diff --git a/.github/workflows/actions.lock b/.github/workflows/actions.lock new file mode 100644 index 000000000..acbdeff60 --- /dev/null +++ b/.github/workflows/actions.lock @@ -0,0 +1,188 @@ +# This file is machine-generated by `gh actions-lock`. +# Do not edit by hand; run `gh actions-lock` to update. +# Docs: https://gh.io/actions-lockfile +version: 'v0.0.3' +workflows: + '.github/workflows/ai-core.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/atomic.yml': + - 'actions/checkout@v4.4.0' + - 'actions/checkout@v7.0.1' + - 'dtolnay/rust-toolchain@stable' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/audio.yml': + - 'actions/cache@v6.1.0' + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + '.github/workflows/cms.yml': + - 'actions/checkout@v4.4.0' + - 'dtolnay/rust-toolchain@stable' + '.github/workflows/crw.yml': + - 'actions/cache@v6.1.0' + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + '.github/workflows/dialogs.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/doc.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/docpath.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/ecproj-format.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/font-store.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/formats.yml': + - 'actions/checkout@v7.0.1' + - 'dtolnay/rust-toolchain@stable' + - 'step-security/harden-runner@v2.22.4' + - 'swatinem/rust-cache@v2.9.3' + '.github/workflows/house-rules.yml': [] + '.github/workflows/i18n.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/jpeg.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + '.github/workflows/keymap.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + '.github/workflows/lockfile.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/media.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + '.github/workflows/menu.yml': + - 'actions/checkout@v7.0.1' + - 'dtolnay/rust-toolchain@stable' + - 'swatinem/rust-cache@v2.9.3' + '.github/workflows/pcraft-format.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/pdf.yml': + - 'actions/checkout@v7.0.1' + - 'actions/setup-python@v7.1.0' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + - 'step-security/taiki-e-install-action@v2.87.32' + '.github/workflows/print.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/proof.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + '.github/workflows/sonarcloud.yml': [] + '.github/workflows/ui.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + '.github/workflows/update.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/vectorcraft-format.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/video.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + '.github/workflows/web-api.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/taiki-e-install-action@v2.87.32' + '.github/workflows/web-host.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + - 'step-security/rust-cache@v2.9.2' + - 'step-security/taiki-e-install-action@v2.87.32' + '.github/workflows/window.yml': + - 'actions/checkout@v7.0.1' + - 'step-security/dtolnay-rust-toolchain@v1.0.0' + - 'step-security/harden-runner@v2.22.4' + '.github/workflows/workspace.yml': [] +dependencies: + 'actions/cache@v6.1.0': + ref: 'v6.1.0' + commit: 'sha1-55cc8345863c7cc4c66a329aec7e433d2d1c52a9' + owner_id: 44036562 + repo_id: 215566462 + 'actions/checkout@v4.4.0': + ref: 'v4.4.0' + commit: 'sha1-11d5960a326750d5838078e36cf38b85af677262' + owner_id: 44036562 + repo_id: 197814629 + 'actions/checkout@v7.0.1': + ref: 'v7.0.1' + commit: 'sha1-3d3c42e5aac5ba805825da76410c181273ba90b1' + owner_id: 44036562 + repo_id: 197814629 + 'actions/setup-python@v7.1.0': + ref: 'v7.1.0' + commit: 'sha1-6d4fd22b20465ccc0f7354476588e8df7c9ca0e0' + owner_id: 44036562 + repo_id: 192625525 + 'dtolnay/rust-toolchain@stable': + ref: 'stable' + commit: 'sha1-b725eda525362e7cf34af9a7243cac1e3f01a7ec' + owner_id: 1940490 + repo_id: 260749683 + 'step-security/dtolnay-rust-toolchain@v1.0.0': + ref: 'v1.0.0' + commit: 'sha1-8dd607977d06adf4ed49b190381e74c17a9dd25f' + owner_id: 88700172 + repo_id: 1249865941 + 'step-security/harden-runner@v2.22.4': + ref: 'v2.22.4' + commit: 'sha1-1d88384fd173af6ecb012dfd0264a306b711241a' + owner_id: 88700172 + repo_id: 422287306 + 'step-security/rust-cache@v2.9.2': + ref: 'v2.9.2' + commit: 'sha1-8871c8ba063416da7c32246932ea6ae08700453e' + owner_id: 88700172 + repo_id: 822636713 + 'step-security/taiki-e-install-action@v2.87.32': + ref: 'v2.87.32' + commit: 'sha1-76747d22a7567cc4438546a84f11bdf18e670ab8' + owner_id: 88700172 + repo_id: 1382753260 + 'swatinem/rust-cache@v2.9.3': + ref: 'v2.9.3' + commit: 'sha1-d1527c8f67ef9af357290475d4af4b0fab8f12e0' + owner_id: 580492 + repo_id: 298565987 diff --git a/.github/workflows/ai-core.yml b/.github/workflows/ai-core.yml index cc8d738ac..5471bf1ec 100644 --- a/.github/workflows/ai-core.yml +++ b/.github/workflows/ai-core.yml @@ -28,11 +28,11 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/.github/workflows/atomic.yml b/.github/workflows/atomic.yml index 1a20fa679..f72a64f1e 100644 --- a/.github/workflows/atomic.yml +++ b/.github/workflows/atomic.yml @@ -28,11 +28,11 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy @@ -50,7 +50,7 @@ jobs: name: Atomic WASM runs-on: ubuntu-latest steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v4.4.0 - uses: dtolnay/rust-toolchain@stable with: targets: wasm32-unknown-unknown diff --git a/.github/workflows/audio.yml b/.github/workflows/audio.yml index 0517efe6f..d6a7ff5e2 100644 --- a/.github/workflows/audio.yml +++ b/.github/workflows/audio.yml @@ -25,18 +25,18 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-audio-lint save-if: ${{ github.ref == 'refs/heads/main' }} @@ -69,15 +69,15 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false # The workspace's rust-version. - run: rustup toolchain install 1.95.0 --profile minimal - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-audio-msrv save-if: ${{ github.ref == 'refs/heads/main' }} @@ -88,16 +88,16 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-audio-features save-if: ${{ github.ref == 'refs/heads/main' }} @@ -113,20 +113,20 @@ jobs: runs-on: ubuntu-26.04 steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-audio-test save-if: ${{ github.ref == 'refs/heads/main' }} - - uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0 + - uses: actions/cache@v6.1.0 with: path: ${{ runner.temp }}/media-fixtures/opus/vectors # The archive's SHA-256 (scripts/fetch-opus-vectors.sh checks it). diff --git a/.github/workflows/cms.yml b/.github/workflows/cms.yml index a15a48c9d..74ce0815f 100644 --- a/.github/workflows/cms.yml +++ b/.github/workflows/cms.yml @@ -16,7 +16,7 @@ jobs: timeout-minutes: 20 runs-on: ubuntu-latest steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v4.4.0 - uses: dtolnay/rust-toolchain@stable with: components: clippy diff --git a/.github/workflows/crw.yml b/.github/workflows/crw.yml index 8fec98661..54be7ba3b 100644 --- a/.github/workflows/crw.yml +++ b/.github/workflows/crw.yml @@ -29,17 +29,17 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-crw-${{ matrix.os }} save-if: ${{ github.ref == 'refs/heads/main' }} @@ -53,17 +53,17 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-crw-wasm save-if: ${{ github.ref == 'refs/heads/main' }} @@ -75,21 +75,21 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-crw-samples save-if: ${{ github.ref == 'refs/heads/main' }} - name: Cache the CC0 sample files - uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0 + uses: actions/cache@v6.1.0 with: path: ${{ runner.temp }}/crw-samples key: crw-samples-${{ hashFiles('crates/crw/tests/samples.json') }} diff --git a/.github/workflows/dialogs.yml b/.github/workflows/dialogs.yml index 97a931aec..c8d43d725 100644 --- a/.github/workflows/dialogs.yml +++ b/.github/workflows/dialogs.yml @@ -23,15 +23,15 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: # The token isn't needed after checkout; cargo runs build scripts and proc macros from # dependencies, so don't leave it in .git/config for that third-party code to read. persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: clippy diff --git a/.github/workflows/doc.yml b/.github/workflows/doc.yml index a0c619c1e..6caa4f0ee 100644 --- a/.github/workflows/doc.yml +++ b/.github/workflows/doc.yml @@ -28,13 +28,13 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/.github/workflows/docpath.yml b/.github/workflows/docpath.yml index 571b65eca..e0115f6f5 100644 --- a/.github/workflows/docpath.yml +++ b/.github/workflows/docpath.yml @@ -28,11 +28,11 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/.github/workflows/ecproj-format.yml b/.github/workflows/ecproj-format.yml index 16bab05c7..24a3350af 100644 --- a/.github/workflows/ecproj-format.yml +++ b/.github/workflows/ecproj-format.yml @@ -28,13 +28,13 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/.github/workflows/font-store.yml b/.github/workflows/font-store.yml index f246f621a..0578f18ec 100644 --- a/.github/workflows/font-store.yml +++ b/.github/workflows/font-store.yml @@ -28,11 +28,11 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/.github/workflows/formats.yml b/.github/workflows/formats.yml index e19902e8c..84d2c0d7e 100644 --- a/.github/workflows/formats.yml +++ b/.github/workflows/formats.yml @@ -72,17 +72,17 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: dtolnay/rust-toolchain@b725eda525362e7cf34af9a7243cac1e3f01a7ec # stable + - uses: dtolnay/rust-toolchain@stable with: components: rustfmt, clippy targets: wasm32-unknown-unknown - - uses: Swatinem/rust-cache@d1527c8f67ef9af357290475d4af4b0fab8f12e0 # v2.9.3 + - uses: Swatinem/rust-cache@v2.9.3 with: key: ${{ matrix.crate }} save-if: ${{ github.ref == 'refs/heads/main' }} diff --git a/.github/workflows/i18n.yml b/.github/workflows/i18n.yml index 5f717aeff..7e6903188 100644 --- a/.github/workflows/i18n.yml +++ b/.github/workflows/i18n.yml @@ -28,11 +28,11 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/.github/workflows/jpeg.yml b/.github/workflows/jpeg.yml index 900baf678..21db0ab79 100644 --- a/.github/workflows/jpeg.yml +++ b/.github/workflows/jpeg.yml @@ -25,18 +25,18 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-jpeg save-if: ${{ github.ref == 'refs/heads/main' }} diff --git a/.github/workflows/keymap.yml b/.github/workflows/keymap.yml index aba535265..41fcc0502 100644 --- a/.github/workflows/keymap.yml +++ b/.github/workflows/keymap.yml @@ -27,24 +27,24 @@ jobs: CRAFT_FONTS_DIR: ${{ github.workspace }}/.test-inputs/craft-fonts steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - name: Checkout test fonts - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + uses: actions/checkout@v7.0.1 with: repository: storytold/craft-fonts ref: b076417b5880bd4154bd22f00695b1d28c6bea83 path: .test-inputs/craft-fonts persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-keymap-linux save-if: ${{ github.ref == 'refs/heads/main' }} @@ -71,23 +71,23 @@ jobs: CRAFT_FONTS_DIR: ${{ github.workspace }}/.test-inputs/craft-fonts steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - name: Checkout test fonts - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + uses: actions/checkout@v7.0.1 with: repository: storytold/craft-fonts ref: b076417b5880bd4154bd22f00695b1d28c6bea83 path: .test-inputs/craft-fonts persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-keymap-${{ matrix.os }} save-if: ${{ github.ref == 'refs/heads/main' }} @@ -98,17 +98,17 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-keymap-wasm save-if: ${{ github.ref == 'refs/heads/main' }} diff --git a/.github/workflows/lockfile.yml b/.github/workflows/lockfile.yml new file mode 100644 index 000000000..e630c9a8f --- /dev/null +++ b/.github/workflows/lockfile.yml @@ -0,0 +1,34 @@ +name: Lockfile + +# Fails when .github/workflows/actions.lock is out of sync with the workflows, or when a locked pin +# no longer verifies upstream, so a moved tag cannot slip a different commit into a run. +# Docs: https://gh.io/actions-lockfile +on: + pull_request: + push: + branches: [main] + workflow_dispatch: + +permissions: + contents: read + +env: + GH_TOKEN: ${{ github.token }} + +jobs: + lockfile: + name: Lockfile is in sync + runs-on: ubuntu-24.04 + timeout-minutes: 10 + steps: + - name: Harden runner + uses: step-security/harden-runner@v2.22.4 + with: + egress-policy: audit + - uses: actions/checkout@v7.0.1 + with: + persist-credentials: false + - name: Install gh-actions-lock + run: gh extension install github/gh-actions-lock --pin v0.4.1 + - name: Verify the lockfile + run: gh actions-lock --verify diff --git a/.github/workflows/media.yml b/.github/workflows/media.yml index cb3912d54..4814e35ca 100644 --- a/.github/workflows/media.yml +++ b/.github/workflows/media.yml @@ -25,18 +25,18 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-media-lint save-if: ${{ github.ref == 'refs/heads/main' }} @@ -56,16 +56,16 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-media-test save-if: ${{ github.ref == 'refs/heads/main' }} diff --git a/.github/workflows/menu.yml b/.github/workflows/menu.yml index 6d0e082cb..8d1058731 100644 --- a/.github/workflows/menu.yml +++ b/.github/workflows/menu.yml @@ -21,14 +21,14 @@ jobs: name: Menu (Linux) runs-on: ubuntu-latest steps: - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: dtolnay/rust-toolchain@b725eda525362e7cf34af9a7243cac1e3f01a7ec # stable + - uses: dtolnay/rust-toolchain@stable with: components: rustfmt, clippy targets: aarch64-apple-darwin, wasm32-unknown-unknown - - uses: Swatinem/rust-cache@d1527c8f67ef9af357290475d4af4b0fab8f12e0 # v2.9.3 + - uses: Swatinem/rust-cache@v2.9.3 with: key: craft-menu-linux - run: cargo fmt -p craft-menu --check diff --git a/.github/workflows/pcraft-format.yml b/.github/workflows/pcraft-format.yml index 177bb4593..9c66b5a10 100644 --- a/.github/workflows/pcraft-format.yml +++ b/.github/workflows/pcraft-format.yml @@ -28,13 +28,13 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: "1.95" components: rustfmt, clippy diff --git a/.github/workflows/pdf.yml b/.github/workflows/pdf.yml index a494d085c..09ddfabca 100644 --- a/.github/workflows/pdf.yml +++ b/.github/workflows/pdf.yml @@ -31,19 +31,19 @@ jobs: RUSTFLAGS: --cfg craft_pdf_unicode_fixtures steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-pdf save-if: ${{ github.ref == 'refs/heads/main' }} - - uses: actions/setup-python@v7 + - uses: actions/setup-python@v7.1.0 with: python-version: "3.12" - name: Provision licensed PDF test fonts @@ -61,14 +61,14 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-pdf-backends save-if: ${{ github.ref == 'refs/heads/main' }} @@ -81,19 +81,19 @@ jobs: CARGO_TARGET_WASM32_UNKNOWN_UNKNOWN_RUNNER: wasm-bindgen-test-runner steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-pdf-wasm save-if: ${{ github.ref == 'refs/heads/main' }} - - uses: actions/setup-python@v7 + - uses: actions/setup-python@v7.1.0 with: python-version: "3.12" - name: Provision licensed PDF test fonts @@ -111,7 +111,7 @@ jobs: - name: Install the matching WASM test runner # A prebuilt, checksum-verified binary instead of building it from source on every run. # A version the pinned action does not list yet falls back to `cargo install --locked`. - uses: step-security/taiki-e-install-action@76747d22a7567cc4438546a84f11bdf18e670ab8 # v2.87.32 + uses: step-security/taiki-e-install-action@v2.87.32 with: tool: wasm-bindgen@${{ steps.wasm-bindgen.outputs.version }} fallback: cargo-install @@ -122,19 +122,19 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: clippy - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-plate-test save-if: ${{ github.ref == 'refs/heads/main' }} - - uses: actions/setup-python@v7 + - uses: actions/setup-python@v7.1.0 with: python-version: "3.12" - name: Provision licensed PDF test fonts diff --git a/.github/workflows/print.yml b/.github/workflows/print.yml index 3838e6c42..8e2108e09 100644 --- a/.github/workflows/print.yml +++ b/.github/workflows/print.yml @@ -24,11 +24,11 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/.github/workflows/proof.yml b/.github/workflows/proof.yml index 0035b1c04..10365a230 100644 --- a/.github/workflows/proof.yml +++ b/.github/workflows/proof.yml @@ -25,27 +25,27 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false # The Language data (craft-i18n `language-data/`), embedded by crates/proof/build.rs. # Pinned to storytold/craft-i18n#114's head (#109's SCOWL dictionaries + judgment/judgement, # CP2.1); TODO(CP2.1): move to #114's merge commit once it merges. - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: repository: storytold/craft-i18n ref: c239f934ace90b7939afaceb23b472ceeafd6b8a path: craft-i18n persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-proof save-if: ${{ github.ref == 'refs/heads/main' }} diff --git a/.github/workflows/ui.yml b/.github/workflows/ui.yml index 0892f5a2c..f15ef6ddc 100644 --- a/.github/workflows/ui.yml +++ b/.github/workflows/ui.yml @@ -30,21 +30,21 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 - name: Install software Vulkan renderer if: runner.os == 'Linux' run: | sudo apt-get update sudo apt-get install -y mesa-vulkan-drivers libvulkan1 echo 'WGPU_BACKEND=vulkan' >> "$GITHUB_ENV" - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-ui save-if: ${{ github.ref == 'refs/heads/main' }} @@ -59,15 +59,15 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: actions/checkout@v7.0.1 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-ui-wasm save-if: ${{ github.ref == 'refs/heads/main' }} diff --git a/.github/workflows/update.yml b/.github/workflows/update.yml index f9c8e1498..9a7139026 100644 --- a/.github/workflows/update.yml +++ b/.github/workflows/update.yml @@ -28,15 +28,15 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: # The token isn't needed after checkout; cargo runs build scripts and proc macros from # dependencies, so don't leave it in .git/config for that third-party code to read. persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/.github/workflows/vectorcraft-format.yml b/.github/workflows/vectorcraft-format.yml index b2b20463a..74144f30f 100644 --- a/.github/workflows/vectorcraft-format.yml +++ b/.github/workflows/vectorcraft-format.yml @@ -28,13 +28,13 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: "1.95" components: rustfmt, clippy diff --git a/.github/workflows/video.yml b/.github/workflows/video.yml index ec1c16eda..fa667bdfc 100644 --- a/.github/workflows/video.yml +++ b/.github/workflows/video.yml @@ -22,18 +22,18 @@ jobs: runs-on: ubuntu-latest steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: "1.95" components: rustfmt, clippy targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-video-lint - run: cargo fmt -p craft-video --check @@ -60,16 +60,16 @@ jobs: name: Video tests (${{ matrix.os }}) steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: "1.95" - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-video-${{ matrix.os }} - name: Install ffmpeg (Linux) diff --git a/.github/workflows/web-api.yml b/.github/workflows/web-api.yml index 2460fb569..4bf1a30fe 100644 --- a/.github/workflows/web-api.yml +++ b/.github/workflows/web-api.yml @@ -26,13 +26,13 @@ jobs: CARGO_TARGET_WASM32_UNKNOWN_UNKNOWN_RUNNER: wasm-bindgen-test-runner steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy @@ -46,7 +46,7 @@ jobs: - name: Install the matching WASM test runner # A prebuilt, checksum-verified binary instead of building it from source on every run. # A version the pinned action does not list yet falls back to `cargo install --locked`. - uses: