu.Unlock() + if readyMax < s.segmentCount-1 { + return false + } + for i := 0; i < s.segmentCount; i++ { + path := filepath.Join(s.tmpDir, "video", fmt.Sprintf("seg-%d.m4s", i)) + fi, err := os.Stat(path) + if err != nil || fi.Size() == 0 { + return false + } + } + return true +} + // waitFFmpeg reaps the ffmpeg process and records its exit error for handlers. // // Auto-restart supervisor: if ffmpeg crashes (non-graceful exit) and the @@ -715,8 +963,10 @@ func (s *HLSSession) restartFromSegment(targetIdx int) error { time.Sleep(50 * time.Millisecond) } - // Build args for the new ffmpeg with -ss offset. - startSec := float64(targetIdx * hlsSegmentDuration) + // Build args for the new ffmpeg with -ss offset. Segments are non-uniform + // (seg-0 is hlsInitSegmentDuration s, the rest are hlsSegmentDuration s), + // so use segmentStartSec for the seek time instead of multiplying. + startSec := segmentStartSec(targetIdx) args := buildHLSFFmpegArgsAt(s.cfg, s.probe, s.tmpDir, targetIdx, startSec) ffCtx, cancel := context.WithCancel(context.Background()) @@ -781,23 +1031,77 @@ func buildHLSFFmpegArgs(cfg HLSSessionConfig, probe *StreamProbe, tmpDir string) return buildHLSFFmpegArgsAt(cfg, probe, tmpDir, 0, 0) } +// EncoderProfile names the codec + preset + decoder hint combination the HLS +// pipeline picks for the given hardware backend + transcode config. Exposed +// so callers can log the chosen encoder before ffmpeg launches and so both +// the demuxer-side `-hwaccel` flag and the encoder-side argv stay in sync +// (otherwise the two switches in buildHLSFFmpegArgsAt could silently drift +// when adding a new backend). +type EncoderProfile struct { + Codec string // ffmpeg encoder name (e.g. "h264_nvenc", "libx264") + Preset string // preset string, or "" when the codec has no preset knob + DecodeHwAccel string // ffmpeg `-hwaccel` value (e.g. "cuda", "qsv", "vaapi"), or "" +} + +// ResolveEncoderProfile mirrors the codec + preset selection inside +// buildHLSFFmpegArgsAt so callers (registry, log lines, diagnostic +// endpoints) can know what ffmpeg will be told to do without parsing argv. +// +// The configured preset is libx264-specific by vocabulary (ultrafast… +// veryslow). Passing it through to NVENC / QSV would have ffmpeg reject +// the argv (NVENC uses p1-p7, QSV uses its own subset). So vendor encoders +// always use their hardcoded vendor preset and ignore configuredPreset. +// VideoToolbox has no preset knob at all. +// +// DecodeHwAccel mirrors the encoder family — `-hwaccel cuda` for NVENC, +// `-hwaccel qsv` for QSV, `-hwaccel vaapi` for VAAPI. We intentionally +// do NOT pass `-hwaccel_output_format vaapi`: that pins decoded frames +// to GPU memory, but our filter chain (scale/format/setparams) runs on +// CPU and can't consume VAAPI surfaces. Keeping output frames on CPU +// makes the filter chain work and the VAAPI encoder still benefits from +// HW-accelerated DECODE on the input side. +func ResolveEncoderProfile(hw HWAccel, configuredPreset string) EncoderProfile { + codec := hw.FFmpegVideoCodec("h264") + switch codec { + case "libx264": + preset := configuredPreset + if preset == "" { + preset = "superfast" + } + return EncoderProfile{Codec: codec, Preset: preset, DecodeHwAccel: ""} + case "h264_nvenc": + return EncoderProfile{Codec: codec, Preset: "p3", DecodeHwAccel: "cuda"} + case "h264_qsv": + return EncoderProfile{Codec: codec, Preset: "veryfast", DecodeHwAccel: "qsv"} + case "h264_vaapi": + return EncoderProfile{Codec: codec, Preset: "", DecodeHwAccel: "vaapi"} + case "h264_videotoolbox": + // No preset knob for VideoToolbox; the speed/quality dial is `-q:v`. + // VideoToolbox uses per-encoder flags rather than a demuxer hint. + return EncoderProfile{Codec: codec, Preset: "", DecodeHwAccel: ""} + } + // Unknown / future codecs: software path. + return EncoderProfile{Codec: codec, Preset: "", DecodeHwAccel: ""} +} + // buildHLSFFmpegArgsAt returns the argv for an HLS encode that starts at the // given segment index (`-ss `) and writes segments numbered from // startIdx so they slot into the existing manifest at the correct position. // `-output_ts_offset` keeps the segment PTS aligned with manifest timeline. func buildHLSFFmpegArgsAt(cfg HLSSessionConfig, probe *StreamProbe, tmpDir string, startIdx int, startSec float64) []string { - hwHint := cfg.Transcode.HWAccel + profile := ResolveEncoderProfile(cfg.Transcode.HWAccel, cfg.Transcode.Preset) args := []string{"-y", "-hide_banner", "-loglevel", "warning"} - switch hwHint { - case HWAccelNVENC: - args = append(args, "-hwaccel", "cuda") - case HWAccelQSV: - args = append(args, "-hwaccel", "qsv") - case HWAccelVAAPI: - args = append(args, "-hwaccel", "vaapi", "-hwaccel_output_format", "vaapi") - case HWAccelNone, HWAccelVideoToolbox: - // No demuxer-side hint. + // Demuxer-side HW-decode hint. Sourced from the profile so a future + // codec/hint mismatch is impossible — the encoder + decode hint are + // computed once and stay coherent. Notably we do NOT add + // `-hwaccel_output_format vaapi` on the VAAPI path: that pins decoded + // frames to GPU memory but our CPU filter chain (scale, format, + // setparams) can't consume VAAPI surfaces. Letting frames flow on CPU + // keeps the filter chain working; the encoder still gets HW-accelerated + // decode on the input side. + if profile.DecodeHwAccel != "" { + args = append(args, "-hwaccel", profile.DecodeHwAccel) } // Seek before -i for fast keyframe-aligned start. The new ffmpeg writes @@ -827,24 +1131,54 @@ func buildHLSFFmpegArgsAt(cfg HLSSessionConfig, probe *StreamProbe, tmpDir strin } args = append(args, "-map", fmt.Sprintf("0:a:%d?", audioIdx)) - // Video encode. - codec := hwHint.FFmpegVideoCodec("h264") + // Video encode. Codec + preset come from the EncoderProfile resolved at + // the top of this function so the demuxer hint, the encoder, and the + // per-session log line all stay consistent. + // + // Defaults are biased for FIRST-START LATENCY over quality — the player + // blocks on seg-0 before the first frame paints, and a slow seg-0 is + // what users notice ("preparando sesión" stuck). Users who want better + // quality can override via `download.transcode.preset` in config.toml. + codec := profile.Codec args = append(args, "-c:v", codec) - // Encoder-specific tuning. Each HW encoder takes a different "preset" - // vocabulary; libx264 uses ultrafast→placebo, NVENC uses p1→p7, QSV uses - // veryfast→veryslow, VAAPI/VideoToolbox don't expose presets. switch codec { case "libx264": - preset := cfg.Transcode.Preset - if preset == "" { - preset = "veryfast" - } - args = append(args, "-preset", preset) + // superfast = ~15-20% faster than veryfast at marginal quality loss + // for the bitrates we target (5-25 Mbps). For 4K software encodes + // this is the difference between ~3 s and ~2.5 s per segment on a + // recent x86 CPU. `-threads 0` is libx264's default but explicit + // helps when the user has set GOMAXPROCS. + args = append(args, "-preset", profile.Preset, "-threads", "0") case "h264_nvenc": - // p4 = balanced quality/speed; p1 fastest, p7 highest quality. - args = append(args, "-preset", "p4", "-rc", "vbr", "-tune", "hq") + // p3 + tune=ll trades ~0.3 dB PSNR for 1.5-2× faster encode vs the + // previous p4 + tune=hq pair — first-segment encode drops from + // ~1.5 s to ~0.8 s on RTX-class hardware. + args = append(args, "-preset", profile.Preset, "-rc", "vbr", "-tune", "ll") case "h264_qsv": - args = append(args, "-preset", "medium", "-look_ahead", "0") + // veryfast is the fastest realistic QSV preset; medium was too + // conservative for first-start. look_ahead=0 keeps the encoder + // truly low-latency (no rate-control look-ahead window). + args = append(args, "-preset", profile.Preset, "-look_ahead", "0") + case "h264_videotoolbox": + // VideoToolbox has no "preset" knob; `-realtime` flips into the + // low-latency path used by FaceTime. We let the `-b:v / -maxrate + // / -bufsize` block (added later in