step-security/taiki-e-install-action@76747d22a7567cc4438546a84f11bdf18e670ab8 # v2.87.32 + uses: step-security/taiki-e-install-action@v2.87.32 with: tool: wasm-bindgen@${{ steps.wasm-bindgen.outputs.version }} fallback: cargo-install diff --git a/.github/workflows/web-host.yml b/.github/workflows/web-host.yml index b780ee76b..76e6701e5 100644 --- a/.github/workflows/web-host.yml +++ b/.github/workflows/web-host.yml @@ -27,18 +27,18 @@ jobs: CARGO_TARGET_WASM32_UNKNOWN_UNKNOWN_RUNNER: wasm-bindgen-test-runner steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 + - uses: actions/checkout@v7.0.1 with: persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy targets: wasm32-unknown-unknown - - uses: step-security/rust-cache@8871c8ba063416da7c32246932ea6ae08700453e # v2.9.2 + - uses: step-security/rust-cache@v2.9.2 with: key: craft-web-host save-if: ${{ github.ref == 'refs/heads/main' }} @@ -55,7 +55,7 @@ jobs: - name: Install the matching WASM test runner # A prebuilt, checksum-verified binary instead of building it from source on every run. # A version the pinned action does not list yet falls back to `cargo install --locked`. - uses: step-security/taiki-e-install-action@76747d22a7567cc4438546a84f11bdf18e670ab8 # v2.87.32 + uses: step-security/taiki-e-install-action@v2.87.32 with: tool: wasm-bindgen@${{ steps.wasm-bindgen.outputs.version }} fallback: cargo-install diff --git a/.github/workflows/window.yml b/.github/workflows/window.yml index f570ef5ce..fdd28e203 100644 --- a/.github/workflows/window.yml +++ b/.github/workflows/window.yml @@ -28,15 +28,15 @@ jobs: runs-on: ${{ matrix.os }} steps: - name: Harden runner - uses: step-security/harden-runner@1d88384fd173af6ecb012dfd0264a306b711241a # v2.22.4 + uses: step-security/harden-runner@v2.22.4 with: egress-policy: audit - - uses: actions/checkout@v7 + - uses: actions/checkout@v7.0.1 with: # The token isn't needed after checkout; cargo runs build scripts and proc macros from # dependencies, so don't leave it in .git/config for that third-party code to read. persist-credentials: false - - uses: step-security/dtolnay-rust-toolchain@8dd607977d06adf4ed49b190381e74c17a9dd25f # v1.0.0 + - uses: step-security/dtolnay-rust-toolchain@v1.0.0 with: toolchain: stable components: rustfmt, clippy diff --git a/Cargo.lock b/Cargo.lock index 3d8cf4a7b..6e48813da 100644 --- a/Cargo.lock +++ b/Cargo.lock @@ -222,6 +222,16 @@ version = "1.0.104" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "330a5ed07fa54e4702c9d6c4174f74427fc0ef6e214bbd677ae50a5099946470" +[[package]] +name = "aotuv_lancer_vorbis_sys" +version = "0.1.6" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "5bc4fd1a61860d2f1198b60bedd30910eaffa978f1ee6214dfb24ac70d589225" +dependencies = [ + "cc", + "ogg_next_sys", +] + [[package]] name = "arbitrary" version = "1.5.0" @@ -1349,6 +1359,7 @@ dependencies = [ "serde_json", "symphonia", "thiserror 2.0.21", + "vorbis_rs", ] [[package]] @@ -1846,18 +1857,22 @@ dependencies = [ name = "craft-image-codecs" version = "0.1.0" dependencies = [ + "avif-serialize", "craft-heif", "craft-jpeg", "craft-metadata", "exr", "flate2", "half", + "hayro-jpeg2000", "image", "image-webp", "jpeg-encoder 0.6.1", "jxl-oxide", "png 0.18.1", "proptest", + "rav1e", + "ravif", "rayon", "serde_json", "thiserror 2.0.21", @@ -2799,6 +2814,7 @@ dependencies = [ name = "craft-xfl" version = "0.1.0" dependencies = [ + "craft-xml", "miniz_oxide 0.8.9", "thiserror 2.0.21", ] @@ -6605,6 +6621,15 @@ dependencies = [ "phf 0.11.3", ] +[[package]] +name = "ogg_next_sys" +version = "0.1.5" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "ed2d7a48e247c2bb07e633aefb65a38648ea58c7eedd4e4408a5861721ab049b" +dependencies = [ + "cc", +] + [[package]] name = "once_cell" version = "1.21.4" @@ -10366,6 +10391,20 @@ version = "0.9.5" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "0b928f33d975fc6ad9f86c8f283853ad26bdd5b10b7f1542aa2fa15e2289105a" +[[package]] +name = "vorbis_rs" +version = "0.5.6" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "49c5da94d280f7a27e8c937e9b73df2da3e23a2583f48471fd8fb4c72f9c1933" +dependencies = [ + "aotuv_lancer_vorbis_sys", + "errno", + "getrandom 0.4.3", + "ogg_next_sys", + "thiserror 2.0.21", + "tinyvec", +] + [[package]] name = "wait-timeout" version = "0.2.1" diff --git a/crates/audio/Cargo.toml b/crates/audio/Cargo.toml index d132f23da..8798272b3 100644 --- a/crates/audio/Cargo.toml +++ b/crates/audio/Cargo.toml @@ -24,13 +24,15 @@ decode-native = ["aac-decode", "alac-decode", "opus-decode", "ac3-decode", "mpa- symphonia = ["decode-native", "dep:symphonia"] # Every encoder: the lossless file writers and the AAC, ALAC, AC-3/E-AC-3, MP3 and Opus encoders. # All but Opus are also file formats of `encode` and `FileEncoder`. -encode = ["lossless-encode", "aac-encode", "alac-encode", "ac3-encode", "eac3-encode", "mp3-encode", "opus-encode"] +encode = ["lossless-encode", "aac-encode", "alac-encode", "ac3-encode", "eac3-encode", "mp3-encode", "opus-encode", "vorbis-encode"] # Encode WAV/BWF/RF64, AIFF/AIFF-C and FLAC files. lossless-encode = ["pcm-encode", "flac-encode"] # Encode WAV/BWF/RF64 and AIFF/AIFF-C files. pcm-encode = [] # Encode FLAC files. flac-encode = ["craft-media/bitstream"] +# Encode Ogg Vorbis files (desktop only: builds the reference libvorbis from its C sources). +vorbis-encode = ["dep:vorbis_rs"] # Codec modules on their own, for packet-level use. aac-decode = ["craft-media/bitstream"] # The AAC encoder (`aac::Encoder`), and AAC in M4A through `encode` and `FileEncoder`. @@ -65,6 +67,9 @@ rayon = { workspace = true, optional = true } # symphonia keeps its AAC decoder for AAC in CAF, a container craft-media does not read. symphonia = { version = "0.6.1", optional = true, default-features = false, features = ["mp3", "flac", "vorbis", "ogg", "wav", "pcm", "aiff", "isomp4", "aac", "alac", "adpcm", "caf", "mkv"] } +[target.'cfg(not(target_arch = "wasm32"))'.dependencies] +vorbis_rs = { version = "0.5", optional = true } + [dev-dependencies] craft-media-testkit = { workspace = true } proptest = { workspace = true } diff --git a/crates/audio/src/lib.rs b/crates/audio/src/lib.rs index 1c6c73dcf..33dcdee3b 100644 --- a/crates/audio/src/lib.rs +++ b/crates/audio/src/lib.rs @@ -62,9 +62,12 @@ mod wav; feature = "alac-encode", feature = "mp3-encode", feature = "ac3-encode", - feature = "eac3-encode" + feature = "eac3-encode", + feature = "vorbis-encode" ))] mod write; +#[cfg(feature = "vorbis-encode")] +pub use write::{OGG_BITRATES, encode_ogg}; /// The FLAC packet encoder (RFC 9639), for a caller that muxes FLAC itself (into MP4, Matroska or /// its own `.flac` writer): planar integer samples in, frames out, and the STREAMINFO to patch in @@ -107,8 +110,6 @@ pub use read::{ pub use read::{RawPcm, aiff_pcm, wav_pcm}; #[cfg(any(feature = "aac-decode", feature = "alac-decode", feature = "opus-decode", feature = "ac3-decode", feature = "mpa-decode"))] pub use track::{LOOKAHEAD_PACKETS, TrackDecoder}; -#[cfg(feature = "pcm-encode")] -pub use write::wav_float_header; #[cfg(any( feature = "pcm-encode", feature = "flac-encode", @@ -118,7 +119,20 @@ pub use write::wav_float_header; feature = "ac3-encode", feature = "eac3-encode" ))] -pub use write::{ENCODE_EXTENSIONS, FileEncoder, encode, encode_format_for}; +pub use write::FileEncoder; +#[cfg(feature = "pcm-encode")] +pub use write::wav_float_header; +#[cfg(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode", + feature = "vorbis-encode" +))] +pub use write::{ENCODE_EXTENSIONS, encode, encode_format_for}; #[cfg(feature = "pcm-encode")] pub use write::{pcm_aiff_file, pcm_aiff_header, pcm_wav_file, pcm_wav_header}; @@ -256,6 +270,8 @@ pub enum EncodeFormat { Ac3, /// An E-AC-3 stream: up to 7.1 at 32, 44.1 or 48 kHz. Eac3, + /// Ogg Vorbis, whole-buffer encoding only. + Ogg, } impl EncodeFormat { @@ -270,6 +286,7 @@ impl EncodeFormat { EncodeFormat::Mp3, EncodeFormat::Ac3, EncodeFormat::Eac3, + EncodeFormat::Ogg, ]; /// The extension a new file of this format gets, lower case and without the dot: `"wav"`, @@ -284,6 +301,7 @@ impl EncodeFormat { EncodeFormat::Mp3 => "mp3", EncodeFormat::Ac3 => "ac3", EncodeFormat::Eac3 => "eac3", + EncodeFormat::Ogg => "ogg", } } @@ -297,6 +315,7 @@ impl EncodeFormat { EncodeFormat::Mp3 => "audio/mpeg", EncodeFormat::Ac3 => "audio/ac3", EncodeFormat::Eac3 => "audio/eac3", + EncodeFormat::Ogg => "audio/ogg", } } @@ -309,6 +328,7 @@ impl EncodeFormat { EncodeFormat::Aac | EncodeFormat::Alac => Container::Mp4, EncodeFormat::Mp3 => Container::MpegAudio, EncodeFormat::Ac3 | EncodeFormat::Eac3 => Container::Ac3, + EncodeFormat::Ogg => Container::Ogg, } } @@ -322,6 +342,7 @@ impl EncodeFormat { EncodeFormat::Mp3 => CodecId::Mp3, EncodeFormat::Ac3 => CodecId::Ac3, EncodeFormat::Eac3 => CodecId::Eac3, + EncodeFormat::Ogg => CodecId::Vorbis, } } @@ -335,10 +356,11 @@ impl EncodeFormat { EncodeFormat::Mp3 => "mp3-encode", EncodeFormat::Ac3 => "ac3-encode", EncodeFormat::Eac3 => "eac3-encode", + EncodeFormat::Ogg => "vorbis-encode", } } - /// Whether [`encode`] and [`FileEncoder`] write this format in this build (its + /// Whether [`encode`] writes this format in this build (its /// [`feature`](Self::feature) is on): what to filter [`EncodeFormat::ALL`] by for a menu. pub fn is_built(self) -> bool { match self { @@ -349,6 +371,7 @@ impl EncodeFormat { EncodeFormat::Mp3 => cfg!(feature = "mp3-encode"), EncodeFormat::Ac3 => cfg!(feature = "ac3-encode"), EncodeFormat::Eac3 => cfg!(feature = "eac3-encode"), + EncodeFormat::Ogg => cfg!(all(feature = "vorbis-encode", not(target_arch = "wasm32"))), } } @@ -359,7 +382,7 @@ impl EncodeFormat { EncodeFormat::Aac => Some(&[7350, 8000, 11_025, 12_000, 16_000, 22_050, 24_000, 32_000, 44_100, 48_000, 64_000, 88_200, 96_000]), EncodeFormat::Mp3 => Some(&[16_000, 22_050, 24_000, 32_000, 44_100, 48_000]), EncodeFormat::Ac3 | EncodeFormat::Eac3 => Some(&[32_000, 44_100, 48_000]), - EncodeFormat::Wav | EncodeFormat::Aiff | EncodeFormat::Flac | EncodeFormat::Alac => None, + EncodeFormat::Wav | EncodeFormat::Aiff | EncodeFormat::Flac | EncodeFormat::Alac | EncodeFormat::Ogg => None, } } @@ -371,7 +394,7 @@ impl EncodeFormat { EncodeFormat::Wav | EncodeFormat::Aiff => 32_767, EncodeFormat::Flac | EncodeFormat::Aac | EncodeFormat::Alac | EncodeFormat::Eac3 => 8, EncodeFormat::Ac3 => 6, - EncodeFormat::Mp3 => 2, + EncodeFormat::Mp3 | EncodeFormat::Ogg => 2, } } @@ -385,7 +408,7 @@ impl EncodeFormat { EncodeFormat::Aac => &[&[0x4], &[0x3], &[0x7], &[0x107], &[0x607], &[0x60F], &[0x70F], &[0xFF]], // Apple's layout tags for 1–8 channels, surrounds as back speakers. EncodeFormat::Alac => &[&[0x4], &[0x3], &[0x7], &[0x107], &[0x37], &[0x3F], &[0x13F], &[0xFF]], - EncodeFormat::Mp3 => &[&[0x4], &[0x3]], + EncodeFormat::Mp3 | EncodeFormat::Ogg => &[&[0x4], &[0x3]], // The coding modes 1/0 to 3/2 with and without an LFE, by channel count. EncodeFormat::Ac3 => &[&[0x4], &[0x3, 0xC], &[0x7, 0x103, 0xB], &[0x107, 0x603, 0xF, 0x10B], &[0x607, 0x10F, 0x60B], &[0x60F]], // The same, and 7.1 with backs and sides. @@ -436,7 +459,7 @@ impl EncodeFormat { /// Whether the format loses audio: AAC, MP3, AC-3 and E-AC-3. These take /// [`EncodeOptions::rate_control`] and ignore the bit depth and dither. pub fn is_lossy(self) -> bool { - matches!(self, EncodeFormat::Aac | EncodeFormat::Mp3 | EncodeFormat::Ac3 | EncodeFormat::Eac3) + matches!(self, EncodeFormat::Aac | EncodeFormat::Mp3 | EncodeFormat::Ac3 | EncodeFormat::Eac3 | EncodeFormat::Ogg) } } @@ -449,7 +472,8 @@ pub enum RateControl { /// set of rates (MP3, AC-3) uses the highest one not above it; a rate outside what the /// stream can carry is moved to the nearest one inside. AAC is written at 6 kbit/s or a /// quarter of the sample rate per channel at least: from about half of that down its encoder - /// codes silence. + /// codes silence. Vorbis treats this as a VBR target (64–500 kbit/s), falling back to + /// a quality preset when that target is unsupported; it does not promise a constant bitrate. Bitrate(u32), /// A constant quality from 0.0 (smallest) to 1.0 (best), at whatever bit rate each frame /// needs. AAC only; values outside the range are clamped, and NaN is an error. diff --git a/crates/audio/src/write/mod.rs b/crates/audio/src/write/mod.rs index 0de55c394..ce4b90773 100644 --- a/crates/audio/src/write/mod.rs +++ b/crates/audio/src/write/mod.rs @@ -9,6 +9,10 @@ mod coded; #[cfg(feature = "flac-encode")] pub(crate) mod flac; mod pcm; +#[cfg(feature = "vorbis-encode")] +mod vorbis; +#[cfg(feature = "vorbis-encode")] +pub use vorbis::{OGG_BITRATES, encode as encode_ogg}; #[cfg(feature = "pcm-encode")] mod raw; #[cfg(feature = "pcm-encode")] @@ -17,6 +21,15 @@ mod wav; use crate::{AudioBuffer, AudioError, ChannelLayout, EncodeFormat, EncodeOptions, Result}; #[cfg(feature = "pcm-encode")] pub use raw::{pcm_aiff_file, pcm_aiff_header, pcm_wav_file, pcm_wav_header}; +#[cfg(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode" +))] use std::io::{Seek, Write}; /// Encode `buf` as the file `opts.format` names. Integer depths clamp samples to [-1, 1]; float @@ -51,6 +64,18 @@ pub fn encode(buf: &AudioBuffer, opts: &EncodeOptions) -> Result> { EncodeFormat::Flac => flac::encode(buf, opts.bit_depth, opts.dither, mask), #[cfg(any(feature = "aac-encode", feature = "alac-encode", feature = "mp3-encode", feature = "ac3-encode", feature = "eac3-encode"))] EncodeFormat::Aac | EncodeFormat::Alac | EncodeFormat::Mp3 | EncodeFormat::Ac3 | EncodeFormat::Eac3 => coded::encode(buf, opts), + #[cfg(feature = "vorbis-encode")] + EncodeFormat::Ogg => { + if opts.layout.is_some_and(|layout| !EncodeFormat::Ogg.supports_layout(&layout)) { + return Err(AudioError::Unsupported("Vorbis requires a mono or stereo speaker layout".into())); + } + let kbps = match opts.rate_control { + None => 192, + Some(crate::RateControl::Bitrate(bps)) => bps / 1000, + Some(crate::RateControl::Quality(_)) => return Err(AudioError::Unsupported("Ogg export takes a target bitrate".into())), + }; + vorbis::encode(buf, kbps) + } #[allow(unreachable_patterns)] f => Err(AudioError::Unsupported(format!("writing {f:?}"))), } @@ -62,8 +87,8 @@ fn check_built(format: EncodeFormat) -> Result<()> { } /// Every extension [`ENCODE_EXTENSIONS`] can list, and how many of them the features built give. -const fn extensions() -> ([&'static str; 12], usize) { - let mut all = [""; 12]; +const fn extensions() -> ([&'static str; 13], usize) { + let mut all = [""; 13]; let mut n = 0; macro_rules! add { ($built:expr, $($extension:literal),+) => { @@ -81,10 +106,11 @@ const fn extensions() -> ([&'static str; 12], usize) { add!(cfg!(feature = "mp3-encode"), "mp3"); add!(cfg!(feature = "ac3-encode"), "ac3"); add!(cfg!(feature = "eac3-encode"), "eac3", "ec3"); + add!(cfg!(all(feature = "vorbis-encode", not(target_arch = "wasm32"))), "ogg"); (all, n) } -const EXTENSIONS: ([&str; 12], usize) = extensions(); +const EXTENSIONS: ([&str; 13], usize) = extensions(); /// The file extensions of the formats [`encode`] can write (those whose features are on). pub const ENCODE_EXTENSIONS: &[&str] = EXTENSIONS.0.split_at(EXTENSIONS.1).0; @@ -99,7 +125,7 @@ pub const ENCODE_EXTENSIONS: &[&str] = EXTENSIONS.0.split_at(EXTENSIONS.1).0; /// use craft_audio::{EncodeFormat, encode_format_for}; /// /// assert_eq!(encode_format_for(".AIFF"), Some(EncodeFormat::Aiff)); -/// assert_eq!(encode_format_for("ogg"), None); +/// assert_eq!(encode_format_for("unknown"), None); /// # } /// ``` pub fn encode_format_for(name: &str) -> Option { @@ -122,6 +148,8 @@ pub fn encode_format_for(name: &str) -> Option { "ac3" => Some(EncodeFormat::Ac3), #[cfg(feature = "eac3-encode")] "eac3" | "ec3" => Some(EncodeFormat::Eac3), + #[cfg(all(feature = "vorbis-encode", not(target_arch = "wasm32")))] + "ogg" => Some(EncodeFormat::Ogg), _ => None, } } @@ -138,6 +166,15 @@ fn wav_mask(layout: &ChannelLayout, channels: usize) -> Result { /// The error a stream encoder returns for every call after one of its writes has failed: the /// sink may already hold a prefix of the failed block, so the encoder's counters can no longer /// describe the bytes on the sink, and later writes or `finish` must not claim success. +#[cfg(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode" +))] fn failed_write_error() -> AudioError { AudioError::Encode("the stream encoder already failed on an earlier write; its output is incomplete".into()) } @@ -163,12 +200,30 @@ fn failed_write_error() -> AudioError { /// assert_eq!(&flac[..4], b"fLaC"); /// # } /// ``` +#[cfg(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode" +))] pub struct FileEncoder { writer: Writer, format: EncodeFormat, channels: usize, } +#[cfg(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode" +))] impl std::fmt::Debug for FileEncoder { fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result { f.debug_struct("FileEncoder").field("format", &self.format).finish_non_exhaustive() @@ -177,6 +232,15 @@ impl std::fmt::Debug for FileEncoder { // The writers differ in size with the features built; the coded one is boxed for its encoder. #[allow(clippy::large_enum_variant)] +#[cfg(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode" +))] enum Writer { #[cfg(feature = "pcm-encode")] Wav(wav::Writer), @@ -188,6 +252,15 @@ enum Writer { Coded(Box>), } +#[cfg(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode" +))] impl FileEncoder { /// Starts a file of `channels` channels at `sample_rate` at the current position of `out`. /// `opts.layout`, when set, must have `channels` channels. @@ -317,9 +390,39 @@ impl EncodeOptions { /// The constant bit rate such a file would be written at ([`FileEncoder::bitrate`]), or the /// error [`EncodeOptions::check`] returns. pub fn bitrate(&self, sample_rate: u32, channels: u16) -> Result> { + #[cfg(feature = "vorbis-encode")] + if self.format == EncodeFormat::Ogg { + // Vorbis is whole-buffer only and VBR has no constant output bitrate. + encode(&AudioBuffer::new(sample_rate, channels, 0), self)?; + return Ok(None); + } // The encoder a file would get, on a sink that is thrown away: what it accepts and the // rate it settles on are then the writer's by construction. - Ok(FileEncoder::new(std::io::Cursor::new(Vec::new()), self, sample_rate, channels)?.bitrate()) + #[cfg(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode" + ))] + { + Ok(FileEncoder::new(std::io::Cursor::new(Vec::new()), self, sample_rate, channels)?.bitrate()) + } + #[cfg(not(any( + feature = "pcm-encode", + feature = "flac-encode", + feature = "aac-encode", + feature = "alac-encode", + feature = "mp3-encode", + feature = "ac3-encode", + feature = "eac3-encode" + )))] + { + let _ = (sample_rate, channels); + Err(AudioError::Unsupported("streaming encoders are not built".into())) + } } } diff --git a/crates/audio/src/write/vorbis.rs b/crates/audio/src/write/vorbis.rs new file mode 100644 index 000000000..f56d01148 --- /dev/null +++ b/crates/audio/src/write/vorbis.rs @@ -0,0 +1,130 @@ +//! Ogg Vorbis encoding (RFC 5334): the reference libvorbis encoder, built from its C sources by +//! `vorbis_rs`. Stereo or mono, variable bitrate at the requested kbit/s (64–500), with a +//! fallback to a quality preset when the bitrate is not supported for the rate and channel count. + +use crate::{AudioBuffer, AudioError, Result}; + +/// The bitrates an Ogg Vorbis export is asked for, in kbit/s. +pub const OGG_BITRATES: (u32, u32) = (64, 500); + +/// The block size the encoder processes at a time: small enough that memory stays bounded +/// regardless of file length. +#[cfg(not(target_arch = "wasm32"))] +const BLOCK: usize = 4096; + +#[cfg(not(target_arch = "wasm32"))] +fn encode_error(e: vorbis_rs::VorbisError) -> AudioError { + AudioError::Encode(e.to_string()) +} + +/// Encode `buf` as an Ogg Vorbis stream at about `kbps` (variable, as Vorbis is): stereo, or mono +/// for a mono buffer. A bitrate the encoder has no setup for at that rate and channel count falls +/// back to a quality that suits it. `Err` says why there is nothing. +#[cfg(not(target_arch = "wasm32"))] +pub fn encode(buf: &AudioBuffer, kbps: u32) -> Result> { + use std::num::{NonZeroU8, NonZeroU32}; + + let channels = buf.channels.len(); + if !(1..=2).contains(&channels) { + return Err(AudioError::Unsupported("Ogg Vorbis export requires mono or stereo audio".into())); + } + if buf.channels.iter().any(|c| c.len() != buf.frames() || c.iter().any(|s| !s.is_finite())) { + return Err(AudioError::Encode("Vorbis samples must be finite and channels equally long".into())); + } + let (Some(rate), Some(count)) = (NonZeroU32::new(buf.sample_rate), NonZeroU8::new(channels as u8)) else { + return Err(AudioError::Encode("there is no audio to encode".into())); + }; + let target = NonZeroU32::new(kbps.clamp(OGG_BITRATES.0, OGG_BITRATES.1) * 1000).unwrap_or(NonZeroU32::MIN); + + let encode = |strategy: vorbis_rs::VorbisBitrateManagementStrategy| -> Result> { + let mut encoder = vorbis_rs::VorbisEncoderBuilder::new_with_serial(rate, count, Vec::new(), 0x4C43_5246) + .bitrate_management_strategy(strategy) + .build() + .map_err(encode_error)?; + let frames = buf.frames(); + let mut at = 0usize; + while at < frames { + let end = (at + BLOCK).min(frames); + let block: Vec<&[f32]> = buf.channels.iter().take(channels).map(|side| side.get(at..end).unwrap_or_default()).collect(); + encoder.encode_audio_block(&block).map_err(encode_error)?; + at = end; + } + encoder.finish().map_err(encode_error) + }; + + // Convert Rust unwinds to an error. This does not catch native crashes or aborts. + let attempt = |strategy| { + std::panic::catch_unwind(std::panic::AssertUnwindSafe(|| encode(strategy))) + .unwrap_or_else(|_| Err(AudioError::Encode("the Vorbis encoder failed".into()))) + }; + match attempt(vorbis_rs::VorbisBitrateManagementStrategy::Vbr { target_bitrate: target }) { + Ok(bytes) => Ok(bytes), + Err(_) => attempt(vorbis_rs::VorbisBitrateManagementStrategy::QualityVbr { target_quality: 0.6 }) + .map_err(|e| AudioError::Encode(format!("the Vorbis encoder refused the audio: {e}"))), + } +} + +/// Ogg Vorbis encoding is unavailable on WASM; returns an unsupported-format error. +#[cfg(target_arch = "wasm32")] +pub fn encode(_buf: &AudioBuffer, _kbps: u32) -> Result> { + Err(AudioError::Unsupported("Ogg Vorbis files are exported by the desktop app".into())) +} + +#[cfg(all(test, not(target_arch = "wasm32")))] +mod tests { + #![allow(clippy::unwrap_used)] + use super::*; + + #[test] + fn rejects_unsupported_or_malformed_channels_without_dropping_audio() { + for channels in [0, 3, 6] { + assert!(matches!(encode(&AudioBuffer::new(48_000, channels, 16), 192), Err(AudioError::Unsupported(_)))); + } + let ragged = AudioBuffer { sample_rate: 48_000, channels: vec![vec![0.0; 16], vec![0.0; 15]] }; + assert!(encode(&ragged, 192).is_err()); + assert!(encode(&AudioBuffer::new(0, 1, 16), 192).is_err()); + let invalid = AudioBuffer { sample_rate: 48_000, channels: vec![vec![f32::NAN]] }; + assert!(encode(&invalid, 192).is_err()); + } + + #[test] + fn generic_and_direct_encoders_write_mono_and_stereo_ogg() { + for channels in [1, 2] { + let audio = AudioBuffer::new(48_000, channels, 512); + let opts = crate::EncodeOptions::new(crate::EncodeFormat::Ogg); + assert!(opts.check(48_000, channels).is_ok()); + assert_eq!(opts.bitrate(48_000, channels).unwrap(), None); + let direct = encode(&audio, 192).unwrap(); + let generic = crate::encode(&audio, &opts).unwrap(); + assert!(direct.starts_with(b"OggS")); + assert_eq!(direct, generic); + #[cfg(feature = "decode")] + { + let (info, decoded) = crate::decode(&generic, Some("ogg")).unwrap(); + assert_eq!(info.channels, channels); + assert_eq!(decoded.num_channels(), channels); + assert_eq!(decoded.frames(), audio.frames()); + } + } + } + #[cfg(feature = "decode")] + #[test] + fn distinct_signals_survive_full_and_partial_blocks() { + for frames in [4096, 4097, 8193] { + let channels: Vec> = [440.0f32, 880.0] + .into_iter() + .map(|hz| (0..frames).map(|i| (std::f32::consts::TAU * hz * i as f32 / 48_000.0).sin() * 0.4).collect()) + .collect(); + let audio = AudioBuffer { sample_rate: 48_000, channels }; + let bytes = encode(&audio, 192).unwrap(); + let (_, decoded) = crate::decode(&bytes, Some("ogg")).unwrap(); + assert_eq!(decoded.frames(), frames); + assert_eq!(decoded.channels.len(), 2); + for (expected, actual) in audio.channels.iter().zip(&decoded.channels) { + let mse = expected.iter().zip(actual).map(|(a, b)| (a - b).powi(2)).sum::() / frames as f32; + // Vorbis is lossy; this tolerates codec error but rejects silence or swapped channels. + assert!(mse.sqrt() < 0.08, "{frames} frames: RMS error {}", mse.sqrt()); + } + } + } +} diff --git a/crates/audio/tests/audio/decode_extensions.rs b/crates/audio/tests/audio/decode_extensions.rs index 27a5573de..fc961cc39 100644 --- a/crates/audio/tests/audio/decode_extensions.rs +++ b/crates/audio/tests/audio/decode_extensions.rs @@ -77,6 +77,6 @@ fn every_encode_format_has_an_extension_it_is_written_and_read_under() { assert!(format.mime_type().starts_with("audio/"), "{format:?}"); } let extensions: Vec<_> = EncodeFormat::ALL.iter().map(|f| f.extension()).collect(); - assert_eq!(extensions, ["wav", "aif", "flac", "m4a", "m4a", "mp3", "ac3", "eac3"]); + assert_eq!(extensions, ["wav", "aif", "flac", "m4a", "m4a", "mp3", "ac3", "eac3", "ogg"]); assert_eq!(EncodeFormat::Aiff.mime_type(), "audio/aiff"); } diff --git a/crates/audio/tests/audio/encode_format.rs b/crates/audio/tests/audio/encode_format.rs index 7713f37a5..e91f93623 100644 --- a/crates/audio/tests/audio/encode_format.rs +++ b/crates/audio/tests/audio/encode_format.rs @@ -14,7 +14,15 @@ fn only_formats_the_encoder_writes_are_named() { for name in craft_audio::ENCODE_EXTENSIONS { assert!(encode_format_for(name).is_some(), "{name}"); } - for other in ["ogg", "opus", "aac", "mp4", "xyz", ""] { + for other in ["opus", "aac", "mp4", "xyz", ""] { assert_eq!(encode_format_for(other), None, "{other}"); } } + +#[test] +fn ogg_names_follow_the_native_feature() { + let built = cfg!(all(feature = "vorbis-encode", not(target_arch = "wasm32"))); + assert_eq!(EncodeFormat::Ogg.is_built(), built); + assert_eq!(encode_format_for("ogg"), built.then_some(EncodeFormat::Ogg)); + assert_eq!(craft_audio::ENCODE_EXTENSIONS.contains(&"ogg"), built); +} diff --git a/crates/audio/tests/codecs/file_formats.rs b/crates/audio/tests/codecs/file_formats.rs index 53947df12..84169f665 100644 --- a/crates/audio/tests/codecs/file_formats.rs +++ b/crates/audio/tests/codecs/file_formats.rs @@ -397,12 +397,15 @@ fn the_formats_are_named() { for name in ENCODE_EXTENSIONS { assert!(encode_format_for(name).is_some(), "{name}"); } - assert_eq!(ENCODE_EXTENSIONS.len(), 12); + assert_eq!(ENCODE_EXTENSIONS.len(), 13); for format in CODED { assert!(EncodeFormat::ALL.contains(&format) && format.is_built()); } let features: Vec<_> = EncodeFormat::ALL.iter().map(|f| f.feature()).collect(); - assert_eq!(features, ["pcm-encode", "pcm-encode", "flac-encode", "alac-encode", "aac-encode", "mp3-encode", "ac3-encode", "eac3-encode"]); + assert_eq!( + features, + ["pcm-encode", "pcm-encode", "flac-encode", "alac-encode", "aac-encode", "mp3-encode", "ac3-encode", "eac3-encode", "vorbis-encode"] + ); } /// ffmpeg's decode of `file` as interleaved `f32`, split into `channels` channels. @@ -495,6 +498,10 @@ fn the_format_queries_agree_with_the_writer() { let layouts = masks.iter().map(|&mask| ChannelLayout::from_mask(mask)).chain((1..=9).map(ChannelLayout::unordered)); for layout in layouts { for &format in EncodeFormat::ALL { + // Vorbis is whole-buffer-only; its direct/generic API and options are tested separately. + if format == EncodeFormat::Ogg { + continue; + } // 24 bits is a depth every format has. let opts = EncodeOptions::new(format).with_layout(layout); for &sample_rate in &rates { @@ -520,7 +527,7 @@ fn the_format_queries_agree_with_the_writer() { // The most channels is the most: one more is refused, by the layouts of a format that has // a fixed set and by the writer of one that has not. let most: Vec = EncodeFormat::ALL.iter().map(|f| f.max_channels()).collect(); - assert_eq!(most, [32_767, 32_767, 8, 8, 8, 2, 6, 8]); + assert_eq!(most, [32_767, 32_767, 8, 8, 8, 2, 6, 8, 2]); for &format in EncodeFormat::ALL { let most = format.max_channels(); match format.fixed_layouts(most) { @@ -569,3 +576,16 @@ fn the_bit_rate_used_can_be_asked_for() { assert_eq!(encode(&buf, &opts).unwrap().len() as u32, 43 * 1152 / 8 * opts.bitrate(48_000, 2).unwrap().unwrap() / 48_000); assert!(is_unsupported(EncodeOptions::new(EncodeFormat::Mp3).bitrate(48_000, 6))); } + +#[test] +fn vorbis_is_whole_buffer_only_and_validates_layouts() { + let opts = EncodeOptions::new(EncodeFormat::Ogg).with_layout(ChannelLayout::STEREO); + assert!(opts.check(48_000, 2).is_ok()); + assert_eq!(opts.bitrate(48_000, 2).unwrap(), None); + assert!(FileEncoder::new(Cursor::new(Vec::new()), &opts, 48_000, 2).is_err()); + let buf = craft_audio::AudioBuffer::new(48_000, 2, 512); + assert!(craft_audio::encode(&buf, &opts).unwrap().starts_with(b"OggS")); + let wrong = opts.with_layout(ChannelLayout::from_mask(0xC)); + assert!(wrong.check(48_000, 2).is_err()); + assert!(craft_audio::encode(&buf, &wrong).is_err()); +} diff --git a/crates/docpath/tests/long_path.rs b/crates/docpath/tests/long_path.rs new file mode 100644 index 000000000..2ecf08ba1 --- /dev/null +++ b/crates/docpath/tests/long_path.rs @@ -0,0 +1,29 @@ +//! Paths long enough that a per-level recursive descent would exhaust the stack. Parsing is +//! iterative — `resolve` and `classify` reach `parse_share` at a fixed depth — so these come +//! back; a refusal is fine, a crash is not. + +use craft_docpath::{classify, resolve}; + +/// Half a megabyte of components under a UNC share: one component per level for a parser that +/// recursed on each separator. +#[test] +fn a_half_megabyte_unc_path_comes_back() { + let mut path = String::from(r"\\?\UNC\host\share"); + for _ in 0..(1 << 18) { + path.push_str(r"\d"); + } + let _ = resolve(&path, None); + let _ = classify(&path); +} + +/// Sixteen thousand nested verbatim prefixes: each level was a candidate recursive call when the +/// UNC prefixes were first reviewed. +#[test] +fn nested_verbatim_prefixes_come_back() { + let mut path = String::new(); + for _ in 0..(1 << 14) { + path.push_str(r"\\?\UNC\"); + } + let _ = resolve(&path, None); + let _ = classify(&path); +} diff --git a/crates/dwg/README.md b/crates/dwg/README.md index ed52bdcfc..80088a89e 100644 --- a/crates/dwg/README.md +++ b/crates/dwg/README.md @@ -11,7 +11,9 @@ allocation failure inside the decoder. On native applications, call `use_own_wor startup and call `serve_if_asked()` before normal argument handling. Subsequent `decode` and `encode_*` calls then run in a child process with a 30-minute timeout; a crash or stack overflow is contained to that child. The worker has no operating-system memory limit yet. Without a -configured worker, conversions run in the current process. +configured worker, conversions are refused rather than run in the current process: call +`allow_in_process()` only for input that is known to be trusted (tests, or bytes this process +produced itself). Coordinates, once a worker is configured but cannot start, also fail closed. The bridge is not built for `wasm32` or FreeBSD/aarch64. Those targets retain the public API and return a clear unsupported-build error. The `acadrust` 0.6.6 dependency is used unmodified under diff --git a/crates/dwg/src/lib.rs b/crates/dwg/src/lib.rs index 2db7afeae..05c1a673c 100644 --- a/crates/dwg/src/lib.rs +++ b/crates/dwg/src/lib.rs @@ -21,6 +21,11 @@ //! These bound the input- and output-size variants. A small file that declares a huge internal //! table or section can still make the dependency allocate before any size is known, so the app //! runs both directions in a child process ([`worker`]): what that kills is the child. +//! +//! Isolation is not optional for untrusted input. With no worker configured the conversion is +//! refused rather than run here, and a worker that cannot start is an error, so a missing worker +//! never turns into decoding a hostile file in the caller's process. [`worker::allow_in_process`] +//! is the explicit exception, for input the caller knows is trusted. #![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)] #![forbid(unsafe_code)] @@ -65,7 +70,10 @@ fn mb(n: u64) -> String { #[cfg(not(no_dwg_bridge))] pub mod worker; #[cfg(not(no_dwg_bridge))] -pub use worker::{WORKER_ARG, crash_worker_for_test, decode, encode_as, encode_kept_noted, encode_noted, serve_if_asked, use_own_worker, use_worker}; +pub use worker::{ + WORKER_ARG, allow_in_process, crash_worker_for_test, decode, encode_as, encode_kept_noted, encode_noted, serve_if_asked, use_own_worker, + use_worker, +}; #[cfg(not(no_dwg_bridge))] mod proxy_transport; @@ -1569,6 +1577,8 @@ pub fn serve_if_asked() -> Option { #[cfg(no_dwg_bridge)] pub fn use_own_worker() {} #[cfg(no_dwg_bridge)] +pub fn allow_in_process() {} +#[cfg(no_dwg_bridge)] pub fn decode(bytes: &[u8], limits: Limits) -> Result<(Vec, Vec), String> { dwg_to_dxf_noted(bytes, limits) } diff --git a/crates/dwg/src/worker.rs b/crates/dwg/src/worker.rs index 0c273257b..1d16f5701 100644 --- a/crates/dwg/src/worker.rs +++ b/crates/dwg/src/worker.rs @@ -6,9 +6,11 @@ //! that reads one request from standard input and writes one reply to standard output. A file that //! kills the worker then costs that one open or save, and every other open drawing stays safe. //! -//! Without a worker (libraries, tests, a build that can't start one) [`decode`] and [`encode_as`] -//! run in the calling process, as they always did. There's no memory limit on the worker yet: that -//! takes per-platform system calls, which `unsafe_code = "forbid"` keeps out of this crate. +//! Without a worker [`decode`] and [`encode_as`] refuse rather than run the conversion in the +//! calling process, because the third-party code is the part that can end a process. +//! [`allow_in_process`] is the explicit exception, for input the caller knows is trusted (tests, or +//! bytes this process produced itself). There's no memory limit on the worker yet: that takes +//! per-platform system calls, which `unsafe_code = "forbid"` keeps out of this crate. //! //! A reply comes from a process that just handled a possibly hostile file, so it is parsed as //! untrusted input. @@ -18,6 +20,7 @@ use std::io::{Read, Write}; use std::path::PathBuf; use std::process::{Command, Stdio}; use std::sync::Mutex; +use std::sync::atomic::{AtomicBool, Ordering}; use std::sync::mpsc::{RecvTimeoutError, channel}; use std::time::Duration; @@ -38,6 +41,9 @@ const MAX_NOTES_BYTES: u64 = 64 << 20; const MAX_DWG_REPLY: u64 = 2 << 30; /// Most bytes one request holds: the largest input, whichever the direction, and its header. const MAX_REQUEST: u64 = (2 << 30) + 256; +/// What a conversion says when it has no worker and is not allowed to run here. +const NO_WORKER: &str = "DWG: no worker is configured, so this file was not converted; \ + call use_worker or use_own_worker at startup, or allow_in_process for input that is trusted"; struct Worker { program: PathBuf, @@ -45,6 +51,8 @@ struct Worker { } static WORKER: Mutex> = Mutex::new(None); +/// Whether this process may run the conversion itself, set only by [`allow_in_process`]. +static IN_PROCESS: AtomicBool = AtomicBool::new(false); /// Start `program args` as the worker from now on. `program` must call [`serve_if_asked`]. pub fn use_worker(program: PathBuf, args: Vec) { @@ -62,12 +70,27 @@ pub fn use_own_worker() { } } +/// Let [`decode`] and [`encode_as`] run the conversion in this process when no worker is set. +/// +/// This opts out of the isolation the crate exists for: the third-party code allocates from sizes +/// a hostile file declares, and a failed allocation ends the process. Call it for input the caller +/// knows is trusted (tests, or bytes this process produced), never for a file that arrived from +/// outside. +pub fn allow_in_process() { + IN_PROCESS.store(true, Ordering::Relaxed); +} + +fn in_process_allowed() -> bool { + IN_PROCESS.load(Ordering::Relaxed) +} + /// Serve one request and return the exit code if this process was started as a worker. pub fn serve_if_asked() -> Option { (std::env::args_os().nth(1)? == WORKER_ARG).then(|| serve(std::io::stdin().lock(), std::io::stdout().lock())) } -/// [`super::dwg_to_dxf_noted`] in the worker, when there is one. +/// [`super::dwg_to_dxf_noted`] in the worker. With no worker configured this is refused unless +/// [`allow_in_process`] was called; a configured worker that cannot start is an error. pub fn decode(bytes: &[u8], limits: Limits) -> Result<(Vec, Vec), String> { // Before starting a process, and so the same limit holds with or without one. if bytes.len() > limits.max_dwg_bytes { @@ -76,17 +99,19 @@ pub fn decode(bytes: &[u8], limits: Limits) -> Result<(Vec, Vec), St let reply_limit = (limits.max_dxf_bytes as u64).saturating_add(MAX_NOTES_BYTES); match call(&request(DECODE, limits, ""), bytes, reply_limit, "reading this file") { Some(reply) => reply?, - None => super::dwg_to_dxf_noted(bytes, limits), + None if in_process_allowed() => super::dwg_to_dxf_noted(bytes, limits), + None => Err(NO_WORKER.to_string()), } } -/// [`super::dxf_to_dwg_as`] in the worker, when there is one. +/// [`super::dxf_to_dwg_as`] in the worker. Refused without a worker unless [`allow_in_process`] +/// was called. pub fn encode_as(dxf: &[u8], version: &str) -> Result, String> { encode_noted(dxf, version).map(|(dwg, _)| dwg) } -/// [`super::dxf_to_dwg_noted`] in the worker, when there is one. What the writer left out comes -/// back as notes of `TYPEcount`. +/// [`super::dxf_to_dwg_noted`] in the worker. What the writer left out comes back as notes of +/// `TYPEcount`. pub fn encode_noted(dxf: &[u8], version: &str) -> Result<(Vec, super::Unwritten), String> { encode_kept_noted(dxf, version, None) } @@ -103,7 +128,8 @@ pub fn encode_kept_noted(dxf: &[u8], version: &str, kept: Option<&str>) -> Resul push_str(&mut header, kept.unwrap_or("")); match call(&header, dxf, MAX_DWG_REPLY, "writing this file") { Some(reply) => reply?.map(|(dwg, notes)| (dwg, unwritten_from(¬es))), - None => encode_here(dxf, version, kept), + None if in_process_allowed() => encode_here(dxf, version, kept), + None => Err(NO_WORKER.to_string()), } } @@ -147,8 +173,9 @@ fn push_str(r: &mut Vec, text: &str) { type Reply = Result<(Vec, Vec), String>; -/// Run one request in a worker. `None`: none could be started, so the caller does the work. -/// `Err`: the worker went wrong, and `doing` says at what, for the user. +/// Run one request in a worker. `None`: no worker is configured, so the caller decides what to do +/// (convert here only with [`allow_in_process`], or refuse). `Err`: a configured worker could not +/// start or went wrong, and `doing` says at what, for the user. fn call(header: &[u8], payload: &[u8], reply_limit: u64, doing: &str) -> Option> { let mut command = { let worker = WORKER.lock().ok()?; @@ -167,14 +194,15 @@ fn call(header: &[u8], payload: &[u8], reply_limit: u64, doing: &str) -> Option< let mut child = match command.spawn() { Ok(c) => c, Err(e) => { + // A worker is configured but cannot run: never fall back to decoding here. log::warn!