this function) drive rate + // control — adding `-q:v` here would conflict because ffmpeg's + // videotoolbox encoder treats `-b:v` as authoritative and + // silently ignores `-q:v`, so the constant-quality knob never + // took effect anyway. + args = append(args, "-realtime", "1") + case "h264_vaapi": + // h264_vaapi has no preset knob. Bitrate args (set later) drive + // rate control. Add `-vaapi_device /dev/dri/renderD128` so the + // encoder doesn't fall back to a NULL device on multi-GPU hosts + // where the default render node is a non-VAAPI GPU (an Nvidia + // dGPU's render node, etc.). The filter chain below switches to + // `format=nv12,hwupload` so frames land on the right VAAPI + // surface before the encoder; we intentionally avoid scale_vaapi + // because mesa 25 + Raphael iGPU emits "Cannot allocate memory" + // per session start, polluting logs even though encode succeeds. + args = append(args, "-vaapi_device", "/dev/dri/renderD128") } // Derive H.264 level from the actual output height. A fixed "4.0" caps the // encoder at 1080p — anything taller (1440p, 4K source on quality=original) @@ -895,14 +1229,32 @@ func buildHLSFFmpegArgsAt(cfg HLSSessionConfig, probe *StreamProbe, tmpDir strin if maxH == 0 { maxH = cfg.Transcode.MaxHeight } + // VAAPI needs frames as nv12 VAAPI surfaces before the encoder. We do + // scale + format conversion on CPU then `hwupload` once at the end — + // skips the mesa 25 + Raphael iGPU "Cannot allocate memory" log spam + // that scale_vaapi triggers per-session-start while still delivering + // the encoder a GPU surface. setparams is dropped because VAAPI + // surfaces don't expose VUI fields the way libx264 does; the encoder + // records its own color metadata via the source PTS chain. + pixFormat := "yuv420p" + hwUploadTail := "" + colorTail := ",setparams=colorspace=bt709:color_trc=bt709:color_primaries=bt709:range=tv" + if codec == "h264_vaapi" { + pixFormat = "nv12" + hwUploadTail = ",hwupload" + colorTail = "" + } var filterChain string if maxH > 0 && probe.Height > maxH { filterChain = fmt.Sprintf( - "scale=-2:%d:force_original_aspect_ratio=decrease,scale=trunc(iw/2)*2:trunc(ih/2)*2,format=yuv420p,setparams=colorspace=bt709:color_trc=bt709:color_primaries=bt709:range=tv", - maxH, + "scale=-2:%d:force_original_aspect_ratio=decrease,scale=trunc(iw/2)*2:trunc(ih/2)*2,format=%s%s%s", + maxH, pixFormat, colorTail, hwUploadTail, ) } else { - filterChain = "scale=trunc(iw/2)*2:trunc(ih/2)*2,format=yuv420p,setparams=colorspace=bt709:color_trc=bt709:color_primaries=bt709:range=tv" + filterChain = fmt.Sprintf( + "scale=trunc(iw/2)*2:trunc(ih/2)*2,format=%s%s%s", + pixFormat, colorTail, hwUploadTail, + ) } args = append(args, "-vf", filterChain) @@ -975,6 +1327,10 @@ func (s *HLSSession) extractSubtitles(ctx context.Context) { // renderVideoPlaylist builds the VOD media playlist for the video stream. // Segment count is derived from the source duration — the player learns the // total timeline from the manifest before any segment is fetched. +// +// seg-0 is the short init segment (hlsInitSegmentDuration s); seg-1 onward +// are hlsSegmentDuration s each. The last segment may be shorter than the +// nominal duration when (duration - init) doesn't divide evenly. func renderVideoPlaylist(durationSec float64, segCount int) string { var b strings.Builder b.WriteString("#EXTM3U\n") @@ -985,7 +1341,7 @@ func renderVideoPlaylist(durationSec float64, segCount int) string { b.WriteString(`#EXT-X-MAP:URI="init.mp4"` + "\n") remaining := durationSec for i := 0; i < segCount; i++ { - segDur := float64(hlsSegmentDuration) + segDur := float64(segmentDurationFor(i)) if remaining < segDur { segDur = remaining } diff --git a/internal/engine/hls_cache.go b/internal/engine/hls_cache.go new file mode 100644 index 0000000..f1bf918 --- /dev/null +++ b/internal/engine/hls_cache.go @@ -0,0 +1,410 @@ +package engine + +import ( + "context" + "crypto/sha256" + "encoding/hex" + "errors" + "fmt" + "log" + "os" + "path/filepath" + "sort" + "sync" + "sync/atomic" + "time" +) + +// HLSCache persists transcoded HLS segments per (source, quality, audio) so a +// second play of the same file at the same quality skips ffmpeg entirely. +// +// Layout on disk: +// +// {root}/{key}/init.mp4 +// {root}/{key}/seg-0.m4s +// {root}/{key}/seg-N.m4s +// {root}/{key}/.complete +// +// Atomicity: the .complete marker is written only when ffmpeg exits 0 AND all +// segments are on disk. A dir without .complete is treated as a partial run — +// next session can reuse the segments already present, ffmpeg fills the gaps. +// +// Concurrency: Pin/Unpin increments a ref counter per key so the LRU sweeper +// never evicts a directory that an active session is reading from. +type HLSCache struct { + root string + maxBytes int64 + + mu sync.Mutex + refs map[string]int + writers map[string]bool // exclusive ffmpeg writer per key; nil entries are absent + + // Counters surfaced via Stats() — useful for /api/internal/agent/cache-stats + // and for the sweeper's daily log line. atomic so RecordHit/RecordMiss are + // safe to call from any goroutine without taking the cache mutex. + hits atomic.Uint64 + misses atomic.Uint64 +} + +const ( + hlsCacheCompleteMarker = ".complete" + // hlsCacheMinBudgetGB clamps absurd / zero / negative SizeGB values to + // a sane floor. NOT a guarantee that any single encode fits — a long + // 4K HEVC re-encode can exceed it. Operators should set size_gb based + // on their actual workload. + hlsCacheMinBudgetGB = 1 + // hlsCacheStartupOrphanAge: directories without .complete older than + // this are removed on cache startup. Long enough that a daemon crash + // during an in-progress encode (which legitimately leaves a partial + // dir) doesn't get nuked too aggressively if the daemon restarts fast. + hlsCacheStartupOrphanAge = 10 * time.Minute +) + +// NewHLSCache creates the cache rooted at the given dir with a size budget in +// gigabytes. A budget < hlsCacheMinBudgetGB is clamped up so a single play +// doesn't get instantly evicted mid-stream. +func NewHLSCache(root string, sizeGB int) (*HLSCache, error) { + if root == "" { + return nil, errors.New("hls_cache: empty root") + } + if sizeGB < hlsCacheMinBudgetGB { + sizeGB = hlsCacheMinBudgetGB + } + if err := os.MkdirAll(root, 0o755); err != nil { + return nil, fmt.Errorf("hls_cache: mkdir root: %w", err) + } + c := &HLSCache{ + root: root, + maxBytes: int64(sizeGB) * 1024 * 1024 * 1024, + refs: make(map[string]int), + writers: make(map[string]bool), + } + // Reap dirs left over from a crashed encode. A dir without .complete that + // hasn't been touched recently was almost certainly orphaned by an + // ungraceful daemon exit — keeping it just feeds the unbounded growth + // pattern the hourly LRU is too slow to contain. + if removed, err := c.cleanStartupOrphans(); err != nil { + log.Printf("[hls_cache] startup orphan cleanup: %v", err) + } else if removed > 0 { + log.Printf("[hls_cache] startup: removed %d orphan dir(s) without .complete", removed) + } + return c, nil +} + +// cleanStartupOrphans removes cache subdirectories that lack a .complete +// marker AND haven't been modified within hlsCacheStartupOrphanAge. Called +// once at construction. Safe at startup because no sessions are active yet, +// so Pin can't race with us. +func (c *HLSCache) cleanStartupOrphans() (int, error) { + entries, err := os.ReadDir(c.root) + if err != nil { + if os.IsNotExist(err) { + return 0, nil + } + return 0, err + } + cutoff := time.Now().Add(-hlsCacheStartupOrphanAge) + removed := 0 + for _, e := range entries { + if !e.IsDir() { + continue + } + dir := filepath.Join(c.root, e.Name()) + if _, err := os.Stat(filepath.Join(dir, hlsCacheCompleteMarker)); err == nil { + continue // sealed, keep + } + info, err := e.Info() + if err != nil { + continue + } + if info.ModTime().After(cutoff) { + continue // too recent — might be a daemon that just restarted mid-encode + } + if err := os.RemoveAll(dir); err == nil { + removed++ + } + } + return removed, nil +} + +// TryAcquireWriter attempts to claim exclusive ffmpeg-write access to a key. +// Returns true on success — the caller is then responsible for ReleaseWriter +// when ffmpeg exits / fails. Returns false if another session is already +// writing this key, in which case the caller must fall back to a private +// per-session tmpdir (no caching for that session). +func (c *HLSCache) TryAcquireWriter(key string) bool { + c.mu.Lock() + defer c.mu.Unlock() + if c.writers[key] { + return false + } + c.writers[key] = true + return true +} + +// ReleaseWriter releases the writer claim acquired via TryAcquireWriter. +// Idempotent on unknown keys. +func (c *HLSCache) ReleaseWriter(key string) { + c.mu.Lock() + delete(c.writers, key) + c.mu.Unlock() +} + +// KeyFor derives a stable cache key for (source, quality, audioIndex). Using +// the absolute source path means renaming a file invalidates the cache, which +// is correct — segment content is tied to the encoded source. +func (c *HLSCache) KeyFor(sourcePath, quality string, audioIndex int) string { + abs, err := filepath.Abs(sourcePath) + if err != nil { + abs = sourcePath + } + h := sha256.Sum256([]byte(fmt.Sprintf("%s|%s|%d", abs, quality, audioIndex))) + return hex.EncodeToString(h[:8]) // 16 hex chars — collision-safe enough for per-host cache +} + +// DirFor returns the on-disk directory for a cache key. Caller is responsible +// for creating it. +func (c *HLSCache) DirFor(key string) string { + return filepath.Join(c.root, key) +} + +// HasComplete returns true when the .complete marker is present, meaning the +// directory holds a full set of segments from a successful encode. +func (c *HLSCache) HasComplete(key string) bool { + if _, err := os.Stat(filepath.Join(c.DirFor(key), hlsCacheCompleteMarker)); err == nil { + return true + } + return false +} + +// MarkComplete writes the .complete marker. Call only after verifying ffmpeg +// exited cleanly AND every expected segment is on disk. The dir must already +// exist — StartHLSSession created it on the writer path. +func (c *HLSCache) MarkComplete(key string) error { + return os.WriteFile(filepath.Join(c.DirFor(key), hlsCacheCompleteMarker), nil, 0o644) +} + +// RecordHit increments the hit counter; called by StartHLSSession on a +// cache-HIT path. +func (c *HLSCache) RecordHit() { c.hits.Add(1) } + +// RecordMiss increments the miss counter; called when a session has to +// encode from scratch (or fails an integrity check on a stale HIT). +func (c *HLSCache) RecordMiss() { c.misses.Add(1) } + +// CacheStats is a snapshot of the cache's runtime counters + on-disk size. +// The size fields are best-effort (computed via dirSize) so callers paying +// for them should cache the result, not poll in a hot loop. +type CacheStats struct { + Hits uint64 + Misses uint64 + EntryCount int + TotalBytes int64 +} + +// Stats returns a snapshot of the cache counters and size. Walks the root +// to total disk usage — O(N segments). Call at most every few minutes. +func (c *HLSCache) Stats() CacheStats { + s := CacheStats{ + Hits: c.hits.Load(), + Misses: c.misses.Load(), + } + entries, err := os.ReadDir(c.root) + if err != nil { + return s + } + for _, e := range entries { + if !e.IsDir() { + continue + } + size, err := dirSize(filepath.Join(c.root, e.Name())) + if err != nil { + continue + } + s.EntryCount++ + s.TotalBytes += size + } + return s +} + +// hitRatePercent returns the current hit/(hit+miss) percentage rounded to +// the nearest int; 0 when no calls have been recorded. +func (c *HLSCache) hitRatePercent() int { + h := c.hits.Load() + m := c.misses.Load() + total := h + m + if total == 0 { + return 0 + } + return int((h*100 + total/2) / total) +} + +// VerifyComplete checks that the .complete marker is present AND the +// essential files (init.mp4 + last segment) exist with non-zero size. A +// dir that passes HasComplete but fails VerifyComplete is treated as +// corrupted — typically external `rm` or a partial-disk-failure scenario. +// When it returns false, callers should Invalidate and re-encode. +func (c *HLSCache) VerifyComplete(key string, segmentCount int) bool { + if !c.HasComplete(key) { + return false + } + dir := c.DirFor(key) + if fi, err := os.Stat(filepath.Join(dir, "video", "init.mp4")); err != nil || fi.Size() == 0 { + return false + } + if segmentCount > 0 { + lastSeg := filepath.Join(dir, "video", fmt.Sprintf("seg-%d.m4s", segmentCount-1)) + if fi, err := os.Stat(lastSeg); err != nil || fi.Size() == 0 { + return false + } + } + return true +} + +// Pin increments the ref counter for a key. The sweeper checks this before +// evicting, so a pinned dir is safe even if its mtime is old. +func (c *HLSCache) Pin(key string) { + c.mu.Lock() + c.refs[key]++ + c.mu.Unlock() +} + +// Unpin decrements; safe to call on unknown keys (no-op). +func (c *HLSCache) Unpin(key string) { + c.mu.Lock() + if c.refs[key] > 0 { + c.refs[key]-- + if c.refs[key] == 0 { + delete(c.refs, key) + } + } + c.mu.Unlock() +} + +func (c *HLSCache) isPinned(key string) bool { + c.mu.Lock() + defer c.mu.Unlock() + return c.refs[key] > 0 +} + +// Touch updates the directory mtime so LRU picks fresher entries as recently +// used. Called when a session starts reading from a cached dir. +func (c *HLSCache) Touch(key string) error { + dir := c.DirFor(key) + now := time.Now() + return os.Chtimes(dir, now, now) +} + +// Sweep enforces the size budget by deleting the least-recently-used cache +// dirs (ignoring pinned ones) until the total size is at or below maxBytes. +// Returns the number of bytes freed. +func (c *HLSCache) Sweep() (int64, error) { + entries, err := os.ReadDir(c.root) + if err != nil { + if os.IsNotExist(err) { + return 0, nil + } + return 0, fmt.Errorf("hls_cache: read root: %w", err) + } + + type item struct { + key string + path string + size int64 + mtime time.Time + } + items := make([]item, 0, len(entries)) + var total, pinned int64 + for _, e := range entries { + if !e.IsDir() { + continue + } + info, err := e.Info() + if err != nil { + continue + } + key := e.Name() + path := filepath.Join(c.root, key) + size, err := dirSize(path) + if err != nil { + continue + } + items = append(items, item{key: key, path: path, size: size, mtime: info.ModTime()}) + total += size + if c.isPinned(key) { + pinned += size + } + } + + if total <= c.maxBytes { + return 0, nil + } + if pinned >= c.maxBytes { + // Every pinned byte already exceeds the budget — even evicting + // every unpinned dir won't bring us under. Warn loudly so the + // operator knows to bump size_gb (or kill the long-running session). + log.Printf("[hls_cache] warn: pinned bytes (%.1f MB) exceed budget (%.1f MB) — cannot enforce limit until sessions release", + float64(pinned)/(1024*1024), float64(c.maxBytes)/(1024*1024)) + return 0, nil + } + + // Oldest first. + sort.Slice(items, func(i, j int) bool { + return items[i].mtime.Before(items[j].mtime) + }) + + var freed int64 + for _, it := range items { + if total-freed <= c.maxBytes { + break + } + if c.isPinned(it.key) { + continue + } + if err := os.RemoveAll(it.path); err != nil { + log.Printf("[hls_cache] evict %s failed: %v", it.key, err) + continue + } + log.Printf("[hls_cache] evicted %s (%.1f MB, age %s)", + it.key, float64(it.size)/(1024*1024), time.Since(it.mtime).Round(time.Second)) + freed += it.size + } + return freed, nil +} + +// StartSweeper kicks off the LRU sweeper goroutine. Cancels on ctx done. +// In addition to enforcing the size budget, logs a daily summary of hit-rate +// + disk usage so operators can see the cache's value at a glance. +func (c *HLSCache) StartSweeper(ctx context.Context, interval time.Duration) { + if interval <= 0 { + interval = time.Hour + } + go func() { + t := time.NewTicker(interval) + defer t.Stop() + statsTick := time.NewTicker(24 * time.Hour) + defer statsTick.Stop() + for { + select { + case <-ctx.Done(): + return + case <-t.C: + if _, err := c.Sweep(); err != nil { + log.Printf("[hls_cache] sweep error: %v", err) + } + case <-statsTick.C: + s := c.Stats() + log.Printf("[hls_cache] day-stats: hits=%d misses=%d ratio=%d%% entries=%d size=%.1fMB", + s.Hits, s.Misses, c.hitRatePercent(), s.EntryCount, + float64(s.TotalBytes)/(1024*1024)) + } + } + }() +} + +// Invalidate removes a cache entry — used when ffmpeg fails to encode the +// source so we don't reuse a half-written dir next time. +func (c *HLSCache) Invalidate(key string) error { + return os.RemoveAll(c.DirFor(key)) +} + diff --git a/internal/engine/hls_cache_smoke_test.go b/internal/engine/hls_cache_smoke_test.go new file mode 100644 index 0000000..a086cb2 --- /dev/null +++ b/internal/engine/hls_cache_smoke_test.go @@ -0,0 +1,134 @@ +//go:build smoke + +package engine + +import ( + "context" + "os/exec" + "path/filepath" + "testing" + "time" +) + +// TestHLSCacheSmoke exercises the end-to-end cache flow against real ffmpeg: +// - First session encodes a 5s test pattern; expect MISS, ffmpeg runs, +// .complete written, MarkComplete logs. +// - Second session for identical (source, quality, audio); expect HIT, +// no ffmpeg, instant Start. +// +// Build tag `smoke` keeps it out of the default `go test ./...` run because +// it depends on a working ffmpeg/ffprobe and takes ~5–10 s. +// +// go test -tags=smoke -run TestHLSCacheSmoke -v ./internal/engine/ +func TestHLSCacheSmoke(t *testing.T) { + ffmpeg, err := exec.LookPath("ffmpeg") + if err != nil { + t.Skipf("ffmpeg not on PATH: %v", err) + } + ffprobe, err := exec.LookPath("ffprobe") + if err != nil { + t.Skipf("ffprobe not on PATH: %v", err) + } + + tmp := t.TempDir() + source := filepath.Join(tmp, "source.mp4") + t.Logf("generating 5 s test pattern → %s", source) + if out, err := exec.Command(ffmpeg, + "-y", "-loglevel", "error", + "-f", "lavfi", "-i", "testsrc=duration=5:size=640x480:rate=30", + "-f", "lavfi", "-i", "sine=frequency=1000:duration=5", + "-c:v", "libx264", "-preset", "ultrafast", "-pix_fmt", "yuv420p", + "-c:a", "aac", + source, + ).CombinedOutput(); err != nil { + t.Fatalf("ffmpeg generate: %v\n%s", err, out) + } + + cacheRoot := filepath.Join(tmp, "cache") + cache, err := NewHLSCache(cacheRoot, 1) + if err != nil { + t.Fatalf("NewHLSCache: %v", err) + } + + cfg := HLSSessionConfig{ + SessionID: "smoke1", + SourcePath: source, + FileName: "source.mp4", + Quality: "720p", + AudioIndex: 0, + Transcode: TranscodeRuntime{ + FFmpegPath: ffmpeg, + FFprobePath: ffprobe, + Preset: "ultrafast", + }, + Cache: cache, + } + + // First run — expect MISS, ffmpeg runs. + t.Log("session 1: expect MISS") + t0 := time.Now() + s1, err := StartHLSSession(context.Background(), cfg) + if err != nil { + t.Fatalf("StartHLSSession #1: %v", err) + } + if s1.fromCache { + t.Fatal("session 1 reported cache HIT on a fresh cache") + } + + // Wait for all segments to land. 5 s source @ 4 s segments → 2 segments. + deadline := time.Now().Add(60 * time.Second) + for { + s1.readyMu.Lock() + ready := s1.readyMax + exited := s1.exited + s1.readyMu.Unlock() + if ready >= s1.segmentCount-1 && exited { + break + } + if time.Now().After(deadline) { + _ = s1.Close() + t.Fatalf("session 1 didn't finish in 60 s (readyMax=%d/%d, exited=%v)", + ready, s1.segmentCount-1, exited) + } + time.Sleep(100 * time.Millisecond) + } + if err := s1.Close(); err != nil { + t.Fatalf("Close #1: %v", err) + } + encodeDur := time.Since(t0) + t.Logf("session 1: MISS completed in %s", encodeDur.Round(time.Millisecond)) + + key := cache.KeyFor(source, "720p", 0) + if !cache.HasComplete(key) { + t.Fatalf("cache.HasComplete(%s) is false after successful encode", key) + } + + // Second run — expect HIT, no ffmpeg. + t.Log("session 2: expect HIT") + cfg.SessionID = "smoke2" + t1 := time.Now() + s2, err := StartHLSSession(context.Background(), cfg) + if err != nil { + t.Fatalf("StartHLSSession #2: %v", err) + } + if !s2.fromCache { + t.Fatal("session 2 should have reported cache HIT") + } + if s2.cmd != nil { + t.Fatal("session 2 should not have spawned ffmpeg (s.cmd != nil)") + } + hitDur := time.Since(t1) + t.Logf("session 2: HIT in %s (%.1f× faster than MISS)", + hitDur.Round(time.Millisecond), float64(encodeDur)/float64(hitDur)) + if hitDur > 500*time.Millisecond { + t.Errorf("HIT path too slow: %s — expected <500 ms", hitDur) + } + if err := s2.Close(); err != nil { + t.Fatalf("Close #2: %v", err) + } + + // After the HIT session closes, the cache dir + .complete must still exist. + if !cache.HasComplete(key) { + t.Fatal(".complete disappeared after HIT session closed") + } +} diff --git a/internal/engine/hls_cache_test.go b/internal/engine/hls_cache_test.go new file mode 100644 index 0000000..cb70ec1 --- /dev/null +++ b/internal/engine/hls_cache_test.go @@ -0,0 +1,361 @@ +package engine + +import ( + "context" + "os" + "path/filepath" + "sync" + "testing" + "time" +) + +func newTestCache(t *testing.T, sizeGB int) *HLSCache { + t.Helper() + root := t.TempDir() + c, err := NewHLSCache(root, sizeGB) + if err != nil { + t.Fatalf("NewHLSCache: %v", err) + } + return c +} + +func TestKeyForStable(t *testing.T) { + c := newTestCache(t, 1) + k1 := c.KeyFor("/a/b/movie.mkv", "1080p", 0) + k2 := c.KeyFor("/a/b/movie.mkv", "1080p", 0) + if k1 != k2 { + t.Fatalf("expected stable keys, got %q vs %q", k1, k2) + } + if c.KeyFor("/a/b/movie.mkv", "720p", 0) == k1 { + t.Fatal("quality should change key") + } + if c.KeyFor("/a/b/movie.mkv", "1080p", 1) == k1 { + t.Fatal("audio index should change key") + } + if c.KeyFor("/x/y/other.mkv", "1080p", 0) == k1 { + t.Fatal("path should change key") + } +} + +func TestMarkCompleteAndHas(t *testing.T) { + c := newTestCache(t, 1) + key := "abc123" + if c.HasComplete(key) { + t.Fatal("fresh cache should not report complete") + } + // Production callers create the dir during StartHLSSession; MarkComplete + // trusts that invariant and fails if the dir was wiped meanwhile. + if err := os.MkdirAll(c.DirFor(key), 0o755); err != nil { + t.Fatalf("mkdir: %v", err) + } + if err := c.MarkComplete(key); err != nil { + t.Fatalf("MarkComplete: %v", err) + } + if !c.HasComplete(key) { + t.Fatal("after MarkComplete, HasComplete must be true") + } +} + +func TestMarkCompleteFailsWithoutDir(t *testing.T) { + c := newTestCache(t, 1) + if err := c.MarkComplete("never-created"); err == nil { + t.Fatal("MarkComplete should error when dir doesn't exist") + } +} + +func TestPinPreventsEviction(t *testing.T) { + c := newTestCache(t, 1) // 1 GB budget, but min clamp keeps it usable + c.maxBytes = 1024 // squeeze budget for the test + + // Write two entries past the budget. + for i, key := range []string{"old", "new"} { + dir := c.DirFor(key) + if err := os.MkdirAll(dir, 0o755); err != nil { + t.Fatalf("mkdir %s: %v", dir, err) + } + path := filepath.Join(dir, "seg.bin") + if err := os.WriteFile(path, make([]byte, 800), 0o644); err != nil { + t.Fatalf("write %s: %v", path, err) + } + now := time.Now().Add(time.Duration(i) * time.Hour) // "old" mtime < "new" + _ = os.Chtimes(dir, now, now) + } + + c.Pin("old") // protect the older one + freed, err := c.Sweep() + if err != nil { + t.Fatalf("Sweep: %v", err) + } + if freed == 0 { + t.Fatal("expected some eviction") + } + if _, err := os.Stat(c.DirFor("old")); err != nil { + t.Fatal("pinned 'old' was evicted") + } + if _, err := os.Stat(c.DirFor("new")); err == nil { + t.Fatal("'new' should have been evicted to make room") + } +} + +func TestSweepNoOpUnderBudget(t *testing.T) { + c := newTestCache(t, 1) + dir := c.DirFor("small") + _ = os.MkdirAll(dir, 0o755) + _ = os.WriteFile(filepath.Join(dir, "x"), []byte("tiny"), 0o644) + freed, err := c.Sweep() + if err != nil { + t.Fatalf("Sweep: %v", err) + } + if freed != 0 { + t.Fatalf("expected 0 freed under budget, got %d", freed) + } + if _, err := os.Stat(dir); err != nil { + t.Fatal("under-budget entry was wrongly evicted") + } +} + +func TestSweepEmptyRoot(t *testing.T) { + c := newTestCache(t, 1) + freed, err := c.Sweep() + if