("DWG worker: can't start it: {e}"); - return None; + return Some(Err(format!("the DWG worker could not start ({e}); nothing else is affected"))); } }; let (Some(mut to_child), Some(from_child)) = (child.stdin.take(), child.stdout.take()) else { let _ = child.kill(); let _ = child.wait(); - return None; + return Some(Err("the DWG worker could not start (no pipes); nothing else is affected".to_string())); }; let received = std::thread::scope(|scope| { // Written beside the read: a worker that replies before it has read everything, or dies, @@ -367,6 +395,25 @@ mod tests { } } + /// A conversion is refused rather than run here, and a worker that cannot start is an error: + /// neither case may turn into decoding a (possibly hostile) file in this process. + #[test] + fn a_conversion_is_refused_without_a_worker_and_never_falls_back_to_this_process() { + // No worker is configured and nothing opted in. + assert_eq!(decode(b"AC1032", Limits::DEFAULT).unwrap_err(), NO_WORKER); + assert_eq!(encode_as(b"not a dxf", "AC1032").unwrap_err(), NO_WORKER); + // Trusted input may run here, and the reader still reports a file it cannot read. + allow_in_process(); + assert!(in_process_allowed()); + assert!(decode(b"AC1032", Limits::DEFAULT).is_err()); + assert!(encode_as(b"not a dxf", "AC1032").is_err()); + // A configured worker that cannot start is an error even after the opt-in: the conversion + // never runs here instead. + use_worker(PathBuf::from("/nonexistent/craft-dwg-worker"), Vec::new()); + assert!(decode(b"AC1032", Limits::DEFAULT).unwrap_err().contains("could not start")); + assert!(encode_as(b"not a dxf", "AC1032").unwrap_err().contains("could not start")); + } + #[test] fn serving_reports_a_decode_failure_as_a_reply() { let mut request = self::request(DECODE, Limits::DEFAULT, ""); diff --git a/crates/gif/src/write.rs b/crates/gif/src/write.rs index bc8a4be97..259aa78f1 100644 --- a/crates/gif/src/write.rs +++ b/crates/gif/src/write.rs @@ -194,6 +194,45 @@ pub fn animation(width: u32, height: u32, frames: &[RgbaFrame<'_>], options: Opt writer.finish() } +/// Encode a single pre-quantized indexed-color frame as a GIF89a image. +/// +/// `palette` holds 1 to 256 RGB triplets; `indices` is `width * height` row-major palette indices. +/// When `interlaced` is true, rows are reordered into GIF's 4-pass interlaced order before writing. +pub fn encode_indexed( + width: u32, + height: u32, + palette: &[[u8; 3]], + indices: &[u8], + transparent: Option, + interlaced: bool, +) -> Result, Error> { + let (w, h) = dimensions(width, height)?; + let expected = usize::from(w) * usize::from(h); + if indices.len() != expected { + return Err(Error::FrameLength { index: 0, actual: indices.len(), expected }); + } + if palette.is_empty() || palette.len() > 256 { + return Err(Error::Encode("GIF palette must hold 1 to 256 entries".into())); + } + let flat_palette: Vec = palette.iter().flat_map(|c| *c).collect(); + let buffer = if interlaced { std::borrow::Cow::Owned(interlace(indices, usize::from(w))) } else { std::borrow::Cow::Borrowed(indices) }; + let mut enc = gif::Encoder::new(Vec::new(), w, h, &flat_palette).map_err(Error::from)?; + let frame = gif::Frame { width: w, height: h, buffer, transparent, interlaced, ..gif::Frame::default() }; + enc.write_frame(&frame).map_err(Error::from)?; + enc.into_inner().map_err(Error::from) +} + +/// The rows of a `width`-wide image in GIF's interlaced order: every 8th row from 0, every 8th +/// from 4, every 4th from 2, then every 2nd from 1. +fn interlace(px: &[u8], width: usize) -> Vec { + let rows: Vec<&[u8]> = px.chunks(width.max(1)).collect(); + [(0, 8), (4, 8), (2, 4), (1, 2)] + .into_iter() + .flat_map(|(start, step)| rows.iter().skip(start).step_by(step)) + .flat_map(|r| r.iter().copied()) + .collect() +} + #[cfg(test)] mod tests { #![allow(clippy::unwrap_used)] @@ -288,4 +327,17 @@ mod tests { assert_eq!(decoded.frames[0].rgba, red); assert_eq!(decoded.frames[1].rgba, blue); } + + #[test] + fn encode_indexed_validates_and_writes_gif() { + let palette = [[255, 0, 0], [0, 255, 0]]; + let indices = [0u8, 1, 1, 0, 0, 1, 1, 0, 0]; + let bytes = super::encode_indexed(3, 3, &palette, &indices, Some(1), true).unwrap(); + let info = crate::scan(&bytes, 4).unwrap(); + assert_eq!((info.width, info.height), (3, 3)); + assert_eq!(info.frames.len(), 1); + assert_eq!(info.frames[0].transparent, Some(1)); + assert!(super::encode_indexed(2, 2, &[], &[0; 4], None, false).is_err()); + assert!(super::encode_indexed(2, 2, &palette, &[0; 3], None, false).is_err()); + } } diff --git a/crates/image-codecs/Cargo.toml b/crates/image-codecs/Cargo.toml index b893c347a..2dc9426ac 100644 --- a/crates/image-codecs/Cargo.toml +++ b/crates/image-codecs/Cargo.toml @@ -13,7 +13,7 @@ workspace = true [features] corpus = [] default = [] -avif = ["image/avif"] +avif = ["image/avif", "dep:ravif", "dep:rav1e", "dep:avif-serialize"] heif = ["dep:craft-heif"] [dependencies] @@ -33,6 +33,10 @@ craft-jpeg = { workspace = true } craft-metadata = { workspace = true } zune-jpegxl = { version = "=0.5.2", default-features = false, features = ["std"] } craft-heif = { workspace = true, optional = true } +ravif = { version = "0.13", default-features = false, optional = true } +rav1e = { version = "0.8", default-features = false, optional = true } +avif-serialize = { version = "0.8", optional = true } +hayro-jpeg2000 = { version = "=0.3.5", default-features = false, features = ["std", "simd"] } [target.'cfg(not(target_arch = "wasm32"))'.dependencies] rayon = { workspace = true } diff --git a/crates/image-codecs/src/codecs.rs b/crates/image-codecs/src/codecs.rs index 4c4c22388..0288ae22b 100644 --- a/crates/image-codecs/src/codecs.rs +++ b/crates/image-codecs/src/codecs.rs @@ -1,3 +1,4 @@ +pub(crate) mod avif; pub(crate) mod dds; pub(crate) mod dds_bc; pub(crate) mod deep_exr; @@ -10,6 +11,7 @@ pub(crate) mod gif_frames; pub(crate) mod gif_palette; pub(crate) mod heif; pub(crate) mod iff; +pub(crate) mod jp2; pub(crate) mod jpeg; pub(crate) mod jxl; pub(crate) mod pcx; diff --git a/crates/image-codecs/src/codecs/avif.rs b/crates/image-codecs/src/codecs/avif.rs new file mode 100644 index 000000000..d27112fa1 --- /dev/null +++ b/crates/image-codecs/src/codecs/avif.rs @@ -0,0 +1,211 @@ +//! AVIF encoding helpers (8/16-bit SDR via `ravif` and 10-bit 4:4:4 BT.2020 HDR via `rav1e` + `avif-serialize`). + +use crate::{CodecError, Format, Image}; + +/// Transfer function for [`encode_avif_hdr`]. +#[derive(Clone, Copy, Debug, PartialEq, Eq)] +pub enum HdrTransfer { + Pq, + Hlg, +} + +/// Absolute luminance of SDR white in an HDR file (ITU-R BT.2408 reference white), cd/m^2. +pub const HDR_REFERENCE_WHITE_NITS: f32 = 203.0; + +#[cfg(all(feature = "avif", not(target_arch = "wasm32")))] +fn pq_encode(nits: f32) -> f32 { + const M1: f32 = 2610.0 / 16384.0; + const M2: f32 = 2523.0 / 4096.0 * 128.0; + const C1: f32 = 3424.0 / 4096.0; + const C2: f32 = 2413.0 / 4096.0 * 32.0; + const C3: f32 = 2392.0 / 4096.0 * 32.0; + let y = (nits / 10_000.0).clamp(0.0, 1.0).powf(M1); + ((C1 + C2 * y) / (1.0 + C3 * y)).powf(M2) +} + +#[cfg(all(feature = "avif", not(target_arch = "wasm32")))] +fn hlg_signal(rgb: [f32; 3]) -> [f32; 3] { + const A: f32 = 0.17883277; + const B: f32 = 0.28466892; + const C: f32 = 0.559_910_7; + rgb.map(|v| { + let s = if v.is_finite() { v.clamp(0.0, 12.0) / 12.0 } else { 0.0 }; + if s <= 1.0 / 12.0 { (3.0 * s).sqrt() } else { A * (12.0 * s - B).ln() + C } + }) +} + +/// Encode an 8-bit or 16-bit SDR `Image` as AVIF (`4:4:4`, 10-bit AV1) with `quality` (`1..=100`) +/// and `speed` (`1..=10`). +pub fn encode_avif_with_speed(img: &Image, quality: u8, speed: u8) -> Result, CodecError> { + #[cfg(all(feature = "avif", not(target_arch = "wasm32")))] + { + let mut enc = + ravif::Encoder::new().with_quality(f32::from(quality.clamp(1, 100))).with_speed(speed.clamp(1, 10)).with_bit_depth(ravif::BitDepth::Ten); + let alpha_mode = if img.layout().has_alpha() { ravif::AlphaColorMode::UnassociatedClean } else { ravif::AlphaColorMode::UnassociatedDirty }; + enc = enc.with_alpha_color_mode(alpha_mode); + let w = img.width() as usize; + let h = img.height() as usize; + if let Some(s16) = img.to_u16_samples() { + let ch = img.layout().channels(); + let mut planes = Vec::with_capacity(w * h); + let mut a_plane = if img.layout().has_alpha() { Some(Vec::with_capacity(w * h)) } else { None }; + for px in s16.chunks_exact(ch) { + let (r, g, b, a) = match ch { + 1 => (px[0], px[0], px[0], u16::MAX), + 2 => (px[0], px[0], px[0], px[1]), + 3 => (px[0], px[1], px[2], u16::MAX), + _ => (px[0], px[1], px[2], px[3]), + }; + let (rf, gf, bf) = (f32::from(r) / 65535.0, f32::from(g) / 65535.0, f32::from(b) / 65535.0); + let y = 0.299 * rf + 0.587 * gf + 0.114 * bf; + let cb = (bf - y) * (0.5 / (1.0 - 0.114)) + 0.5; + let cr = (rf - y) * (0.5 / (1.0 - 0.299)) + 0.5; + let q10 = |v: f32| (v * 1023.0).round().clamp(0.0, 1023.0) as u16; + planes.push([q10(y), q10(cb), q10(cr)]); + if let Some(ap) = &mut a_plane { + ap.push(q10(f32::from(a) / 65535.0)); + } + } + let alpha = a_plane.map(Vec::into_iter); + let res = enc + .encode_raw_planes_10_bit(w, h, planes, alpha, ravif::PixelRange::Full, ravif::MatrixCoefficients::BT601) + .map_err(|e| CodecError::encode(Format::Avif, e))?; + return Ok(res.avif_file); + } + let rgba = img.to_rgba8(); + let px: Vec = rgba.as_chunks::<4>().0.iter().map(|c| ravif::RGBA8::new(c[0], c[1], c[2], c[3])).collect(); + let res = enc.encode_rgba(ravif::Img::new(&px[..], w, h)).map_err(|e| CodecError::encode(Format::Avif, e))?; + Ok(res.avif_file) + } + #[cfg(not(all(feature = "avif", not(target_arch = "wasm32"))))] + { + let _ = (img, quality, speed); + Err(CodecError::unsupported(Format::Avif, "AVIF encoding is not available in this build")) + } +} + +/// Encode linear BT.2020 HDR `rgb` (`width * height * 3` floats, `1.0` = `203` cd/m^2) as +/// 10-bit 4:4:4 BT.2020 PQ or HLG AVIF. +pub fn encode_avif_hdr( + width: u32, + height: u32, + rgb: &[f32], + quality: u8, + speed: u8, + exif: Option<&[u8]>, + transfer: HdrTransfer, +) -> Result, CodecError> { + let (w, h) = (width as usize, height as usize); + let n = w.checked_mul(h).ok_or_else(|| CodecError::encode(Format::Avif, "image too large"))?; + if n == 0 || n.checked_mul(3).is_none_or(|need| rgb.len() < need) { + return Err(CodecError::encode(Format::Avif, "HDR AVIF: sample buffer too short")); + } + #[cfg(all(feature = "avif", not(target_arch = "wasm32")))] + { + use rav1e::prelude::*; + const K: [f32; 3] = [0.2627, 0.6780, 0.0593]; + let mut max_nits = 0f32; + let mut sum_nits = 0f64; + let planes: Vec<[u16; 3]> = rgb + .as_chunks::<3>() + .0 + .iter() + .take(n) + .map(|c| { + let nits = [c[0], c[1], c[2]].map(|v| (if v.is_finite() { v.max(0.0) } else { 0.0 }) * HDR_REFERENCE_WHITE_NITS); + let y_nits = K[0] * nits[0] + K[1] * nits[1] + K[2] * nits[2]; + max_nits = max_nits.max(nits[0].max(nits[1]).max(nits[2])); + sum_nits += f64::from(y_nits); + let [r, g, b] = match transfer { + HdrTransfer::Pq => nits.map(pq_encode), + HdrTransfer::Hlg => hlg_signal([c[0], c[1], c[2]]), + }; + let y = K[0] * r + K[1] * g + K[2] * b; + let cb = (b - y) / (2.0 * (1.0 - K[2])) + 0.5; + let cr = (r - y) / (2.0 * (1.0 - K[0])) + 0.5; + [y, cb, cr].map(|v| (v * 1023.0).round().clamp(0.0, 1023.0) as u16) + }) + .collect(); + let q = f32::from(quality.clamp(1, 100)) / 100.0; + let x = if q >= 0.82 { + (1.0 - q) * 2.6 + } else if q > 0.25 { + 1.0 - 0.125 - q * 0.5 + } else { + 1.0 - q + }; + let quantizer = (x * 255.0).round() as usize; + let mut ec = EncoderConfig::with_speed_preset(speed.clamp(1, 10)); + ec.width = w; + ec.height = h; + ec.time_base = Rational::new(1, 1); + ec.bit_depth = 10; + ec.chroma_sampling = ChromaSampling::Cs444; + ec.pixel_range = PixelRange::Full; + ec.color_description = Some(ColorDescription { + color_primaries: ColorPrimaries::BT2020, + transfer_characteristics: match transfer { + HdrTransfer::Pq => TransferCharacteristics::SMPTE2084, + HdrTransfer::Hlg => TransferCharacteristics::HLG, + }, + matrix_coefficients: MatrixCoefficients::BT2020NCL, + }); + let cll = (max_nits.ceil().clamp(1.0, 10_000.0) as u16, (sum_nits / n as f64).ceil().clamp(1.0, 10_000.0) as u16); + ec.content_light = + (transfer == HdrTransfer::Pq).then_some(ContentLight { max_content_light_level: cll.0, max_frame_average_light_level: cll.1 }); + ec.still_picture = true; + ec.quantizer = quantizer; + ec.min_quantizer = quantizer.min(255) as u8; + ec.speed_settings.multiref = false; + ec.speed_settings.rdo_lookahead_frames = 1; + ec.speed_settings.scene_detection_mode = SceneDetectionSpeed::None; + let mut ctx: Context = Config::new().with_encoder_config(ec).new_context().map_err(|e| CodecError::encode(Format::Avif, e))?; + let mut frame = ctx.new_frame(); + { + let mut it = planes.iter(); + let mut fp = frame.planes.iter_mut(); + let (Some(py), Some(pu), Some(pv)) = (fp.next(), fp.next(), fp.next()) else { + return Err(CodecError::encode(Format::Avif, "AV1: frame has no planes")); + }; + let (mut sy, mut su, mut sv) = (py.mut_slice(Default::default()), pu.mut_slice(Default::default()), pv.mut_slice(Default::default())); + for ((ry, ru), rv) in sy.rows_iter_mut().zip(su.rows_iter_mut()).zip(sv.rows_iter_mut()).take(h) { + for ((y, u), v) in ry.iter_mut().zip(ru.iter_mut()).zip(rv.iter_mut()).take(w) { + let Some(p) = it.next() else { break }; + (*y, *u, *v) = (p[0], p[1], p[2]); + } + } + } + ctx.send_frame(frame).map_err(|e| CodecError::encode(Format::Avif, e))?; + ctx.flush(); + let mut av1 = Vec::new(); + loop { + match ctx.receive_packet() { + Ok(mut p) if p.frame_type == FrameType::KEY => av1.append(&mut p.data), + Ok(_) => {} + Err(EncoderStatus::Encoded | EncoderStatus::LimitReached) => break, + Err(e) => return Err(CodecError::encode(Format::Avif, e)), + } + } + use avif_serialize::constants as c; + let mut a = avif_serialize::Aviffy::new(); + a.set_matrix_coefficients(c::MatrixCoefficients::Bt2020Ncl) + .set_transfer_characteristics(match transfer { + HdrTransfer::Pq => c::TransferCharacteristics::Smpte2084, + HdrTransfer::Hlg => c::TransferCharacteristics::Hlg, + }) + .set_color_primaries(c::ColorPrimaries::Bt2020) + .set_full_color_range(true); + if transfer == HdrTransfer::Pq { + a.set_content_light_level(cll.0, cll.1); + } + if let Some(exif) = exif { + a.set_exif(exif.to_vec()); + } + Ok(a.to_vec(&av1, None, width, height, 10)) + } + #[cfg(not(all(feature = "avif", not(target_arch = "wasm32"))))] + { + let _ = (width, height, rgb, quality, speed, exif, transfer); + Err(CodecError::unsupported(Format::Avif, "AVIF encoding is not available in this build")) + } +} diff --git a/crates/image-codecs/src/codecs/jp2.rs b/crates/image-codecs/src/codecs/jp2.rs new file mode 100644 index 000000000..763ac9c9c --- /dev/null +++ b/crates/image-codecs/src/codecs/jp2.rs @@ -0,0 +1,242 @@ +//! JPEG 2000 (JP2 container and raw J2K codestream) decoding via `hayro-jpeg2000`. +//! +//! Decodes within `Limits`, rejecting unsupported JPX multi-layer/composition boxes and +//! checking declared SIZ/tile/palette/channel-definition bounds before pixel allocation. + +use hayro_jpeg2000::{ColorSpace, DecodeSettings, DecoderContext, Image as Jp2Image}; + +use crate::Format; +use crate::error::CodecError; +use crate::image::{ChannelLayout, Image, SampleType}; +use crate::options::Limits; + +const F: Format = Format::Jp2; +const JP2_SIGNATURE: &[u8] = b"\0\0\0\x0cjP \r\n\x87\n"; +const MAX_PROFILE: usize = 4 * 1024 * 1024; +const MAX_TILES: u64 = 65_535; +const TILE_COMPONENT_BYTES: u64 = 256; + +/// Boxes at one JP2 nesting level. Invalid lengths stop the walk at that level. +fn boxes(b: &[u8]) -> impl Iterator { + let mut at = 0usize; + std::iter::from_fn(move || { + let header = b.get(at..at.checked_add(8)?)?; + let size = u32::from_be_bytes(header.get(..4)?.try_into().ok()?); + let (length, head) = match size { + 0 => (b.len().checked_sub(at)?, 8usize), + 1 => (usize::try_from(u64::from_be_bytes(b.get(at.checked_add(8)?..at.checked_add(16)?)?.try_into().ok()?)).ok()?, 16usize), + n => (n as usize, 8usize), + }; + if length < head { + return None; + } + let end = at.checked_add(length)?; + let data = b.get(at.checked_add(head)?..end)?; + let ty = header.get(4..8)?; + at = end; + Some((ty, data)) + }) +} + +fn unsupported_jpx_content(bytes: &[u8]) -> bool { + fn association_has_content(data: &[u8], depth: usize) -> bool { + if depth >= 16 { + return true; + } + boxes(data).any(|(ty, child)| { + matches!(ty, b"jp2c" | b"jpch" | b"jplh" | b"comp" | b"ftbl") || (ty == b"asoc" && association_has_content(child, depth + 1)) + }) + } + let mut codestreams = 0u8; + for (ty, data) in boxes(bytes) { + match ty { + b"jp2c" => { + codestreams = codestreams.saturating_add(1); + if codestreams > 1 { + return true; + } + } + b"jpch" | b"jplh" | b"comp" | b"ftbl" => return true, + b"asoc" if association_has_content(data, 0) => return true, + _ => {} + } + } + false +} + +fn codestream_size(bytes: &[u8]) -> Option<(u64, u32, u32, u64)> { + let stream = if bytes.starts_with(b"\xff\x4f") { bytes } else { boxes(bytes).find_map(|(ty, data)| (ty == b"jp2c").then_some(data))? }; + if stream.get(..4)? != b"\xff\x4f\xff\x51" { + return None; + } + let length = u16::from_be_bytes(stream.get(4..6)?.try_into().ok()?) as usize; + let count = u16::from_be_bytes(stream.get(40..42)?.try_into().ok()?) as usize; + if count == 0 || length != 38usize.checked_add(count.checked_mul(3)?)? || stream.get(4..4usize.checked_add(length)?)?.len() != length { + return None; + } + let be32 = |range: std::ops::Range| stream.get(range).and_then(|v| v.try_into().ok()).map(u32::from_be_bytes); + let width = be32(8..12)?.checked_sub(be32(16..20)?)?; + let height = be32(12..16)?.checked_sub(be32(20..24)?)?; + let tiles = |grid: u32, offset: u32, tile: u32| match (grid.checked_sub(offset), tile) { + (Some(span), 1..) => Some(u64::from(span).div_ceil(u64::from(tile))), + _ => None, + }; + let across = tiles(be32(8..12)?, be32(32..36)?, be32(24..28)?)?; + let down = tiles(be32(12..16)?, be32(36..40)?, be32(28..32)?)