err != nil { + t.Fatalf("Sweep empty: %v", err) + } + if freed != 0 { + t.Fatalf("freed=%d, want 0", freed) + } +} + +func TestInvalidateRemovesDir(t *testing.T) { + c := newTestCache(t, 1) + key := "drop" + dir := c.DirFor(key) + _ = os.MkdirAll(dir, 0o755) + _ = os.WriteFile(filepath.Join(dir, "x"), []byte("y"), 0o644) + if err := c.Invalidate(key); err != nil { + t.Fatalf("Invalidate: %v", err) + } + if _, err := os.Stat(dir); err == nil { + t.Fatal("dir still present after Invalidate") + } +} + +func TestTouchUpdatesMtime(t *testing.T) { + c := newTestCache(t, 1) + key := "touch" + dir := c.DirFor(key) + _ = os.MkdirAll(dir, 0o755) + old := time.Now().Add(-2 * time.Hour) + _ = os.Chtimes(dir, old, old) + + if err := c.Touch(key); err != nil { + t.Fatalf("Touch: %v", err) + } + info, err := os.Stat(dir) + if err != nil { + t.Fatalf("stat: %v", err) + } + if !info.ModTime().After(old.Add(time.Minute)) { + t.Fatalf("mtime not refreshed: %v", info.ModTime()) + } +} + +func TestPinUnpinSymmetry(t *testing.T) { + c := newTestCache(t, 1) + c.Pin("k") + c.Pin("k") + if !c.isPinned("k") { + t.Fatal("Pin twice should leave pinned") + } + c.Unpin("k") + if !c.isPinned("k") { + t.Fatal("Unpin once should keep pinned (refs=1)") + } + c.Unpin("k") + if c.isPinned("k") { + t.Fatal("Unpin twice should drop pin") + } + c.Unpin("k") // safe no-op +} + +func TestConcurrentPinUnpin(t *testing.T) { + c := newTestCache(t, 1) + var wg sync.WaitGroup + for i := 0; i < 100; i++ { + wg.Add(1) + go func() { + defer wg.Done() + c.Pin("race") + time.Sleep(time.Microsecond) + c.Unpin("race") + }() + } + wg.Wait() + if c.isPinned("race") { + t.Fatal("refs leaked") + } +} + +func TestSweeperLoopExits(t *testing.T) { + c := newTestCache(t, 1) + ctx, cancel := context.WithCancel(context.Background()) + c.StartSweeper(ctx, 10*time.Millisecond) + time.Sleep(30 * time.Millisecond) + cancel() + // If StartSweeper doesn't exit on cancel the test would leak a goroutine; + // the leak detector in the test runner will surface it. + time.Sleep(20 * time.Millisecond) +} + +func TestMinBudgetClamp(t *testing.T) { + root := t.TempDir() + c, err := NewHLSCache(root, 0) // below floor + if err != nil { + t.Fatalf("NewHLSCache: %v", err) + } + if c.maxBytes != int64(hlsCacheMinBudgetGB)*1024*1024*1024 { + t.Fatalf("budget not clamped to min: got %d", c.maxBytes) + } +} + +func TestTryAcquireWriterExclusive(t *testing.T) { + c := newTestCache(t, 1) + if !c.TryAcquireWriter("k") { + t.Fatal("first acquire should succeed") + } + if c.TryAcquireWriter("k") { + t.Fatal("second acquire for same key must fail") + } + if !c.TryAcquireWriter("other") { + t.Fatal("different key should not conflict") + } + c.ReleaseWriter("k") + if !c.TryAcquireWriter("k") { + t.Fatal("acquire after release should succeed") + } + c.ReleaseWriter("k") + c.ReleaseWriter("k") // idempotent +} + +func TestStartupOrphanCleanup(t *testing.T) { + root := t.TempDir() + + // Pre-seed: one sealed dir + one orphan old enough + one orphan fresh. + sealed := filepath.Join(root, "sealed") + _ = os.MkdirAll(sealed, 0o755) + _ = os.WriteFile(filepath.Join(sealed, hlsCacheCompleteMarker), nil, 0o644) + + staleOrphan := filepath.Join(root, "stale_orphan") + _ = os.MkdirAll(staleOrphan, 0o755) + old := time.Now().Add(-2 * hlsCacheStartupOrphanAge) + _ = os.Chtimes(staleOrphan, old, old) + + freshOrphan := filepath.Join(root, "fresh_orphan") + _ = os.MkdirAll(freshOrphan, 0o755) + + if _, err := NewHLSCache(root, 1); err != nil { + t.Fatalf("NewHLSCache: %v", err) + } + + if _, err := os.Stat(sealed); err != nil { + t.Fatal("sealed dir was wrongly removed") + } + if _, err := os.Stat(staleOrphan); err == nil { + t.Fatal("stale orphan should have been removed at startup") + } + if _, err := os.Stat(freshOrphan); err != nil { + t.Fatal("fresh orphan should be kept (might be a mid-restart encode)") + } +} + +func TestHitMissCounters(t *testing.T) { + c := newTestCache(t, 1) + if s := c.Stats(); s.Hits != 0 || s.Misses != 0 { + t.Fatalf("fresh cache stats not zero: %+v", s) + } + c.RecordHit() + c.RecordHit() + c.RecordMiss() + s := c.Stats() + if s.Hits != 2 || s.Misses != 1 { + t.Fatalf("counters wrong: %+v", s) + } + // 2/3 = 67% + if got := c.hitRatePercent(); got != 67 { + t.Fatalf("hitRatePercent=%d, want 67", got) + } +} + +func TestStatsEntryCount(t *testing.T) { + c := newTestCache(t, 1) + for _, k := range []string{"a", "b", "c"} { + dir := c.DirFor(k) + _ = os.MkdirAll(dir, 0o755) + _ = os.WriteFile(filepath.Join(dir, "x"), []byte("hello"), 0o644) + } + s := c.Stats() + if s.EntryCount != 3 { + t.Fatalf("EntryCount=%d, want 3", s.EntryCount) + } + if s.TotalBytes != 15 { + t.Fatalf("TotalBytes=%d, want 15", s.TotalBytes) + } +} + +func TestVerifyCompleteRejectsMissingFiles(t *testing.T) { + c := newTestCache(t, 1) + key := "v" + dir := c.DirFor(key) + _ = os.MkdirAll(filepath.Join(dir, "video"), 0o755) + + // No .complete yet → reject. + if c.VerifyComplete(key, 2) { + t.Fatal("VerifyComplete should reject without .complete") + } + + // Mark complete but no files → reject. + if err := c.MarkComplete(key); err != nil { + t.Fatalf("MarkComplete: %v", err) + } + if c.VerifyComplete(key, 2) { + t.Fatal("VerifyComplete should reject when init.mp4 missing") + } + + // Write init.mp4, last seg missing → reject. + _ = os.WriteFile(filepath.Join(dir, "video", "init.mp4"), []byte("..."), 0o644) + if c.VerifyComplete(key, 2) { + t.Fatal("VerifyComplete should reject when last segment missing") + } + + // Write last seg → pass. + _ = os.WriteFile(filepath.Join(dir, "video", "seg-1.m4s"), []byte("..."), 0o644) + if !c.VerifyComplete(key, 2) { + t.Fatal("VerifyComplete should pass with all files present") + } + + // Zero-size last seg → reject. + _ = os.WriteFile(filepath.Join(dir, "video", "seg-1.m4s"), nil, 0o644) + if c.VerifyComplete(key, 2) { + t.Fatal("VerifyComplete should reject zero-size last segment") + } +} + +func TestSweepRespectsPinnedExceedsBudget(t *testing.T) { + c := newTestCache(t, 1) + c.maxBytes = 256 // squeeze + + pinned := c.DirFor("pinned") + _ = os.MkdirAll(pinned, 0o755) + _ = os.WriteFile(filepath.Join(pinned, "x"), make([]byte, 1024), 0o644) + c.Pin("pinned") + + freed, err := c.Sweep() + if err != nil { + t.Fatalf("Sweep: %v", err) + } + if freed != 0 { + t.Fatalf("nothing should have been freed: got %d", freed) + } + if _, err := os.Stat(pinned); err != nil { + t.Fatal("pinned dir wrongly removed despite over-budget pin") + } +} diff --git a/internal/engine/hls_test.go b/internal/engine/hls_test.go index 7c7cfa4..7d4cde3 100644 --- a/internal/engine/hls_test.go +++ b/internal/engine/hls_test.go @@ -115,10 +115,11 @@ func TestRenderVideoPlaylist(t *testing.T) { } func TestRenderVideoPlaylistShortFinalSegment(t *testing.T) { - // 9.5s total, 4s segments → 3 segs of 4/4/1.5 - out := renderVideoPlaylist(9.5, 3) + // 9.5s total, 2s segments → 5 segs of 2/2/2/2/1.5 + segCount := segmentCountForDuration(9.5) + out := renderVideoPlaylist(9.5, segCount) if !strings.Contains(out, "#EXTINF:1.500,") { - t.Errorf("expected final segment 1.5s in playlist, got:\n%s", out) + t.Errorf("expected final segment 1.5s in playlist (segCount=%d), got:\n%s", segCount, out) } } diff --git a/internal/engine/hwaccel.go b/internal/engine/hwaccel.go index 7108379..d7d1bd4 100644 --- a/internal/engine/hwaccel.go +++ b/internal/engine/hwaccel.go @@ -86,6 +86,117 @@ func listFFmpegEncoders(ctx context.Context, ffmpegPath string) string { return string(out) } +// HWAccelDiagnostic bundles what we know about the host's ffmpeg + HW encode +// capabilities so the daemon can log a single coherent line at startup and the +// web side can surface "this agent is software-only" without re-running probes. +type HWAccelDiagnostic struct { + Pick HWAccel // backend selected by DetectHWAccel + FFmpegPath string // resolved ffmpeg binary + FFmpegVersion string // first line of `ffmpeg -version` (e.g. "ffmpeg version 6.1.1") + Encoders []string // HW + libsvtav1/libvpx9-class