?; + Some((count as u64, width, height, across.checked_mul(down)?)) +} + +fn single_header_boxes(bytes: &[u8]) -> bool { + let mut headers = boxes(bytes).filter(|(ty, _)| *ty == b"jp2h"); + let (Some((_, header)), None) = (headers.next(), headers.next()) else { + return false; + }; + [b"pclr", b"cmap", b"cdef"].iter().all(|kind| boxes(header).filter(|(ty, _)| ty == kind).count() <= 1) +} + +fn channel_definitions(bytes: &[u8]) -> Option { + if bytes.starts_with(b"\xff\x4f") { + return None; + } + let header = boxes(bytes).find_map(|(ty, data)| (ty == b"jp2h").then_some(data))?; + let cdef = boxes(header).find_map(|(ty, data)| (ty == b"cdef").then_some(data))?; + let count = cdef.get(..2).and_then(|n| n.try_into().ok()).map_or(0, |n: [u8; 2]| usize::from(u16::from_be_bytes(n))); + let Some(entries) = cdef.get(2..).filter(|rest| count > 0 && rest.len() == count * 6) else { + return Some(0); + }; + let mut defs: Vec<[u16; 3]> = + entries.as_chunks::<6>().0.iter().map(|&[a, b, c, d, e, f]| [[a, b], [c, d], [e, f]].map(u16::from_be_bytes)).collect(); + defs.sort_unstable(); + let opacity = defs.last().is_some_and(|&[_, kind, association]| kind == 1 && association == 0); + let colors = defs.len() - usize::from(opacity); + let mut seen = vec![false; colors]; + let layout_ok = defs.iter().enumerate().all(|(i, &[channel, kind, association])| { + usize::from(channel) == i + && (i >= colors + || (kind == 0 + && usize::from(association).checked_sub(1).and_then(|slot| seen.get_mut(slot)).is_some_and(|s| !std::mem::replace(s, true)))) + }); + Some(if layout_ok { defs.len() as u64 } else { 0 }) +} + +fn palette_columns(bytes: &[u8], components: u64) -> Result, ()> { + if bytes.starts_with(b"\xff\x4f") { + return Ok(None); + } + let Some(header) = boxes(bytes).find_map(|(ty, data)| (ty == b"jp2h").then_some(data)) else { + return Ok(None); + }; + let Some(palette) = boxes(header).find_map(|(ty, data)| (ty == b"pclr").then_some(data)) else { + return Ok(None); + }; + let columns = palette.get(2).copied().ok_or(())?; + let depths = palette.get(3..3 + usize::from(columns)).ok_or(())?; + if columns == 0 || depths.iter().any(|&d| d & 0x80 != 0 || (d & 0x7f) + 1 > 31) { + return Err(()); + } + let Some(mapping) = boxes(header).find_map(|(ty, data)| (ty == b"cmap").then_some(data)) else { + return Ok(Some(u64::from(columns))); + }; + if mapping.len() != usize::from(columns) * 4 { + return Err(()); + } + for &[c0, c1, kind, column] in mapping.as_chunks::<4>().0 { + let valid = u64::from(u16::from_be_bytes([c0, c1])) < components + && match kind { + 0 => true, + 1 => column < columns, + _ => false, + }; + if !valid { + return Err(()); + } + } + Ok(Some(u64::from(columns))) +} + +pub(crate) fn decode(bytes: &[u8], limits: &Limits) -> Result { + let malformed = |msg: &str| CodecError::malformed(F, msg); + if bytes.starts_with(JP2_SIGNATURE) && unsupported_jpx_content(bytes) { + return Err(CodecError::unsupported(F, "unsupported color space or JPX content")); + } + if bytes.starts_with(JP2_SIGNATURE) && !single_header_boxes(bytes) { + return Err(malformed("invalid JP2 header boxes")); + } + let (codestream_channels, declared_w, declared_h, tile_count) = codestream_size(bytes).ok_or_else(|| malformed("invalid codestream SIZ"))?; + if tile_count == 0 || tile_count > MAX_TILES { + return Err(malformed("invalid tile count")); + } + let max_alloc = limits.max_alloc; + let declared_pixels = u64::from(declared_w).checked_mul(u64::from(declared_h)); + if declared_w == 0 + || declared_h == 0 + || declared_w > limits.max_width + || declared_h > limits.max_height + || declared_pixels.is_none_or(|n| n > limits.max_pixels) + || declared_pixels.and_then(|n| n.checked_mul(codestream_channels)).and_then(|n| n.checked_mul(4)).is_none_or(|n| n > max_alloc) + || tile_count.checked_mul(codestream_channels).and_then(|n| n.checked_mul(TILE_COMPONENT_BYTES)).is_none_or(|n| n > max_alloc) + { + return Err(CodecError::LimitExceeded(format!("dimensions {declared_w}x{declared_h} exceed decode limits"))); + } + let header = boxes(bytes).find_map(|(ty, data)| (ty == b"jp2h").then_some(data)).unwrap_or_default(); + if boxes(header).any(|(ty, data)| ty == b"colr" && data.len() > MAX_PROFILE + 3) { + return Err(CodecError::unsupported(F, "ICC profile exceeds 4 MiB limit")); + } + let palette = palette_columns(bytes, codestream_channels).map_err(|()| malformed("invalid JP2 palette"))?; + let palette_rows = boxes(header).find_map(|(ty, data)| (ty == b"pclr").then(|| data.get(..2)).flatten()); + if let (Some(columns), Some(&[r0, r1])) = (palette, palette_rows) + && u64::from(u16::from_be_bytes([r0, r1])) * (columns * 8 + 24) > max_alloc + { + return Err(CodecError::LimitExceeded("JP2 palette exceeds max_alloc".into())); + } + let image = Jp2Image::new(bytes, &DecodeSettings::default()).map_err(|e| malformed(&format!("{e:?}")))?; + let (w, h) = (image.width(), image.height()); + let color = image.color_space(); + let (channels, icc) = match color { + ColorSpace::Gray => (1u8, None), + ColorSpace::RGB => (3u8, None), + ColorSpace::CMYK => (4u8, None), + ColorSpace::Icc { num_channels: ch @ (1 | 3 | 4), profile } if profile.len() <= MAX_PROFILE => (*ch, Some(profile.to_vec())), + _ => return Err(CodecError::unsupported(F, "unsupported JPEG 2000 color space")), + }; + let has_alpha = image.has_alpha(); + let stride = u64::from(channels) + u64::from(has_alpha); + if palette.is_some_and(|columns| columns != stride) { + return Err(malformed("JP2 palette columns do not match output channels")); + } + if channel_definitions(bytes).is_some_and(|count| count != stride || count != palette.unwrap_or(codestream_channels)) { + return Err(malformed("JP2 channel definitions do not match output channels")); + } + let layout = match (channels, has_alpha) { + (1, false) => ChannelLayout::Gray, + (1, true) => ChannelLayout::GrayA, + (3, false) => ChannelLayout::Rgb, + (3, true) => ChannelLayout::Rgba, + (4, false) => ChannelLayout::Cmyk, + (4, true) => ChannelLayout::CmykA, + _ => return Err(CodecError::unsupported(F, "unsupported channel count")), + }; + limits.check(w, h, layout, SampleType::U8)?; + let plane_channels = codestream_channels.max(stride).max(palette.unwrap_or(0)); + let pixels = u64::from(w).checked_mul(u64::from(h)).ok_or_else(|| CodecError::LimitExceeded("pixel count overflow".into()))?; + let size = pixels.checked_mul(stride).ok_or_else(|| CodecError::LimitExceeded("sample count overflow".into()))?; + let plane_bytes = + pixels.checked_mul(plane_channels).and_then(|n| n.checked_mul(4)).ok_or_else(|| CodecError::LimitExceeded("plane overflow".into()))?; + let scratch_bytes = plane_bytes + .checked_mul(2) + .and_then(|n| n.checked_add(pixels.checked_mul(8)?)) + .ok_or_else(|| CodecError::LimitExceeded("scratch overflow".into()))?; + let working = size + .checked_add(plane_bytes) + .and_then(|n| n.checked_add(scratch_bytes)) + .ok_or_else(|| CodecError::LimitExceeded("memory overflow".into()))?; + if working > max_alloc { + return Err(CodecError::LimitExceeded(format!("working set {working} exceeds max_alloc {max_alloc}"))); + } + let len = usize::try_from(size).map_err(|_| CodecError::LimitExceeded("buffer does not fit in memory".into()))?; + let mut samples = vec![0u8; len]; + image.decode_into(&mut samples, &mut DecoderContext::default()).map_err(|e| malformed(&format!("{e:?}")))?; + let mut out = Image::from_raw(w, h, layout, SampleType::U8, samples)?; + out.icc = icc; + Ok(out) +} diff --git a/crates/image-codecs/src/codecs/jpeg.rs b/crates/image-codecs/src/codecs/jpeg.rs index 173b4b740..bcd86ad5a 100644 --- a/crates/image-codecs/src/codecs/jpeg.rs +++ b/crates/image-codecs/src/codecs/jpeg.rs @@ -479,6 +479,9 @@ pub(crate) fn encode(src: &Image, plan: Plan, opts: &EncodeOptions) -> Result "Cineon", Format::Dicom => "DICOM", Format::Iff => "Amiga IFF", + Format::Jp2 => "JPEG 2000", } } @@ -201,6 +206,7 @@ impl Format { Format::Cineon => &["cin"], Format::Dicom => &["dcm", "dicom", "dc3"], Format::Iff => &["iff", "lbm", "ilbm"], + Format::Jp2 => &["jp2", "j2k", "j2c", "jpf", "jpx"], } } @@ -229,6 +235,7 @@ impl Format { Format::Cineon => "image/x-cin", Format::Dicom => "application/dicom", Format::Iff => "image/x-iff", + Format::Jp2 => "image/jp2", } } @@ -303,6 +310,7 @@ pub fn caps(format: Format) -> FormatCaps { Format::Cineon => FormatCaps { depths: &[S::U16], layouts: &[L::Gray, L::Rgb], alpha: false, ..base }, Format::Dicom => FormatCaps { read: true, write: false, depths: &[S::U8, S::U16], layouts: &[L::Gray, L::Rgb], alpha: false, ..base }, Format::Iff => FormatCaps { depths: &[S::U8], layouts: &[L::Rgb, L::Rgba], alpha: true, ..base }, + Format::Jp2 => FormatCaps { read: true, write: false, layouts: &[L::Gray, L::GrayA, L::Rgb, L::Rgba, L::Cmyk, L::CmykA], icc: true, ..base }, } } @@ -320,6 +328,10 @@ pub fn detect(bytes: &[u8]) -> Option { if b.starts_with(&[0xFF, 0x0A]) || b.starts_with(b"\0\0\0\x0cJXL \r\n\x87\n") { return Some(Format::Jxl); } + // JPEG 2000 container signature box or raw J2K codestream (SOC + SIZ). + if b.starts_with(b"\0\0\0\x0cjP \r\n\x87\n") || b.starts_with(b"\xff\x4f\xff\x51") { + return Some(Format::Jp2); + } if b.starts_with(b"II*\0") || b.starts_with(b"MM\0*") || b.starts_with(b"II+\0") || b.starts_with(b"MM\0+") { return Some(Format::Tiff); } diff --git a/crates/image-codecs/src/lib.rs b/crates/image-codecs/src/lib.rs index b4358bba1..61109faa1 100644 --- a/crates/image-codecs/src/lib.rs +++ b/crates/image-codecs/src/lib.rs @@ -33,6 +33,7 @@ pub mod web; pub use crate::allocation::probe_allocation_size; pub use crate::cicp::Cicp; pub use crate::cicp::ContentLightLevel; +pub use crate::codecs::avif::{HDR_REFERENCE_WHITE_NITS, HdrTransfer, encode_avif_hdr, encode_avif_with_speed}; pub use crate::codecs::exr_cryptomatte::{cryptomatte_id, cryptomatte_key, cryptomatte_preview_color}; pub use crate::codecs::gif_frames::{AnimationFrame, MAX_GIF_FRAMES}; pub use crate::codecs::png::{ @@ -61,7 +62,7 @@ pub use crate::resolution::{exif_resolution, export_exif, export_xmp, photoshop_ pub use crate::source_settings::{WebpEncoding, jpeg_quality, webp_encoding}; pub use half::f16; -use crate::codecs::{dds, dicom, dpx, exr, fits, heif, iff, jpeg, jxl, pcx, png, pnm, tiff, via_image, webp, xpm}; +use crate::codecs::{dds, dicom, dpx, exr, fits, heif, iff, jp2, jpeg, jxl, pcx, png, pnm, tiff, via_image, webp, xpm}; /// Detect the format and decode with default [`Limits`]. pub fn decode(bytes: &[u8]) -> Result { @@ -102,6 +103,7 @@ pub fn decode_as_with(format: Format, bytes: &[u8], opts: &DecodeOptions) -> Res Format::Iff => iff::decode(bytes, l), Format::OpenExr => exr::decode(bytes, l), Format::Heif => heif::decode(bytes, l, opts.keep_orientation), + Format::Jp2 => jp2::decode(bytes, l), Format::Jxl => Err(CodecError::unsupported(format, "JPEG XL reading is not yet available; export is supported")), Format::Gif | Format::Bmp | Format::Tga | Format::Ico | Format::Qoi | Format::Hdr | Format::Avif => via_image::decode(format, bytes, l), }?; @@ -446,6 +448,7 @@ pub fn encode(image: &Image, format: Format, opts: &EncodeOptions) -> Result iff::encode(image, plan), Format::OpenExr => exr::encode(image, plan, opts), Format::Heif => Err(CodecError::unsupported(format, "encoding is not available in this build")), + Format::Jp2 => Err(CodecError::unsupported(format, "JPEG 2000 is import-only; save as PNG or TIFF instead")), Format::Jxl => jxl::encode(image, plan), Format::Gif | Format::Bmp | Format::Tga | Format::Ico | Format::Qoi | Format::Hdr | Format::Avif => { via_image::encode(format, image, plan, opts) diff --git a/crates/image-codecs/src/options.rs b/crates/image-codecs/src/options.rs index 9674b4edf..ed96b8733 100644 --- a/crates/image-codecs/src/options.rs +++ b/crates/image-codecs/src/options.rs @@ -299,6 +299,9 @@ pub struct EncodeOptions { pub jpeg_chroma_subsampling: bool, /// JPEG scan layout (baseline, baseline optimized or progressive). pub jpeg_format: JpegFormat, + /// Optional number of progressive JPEG scans (typically 3..=5). Ignored unless + /// `jpeg_format` is [`JpegFormat::Progressive`]. + pub jpeg_progressive_scans: Option, pub png_compression: PngCompression, /// Write Adam7-interlaced PNG. pub png_interlaced: bool, @@ -336,6 +339,7 @@ impl Default for EncodeOptions { jpeg_quality: 90, jpeg_chroma_subsampling: true, jpeg_format: JpegFormat::Baseline, + jpeg_progressive_scans: None, png_compression: PngCompression::Default, png_interlaced: false, png_stereo: None, diff --git a/crates/pdf/README.md b/crates/pdf/README.md index 8cd2c3f25..210b79849 100644 --- a/crates/pdf/README.md +++ b/crates/pdf/README.md @@ -202,9 +202,12 @@ PDF files. The reader decrypts on its own and does not use it. It covers revisio block is ignored and bad padding is left in. - `permissions()` gives the `/P` flags; for R5/R6 a permission counts only when `/Perms` grants it too. Owners have every permission. -- `create(&NewEncryption, id0)` builds the `/Encrypt` values for RC4-40, RC4-128, AES-128 or - AES-256. The caller supplies 32 bytes of entropy (`seed`), so the module needs no OS access - and builds for the web. `file_id` derives a file identifier from the same seed. +- `create(&NewEncryption, id0)` builds the `/Encrypt` values for AES-128 or AES-256. It refuses + the legacy RC4-40 and RC4-128 schemes, because RC4 is broken (RFC 7465); `create_legacy` + writes those, for a file that must stay readable by an old reader. `Document::set_encryption` + refuses them too, with `set_encryption_legacy` as the explicit way in. The caller supplies 32 + bytes of entropy (`seed`), so the module needs no OS access and builds for the web. `file_id` + derives a file identifier from the same seed. - `EncryptDict` holds the dictionary's values. The caller reads them from, and writes them to, its own object model; the module does not parse PDF syntax. Keys and prepared passwords are zeroized when dropped. diff --git a/crates/pdf/examples/cos_protect.rs b/crates/pdf/examples/cos_protect.rs index c40bccdce..b12c700cf 100644 --- a/crates/pdf/examples/cos_protect.rs +++ b/crates/pdf/examples/cos_protect.rs @@ -22,7 +22,8 @@ fn main() { _ => craft_pdf::cos::Algorithm::Aes256, }; let mut doc = craft_pdf::cos::Document::open(std::sync::Arc::new(std::fs::read(input).unwrap())).unwrap(); - doc.set_encryption(&craft_pdf::cos::NewEncryption { + // `rc4-40` and `rc4-128` are the legacy schemes, which `set_encryption` refuses. + doc.set_encryption_legacy(&craft_pdf::cos::NewEncryption { algorithm, user_password: user, owner_password: owner, diff --git a/crates/pdf/src/cos/document/open.rs b/crates/pdf/src/cos/document/open.rs index 46a6423a9..4991412ce 100644 --- a/crates/pdf/src/cos/document/open.rs +++ b/crates/pdf/src/cos/document/open.rs @@ -414,7 +414,20 @@ impl Document { /// Protect the document with a password (§7.6.4). Takes effect on the next save, which is /// always a full rewrite. Returns the handler (authenticated as owner). + /// + /// The legacy RC4 schemes are refused (RC4 is broken, `RFC 7465`): [`Self::set_encryption_legacy`] + /// writes them for a file an old reader must open. pub fn set_encryption(&mut self, params: &crate::crypt::NewEncryption) -> Result<(), CosError> { + self.set_encryption_with(params, false) + } + + /// [`Self::set_encryption`] with the legacy RC4-40 and RC4-128 schemes allowed. Pass it only + /// for a file that must stay readable by a reader without AES. + pub fn set_encryption_legacy(&mut self, params: &crate::crypt::NewEncryption) -> Result<(), CosError> { + self.set_encryption_with(params, true) + } + + fn set_encryption_with(&mut self, params: &crate::crypt::NewEncryption, legacy: bool) -> Result<(), CosError> { // The file identifier is part of the key; make sure it exists and keep it. let id = self.trailer.get(b"ID").map(|id| self.resolve(id)); let id0 = match id.as_deref() { @@ -426,7 +439,8 @@ impl Document { id } }; - let h = crate::crypt::create(params, &id0).map_err(|e| CosError::Security(e.to_string()))?; + let built = if legacy { crate::crypt::create_legacy(params, &id0) } else { crate::crypt::create(params, &id0) }; + let h = built.map_err(|e| CosError::Security(e.to_string()))?; let d = h.dict(); let s = |b: &[u8]| Object::String(crate::cos::PdfString { bytes: b.to_vec(), hex: true }); let mut e = Dict::new(); diff --git a/crates/pdf/src/crypt.rs b/crates/pdf/src/crypt.rs index 3d541704f..c516fa86e 100644 --- a/crates/pdf/src/crypt.rs +++ b/crates/pdf/src/crypt.rs @@ -658,9 +658,9 @@ fn key_r6(d: &EncryptDict, pw: &[u8], owner: bool) -> Result bool { + matches!(self, Algorithm::Rc4_40 | Algorithm::Rc4_128) + } +} + /// A password as [`create`] prepares it: for R2–R4 its PDFDocEncoding (characters it lacks /// dropped), at most 32 bytes; for R6 its SASLprep (RFC 4013: soft hyphens and the like removed, /// other spaces made plain, then NFKC; the password as given if SASLprep refuses it), in UTF-8, @@ -732,7 +740,23 @@ pub fn file_id(seed: &[u8; 32]) -> Vec { } /// Build the `/Encrypt` values and an authenticated (owner) handler for a new encryption. +/// +/// The legacy RC4 schemes are refused, because RC4 is broken (`RFC 7465`) and a file written with +/// it protects nothing: [`create_legacy`] is the explicit way to write them. pub fn create(params: &NewEncryption, id0: &[u8]) -> Result { + if params.algorithm.is_legacy() { + return Err(CryptError::Unsupported("RC4 encryption is legacy (RFC 7465); use create_legacy for a file an old reader must open".into())); + } + create_any(params, id0) +} + +/// [`create`] without the RC4 gate: writes RC4-40 and RC4-128 too. For a file that must stay +/// readable by a reader without AES, and for tests that read and write every algorithm. +pub fn create_legacy(params: &NewEncryption, id0: &[u8]) -> Result { + create_any(params, id0) +} + +fn create_any(params: &NewEncryption, id0: &[u8]) -> Result { // R2 reserves bits 7-32; later revisions reserve bits 7-8 and 13-32. // Reserved bits must be 1, and bits 1-2 must be 0 (§7.6.4.2). let reserved = if params.algorithm == Algorithm::Rc4_40 { !0x3F } else { !0xFFF | 0xC0 }; diff --git a/crates/pdf/src/crypt/tests.rs b/crates/pdf/src/crypt/tests.rs index 7de881273..08878c8be 100644 --- a/crates/pdf/src/crypt/tests.rs +++ b/crates/pdf/src/crypt/tests.rs @@ -52,13 +52,35 @@ fn aes_padding_is_added_and_tolerantly_removed() { const ID: &[u8] = b"0123456789abcdef"; fn new(alg: Algorithm, user: &str, owner: &str, perms: i32) -> SecurityHandler { - create( + // These tests read and write every algorithm, RC4 included. + create_legacy( &NewEncryption { algorithm: alg, user_password: user, owner_password: owner, permissions: perms, encrypt_metadata: true, seed: [42; 32] }, ID, ) .unwrap() } +/// New files are not written with RC4: `create` refuses it, and the explicitly named +/// `create_legacy` is the only way to write it. +#[test] +fn create_refuses_rc4_unless_asked() { + let params = |algorithm| NewEncryption { + algorithm, + user_password: "user", + owner_password: "owner", + permissions: -1, + encrypt_metadata: true, + seed: [42; 32], + }; + for (alg, revision) in [(Algorithm::Rc4_40, 2), (Algorithm::Rc4_128, 3)] { + assert!(matches!(create(¶ms(alg), ID), Err(CryptError::Unsupported(_))), "{alg:?}"); + assert_eq!(create_legacy(¶ms(alg), ID).unwrap().dict().r, revision, "{alg:?}"); + } + for alg in [Algorithm::Aes128, Algorithm::Aes256] { + assert!(create(¶ms(alg), ID).is_ok(), "AES is not legacy: {alg:?}"); + } +} + /// Every verifier check goes through this comparison: it reads both sides in full and a value /// that merely starts the same never matches. #[test] diff --git a/crates/pdf/src/write/svg.rs b/crates/pdf/src/write/svg.rs index 2eaf055ec..eb6542c0a 100644 --- a/crates/pdf/src/write/svg.rs +++ b/crates/pdf/src/write/svg.rs @@ -112,6 +112,7 @@ fn shared_limits(bytes: usize, limits: Limits) -> craft_svg::tree::Limits { craft_svg::tree::Limits { bytes, depth: limits.depth, + tree_depth: limits.depth, refuse_entity_markup: true, xml_nodes: limits.xml_nodes, embedded_bytes: limits.embedded_svg_bytes, @@ -161,3 +162,15 @@ fn image_resolver(limits: Limits) -> usvg::ImageHrefResolver<'static> { resolve_string: Box::new(|_, _| None), } } + +#[cfg(test)] +mod tests { + use super::{Limits, shared_limits}; + + #[test] + fn shared_depth_limit_also_bounds_svg_references() { + let limits = shared_limits(1024, Limits { depth: 7, ..Limits::default() }); + assert_eq!(limits.depth, 7); + assert_eq!