encoders found in -encoders output + Devices []string // device files / drivers detected at probe time +} + +// DetectHWAccelDiagnostic returns the full diagnostic picture for the host's +// transcode pipeline. Unlike DetectHWAccel, this is NOT cached — callers pay +// for an ffmpeg subprocess on each call (one `-encoders`, one `-version`). +// Daemon startup is the natural caller; per-session lookups should keep using +// DetectHWAccel (cached) and only re-probe diagnostics if the user runs an +// explicit doctor command. +func DetectHWAccelDiagnostic(ctx context.Context, ffmpegPath string) HWAccelDiagnostic { + d := HWAccelDiagnostic{Pick: HWAccelNone, FFmpegPath: ffmpegPath} + if ffmpegPath == "" { + return d + } + d.FFmpegVersion = ffmpegVersionLine(ctx, ffmpegPath) + encoders := listFFmpegEncoders(ctx, ffmpegPath) + for _, name := range hwEncoderNames { + if strings.Contains(encoders, name) { + d.Encoders = append(d.Encoders, name) + } + } + // Device-file checks mirror the picks below so the log line tells the + // reader why a present encoder might still have been rejected (e.g. NVENC + // compiled in but /dev/nvidia0 missing inside a container). + if fileExists("/dev/nvidia0") { + d.Devices = append(d.Devices, "/dev/nvidia0") + } + if fileExists("/dev/dri/renderD128") { + d.Devices = append(d.Devices, "/dev/dri/renderD128") + } + if hasNvidiaDriver() { + d.Devices = append(d.Devices, "nvidia-smi") + } + d.Pick = DetectHWAccel(ctx, ffmpegPath) + return d +} + +// LogLine returns a one-line human-readable summary of the diagnostic, +// suitable for daemon startup output. Format: +// +// "[transcode] ffmpeg 6.1.1 at /usr/bin/ffmpeg, HW=nvenc (h264_nvenc), devices=/dev/nvidia0,nvidia-smi" +// "[transcode] ffmpeg 6.1.1 at /home/linuxbrew/.../ffmpeg, HW=none (software libx264) — no HW encoders compiled in" +func (d HWAccelDiagnostic) LogLine() string { + var b strings.Builder + b.WriteString("[transcode] ") + if d.FFmpegVersion != "" { + b.WriteString(d.FFmpegVersion) + } else { + b.WriteString("ffmpeg") + } + if d.FFmpegPath != "" { + b.WriteString(" at ") + b.WriteString(d.FFmpegPath) + } + b.WriteString(", HW=") + b.WriteString(string(d.Pick)) + if d.Pick == HWAccelNone { + if len(d.Encoders) == 0 { + b.WriteString(" (software libx264) — no HW encoders compiled in") + } else { + b.WriteString(" (software libx264) — encoders found but no matching device: ") + b.WriteString(strings.Join(d.Encoders, ",")) + } + } else { + b.WriteString(" (") + b.WriteString(d.Pick.FFmpegVideoCodec("h264")) + b.WriteString(")") + if len(d.Devices) > 0 { + b.WriteString(", devices=") + b.WriteString(strings.Join(d.Devices, ",")) + } + } + return b.String() +} + +// hwEncoderNames lists the HW-accelerated encoders we care about for the +// startup log. Kept in lookup order so the output reads predictably across +// hosts. +var hwEncoderNames = []string{ + "h264_nvenc", "hevc_nvenc", + "h264_qsv", "hevc_qsv", + "h264_vaapi", "hevc_vaapi", + "h264_videotoolbox", "hevc_videotoolbox", +} + +// ffmpegVersionLine extracts the "ffmpeg version X.Y.Z" prefix from +// `ffmpeg -version`. Bounded to avoid hanging the daemon on a misbehaving +// binary. +func ffmpegVersionLine(ctx context.Context, ffmpegPath string) string { + cmd := exec.CommandContext(ctx, ffmpegPath, "-hide_banner", "-version") + out, err := cmd.CombinedOutput() + if err != nil || len(out) == 0 { + return "" + } + line, _, _ := strings.Cut(string(out), "\n") + // "ffmpeg version 6.1.1-some-build-suffix Copyright..." → keep up to first + // space after "version 6.x" to avoid spamming build flags into the log. + if idx := strings.Index(line, "Copyright"); idx > 0 { + line = strings.TrimSpace(line[:idx]) + } + return strings.TrimSpace(line) +} + func fileExists(path string) bool { _, err := os.Stat(path) return err == nil diff --git a/internal/engine/hwaccel_test.go b/internal/engine/hwaccel_test.go index f022d29..cf3bec2 100644 --- a/internal/engine/hwaccel_test.go +++ b/internal/engine/hwaccel_test.go @@ -1,6 +1,9 @@ package engine -import "testing" +import ( + "strings" + "testing" +) func TestHWAccelFFmpegVideoCodec(t *testing.T) { cases := []struct { @@ -32,3 +35,122 @@ func TestDetectHWAccelEmptyPathReturnsNone(t *testing.T) { t.Errorf("got %s, want %s", got, HWAccelNone) } } + +func TestResolveEncoderProfileDefaults(t *testing.T) { + cases := []struct { + hw HWAccel + configured string + wantCodec string + wantPreset string + wantHint string + }{ + // Empty configured preset → pick latency-biased default per backend. + // DecodeHwAccel matches the encoder family for HW encoders; libx264 + + // VideoToolbox have no demuxer hint. + {HWAccelNone, "", "libx264", "superfast", ""}, + {HWAccelNVENC, "", "h264_nvenc", "p3", "cuda"}, + {HWAccelQSV, "", "h264_qsv", "veryfast", "qsv"}, + // VAAPI: decoder hint set, no preset, no `-hwaccel_output_format vaapi` + // (so the CPU filter chain can consume the decoded frames). + {HWAccelVAAPI, "", "h264_vaapi", "", "vaapi"}, + // VideoToolbox has no preset knob — Preset should be "" regardless of input. + // VideoToolbox uses per-encoder flags, not a demuxer `-hwaccel` hint. + {HWAccelVideoToolbox, "p4", "h264_videotoolbox", "", ""}, + {HWAccelVideoToolbox, "", "h264_videotoolbox", "", ""}, + } + for _, tc := range cases { + got := ResolveEncoderProfile(tc.hw, tc.configured) + if got.Codec != tc.wantCodec || got.Preset != tc.wantPreset || got.DecodeHwAccel != tc.wantHint { + t.Errorf("ResolveEncoderProfile(%s, %q) = {codec=%s preset=%s hint=%s}, want {codec=%s preset=%s hint=%s}", + tc.hw, tc.configured, + got.Codec, got.Preset, got.DecodeHwAccel, + tc.wantCodec, tc.wantPreset, tc.wantHint) + } + } +} + +func TestResolveEncoderProfileHonoursConfiguredPreset(t *testing.T) { + // Only libx264 honours the configured preset — the libx264 vocabulary + // (ultrafast…veryslow) doesn't apply to vendor encoders. NVENC has its + // own p1-p7 scale; QSV uses a different subset; VideoToolbox has no + // preset knob. Passing a libx264 preset to them would have ffmpeg reject + // the argv, so ResolveEncoderProfile always falls back to the hardcoded + // vendor preset for non-libx264 codecs. + cases := []struct { + hw HWAccel + configured string + wantPreset string + }{ + {HWAccelNone, "ultrafast", "ultrafast"}, // libx264 honours + {HWAccelNone, "medium", "medium"}, // libx264 honours + {HWAccelNVENC, "p1", "p3"}, // NVENC ignores, sticks to p3 + {HWAccelNVENC, "veryfast", "p3"}, // NVENC ignores libx264 vocab + {HWAccelQSV, "veryslow", "veryfast"}, // QSV ignores, sticks to veryfast + {HWAccelVideoToolbox, "veryfast", ""}, // VideoToolbox has no preset + } + for _, tc := range cases { + got := ResolveEncoderProfile(tc.hw, tc.configured) + if got.Preset != tc.wantPreset { + t.Errorf("ResolveEncoderProfile(%s, %q).Preset = %q, want %q", + tc.hw, tc.configured, got.Preset, tc.wantPreset) + } + } +} + +func TestHWAccelDiagnosticLogLineNone(t *testing.T) { + d := HWAccelDiagnostic{ + Pick: HWAccelNone, + FFmpegPath: "/usr/local/bin/ffmpeg", + FFmpegVersion: "ffmpeg version 6.1.1", + Encoders: nil, + Devices: nil, + } + line := d.LogLine() + wantSubstrings := []string{ + "ffmpeg version 6.1.1", + "/usr/local/bin/ffmpeg", + "HW=none", + "software libx264", + "no HW encoders compiled in", + } + for _, want := range wantSubstrings { + if !strings.Contains(line, want) { + t.Errorf("expected substring %q in log line; got %q", want, line) + } + } +} + +func TestHWAccelDiagnosticLogLineNVENCWithDevices(t *testing.T) { + d := HWAccelDiagnostic{ + Pick: HWAccelNVENC, + FFmpegPath: "/usr/bin/ffmpeg", + FFmpegVersion: "ffmpeg version 6.0", + Encoders: []string{"h264_nvenc", "hevc_nvenc", "h264_qsv"}, + Devices: []string{"/dev/nvidia0", "nvidia-smi"}, + } + line := d.LogLine() + for _, want := range []string{"HW=nvenc", "h264_nvenc", "/dev/nvidia0", "nvidia-smi"} { + if !strings.Contains(line, want) { + t.Errorf("expected substring %q in log line; got %q", want, line) + } + } +} + +func TestHWAccelDiagnosticLogLineSoftwareButEncodersFound(t *testing.T) { + // Edge case: ffmpeg compiled WITH nvenc but no /dev/nvidia0 (container w/o GPU). + // LogLine should flag the encoders so the user knows where the gap is. + d := HWAccelDiagnostic{ + Pick: HWAccelNone, + FFmpegPath: "/usr/bin/ffmpeg", + FFmpegVersion: "ffmpeg version 6.0", + Encoders: []string{"h264_nvenc"}, + Devices: nil, + } + line := d.LogLine() + for _, want := range []string{"HW=none", "encoders found but no matching device", "h264_nvenc"} { + if !strings.Contains(line, want) { + t.Errorf("expected substring %q in log line; got %q", want, line) + } + } +} + diff --git a/internal/engine/probe.go b/internal/engine/probe.go index 930b669..c29c81a 100644 --- a/internal/engine/probe.go +++ b/internal/engine/probe.go @@ -88,7 +88,15 @@ const ( ) // ProbeFile runs ffprobe and returns a StreamProbe view of the file. +// +// Result is memoised by (path, mtime, size) for probeCacheTTL — repeat plays +// of the same file at the same quality (the HLS cache HIT path) skip ffprobe +// entirely. ffprobe on a 50 GB MKV can cost 1-3 s; first-segment latency +// shrinks by the same amount on the second play. func ProbeFile(ctx context.Context, ffprobePath, filePath string) (*StreamProbe, error) { + if cached, ok := lookupProbeCache(filePath); ok { + return cached, nil + } mi, err := mediainfo.ExtractMediaInfo(ctx, ffprobePath, filePath) if err != nil { return nil, fmt.Errorf("probe: %w", err) @@ -136,6 +144,7 @@ func ProbeFile(ctx context.Context, ffprobePath, filePath string) (*StreamProbe, }) } } + storeProbeCache(filePath, probe) return probe, nil } diff --git a/internal/engine/probe_cache.go b/internal/engine/probe_cache.go new file mode 100644 index 0000000..fcc7dec --- /dev/null +++ b/internal/engine/probe_cache.go @@ -0,0 +1,141 @@ +package engine + +import ( + "os" + "sync" + "time" +) + +// probeCacheTTL is how long a cached probe stays usable. The cache key +// already incorporates mtime + size, so the TTL is a defense against +// runaway memory growth from stale paths, not a freshness guarantee — a +// rename + recreate at the same inode (rare) would still be caught by the +// mtime delta. +const probeCacheTTL = 30 * time.Minute + +// probeCacheJanitorInterval is how often the background sweeper wakes to +// drop expired entries. Lookup-time eviction handles hot paths, but a +// user who browses 5k files and then stops would leak entries until each +// is individually re-touched. 5 min ≈ 6 sweeps per TTL window — enough +// to keep memory bounded without burning CPU. +const probeCacheJanitorInterval = 5 * time.Minute + +type probeCacheEntry struct { + probe *StreamProbe + expires time.Time +} + +type probeCacheKey struct { + path string + mtime int64 // ModTime().UnixNano() + size int64 +} + +var ( + probeCacheMu sync.RWMutex + probeCache = make(map[probeCacheKey]probeCacheEntry) + probeCacheJanitor sync.Once +) + +// startProbeCacheJanitor launches the background sweeper exactly once per +// process. Lazy — fired on first storeProbeCache. Drops expired entries +// every probeCacheJanitorInterval. Idempotent (sync.Once). +func startProbeCacheJanitor() { + probeCacheJanitor.Do(func() { + go func() { + ticker := time.NewTicker(probeCacheJanitorInterval) + defer ticker.Stop() + for range ticker.C { + sweepProbeCache(time.Now()) + } + }() + }) +} + +// sweepProbeCache removes every entry whose expiry is at or before `now`. +// Exposed for tests; production code calls it indirectly via the janitor +// goroutine. +func sweepProbeCache(now time.Time) int { + probeCacheMu.Lock() + defer probeCacheMu.Unlock() + removed := 0 + for k, e := range probeCache { + if !now.Before(e.expires) { + delete(probeCache, k) + removed++ + } + } + return removed +} + +// lookupProbeCache returns the cached StreamProbe for the given path if its +// mtime + size still match the value recorded at insert time, AND the cache +// entry hasn't expired. Any stat failure / mismatch returns (nil, false) so +// the caller falls through to a fresh ffprobe run. +func lookupProbeCache(path string) (*StreamProbe, bool) { + fi, err := os.Stat(path) + if err != nil { + return nil, false + } + key := probeCacheKey{ + path: path, + mtime: fi.ModTime().UnixNano(), + size: fi.Size(), + } + probeCacheMu.RLock() + entry, ok := probeCache[key] + probeCacheMu.RUnlock() + if !ok { + return nil, false + } + if time.Now().After(entry.expires) { + // Re-check under the write lock so a concurrent re-insert (same key, + // fresh expiry) isn't accidentally evicted. + probeCacheMu.Lock() + if cur, stillThere := probeCache[key]; stillThere && time.Now().After(cur.expires) { + delete(probeCache, key) + } + probeCacheMu.Unlock() + return nil, false + } + return entry.probe, true +} + +// storeProbeCache stashes a fresh probe result under the (path, mtime, size) +// key. A subsequent ffprobe-skipping HIT requires the file to still have the +// same mtime + size — anything else (re-encoded, renamed+recreated at the +// same path, truncated) misses and triggers a re-probe. +func storeProbeCache(path string, probe *StreamProbe) { + fi, err := os.Stat(path) + if err != nil { + return + } + key := probeCacheKey{ + path: path, + mtime: fi.ModTime().UnixNano(), + size: fi.Size(), + } + probeCacheMu.Lock() + probeCache[key] = probeCacheEntry{ + probe: probe, + expires: time.Now().Add(probeCacheTTL), + } + probeCacheMu.Unlock() + // Lazy janitor — fires once per process. No-op after first call. + startProbeCacheJanitor() +} + +// ResetProbeCache clears the in-memory probe cache. Test-only. +func ResetProbeCache() { + probeCacheMu.Lock() + probeCache = make(map[probeCacheKey]probeCacheEntry) + probeCacheMu.Unlock() +} + +// ProbeCacheSize returns the number of entries currently cached. Exposed +// for diagnostics + tests. +func ProbeCacheSize() int { + probeCacheMu.RLock() + defer probeCacheMu.RUnlock() + return len(probeCache) +} diff --git a/internal/engine/probe_cache_test.go b/internal/engine/probe_cache_test.go new file mode 100644 index 0000000..76c79da --- /dev/null +++ b/internal/engine/probe_cache_test.go @@ -0,0 +1,202 @@ +package engine + +import ( + "os" + "path/filepath" + "testing" + "time" +) + +func TestProbeCache_LookupMissNonexistent(t *testing.T) { + ResetProbeCache() + t.Cleanup(ResetProbeCache) + + if _, ok := lookupProbeCache("/path/that/does/not/exist"); ok { + t.Fatal("expected MISS for non-existent path") + } +} + +func TestProbeCache_StoreThenLookupHit(t *testing.T) { + ResetProbeCache() + t.Cleanup(ResetProbeCache) + + dir := t.TempDir() + path := filepath.Join(dir, "movie.mkv") + if err := os.WriteFile(path, []byte("fake content"), 0o644); err != nil { + t.Fatalf("write tmp file: %v", err) + } + + probe := &StreamProbe{VideoCodec: "h264", Width: 1920, Height: 1080, DurationSec: 5400} + storeProbeCache(path, probe) + + got, ok := lookupProbeCache(path) + if !ok { + t.Fatal("expected HIT after store") + } + if got != probe { + t.Fatalf("expected pointer-identical probe; got different") + } +} + +func TestProbeCache_MtimeChangeInvalidates(t *testing.T) { + ResetProbeCache() + t.Cleanup(ResetProbeCache) + + dir := t.TempDir() + path := filepath.Join(dir, "movie.mkv") + if err := os.WriteFile(path, []byte("original"), 0o644); err != nil { + t.Fatalf("write: %v", err) + } + + probe := &StreamProbe{VideoCodec: "h264", DurationSec: 100} + storeProbeCache(path, probe) + + // Force mtime change. WriteFile doesn't guarantee a different mtime if + // the filesystem timestamp resolution is coarse, so set it explicitly + // to a value 1 hour in the future. + future := time.Now().Add(1 * time.Hour) + if err := os.Chtimes(path, future, future); err != nil { + t.Fatalf("chtimes: %v", err) + } + + if _, ok := lookupProbeCache(path); ok { + t.Fatal("expected MISS after mtime change") + } +} + +func TestProbeCache_SizeChangeInvalidates(t *testing.T) { + ResetProbeCache() + t.Cleanup(ResetProbeCache) + + dir := t.TempDir() + path := filepath.Join(dir, "movie.mkv") + if err := os.WriteFile(path, []byte("aaaaa"), 0o644); err != nil { + t.Fatalf("write: %v", err) + } + originalMtime := time.Now().Add(-1 * time.Hour) // stable, in the past + if err := os.Chtimes(path, originalMtime, originalMtime); err != nil { + t.Fatalf("chtimes original: %v", err) + } + + probe := &StreamProbe{VideoCodec: "h264", DurationSec: 100} + storeProbeCache(path, probe) + + // Truncate to a different size, then reset mtime to the original so + // only `size` differs between store and lookup keys — isolates the + // size-check path. Without the Chtimes, WriteFile bumps mtime and the + // test would pass via mtime invalidation regardless of size logic. + if err := os.WriteFile(path, []byte("a"), 0o644); err != nil { + t.Fatalf("rewrite: %v", err) + } + if err := os.Chtimes(path, originalMtime, originalMtime); err != nil { + t.Fatalf("chtimes restore: %v", err) + } + + if _, ok := lookupProbeCache(path); ok { + t.Fatal("expected MISS after size change") + } +} + +func TestProbeCache_ExpiryDropsEntry(t *testing.T) { + ResetProbeCache() + t.Cleanup(ResetProbeCache) + + dir := t.TempDir() + path := filepath.Join(dir, "movie.mkv") + if err := os.WriteFile(path, []byte("content"), 0o644); err != nil { + t.Fatalf("write: %v", err) + } + + // Stash an entry whose expires is already in the past — simulates TTL + // having elapsed without sleeping for 30 min. + fi, err := os.Stat(path) + if err != nil { + t.Fatalf("stat: %v", err) + } + key := probeCacheKey{path: path, mtime: fi.ModTime().UnixNano(), size: fi.Size()} + probeCacheMu.Lock() + probeCache[key] = probeCacheEntry{ + probe: &StreamProbe{VideoCodec: "h264"}, + expires: time.Now().Add(-1 * time.Minute), + } + probeCacheMu.Unlock() + + if _, ok := lookupProbeCache(path); ok { + t.Fatal("expected MISS for expired entry") + } + // Side-effect: lookup should have evicted the stale entry. + if ProbeCacheSize() != 0 { + t.Fatalf("expected cache size 0 after expiry eviction; got %d", ProbeCacheSize()) + } +} + +func TestProbeCache_ResetClears(t *testing.T) { + ResetProbeCache() + + dir := t.TempDir() + path := filepath.Join(dir, "movie.mkv") + if err := os.WriteFile(path, []byte("x"), 0o644); err != nil { + t.Fatalf("write: %v", err) + } + + storeProbeCache(path, &StreamProbe{VideoCodec: "h264"}) + if ProbeCacheSize() != 1 { + t.Fatalf("expected size 1 after store; got %d", ProbeCacheSize()) + } + + ResetProbeCache() + if ProbeCacheSize() != 0 { + t.Fatalf("expected size 0 after reset; got %d", ProbeCacheSize()) + } +} + +func TestProbeCache_StoreNonexistentNoOp(t *testing.T) { + ResetProbeCache() + t.Cleanup(ResetProbeCache) + + // Store on a non-existent path should silently do nothing (stat fails), + // not panic, and not poison the cache with a zero key. + storeProbeCache("/nope/never/exists.mkv", &StreamProbe{VideoCodec: "h264"}) + if ProbeCacheSize() != 0 { + t.Fatalf("expected 0 entries; got %d", ProbeCacheSize()) + } +} + +func TestProbeCache_SweepDropsExpired(t *testing.T) { + ResetProbeCache() + t.Cleanup(ResetProbeCache) + + dir := t.TempDir() + // Two entries: one expired, one fresh. + expiredPath := filepath.Join(dir, "old.mkv") + freshPath := filepath.Join(dir, "new.mkv") + if err := os.WriteFile(expiredPath, []byte("a"), 0o644); err != nil { + t.Fatalf("write expired: %v", err) + } + if err := os.WriteFile(freshPath, []byte("b"), 0o644); err != nil { + t.Fatalf("write fresh: %v", err) + } + + now := time.Now() + fiExp, _ := os.Stat(expiredPath) + fiFresh, _ := os.Stat(freshPath) + + probeCacheMu.Lock() + probeCache[probeCacheKey{path: expiredPath, mtime: fiExp.ModTime().UnixNano(), size: fiExp.Size()}] = probeCacheEntry{ + probe: &StreamProbe{VideoCodec: "h264"}, + expires: now.Add(-1 * time.Minute), // expired + } + probeCache[probeCacheKey{path: freshPath, mtime: fiFresh.ModTime().UnixNano(), size: fiFresh.Size()}] = probeCacheEntry{ + probe: &StreamProbe{VideoCodec: "h264"}, + expires: now.Add(10 * time.Minute), // fresh + } + probeCacheMu.Unlock() + + removed := sweepProbeCache(now) + if removed != 1 { + t.Fatalf("expected 1 expired entry removed; got %d", removed) + } + if ProbeCacheSize() != 1 { + t.Fatalf("expected 1 fresh entry kept; got %d", ProbeCacheSize()) + } +} diff --git a/internal/engine/vaapi_args_test.go b/internal/engine/vaapi_args_test.go new file mode 100644 index 0000000..33d0786 --- /dev/null +++ b/internal/engine/vaapi_args_test.go @@ -0,0 +1,97 @@ +package engine + +import ( + "strings" + "testing" +) + +func TestBuildHLSFFmpegArgsVAAPI(t *testing.T) { + cfg := HLSSessionConfig{ + SessionID: "test", + SourcePath: "/tmp/test.mkv", + Quality: "720p", + AudioIndex: 0, + Transcode: TranscodeRuntime{ + FFmpegPath: "/usr/bin/ffmpeg", + FFprobePath: "/usr/bin/ffprobe", + HWAccel: HWAccelVAAPI, + }, + } + probe := &StreamProbe{Width: 1920, Height: 1080, DurationSec: 100} + args := buildHLSFFmpegArgsAt(cfg, probe, "/tmp/tmpdir", 0, 0) + got := strings.Join(args, " ") + + wants := []string{ + "-hwaccel vaapi", + "-vaapi_device /dev/dri/renderD128", + "-c:v h264_vaapi", + "format=nv12", + "hwupload", + } + for _, want := range wants { + if !strings.Contains(got, want) { + t.Errorf("argv missing %q\n%s", want, got) + } + } + if strings.Contains(got, "scale_vaapi") { + t.Errorf("argv unexpectedly contains scale_vaapi (mesa bug): %s", got) + } + if strings.Contains(got, "format=yuv420p") { + t.Errorf("argv contains format=yuv420p (libx264 path) for VAAPI codec: %s", got) + } +} + +func TestBuildHLSFFmpegArgsLibx264NoRegression(t *testing.T) { + cfg := HLSSessionConfig{ + SessionID: "test", + SourcePath: "/tmp/test.mkv", + Quality: "720p", + AudioIndex: 0, + Transcode: TranscodeRuntime{ + FFmpegPath: "/usr/bin/ffmpeg", + FFprobePath: "/usr/bin/ffprobe", + HWAccel: HWAccelNone, + }, + } + probe := &StreamProbe{Width: 1920, Height: 1080, DurationSec: 100} + args := buildHLSFFmpegArgsAt(cfg, probe, "/tmp/tmpdir", 0, 0) + got := strings.Join(args, " ") + for _, want := range []string{"-c:v libx264", "format=yuv420p", "setparams=colorspace=bt709"} { + if !strings.Contains(got, want) { + t.Errorf("libx264 argv missing %q: %s", want, got) + } + } + for _, bad := range []string{"-vaapi_device", "format=nv12", "hwupload"} { + if strings.Contains(got, bad) { + t.Errorf("libx264 argv unexpectedly contains %q: %s", bad, got) + } + } +} + +// TestBuildHLSFFmpegArgsVAAPIDump prints the full argv buildHLSFFmpegArgsAt +// emits for a typical VAAPI session. Mimics the daemon spawn step so the +// operator can verify the ffmpeg command-line shape without booting the +// stack — equivalent to `journalctl --user -u unarr-dev | grep ffmpeg` +// but without waiting for a real player session. +func TestBuildHLSFFmpegArgsVAAPIDump(t *testing.T) { + cfg := HLSSessionConfig{ + SessionID: "vaapi-smoke", + SourcePath: "/mnt/nas/peliculas/sample.mkv", + Quality: "720p", + AudioIndex: -1, + Transcode: TranscodeRuntime{ + FFmpegPath: "/usr/bin/ffmpeg", + FFprobePath: "/usr/bin/ffprobe", + HWAccel: HWAccelVAAPI, + }, + } + probe := &StreamProbe{ + VideoCodec: "hevc", + Width: 3840, + Height: 2160, + DurationSec: 5400, + AudioTracks: []ProbeAudioTrack{{Index: 0, Lang: "en", Codec: "ac3"}}, + } + args := buildHLSFFmpegArgsAt(cfg, probe, "/tmp/smoke-tmpdir", 0, 0) + t.Logf("ffmpeg %s", strings.Join(args, " ")) +} diff --git a/internal/engine/validate.go b/internal/engine/validate.go index dd07516..0efd4de 100644 --- a/internal/engine/validate.go +++ b/internal/engine/validate.go @@ -21,12 +21,27 @@ var validSessionID = regexp.MustCompile(`^[a-zA-Z0-9_-]{1,128}$`) // 127.0.0.1 is listed in addition to localhost because some browsers treat // them as distinct origins for CORS. // +// Mirrors (`.to`, `staging.torrentclaw.com`, `www.`) are listed so a user +// playing from any official mirror succeeds the HEAD probe; without these +// the browser drops the response for "missing ACAO" and the player reports +// "404 todos los canales" even though the daemon returned 200. +// // Note: media tags (