(limits.tree_depth, 7); + } +} diff --git a/crates/pdf/tests/cos/bounded_writer.rs b/crates/pdf/tests/cos/bounded_writer.rs index b8fe79370..5677ca542 100644 --- a/crates/pdf/tests/cos/bounded_writer.rs +++ b/crates/pdf/tests/cos/bounded_writer.rs @@ -139,7 +139,8 @@ fn full_rewrites_preserve_graph_across_multiple_object_streams_and_evictions() { } fn protect(doc: &mut Document, algorithm: Algorithm) { - doc.set_encryption(&NewEncryption { + // Every algorithm, RC4 included, has to survive a rewrite. + doc.set_encryption_legacy(&NewEncryption { algorithm, user_password: "user", owner_password: "owner", diff --git a/crates/pdf/tests/cos/encryption.rs b/crates/pdf/tests/cos/encryption.rs index 923b04e39..2ac53cb69 100644 --- a/crates/pdf/tests/cos/encryption.rs +++ b/crates/pdf/tests/cos/encryption.rs @@ -39,9 +39,11 @@ fn fixture_with(extra: &str) -> Vec { out } +/// Every algorithm, RC4-40 and RC4-128 included: those two are the legacy schemes, which +/// `set_encryption` refuses. fn protect(alg: Algorithm, user: &str, owner: &str, permissions: i32) -> Vec { let mut doc = Document::open(Arc::new(fixture())).unwrap(); - doc.set_encryption(&NewEncryption { + doc.set_encryption_legacy(&NewEncryption { algorithm: alg, user_password: user, owner_password: owner, @@ -287,7 +289,7 @@ fn signature_contents_stay_unencrypted() { c.set(b"Test".to_vec(), Object::Array(vec![Object::Ref(sig), Object::Ref(untyped), Object::Ref(note)])) }) .unwrap(); - doc.set_encryption(&NewEncryption { + doc.set_encryption_legacy(&NewEncryption { algorithm: alg, user_password: "", owner_password: "owner", @@ -423,7 +425,7 @@ fn an_earlier_revision_opens_with_the_documents_own_credentials() { // Another file, encrypted with other keys: the document's key is no use, and it says so. let other = { let mut doc = Document::open(Arc::new(fixture())).unwrap(); - doc.set_encryption(&NewEncryption { + doc.set_encryption_legacy(&NewEncryption { algorithm: alg, user_password: "user", owner_password: "owner", @@ -444,7 +446,7 @@ fn an_earlier_revision_opens_with_the_documents_own_credentials() { /// instead of as an indirect object. fn protect_inline(alg: Algorithm, user: &str, owner: &str) -> Vec { let mut doc = Document::open(Arc::new(fixture())).unwrap(); - doc.set_encryption(&NewEncryption { + doc.set_encryption_legacy(&NewEncryption { algorithm: alg, user_password: user, owner_password: owner, diff --git a/crates/pdf/tests/cos/file_identifier.rs b/crates/pdf/tests/cos/file_identifier.rs index 551db4ebd..5ca0f0f54 100644 --- a/crates/pdf/tests/cos/file_identifier.rs +++ b/crates/pdf/tests/cos/file_identifier.rs @@ -122,7 +122,8 @@ fn caller_entropy_makes_the_new_identifier_reproducible() { fn an_encrypted_rewrite_keeps_the_first_identifier_string_its_key_depends_on() { for algorithm in [Algorithm::Rc4_128, Algorithm::Aes128, Algorithm::Aes256] { let mut doc = reopen(source(true)); - doc.set_encryption(&NewEncryption { + // The loop covers RC4-128, which `set_encryption` refuses. + doc.set_encryption_legacy(&NewEncryption { algorithm, user_password: "user", owner_password: "owner", @@ -160,7 +161,7 @@ fn unusable_ids() -> Vec<(&'static str, Option)> { /// An encrypted file whose key is derived from the empty string, with its trailer `/ID` replaced by `id`. fn protected_with_id(algorithm: Algorithm, id: Option) -> Vec { let mut doc = reopen(source_with(Some("/ID [<> <>]"))); - doc.set_encryption(&NewEncryption { + doc.set_encryption_legacy(&NewEncryption { algorithm, user_password: "user", owner_password: "owner", diff --git a/crates/pdf/tests/crypt/permissions_revision.rs b/crates/pdf/tests/crypt/permissions_revision.rs index 06a5a5b98..a58868637 100644 --- a/crates/pdf/tests/crypt/permissions_revision.rs +++ b/crates/pdf/tests/crypt/permissions_revision.rs @@ -3,7 +3,7 @@ //! //! Bit n has the mask `1 << (n - 1)`. Readers use bits 3 to 6 and 9 to 12 and ignore the rest. -use craft_pdf::crypt::{Algorithm, Auth, NewEncryption, Permissions, SecurityHandler, create}; +use craft_pdf::crypt::{Algorithm, Auth, NewEncryption, Permissions, SecurityHandler, create_legacy}; const ID: &[u8] = b"0123456789abcdef"; const T: bool = true; @@ -18,7 +18,7 @@ fn cells(p: &Permissions) -> [bool; 8] { /// Creates a file with `flags` at revision `revision`, opens it with the user password, and /// returns the user's permissions. fn user_view(alg: Algorithm, revision: i64, flags: i32) -> Permissions { - let made = create( + let made = create_legacy( &NewEncryption { algorithm: alg, user_password: "user", owner_password: "owner", permissions: flags, encrypt_metadata: true, seed: [42; 32] }, ID, ) @@ -56,7 +56,7 @@ fn revision_three_rows_hold_through_a_created_and_reopened_file() { #[test] fn owner_access_gives_every_flag() { - let made = create( + let made = create_legacy( &NewEncryption { algorithm: Algorithm::Rc4_128, user_password: "user", diff --git a/crates/pdf/tests/crypt/proptests.rs b/crates/pdf/tests/crypt/proptests.rs index a81808ab9..7b9ede4d3 100644 --- a/crates/pdf/tests/crypt/proptests.rs +++ b/crates/pdf/tests/crypt/proptests.rs @@ -19,7 +19,7 @@ use craft_pdf::crypt::{ Algorithm, Auth, CryptError, Method, NewEncryption, PADDING, SecurityHandler, StreamKind, aes_cbc_decrypt, aes_cbc_encrypt, create, - effective_password, pdfdoc_byte, rc4, + create_legacy, effective_password, pdfdoc_byte, rc4, }; use proptest::prelude::*; @@ -92,8 +92,12 @@ fn expected_auth(algorithm: Algorithm, user_target: &[u8], owner_target: &[u8], } fn make(algorithm: Algorithm, user: &str, owner: &str) -> SecurityHandler { - create(&NewEncryption { algorithm, user_password: user, owner_password: owner, permissions: -1, encrypt_metadata: true, seed: [0x5e; 32] }, ID) - .unwrap() + // These properties read and write every algorithm, RC4 included. + create_legacy( + &NewEncryption { algorithm, user_password: user, owner_password: owner, permissions: -1, encrypt_metadata: true, seed: [0x5e; 32] }, + ID, + ) + .unwrap() } proptest! { @@ -457,7 +461,7 @@ proptest! { } // Revisions below 4 always encrypt metadata, whatever the caller asked for. - let legacy = create(&NewEncryption { algorithm: Algorithm::Rc4_128, user_password: "user", owner_password: "owner", permissions: -1, encrypt_metadata: false, seed: [0x5e; 32] }, ID).unwrap(); + let legacy = create_legacy(&NewEncryption { algorithm: Algorithm::Rc4_128, user_password: "user", owner_password: "owner", permissions: -1, encrypt_metadata: false, seed: [0x5e; 32] }, ID).unwrap(); prop_assert!(legacy.dict().encrypt_metadata && legacy.encrypts_metadata(), "V<4 always encrypts metadata, as create() documents"); } diff --git a/crates/prepress/src/pdfx1a.rs b/crates/prepress/src/pdfx1a.rs index 0755424be..92110166b 100644 --- a/crates/prepress/src/pdfx1a.rs +++ b/crates/prepress/src/pdfx1a.rs @@ -25,6 +25,10 @@ const CONFORMANCE: &str = "PDF/X-1a:2003"; /// `/RegistryName` of the ICC's registry of characterized printing conditions. An output intent /// naming one of them needs no `/DestOutputProfile` in PDF/X. const ICC_REGISTRY: &str = "http://www.color.org"; +const REGISTERED_CONDITIONS: &[&str] = &["FOGRA39"]; +/// Largest classic cross-reference size accepted by the public finalizer. The rewritten table +/// emits one entry per object number, so this is also an output-allocation bound. +const MAX_XREF_SIZE: usize = 1 << 20; /// The graphics placed on one output page, to name in errors: the page's name and the graphics' /// names. @@ -114,7 +118,6 @@ pub fn check_pdfx1a(pdf: &[u8]) -> Vec { /// [`check_pdfx1a`]'s findings: transparency and color spaces first, then the rest. fn checks(pdf: &[u8]) -> (Vec, Vec) { let mut issues = Vec::new(); - let text = String::from_utf8_lossy(pdf); let file = craft_pdf::syntax::Pdf::new(pdf.to_vec()).ok(); if !pdf.starts_with(b"%PDF-1.") || pdf.get(7).is_none_or(|v| *v > b'4') { issues.push("the PDF version must be 1.4 or lower".into()); @@ -122,41 +125,46 @@ fn checks(pdf: &[u8]) -> (Vec, Vec) { if file.as_ref().is_none_or(|f| pdfx_intents(f).is_empty()) { issues.push("no PDF/X output intent".into()); } - let info = |key: &[u8]| file.as_ref().and_then(|f| info_text(pdf, f, key)); + let info = |key: &[u8]| file.as_ref().and_then(|f| info_text(f, key)); if info(b"GTS_PDFXVersion").as_deref() != Some(VERSION) { issues.push(format!("the document information doesn't name {VERSION}")); } if info(b"GTS_PDFXConformance").as_deref() != Some(CONFORMANCE) { issues.push(format!("the document information doesn't name the conformance level {CONFORMANCE}")); } - if !text.contains("/Trapped/False") && !text.contains("/Trapped/True") && !text.contains("/Trapped /False") && !text.contains("/Trapped /True") { + let trapped = file.as_ref().and_then(|f| f.info()).and_then(|info| info.get::>(b"Trapped")); + if !trapped.as_ref().is_some_and(|value| matches!(value.as_ref(), b"False" | b"True")) { issues.push("the document information needs /Trapped True or False".into()); } - if text.contains("/Interpolate true") { - issues.push("an image asks for interpolation".into()); - } - if text.contains("/Encrypt") { + if file.as_ref().is_some_and(|f| f.is_encrypted()) { issues.push("the file is encrypted".into()); } - let pages = text.matches("/Type/Page/").count() + text.matches("/Type/Page>").count(); - let trims = text.matches("/TrimBox").count(); - if pages > 0 && trims < pages { - issues.push("every page needs a trim box".into()); + if let Some(file) = file.as_ref() { + for (index, page) in file.pages().iter().enumerate() { + let raw = page.raw(); + let trim = raw.contains_key(b"TrimBox"); + let art = raw.contains_key(b"ArtBox"); + if trim == art { + issues.push(format!("page {} needs either a trim box or an art box, but not both", index + 1)); + } + } } - let (content, fonts) = content(file.as_ref()); + let (content, fonts, scan_issues) = content(file.as_ref()); issues.extend(fonts); + issues.extend(scan_issues); (content, issues) } /// Transparency and color spaces, then the output intent and fonts, walked through every page's /// content and what it uses. -fn content(file: Option<&craft_pdf::syntax::Pdf>) -> (Vec, Vec) { +fn content(file: Option<&craft_pdf::syntax::Pdf>) -> (Vec, Vec, Vec) { let Some(file) = file else { - return (vec!["the file couldn't be read to check its transparency and color spaces".into()], Vec::new()); + return (vec!["the file couldn't be read to check its transparency and color spaces".into()], Vec::new(), Vec::new()); }; let mut fonts = intent_issues(file); let mut found: Vec<(String, Vec)> = Vec::new(); let mut font_found: Vec<(String, Vec)> = Vec::new(); + let mut scan_found: Vec<(String, Vec)> = Vec::new(); for (i, page) in file.pages().iter().enumerate() { let mut scan = Scan::default(); let res = page.resources(); @@ -166,7 +174,7 @@ fn content(file: Option<&craft_pdf::syntax::Pdf>) -> (Vec, Vec) if let Some(content) = page.page_stream() { scan.content(content, res, 0); } - for (list, items) in [(&mut found, scan.found), (&mut font_found, scan.font_issues)] { + for (list, items) in [(&mut found, scan.found), (&mut font_found, scan.font_issues), (&mut scan_found, scan.other_issues)] { for what in items { match list.iter_mut().find(|(w, _)| *w == what) { Some((_, pages)) => pages.push(i + 1), @@ -182,7 +190,7 @@ fn content(file: Option<&craft_pdf::syntax::Pdf>) -> (Vec, Vec) format!("{what} on page{s} {}{more}", list.join(", ")) }; fonts.extend(font_found.into_iter().map(on_pages)); - (found.into_iter().map(on_pages).collect(), fonts) + (found.into_iter().map(on_pages).collect(), fonts, scan_found.into_iter().map(on_pages).collect()) } /// A `GTS_PDFX` output intent, as PDF/X-1a judges it. @@ -230,7 +238,8 @@ fn pdfx_intents(file: &craft_pdf::syntax::Pdf) -> Vec { Intent { obj: oi.obj_id().and_then(|id| number(id.obj_number)), name: text(b"OutputCondition").or_else(|| identifier.clone()).unwrap_or_else(|| "?".into()), - registered: registry.is_some_and(|r| r.trim_end_matches('/') == ICC_REGISTRY) && identifier.as_deref().is_some_and(|i| i != "Custom"), + registered: registry.is_some_and(|r| r.trim_end_matches('/') == ICC_REGISTRY) + && identifier.as_deref().is_some_and(|identifier| REGISTERED_CONDITIONS.contains(&identifier)), has_profile: oi.contains_key(b"DestOutputProfile"), profile_obj: oi.get_ref(b"DestOutputProfile").and_then(|r| number(r.obj_number)), channels: profile.as_ref().and_then(|p| p.dict().get::(b"N")), @@ -247,14 +256,8 @@ fn raw_pdfx_intents<'a>(file: &'a craft_pdf::syntax::Pdf) -> Vec> { intents.iter::>().filter(|oi| oi.get::>(b"S").is_some_and(|s| s.as_ref() == b"GTS_PDFX")).collect() } -fn info_text(pdf: &[u8], file: &craft_pdf::syntax::Pdf, key: &[u8]) -> Option { - let layout = Layout::read(pdf)?; - let at = layout.trailer.find("/Info")? + "/Info".len(); - let tail = layout.trailer.get(at..)?.trim_start(); - let (number, _) = tail.split_once(' ')?; - let id = number.parse::().ok()?; - let dict = file.xref().get::>(craft_pdf::syntax::object::ObjectIdentifier { obj_number: id, gen_number: 0 })?; - dict.get::>(key).map(|s| String::from_utf8_lossy(s.as_bytes()).trim().to_string()) +fn info_text(file: &craft_pdf::syntax::Pdf, key: &[u8]) -> Option { + file.info()?.get::>(key).map(|s| String::from_utf8_lossy(s.as_bytes()).trim().to_string()) } /// What PDF/X-1a finds wrong with the file's output intent: a CMYK profile, ICC v2 (PDF 1.4 @@ -285,7 +288,7 @@ pub fn finish_x1a(pdf: &[u8]) -> Result, String> { let layout = Layout::read(pdf).ok_or_else(unreadable)?; let file = craft_pdf::syntax::Pdf::new(pdf.to_vec()).map_err(|_| unreadable())?; let mut bodies: HashMap> = HashMap::new(); - let info = int_after(&layout.trailer, "/Info").ok_or_else(unreadable)?; + let info = file.info().and_then(|info| info.obj_id()).and_then(|id| usize::try_from(id.obj_number).ok()).ok_or_else(unreadable)?; let body = layout.body(pdf, info).ok_or_else(unreadable)?; let (head, close) = body.split_at_checked(rfind(body, b">>").ok_or_else(unreadable)?).ok_or_else(unreadable)?; bodies.insert(info, [head, format!("/GTS_PDFXConformance({CONFORMANCE})").as_bytes(), close].concat()); @@ -404,6 +407,9 @@ impl Layout { return None; } let size = int_after(&trailer, "/Size")?; + if size > MAX_XREF_SIZE { + return None; + } let mut tokens = std::str::from_utf8(section.get(..t)?).ok()?.split_whitespace(); if tokens.next()? != "xref" { return None; @@ -526,8 +532,8 @@ fn disallowed(cs: &Object<'_>, res: &Resources<'_>, depth: usize) -> Option { let _names = items.next(); let alt = items.next()?; - if let Some(why) = disallowed(&alt, res, depth + 1) { - return Some(format!("a spot color whose alternate space is {why}")); + if !is_process_space(&alt, res, depth + 1) { + return Some("a spot color whose alternate space isn't a CMYK or gray process space".into()); } let _tint = items.next(); // NChannel attributes are PDF 1.6, past the 1.4 that PDF/X-1a:2003 allows. @@ -546,6 +552,26 @@ fn disallowed(cs: &Object<'_>, res: &Resources<'_>, depth: usize) -> Option, res: &Resources<'_>, depth: usize) -> bool { + if depth > 4 { + return false; + } + match cs { + Object::Name(name) => match name.as_ref() { + b"DeviceCMYK" | b"CMYK" | b"DeviceGray" | b"G" => true, + other => match res.get_color_space(&Name::new_unescaped(other)) { + Some(Object::Name(alias)) if alias.as_ref() != other => is_process_space(&Object::Name(alias), res, depth + 1), + Some(named) if !matches!(named, Object::Name(_)) => is_process_space(&named, res, depth + 1), + _ => false, + }, + }, + Object::Array(array) => { + array.iter::>().next().is_some_and(|family| matches!(family.as_ref(), b"DeviceCMYK" | b"CMYK" | b"DeviceGray" | b"G")) + } + _ => false, + } +} + /// What one page has that PDF/X-1a doesn't allow, in the order found. #[derive(Default)] struct Scan { @@ -553,6 +579,7 @@ struct Scan { found: Vec, known: HashSet, font_issues: Vec, + other_issues: Vec, seen: HashSet, fonts: HashSet, } @@ -572,6 +599,7 @@ impl Scan { fn content(&mut self, data: &[u8], res: &Resources<'_>, depth: usize) { if depth > MAX_DEPTH { + self.add("content was not fully checked because nested resources exceed the scan limit".into()); return; } let mut ops = UntypedIter::new(data); @@ -617,10 +645,12 @@ impl Scan { } } b"BI" | b"EI" => { - if let Some(Object::Stream(img)) = last - && let Some(cs) = img.dict().get::>(b"CS").or_else(|| img.dict().get::>(b"ColorSpace")) - { - self.space(&cs, res); + if let Some(Object::Stream(img)) = last { + let dict = img.dict(); + if let Some(cs) = dict.get::>(b"CS").or_else(|| dict.get::>(b"ColorSpace")) { + self.space(&cs, res); + } + self.interpolation(dict); } } _ => {} @@ -664,6 +694,7 @@ impl Scan { return; } let d = img.dict(); + self.interpolation(d); if d.get::>(b"SMask").is_some() { self.add(format!("an image with a soft mask (object {}, transparency)", img.obj_id().obj_number)); } @@ -685,16 +716,30 @@ impl Scan { } } + fn interpolation(&mut self, dict: &Dict<'_>) { + if dict.get::(b"Interpolate") == Some(true) { + let message = "an image asks for interpolation".to_string(); + if !self.other_issues.contains(&message) { + self.other_issues.push(message); + } + } + } + fn form(&mut self, form: &Stream<'_>, parent: &Resources<'_>, depth: usize) { - if depth >= MAX_DEPTH || !self.seen.insert(form.obj_id()) { + if depth >= MAX_DEPTH { + self.add("content was not fully checked because nested forms exceed the scan limit".into()); + return; + } + if !self.seen.insert(form.obj_id()) { return; } let res = Resources::from_parent(form.dict().get::>(b"Resources").unwrap_or_default(), parent.clone()); if let Some(group) = form.dict().get::>(b"Group") { self.group(&group); } - if let Ok(data) = form.decoded() { - self.content(&data, &res, depth + 1); + match form.decoded() { + Ok(data) => self.content(&data, &res, depth + 1), + Err(_) => self.add(format!("a form XObject could not be decoded (object {})", form.obj_id().obj_number)), } } @@ -756,6 +801,7 @@ impl Scan { /// A Type 3 font's glyph procedures, with the font's resources over `parent`'s. fn type3(&mut self, font: &Dict<'_>, parent: &Resources<'_>, depth: usize) { if depth >= MAX_DEPTH { + self.add("content was not fully checked because nested Type 3 glyphs exceed the scan limit".into()); return; } let res = Resources::from_parent(font.get::>(b"Resources").unwrap_or_default(), parent.clone()); @@ -763,10 +809,105 @@ impl Scan { for key in procs.keys() { if let Some(glyph) = procs.get::>(key.as_ref()) && self.seen.insert(glyph.obj_id()) - && let Ok(data) = glyph.decoded() { - self.content(&data, &res, depth + 1); + match glyph.decoded() { + Ok(data) => self.content(&data, &res, depth + 1), + Err(_) => self.add(format!("a Type 3 glyph stream could not be decoded (object {})", glyph.obj_id().obj_number)), + } } } } } + +#[cfg(test)] +mod review_regressions { + use super::{Layout, check_pdfx1a, finish_x1a}; + fn pdf(page_boxes: &str, info: &str, content: &[u8], extra: &[(&str, &str)]) -> Vec { + let mut objects = vec![ + b"<< /Type /Catalog /Pages 2 0 R /OutputIntents [3 0 R] >>".to_vec(), + b"<< /Type /Pages /Kids [4 0 R] /Count 1 >>".to_vec(), + b"<< /Type /OutputIntent /S /GTS_PDFX /OutputConditionIdentifier (FOGRA39) /RegistryName (http://www.color.org) >>".to_vec(), + format!("<< /Type /Page /Parent 2 0 R /MediaBox [0 0 100 100] {page_boxes} /Resources <<>> /Contents 5 0 R >>").into_bytes(), + [format!("<< /Length {} >>\nstream\n", content.len()).as_bytes(), content, b"\nendstream"].concat(), + format!("<< {info} >>").into_bytes(), + ]; + for (key, body) in extra { + objects.push(format!("<< {key} {body} >>").into_bytes()); + } + let mut out = b"%PDF-1.4\n% parser regression fixture\n".to_vec(); + let mut offsets = Vec::new(); + for (index, body) in objects.iter().enumerate() { + offsets.push(out.len()); + out.extend_from_slice(format!("{} 0 obj\n", index + 1).as_bytes()); + out.extend_from_slice(body); + out.extend_from_slice(b"\nendobj\n"); + } + let xref = out.len(); + out.extend_from_slice(format!("xref\n0 {}\n0000000000 65535 f \n", objects.len() + 1).as_bytes()); + for offset in offsets { + out.extend_from_slice(format!("{offset:010} 00000 n \n").as_bytes()); + } + out.extend_from_slice(format!("trailer\n<< /Size {} /Root 1 0 R /Info 6 0 R >>\nstartxref\n{xref}\n%%EOF\n", objects.len() + 1).as_bytes()); + out + } + + fn compliant_info() -> &'static str { + "/GTS_PDFXVersion (PDF/X-1:2003) /GTS_PDFXConformance (PDF/X-1a:2003) /Trapped /False" + } + + #[test] + fn parsed_info_values_are_required_even_when_comment_contains_expected_values() { + let bytes = pdf( + "/TrimBox [0 0 100 100]", + "/GTS_PDFXVersion (wrong) /GTS_PDFXConformance (wrong) /Trapped /False", + b"% /GTS_PDFXVersion (PDF/X-1:2003) /GTS_PDFXConformance (PDF/X-1a:2003)\n", + &[], + ); + let issues = check_pdfx1a(&bytes); + assert!(issues.iter().any(|issue| issue.contains("doesn't name PDF/X-1:2003")), "{issues:?}"); + assert!(issues.iter().any(|issue| issue.contains("doesn't name the conformance level")), "{issues:?}"); + } + + #[test] + fn either_trim_or_art_box_is_accepted_but_not_both() { + let art_only = pdf("/ArtBox [0 0 100 100]", compliant_info(), b"", &[]); + let both = pdf("/TrimBox [0 0 100 100] /ArtBox [0 0 100 100]", compliant_info(), b"", &[]); + assert!(!check_pdfx1a(&art_only).iter().any(|issue| issue.contains("either a trim box or an art box"))); + assert!(check_pdfx1a(&both).iter().any(|issue| issue.contains("either a trim box or an art box"))); + } + + #[test] + fn an_unknown_registry_identifier_cannot_replace_a_destination_profile() { + let mut bytes = pdf("/TrimBox [0 0 100 100]", compliant_info(), b"", &[]); + let known = b"FOGRA39"; + let at = bytes.windows(known.len()).position(|window| window == known); + assert!(at.is_some()); + if let Some(at) = at { + bytes[at..at + known.len()].copy_from_slice(b"NOTREAL"); + } + let issues = check_pdfx1a(&bytes); + assert!(issues.iter().any(|issue| issue.contains("doesn't name a printing condition in the ICC registry")), "{issues:?}"); + } + + #[test] + fn finalizer_rejects_unbounded_xref_size_before_allocating_the_table() { + let mut bytes = pdf("/TrimBox [0 0 100 100]", compliant_info(), b"", &[]); + let size = b"/Size 7"; + let at = bytes.windows(size.len()).position(|window| window == size); + assert!(at.is_some()); + if let Some(at) = at { + bytes.splice(at..at + size.len(), b"/Size 1048577".iter().copied()); + } + assert!(Layout::read(&bytes).is_none()); + assert!(finish_x1a(&bytes).is_err()); + } + + #[test] + fn trapped_must_be_a_pdf_name_with_a_boolean_name_value() { + for value in ["(False)", "true", "(True)"] { + let info = format!("/GTS_PDFXVersion (PDF/X-1:2003) /GTS_PDFXConformance (PDF/X-1a:2003) /Trapped {value}"); + let bytes = pdf("/TrimBox [0 0 100 100]", &info, b"", &[]); + assert!(check_pdfx1a(&bytes).iter().any(|issue| issue.contains("needs /Trapped True or False")), "{value}"); + } + } +} diff --git a/crates/psd/src/io.rs b/crates/psd/src/io.rs index d8aa4b51d..8ca8353b9 100644 --- a/crates/psd/src/io.rs +++ b/crates/psd/src/io.rs @@ -440,6 +440,18 @@ pub(crate) fn write_unicode_units(out: &mut Vec, units: &[u16]) { #[cfg(test)] mod tests { + /// A section length that does not fit the file's 32-bit field is refused, not truncated: + /// this is what keeps every `as u32` on a length inside a serialized file from ever wrapping, + /// and it is why an over-large PSD fails the write instead of producing a short length. + #[test] + fn a_length_that_does_not_fit_32_bits_is_refused() { + let mut out = Vec::new(); + assert!(out.put_len(u64::from(u32::MAX), false).is_ok(), "the largest 32-bit length fits"); + assert!(matches!(out.put_len(u64::from(u32::MAX) + 1, false), Err(PsdError::LimitExceeded(_)))); + // A PSB length field is 64 bits, so the same value is written there. + assert!(out.put_len(u64::from(u32::MAX) + 1, true).is_ok(), "PSB has room for it"); + } + use super::*; #[test] diff --git a/crates/stream-codecs/src/ccitt.rs b/crates/stream-codecs/src/ccitt.rs index 2733cf98b..c5a213cea 100644 --- a/crates/stream-codecs/src/ccitt.rs +++ b/crates/stream-codecs/src/ccitt.rs @@ -152,6 +152,76 @@ fn starts_with_end_of_line(data: &[u8]) -> bool { /// Rows of 1-bit samples filled from a decoder's pixels: `columns` a row, each row on a new byte, at /// most `rows` rows. Pixels past a row's end and rows past the last are dropped. +/// Reason a raw CCITT fax decode stopped or failed. +#[derive(Clone, Copy, Debug, PartialEq, Eq)] +pub enum CcittErrorKind { + UnexpectedEof, + Overflow, + InvalidData, +} + +impl std::fmt::Display for CcittErrorKind { + fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result { + match self { + Self::UnexpectedEof => f.write_str("unexpected end of CCITT stream"), + Self::Overflow => f.write_str("CCITT stream overflow"), + Self::InvalidData => f.write_str("invalid CCITT data"), + } + } +} + +/// Result of [`decode_ccitt_raw`], including any partial rows decoded before an error. +#[derive(Clone, Debug, PartialEq, Eq)] +pub struct CcittRawOutput { + pub data: Vec, + pub rows_decoded: u32, + pub bytes_consumed: usize, + pub error: Option, +} + +/// Decodes CCITT Group 3 or Group 4 fax `data` with caller-specified `params` and `max_rows`, +/// returning packed MSB-first rows, rows decoded, input bytes consumed, and any decode error. +pub fn decode_ccitt_raw(data: &[u8], p: &CcittParams, max_rows: u32) -> Result { + if p.columns <= 0 || p.columns > i64::from(MAX_CCITT_COLUMNS) { + return Err(CodecError::Corrupt { filter: "CCITTFaxDecode", detail: format!("CCITT columns {} outside 1..={MAX_CCITT_COLUMNS}", p.columns) }); + } + let columns = u32::try_from(p.columns).map_err(|_| CodecError::Corrupt { filter: "CCITTFaxDecode", detail: "invalid CCITT columns".into() })?; + let rows = if p.rows > 0 { u32::try_from(p.rows).unwrap_or(max_rows).min(max_rows) } else { max_rows }; + let settings = hayro_ccitt::DecodeSettings { + columns, + rows, + end_of_block: p.end_of_block, + end_of_line: p.end_of_line, + rows_are_byte_aligned: p.encoded_byte_align && (p.k < 0 || !starts_with_end_of_line(data)), + encoding: match p.k { + ..0 => hayro_ccitt::EncodingMode::Group4, + 0 => hayro_ccitt::EncodingMode::Group3_1D, + k => hayro_ccitt::EncodingMode::Group3_2D { k: u32::try_from(k).unwrap_or(u32::MAX) }, + }, + invert_black: false, + }; + let mut out = Rows::new(columns as usize, rows as usize, p.black_is_1); + let (bytes_consumed, error) = match hayro_ccitt::decode(data, &mut out, &mut hayro_ccitt::DecoderContext::new(settings)) { + Ok(used) => (used, None), + Err(e) => { + let kind = match e { + hayro_ccitt::DecodeError::UnexpectedEof => CcittErrorKind::UnexpectedEof, + hayro_ccitt::DecodeError::Overflow => CcittErrorKind::Overflow, + _ => CcittErrorKind::InvalidData, + }; + let used = if out.done > 0 { + let mut drop_rows = Rows::new(columns as usize, out.done, p.black_is_1); + let retry = hayro_ccitt::DecodeSettings { rows: u32::try_from(out.done).unwrap_or(u32::MAX), ..settings }; + hayro_ccitt::decode(data, &mut drop_rows, &mut hayro_ccitt::DecoderContext::new(retry)).unwrap_or(data.len()) + } else { + 0 + }; + (used, Some(kind)) + } + }; + Ok(CcittRawOutput { data: out.out, rows_decoded: u32::try_from(out.done).unwrap_or(u32::MAX), bytes_consumed, error }) +} + pub(crate) struct Rows { pub(crate) out: Vec, row: Vec, diff --git a/crates/stream-codecs/src/lib.rs b/crates/stream-codecs/src/lib.rs index d129cacaa..ee275b2ce 100644 --- a/crates/stream-codecs/src/lib.rs +++ b/crates/stream-codecs/src/lib.rs @@ -42,7 +42,7 @@ mod mq; pub mod predictor; pub mod runlength; -pub use ccitt::{CcittParams, MAX_CCITT_COLUMNS, MAX_G4_COLUMNS, decode_ccitt_within, encode_g4}; +pub use ccitt::{CcittErrorKind, CcittParams, CcittRawOutput, MAX_CCITT_COLUMNS, MAX_G4_COLUMNS, decode_ccitt_raw, decode_ccitt_within, encode_g4}; pub use jbig2::{MAX_JBIG2_COLUMNS, MAX_JBIG2_PIXELS, SharedJbig2, decode_jbig2_within, encode_jbig2, encode_jbig2_generic, encode_jbig2_shared}; /// Parameters of the predictors and of LZW, as a PDF `/DecodeParms` dictionary gives them diff --git a/crates/svg/src/tree/draw.rs b/crates/svg/src/tree/draw.rs index a7bdfd80f..f83653284 100644 --- a/crates/svg/src/tree/draw.rs +++ b/crates/svg/src/tree/draw.rs @@ -24,6 +24,8 @@ use std::collections::HashMap; use usvg::Transform; +#[cfg(feature = "tree-render")] +pub use resvg; #[cfg(feature = "tree-render")] pub use resvg::tiny_skia; diff --git a/crates/video/README.md b/crates/video/README.md index 74b8c6414..7de348bb6 100644 --- a/crates/video/README.md +++ b/crates/video/README.md @@ -21,7 +21,7 @@ No codec feature is on by default. | `dnx` | [`dnx`](src/dnx/README.md) | Avid DNxHD / DNxHR (SMPTE ST 2019-1 VC-3): decoder and encoder | | `apv` | [`apv`](src/apv/README.md) | APV (Advanced Professional Video, RFC 9924): decoder (`apv-decode` alone) and encoder (`apv-encode` alone) | | `vp8` | `vp8` | VP8 (RFC 6386): decoder | -| `mpeg2` | [`mpeg2`](src/mpeg2/README.md) | MPEG-2 video (ITU-T H.262) and MPEG-1 video (ISO/IEC 11172-2): decoder (`mpeg2-decode` alone); MPEG-2 encoder with I and P pictures (`mpeg2-encode`, which also turns on the decoder) | +| `mpeg2` | [`mpeg2`](src/mpeg2/README.md) | MPEG-2 video (ITU-T H.262) and MPEG-1 video (ISO/IEC 11172-2): decoder (`mpeg2-decode` alone); MPEG-2 encoder with I, P and B pictures (`mpeg2-encode`, which also turns on the decoder) | | `mpeg4` | [`mpeg4`](src/mpeg4/README.md) | MPEG-4 Part 2 video (ISO/IEC 14496-2, the Xvid/DivX format): Simple profile encoder (`mpeg4-encode`) | | `prores` | `prores` | Apple ProRes (SMPTE RDD 36): decoder and encoder | | `h264` | [`h264`](src/h264/decode/README.md) | ITU-T H.264 / AVC: decoder (`h264-decode` alone) and encoder (`h264-encode` alone) | diff --git a/crates/video/src/av1/encode/mod.rs b/crates/video/src/av1/encode/mod.rs index 0d8e2b04f..10e88d0ad 100644 --- a/crates/video/src/av1/encode/mod.rs +++ b/crates/video/src/av1/encode/mod.rs @@ -600,8 +600,13 @@ impl Encoder { /// Encodes one frame. pub fn encode(&mut self, frame: &Frame) -> Result { + self.encode_with_key(frame, false) + } + + /// Encodes one frame, optionally forcing a keyframe regardless of `keyint`. + pub fn encode_with_key(&mut self, frame: &Frame, force_key: bool) -> Result { self.check_frame(frame)?; - let key = self.frame_num.is_multiple_of(self.cfg.keyint.max(1) as u64); + let key = force_key || self.frame_num.is_multiple_of(self.cfg.keyint.max(1) as u64); let src = self.load_source(frame); let (frame_obu, rf) = match self.cfg.rate { RateControl::ConstantQ(q) => self.encode_frame(&src, key, q.max(1)), diff --git a/crates/video/src/lib.rs b/crates/video/src/lib.rs index d9723a81d..0d8625bbd 100644 --- a/crates/video/src/lib.rs +++ b/crates/video/src/lib.rs @@ -8,7 +8,7 @@ //! | `dnx` | [`dnx`] | Avid DNxHD / DNxHR (SMPTE ST 2019-1): decoder and encoder | //! | `apv` | [`apv`] | APV (RFC 9924): decoder (`apv-decode`) and encoder (`apv-encode`) | //! | `vp8` | [`vp8`] | VP8 (RFC 6386): decoder (`vp8-decode`) and key-frame encoder (`vp8-encode`) | -//! | `mpeg2` | [`mpeg2`] | MPEG-2 video (ITU-T H.262) and MPEG-1 video: decoder (`mpeg2-decode`) and MPEG-2 I/P encoder (`mpeg2-encode`) | +//! | `mpeg2` | [`mpeg2`] | MPEG-2 video (ITU-T H.262) and MPEG-1 video: decoder (`mpeg2-decode`) and MPEG-2 I/P/B encoder (`mpeg2-encode`) | //! | `mpeg4` | [`mpeg4`] | MPEG-4 Part 2 video (ISO/IEC 14496-2, Xvid/DivX): Simple profile encoder (`mpeg4-encode`) | //! | `prores` | [`prores`] | Apple ProRes (SMPTE RDD 36): decoder and encoder | //! | `h264` | [`h264`] | ITU-T H.264 / AVC: decoder (`h264-decode`) and encoder (`h264-encode`) | diff --git a/crates/video/src/mpeg2/README.md b/crates/video/src/mpeg2/README.md index b68dc800b..069fd67bd 100644 --- a/crates/video/src/mpeg2/README.md +++ b/crates/video/src/mpeg2/README.md @@ -1,7 +1,7 @@ # MPEG-2 and MPEG-1 video decoder, MPEG-2 encoder (`craft_video::mpeg2`, features `mpeg2-decode`, `mpeg2-encode`) -Clean-room MPEG-2 video and MPEG-1 video decoder, and the first step of an MPEG-2 video encoder -(I and P pictures, see "Encoder"), in pure Rust. Depends only on +Clean-room MPEG-2 video and MPEG-1 video decoder, and the first steps of an MPEG-2 video encoder +(I, P and B pictures, see "Encoder"), in pure Rust. Depends only on `thiserror` and (optional, feature `threads`) `rayon`. Builds for `wasm32-unknown-unknown`; no `unsafe`. No GPL/LGPL code (FFmpeg, libmpeg2, libavcodec) was consulted; FFmpeg is used only as an external fixture generator and test oracle. @@ -77,8 +77,8 @@ the synthetic-stream tests confirmed against FFmpeg: ## Encoder (feature `mpeg2-encode`) -The encoder is the first step of MPEG-2 encoding in this crate, tracked in -storytold/craft-libs#1133. It writes valid MPEG-2 elementary streams, but not yet DVD or +The encoder is built in steps, tracked in storytold/craft-libs#1133: I and P pictures came in +#1136, B pictures in the second step. It writes valid MPEG-2 elementary streams, but not yet DVD or broadcast deliverables: those need the later steps listed below. ```rust @@ -86,13 +86,15 @@ use craft_video::mpeg2::{Encoder, EncoderConfig, RateControl, YuvFrame}; let mut cfg = EncoderConfig::new(720, 576, 25, 1); cfg.rate = RateControl::Abr { kbps: 6000 }; // or RateControl::Quantiser(1..=31) -cfg.gop_size = 12; // one I picture, then P pictures +cfg.gop_size = 12; // pictures from one I picture to the next +cfg.b_pictures = 2; // I B B P B B P ...; 0 for I and P only +cfg.closed_gop = true; // or open GOPs let mut enc = Encoder::new(cfg)?; for (i, f) in frames.iter().enumerate() { let frame = YuvFrame { y: &f.y, cb: &f.cb, cr: &f.cr, y_stride: 720, c_stride: 360 }; for p in enc.encode(&frame, i as i64)? { /* p.data, p.pts, p.dts, p.keyframe, p.picture_type */ } } -for p in enc.flush() { /* none yet: no B pictures */ } +for p in enc.flush() { /* the pictures held for reordering */ } es.extend_from_slice(enc.end_of_sequence()); ``` @@ -101,9 +103,18 @@ What it does: - Main Profile (4:2:0) at Low, Main, High-1440 and High Level, and 4:2:2 Profile at Main and High Level. The lowest level that holds the size, frame rate and (average) bitrate is chosen; sizes beyond High Level are refused. Any even picture size; edges are padded by repetition. -- Progressive frame pictures, I and P, in closed GOPs. Every GOP starts with a sequence header, - sequence extension and GOP header (with a time code), so every I picture is a random access - point. Packets come out in decoding order, which equals display order without B pictures. +- Progressive frame pictures, I, P and B. Every GOP starts with a sequence header, sequence + extension and GOP header (with a time code), so every I picture is a random access point. +- B pictures (0-4 between anchors) and reordering: pictures go in in display order and packets + come out in decoding order; a B picture is held until the anchor after it is coded, and + `flush` codes the last held picture as a P picture. With B pictures, `dts` is the presentation + times in decoding order one picture interval earlier, so presentation times must step evenly. + A single picture flushed before a second one gets `dts = pts - 1`. With average bitrate, a + GOP that `flush` cuts short gives back the budget of the pictures it did not get. + `low_delay` is set only without B pictures. +- GOPs are closed or open. The B pictures shown before a GOP's I picture (coded after it) predict + backward only in a closed GOP, and from the previous GOP's last anchor too in an open one. + temporal_reference counts display order from the GOP's first picture. - Forward DCT on the transpose of the decoder's IDCT basis; intra quantisation with a small dead zone, non-intra quantisation by truncation; default or custom quantiser matrices (sequence header); linear or non-linear quantiser scale; intra DC precision 8-10 bits; @@ -112,9 +123,14 @@ What it does: predictor, the previous picture's vectors), then half-sample refinement; cost SAD + λ·bits. Each macroblock is coded intra, motion-compensated with or without residual, "no MC" with residual, or skipped. The search range sets f_code (1-4). +- B pictures: a search per direction (candidates include the P picture's vectors scaled to the + B picture's distance), then forward, backward, interpolated or the skip prediction (the + previous macroblock's directions with the predictors), whichever costs least, or intra. + Interpolated predictions are averaged in the decoder's order. - Adaptive quantisation by macroblock activity (Test Model 5 §5.2.3). -- Rate control: a constant quantiser, or single-pass average bitrate with Test Model 5's - target bit allocation at picture level. Streams are signalled as variable bitrate +- Rate control: a constant quantiser (B pictures at 1.25 times it), or single-pass average + bitrate with Test Model 5's target bit allocation at picture level (B pictures weighted by + K_b = 1.4). Streams are signalled as variable bitrate (`vbv_delay` 0xFFFF, the level's maximum bitrate and VBV size in the sequence header); the VBV buffer is not modelled. - The reconstruction uses the decoder's own dequantisation, mismatch control, IDCT and motion @@ -122,7 +138,7 @@ What it does: - Slices (one per macroblock row) are coded in parallel on rayon with `threads`; the stream is identical either way. -Later steps (#1133): B pictures and reordering; CBR and VBR with a VBV guarantee; DVD presets +Later steps (#1133): CBR and VBR with a VBV guarantee; DVD presets (sizes, aspect, GOP length, 9.8 Mb/s, 23.976 soft telecine); interlaced coding (field/frame DCT, field prediction, alternate scan) and 4:2:2 High Level CBR for XDCAM HD422. MPEG-2 essence in the MXF writer and a program stream muxer belong to `craft-media`. @@ -191,9 +207,15 @@ fell to 77–150 fps). These are lower bounds. rate control converges on its target. `encode/tests.rs`: our decoder outputs exactly the encoder's reconstruction (intra only, a 30-picture P GOP, 4:2:2, linear and non-linear quantisers, quantisers 1 and 31, DC precision 8 and 10, custom matrices, odd macroblock counts, - average bitrate); quality and size follow the quantiser; repeated pictures skip; levels and - aspect in the sequence header; bad configurations and short or inconsistent planes are refused. -- `tests/mpeg2/encode.rs`: FFmpeg decodes our streams of `testsrc2` (CIF and 576-line 4:2:2) - without errors and within the decoder suite's tolerance of our reconstruction, ffprobe reports - the profile, level, pixel format and I/P picture types, and average bitrate lands within 10 % of - 800 and 2500 kbit/s targets. + average bitrate, each with and without B pictures; 1-4 B pictures with closed and open GOPs, + GOP lengths that are and are not multiples of the anchor distance, and streams that end inside + a group of B pictures); decoding order, decoding times and temporal_reference; a closed GOP + decodes from its I picture with its leading B pictures and an open one without them; B pictures + are smaller than P pictures; uneven presentation times are refused with B pictures; quality and + size follow the quantiser; repeated pictures skip; levels and aspect in the sequence header; + bad configurations and short or inconsistent planes are refused. +- `tests/mpeg2/encode.rs`: FFmpeg decodes our streams of `testsrc2` (CIF and 576-line 4:2:2; + without B pictures, with 2 and 4, closed and open GOPs) without errors and within the decoder + suite's tolerance of our reconstruction, ffprobe reports the profile, level, pixel format and + picture types in display order, and average bitrate lands within 10 % of 800 and 2500 kbit/s + targets. diff --git a/crates/video/src/mpeg2/encode/mod.rs b/crates/video/src/mpeg2/encode/mod.rs index fcd1bd8d1..af32b271d 100644 --- a/crates/video/src/mpeg2/encode/mod.rs +++ b/crates/video/src/mpeg2/encode/mod.rs @@ -1,9 +1,9 @@ //! MPEG-2 video encoder (ITU-T H.262 | ISO/IEC 13818-2), feature `mpeg2-encode`. //! -//! This is the first step of MPEG-2 encoding in this crate (storytold/craft-libs#1133): I and P -//! pictures, progressive frame pictures, 4:2:0 (Main Profile) and 4:2:2 (4:2:2 Profile), closed -//! GOPs, constant quantiser or single-pass average bitrate. B pictures, a VBV guarantee, -//! interlaced coding and DVD / broadcast presets come later. +//! MPEG-2 encoding in this crate is built in steps (storytold/craft-libs#1133). So far: I, P and B +//! pictures, progressive frame pictures, 4:2:0 (Main Profile) and 4:2:2 (4:2:2 Profile), closed or +//! open GOPs, constant quantiser or single-pass average bitrate. A VBV guarantee, interlaced +//! coding and DVD / broadcast presets come later. //! //! ```no_run //! use craft_video::mpeg2::encode::{Encoder, EncoderConfig, RateControl, YuvFrame}; @@ -36,6 +36,8 @@ mod transform; #[cfg(test)] mod tests; +use std::collections::VecDeque; + use crate::mpeg2::ChromaFormat; use crate::mpeg2::headers::{DEFAULT_INTRA, PictureType, SCAN}; use crate::mpeg2::mc::Frame; @@ -50,6 +52,10 @@ pub enum EncodeError { Config(String), #[error("bad input frame: {0}")] BadFrame(String), + /// With B pictures, presentation times must increase by a constant step (the decoding times + /// are the presentation times one step earlier). + #[error("bad presentation time: {0}")] + BadTimestamp(String), } /// aspect_ratio_information (H.262 Table 6-3). @@ -97,8 +103,15 @@ pub struct EncoderConfig { /// 4:2:0 (Main Profile) or 4:2:2 (4:2:2 Profile). pub chroma: ChromaFormat, pub aspect: AspectRatio, - /// Pictures per GOP: one I picture, then P pictures. 1 codes every picture as I. 1..=1023. + /// Pictures from one I picture to the next (display order). 1 codes every picture as I. + /// 1..=1023, and at most 1024 with `b_pictures` added. pub gop_size: u32, + /// B pictures between two anchor (I or P) pictures, 0..=4. 0 gives I and P pictures only and a + /// low-delay stream (no reordering). + pub b_pictures: u32, + /// Closed GOPs: the B pictures shown before a GOP's I picture predict from it alone, so the + /// GOP decodes on its own. Open GOPs let them predict from the previous GOP too. + pub closed_gop: bool, pub rate: RateControl, /// q_scale_type: the non-linear quantiser scale (finer steps at low quantisers). pub nonlinear_quant: bool, @@ -125,6 +138,8 @@ impl EncoderConfig { chroma: ChromaFormat::Yuv420, aspect: AspectRatio::Square, gop_size: 12, + b_pictures: 2, + closed_gop: true, rate: RateControl::Quantiser(4), nonlinear_quant: true, adaptive_quant: true, @@ -155,9 +170,12 @@ pub struct Packet { /// the picture. pub data: Vec, pub pts: i64, - /// Decoding timestamp; equal to `pts` while there are no B pictures. + /// Decoding timestamp: `pts` without B pictures; with them, the presentation times in + /// decoding order one picture interval earlier, so `dts <= pts`. A stream of one picture has + /// no interval: its `dts` is `pts - 1`. pub dts: i64, - /// An I picture starting a closed GOP (random access point). + /// An I picture (random access point; in an open GOP the B pictures after it in decoding + /// order also need the previous GOP). pub keyframe: bool, pub picture_type: PictureType, /// Average quantiser_scale of the picture's macroblocks. @@ -260,6 +278,13 @@ fn validate(cfg: &EncoderConfig) -> Result<(), EncodeError> { if !(1..=1023).contains(&cfg.gop_size) { return bad(format!("GOP size {} (1..=1023)", cfg.gop_size)); } + if cfg.b_pictures > 4 { + return bad(format!("{} B pictures between anchors (0..=4)", cfg.b_pictures)); + } + // temporal_reference (10 bits) counts a GOP's leading B pictures too + if cfg.gop_size + cfg.b_pictures > 1024 { + return bad(format!("GOP size {} with {} B pictures: more than 1024 pictures to number", cfg.gop_size, cfg.b_pictures)); + } match cfg.rate { RateControl::Quantiser(q) if !(1..=31).contains(&q) => return bad(format!("quantiser_scale_code {q} (1..=31)")), RateControl::Abr { kbps: 0 } => return bad("bitrate 0".into()), @@ -284,6 +309,26 @@ fn put_matrix(w: &mut BitWriter, m: &[u8; 64]) { } } +/// A source picture waiting for the anchor after it (a B picture). +struct Held { + src: Frame, + pts: i64, + /// Display index. + n: u64, +} + +/// A coded picture before rate control and timestamps. +struct Coded { + packet: Packet, + recon: Frame, + mvs: Vec<[i32; 2]>, + avg_q: f64, +} + +/// Quantiser ratio of B to P pictures with a constant quantiser (TM5 uses 1.4 for its rate +/// allocation; a constant quantiser keeps B pictures a little closer to the anchors). +const B_QUANT_FACTOR: f64 = 1.25; + /// A stateful encoder. Feed pictures in display order; packets come out in decoding order. pub struct Encoder { cfg: EncoderConfig, @@ -297,16 +342,32 @@ pub struct Encoder { inter_matrix: [u8; 64], /// Sequence header and sequence extension, sent before every GOP. seq_header: Vec, - /// The reconstruction of the previous picture. - reference: Option, - prev_mvs: Vec<[i32; 2]>, - /// Pictures coded so far. + /// The reconstructions of the last two anchor pictures and their display indices: `latest` + /// is the backward reference of the B pictures held, and the forward reference of the next + /// P picture. + older: Option<(Frame, u64)>, + latest: Option<(Frame, u64)>, + /// Forward vectors of the latest anchor when it is a P picture (search candidates). + anchor_mvs: Vec<[i32; 2]>, + /// B pictures waiting for their backward reference. + held: Vec, + /// Display index of the next picture. count: u64, - /// Position of the next picture in its GOP. - gop_pos: u32, + /// Display indices of the current GOP's I picture and of its first picture in display order + /// (a leading B picture in front of the I picture, or the I picture). + gop_start: u64, + gop_first: u64, + /// The current GOP is closed: its leading B pictures predict backward only. + gop_closed: bool, + /// Presentation times not yet given to a coded picture as its decoding time, in input order. + display_pts: VecDeque, + last_pts: Option, + pts_step: Option, + /// Packets waiting for the presentation time step (the decoding time offset). + out: Vec, rc: Option, recon_capture: bool, - recon: Vec, + recon: Vec<(u64, ReconFrame)>, } impl Encoder { @@ -330,10 +391,18 @@ impl Encoder { intra_matrix: cfg.intra_matrix.unwrap_or(DEFAULT_INTRA), inter_matrix: cfg.non_intra_matrix.unwrap_or([16; 64]), seq_header: Vec::new(), - reference: None, - prev_mvs: Vec::new(), + older: None, + latest: None, + anchor_mvs: Vec::new(), + held: Vec::new(), count: 0, - gop_pos: 0, + gop_start: 0, + gop_first: 0, + gop_closed: true, + display_pts: VecDeque::new(), + last_pts: None, + pts_step: None, + out: Vec::new(), rc, recon_capture: false, recon: Vec::new(), @@ -367,13 +436,32 @@ impl Encoder { self.recon_capture = on; } + /// The reconstructions captured so far, in display order. pub fn take_recon(&mut self) -> Vec { - std::mem::take(&mut self.recon) + let mut r = std::mem::take(&mut self.recon); + r.sort_by_key(|(n, _)| *n); + r.into_iter().map(|(_, f)| f).collect() } - /// Pictures held for reordering (none yet: there are no B pictures). + /// Code the pictures held for reordering (the last of them becomes a P picture) and return + /// every packet not yet returned. The next picture starts a new GOP. pub fn flush(&mut self) -> Vec { - Vec::new() + let mut out = Vec::new(); + if let Some(last) = self.held.pop() { + self.code_anchor(last, PictureType::P, &mut out); + } + // the GOP ends here: its budget covered `gop_size` pictures, so take back the share of + // those that never came, or every flushed GOP would raise the bitrate of the next + if self.latest.is_some() { + let missing = u64::from(self.cfg.gop_size).saturating_sub(self.count - self.gop_start); + if let Some(rc) = self.rc.as_mut() { + rc.end_gop(missing as u32); + } + } + self.older = None; + self.latest = None; + self.anchor_mvs.clear(); + self.release(out, true) } fn sequence_header_bytes(&self) -> Vec { @@ -406,16 +494,15 @@ impl Encoder { w.put((bit_rate >> 18) as u32 & 0xFFF, 12); w.bit(true); // marker_bit w.put((vbv >> 10) as u32 & 0xFF, 8); - w.bit(true); // low_delay: no B pictures + w.bit(c.b_pictures == 0); // low_delay: no B pictures, no reordering w.put(0, 2); // frame_rate_extension_n w.put(0, 5); // frame_rate_extension_d w.finish() } - /// group_of_pictures_header with the time code of picture `self.count`. - fn gop_header(&self, w: &mut BitWriter) { + /// group_of_pictures_header with the time code of display picture `n`. + fn gop_header(&self, w: &mut BitWriter, n: u64, closed: bool) { let fps = u64::from(self.cfg.fps_num.div_ceil(self.cfg.fps_den)).max(1); - let n = self.count; let secs = n / fps; w.start_code(0xB8); w.bit(false); // drop_frame_flag @@ -424,24 +511,40 @@ impl Encoder { w.bit(true); // marker_bit w.put((secs % 60) as u32, 6); w.put((n % fps) as u32, 6); - w.bit(true); // closed_gop + w.bit(closed); // closed_gop w.bit(false); // broken_link } - fn picture_header(&self, w: &mut BitWriter, intra: bool) { + fn picture_header(&self, w: &mut BitWriter, ptype: PictureType, temporal_reference: u64) { w.start_code(0x00); - w.put(self.gop_pos & 0x3FF, 10); // temporal_reference: display order within the GOP - w.put(if intra { 1 } else { 2 }, 3); + w.put((temporal_reference & 0x3FF) as u32, 10); // display order within the GOP + w.put( + match ptype { + PictureType::P => 2, + PictureType::B => 3, + _ => 1, + }, + 3, + ); w.put(0xFFFF, 16); // vbv_delay: variable bitrate - if !intra { - w.bit(false); // full_pel_forward_vector - w.put(7, 3); // forward_f_code (MPEG-2: in the extension) + for dir in [PictureType::P, PictureType::B] { + if ptype == dir || ptype == PictureType::B { + // full_pel_{forward,backward}_vector and {forward,backward}_f_code: MPEG-2 + // carries them in the extension + w.bit(false); + w.put(7, 3); + } } w.bit(false); // extra_bit_picture w.start_code(0xB5); w.put(8, 4); // picture coding extension - let fc = if intra { 15 } else { u32::from(self.f_code) }; - for v in [fc, fc, 15, 15] { + let fc = u32::from(self.f_code); + let f_codes = match ptype { + PictureType::P => [fc, fc, 15, 15], + PictureType::B => [fc; 4], + _ => [15; 4], + }; + for v in f_codes { w.put(v, 4); } w.put(u32::from(self.cfg.intra_dc_bits - 8), 2); @@ -502,28 +605,113 @@ impl Encoder { acts.iter().map(|&a| qcode_for(nl, base * (2.0 * a + avg) / (a + 2.0 * avg))).collect() } - /// Encode one picture. Returns the packets that became ready (one per picture for now). + /// With B pictures, presentation times must step by a constant amount, and the decoding + /// times (one step earlier) must fit in an `i64`. Checked before any state changes. + fn check_pts(&self, pts: i64) -> Result, EncodeError> { + if self.cfg.b_pictures == 0 { + return Ok(None); + } + let no_dts = |pts: i64, step: i64| EncodeError::BadTimestamp(format!("{pts}: no decoding time {step} earlier")); + let Some(last) = self.last_pts else { + // flushed before a second picture, this one decodes at `pts - 1` (see `release`) + return if pts == i64::MIN { Err(no_dts(pts, 1)) } else { Ok(None) }; + }; + let step = pts.checked_sub(last).filter(|&d| d > 0).ok_or_else(|| EncodeError::BadTimestamp(format!("{pts} after {last}")))?; + match self.pts_step { + Some(s) if s != step => Err(EncodeError::BadTimestamp(format!("{pts} after {last}: step {step}, earlier {s}"))), + // the earliest decoding time is the earlier picture's time less the step + None if last.checked_sub(step).is_none() => Err(no_dts(last, step)), + _ => Ok(Some(step)), + } + } + + /// Encode one picture (display order). Returns the packets that became ready, in decoding + /// order: with B pictures, a B picture is held until the anchor after it is coded. pub fn encode(&mut self, frame: &YuvFrame, pts: i64) -> Result, EncodeError> { let src = self.load(frame)?; - let reference = if self.gop_pos == 0 { None } else { self.reference.take() }; - let intra = reference.is_none(); + if let Some(step) = self.check_pts(pts)? { + self.pts_step = Some(step); + } + self.last_pts = Some(pts); + let n = self.count; + self.count += 1; + self.display_pts.push_back(pts); + let held = Held { src, pts, n }; + let mut out = Vec::new(); + let m = u64::from(self.cfg.b_pictures) + 1; + if self.latest.is_none() || n - self.gop_start >= u64::from(self.cfg.gop_size) { + self.code_anchor(held, PictureType::I, &mut out); + } else if (n - self.gop_start).is_multiple_of(m) { + self.code_anchor(held, PictureType::P, &mut out); + } else { + self.held.push(held); + } + Ok(self.release(out, false)) + } + + /// Code an anchor picture, then the B pictures held in front of it. + fn code_anchor(&mut self, h: Held, ptype: PictureType, out: &mut Vec) { + let intra = ptype == PictureType::I; if intra { - self.gop_pos = 0; + self.gop_start = h.n; + self.gop_first = self.held.first().map_or(h.n, |b| b.n); + self.gop_closed = self.cfg.closed_gop || self.held.is_empty(); + let n = u64::from(self.cfg.gop_size); + let m = u64::from(self.cfg.b_pictures) + 1; + let np = (n - 1) / m; if let Some(rc) = self.rc.as_mut() { - rc.start_gop(self.cfg.gop_size); + rc.start_gop(np as u32, (n - 1 - np) as u32); } } + let fwd = if intra { None } else { self.latest.as_ref().map(|r| &r.0) }; + let cand: &[[i32; 2]] = if intra { &[] } else { &self.anchor_mvs }; + let c = self.code_picture(&h, ptype, (fwd, None), [cand, &[]]); + let (recon, mvs) = self.finish(c, h.n, out); + self.older = self.latest.take(); + self.latest = Some((recon, h.n)); + self.anchor_mvs = if intra { Vec::new() } else { mvs }; + for b in std::mem::take(&mut self.held) { + self.code_b(b, out); + } + } + + /// Code a B picture between the two latest anchors. The leading B pictures of a closed GOP + /// (shown before its I picture) predict backward only. + fn code_b(&mut self, b: Held, out: &mut Vec) { + let Some((bwd, bn)) = self.latest.as_ref() else { return }; + let leading = b.n < self.gop_start; + let fwd = self.older.as_ref().filter(|_| !(leading && self.gop_closed)); + // search candidates: the latest anchor's vectors (which span the two anchors) scaled to + // this picture's distance from each + let span = self.older.as_ref().map_or(1, |o| bn.saturating_sub(o.1).max(1)) as i64; + let scaled = |d: i64| -> Vec<[i32; 2]> { + self.anchor_mvs.iter().map(|v| [(i64::from(v[0]) * d / span) as i32, (i64::from(v[1]) * d / span) as i32]).collect() + }; + let cf = fwd.map_or(Vec::new(), |f| scaled(b.n as i64 - f.1 as i64)); + let cb = scaled(b.n as i64 - *bn as i64); + let c = self.code_picture(&b, PictureType::B, (fwd.map(|f| &f.0), Some(bwd)), [&cf, &cb]); + self.finish(c, b.n, out); + } + + /// Code one picture with the given references and search candidates. + fn code_picture(&self, h: &Held, ptype: PictureType, (fwd, bwd): (Option<&Frame>, Option<&Frame>), cand: [&[[i32; 2]]; 2]) -> Coded { let base = match (&self.rc, self.cfg.rate) { - (Some(rc), _) => rc.qscale(intra), - (None, RateControl::Quantiser(q)) => f64::from(qscale(self.cfg.nonlinear_quant, q)), + (Some(rc), _) => rc.qscale(ptype), + (None, RateControl::Quantiser(q)) => { + let qs = f64::from(qscale(self.cfg.nonlinear_quant, q)); + if ptype == PictureType::B { qs * B_QUANT_FACTOR } else { qs } + } (None, _) => 8.0, }; - let mb_qcode = self.mb_quant(&src, base); + let src = &h.src; + let mb_qcode = self.mb_quant(src, base); let mut recon = Frame::new(src.w, src.h, src.cw, src.ch); let ctx = PicCtx { - src: &src, - reference: reference.as_ref(), - prev_mvs: &self.prev_mvs, + ptype, + src, + fwd, + bwd, + cand, mb_w: self.mb_w, shx: self.shx, shy: self.shy, @@ -536,11 +724,12 @@ impl Encoder { mb_qcode: &mb_qcode, }; let rows = encode_rows(&ctx, &mut recon, self.cfg.threads); + let intra = ptype == PictureType::I; let mut w = BitWriter::new(); if intra { - self.gop_header(&mut w); + self.gop_header(&mut w, self.gop_first, self.gop_closed); } - self.picture_header(&mut w, intra); + self.picture_header(&mut w, ptype, h.n - self.gop_first); let mut data = if intra { self.seq_header.clone() } else { Vec::new() }; data.extend_from_slice(&w.finish()); let mut qsum = 0u64; @@ -551,18 +740,47 @@ impl Encoder { mvs.extend_from_slice(&r.mvs); } let avg_q = qsum as f64 / (self.mb_w * self.mb_h).max(1) as f64; + let packet = Packet { data, pts: h.pts, dts: h.pts, keyframe: intra, picture_type: ptype, quantiser_scale: avg_q as f32 }; + Coded { packet, recon, mvs, avg_q } + } + + /// Account for a coded picture (display index `n`): rate control, its decoding time, the + /// reconstruction capture. Returns the reconstruction and the picture's vectors. + fn finish(&mut self, c: Coded, n: u64, out: &mut Vec) -> (Frame, Vec<[i32; 2]>) { + let Coded { mut packet, recon, mvs, avg_q } = c; if let Some(rc) = self.rc.as_mut() { - rc.update(intra, data.len() as u64 * 8, avg_q); + rc.update(packet.picture_type, packet.data.len() as u64 * 8, avg_q); } + // the k-th coded picture decodes at the k-th presentation time (less the offset that + // `release` subtracts); one time is queued per input picture + packet.dts = self.display_pts.pop_front().unwrap_or(packet.pts); if self.recon_capture { - self.recon.push(self.crop(&recon)); + self.recon.push((n, self.crop(&recon))); } - self.reference = Some(recon); - self.prev_mvs = mvs; - self.count += 1; - self.gop_pos = (self.gop_pos + 1) % self.cfg.gop_size; - let picture_type = if intra { PictureType::I } else { PictureType::P }; - Ok(vec![Packet { data, pts, dts: pts, keyframe: intra, picture_type, quantiser_scale: avg_q as f32 }]) + out.push(packet); + (recon, mvs) + } + + /// Hand out packets once their decoding times are known: with B pictures they are the + /// presentation times one step earlier, so packets wait for the second picture's time. + fn release(&mut self, out: Vec, flushing: bool) -> Vec { + self.out.extend(out); + if self.cfg.b_pictures == 0 { + return std::mem::take(&mut self.out); + } + if self.pts_step.is_none() && !flushing { + return Vec::new(); + } + // A lone picture flushed before a second one set the step decodes one unit early, the + // smallest step there is: if encoding goes on, the next picture decodes at this one's + // presentation time whatever the step turns out to be, and that must be later. + let off = self.pts_step.unwrap_or(1); + let mut res = std::mem::take(&mut self.out); + for p in &mut res { + // `check_pts` refused the times this would not fit for + p.dts = p.dts.saturating_sub(off); + } + res } fn crop(&self, f: &Frame) -> ReconFrame { diff --git a/crates/video/src/mpeg2/encode/motion.rs b/crates/video/src/mpeg2/encode/motion.rs index a7bedc80b..8242faafe 100644 --- a/crates/video/src/mpeg2/encode/motion.rs +++ b/crates/video/src/mpeg2/encode/motion.rs @@ -47,7 +47,7 @@ impl Search<'_> { } /// A half-sample vector the bitstream can carry whose prediction stays inside the reference. - fn valid(&self, mv: [i32; 2]) -> bool { + pub(crate) fn valid(&self, mv: [i32; 2]) -> bool { let f = 1i32 << self.f_code.saturating_sub(1); let (ox, oy) = self.origin(); let size = [self.reference.w as i32, self.reference.h as i32]; @@ -58,7 +58,8 @@ impl Search<'_> { }) } - fn mv_cost(&self, mv: [i32; 2]) -> u32 { + /// λ·(bits of `mv` against the predictor). + pub(crate) fn mv_cost(&self, mv: [i32; 2]) -> u32 { let t = enc_tables(); let bits: u32 = (0..2).map(|c| t.motion_bits(wrap_delta(mv[c], self.pmv[c], self.f_code), self.f_code)).sum(); self.lambda * bits diff --git a/crates/video/src/mpeg2/encode/picture.rs b/crates/video/src/mpeg2/encode/picture.rs index e7bb84b7e..0be5dbafd 100644 --- a/crates/video/src/mpeg2/encode/picture.rs +++ b/crates/video/src/mpeg2/encode/picture.rs @@ -1,11 +1,11 @@ //! Slices and macroblocks of one frame picture (H.262 §6.2.4-§6.2.6), with a reconstruction //! that matches the decoder exactly. One slice per macroblock row; slices only depend on the -//! reference picture, so rows can be coded in parallel. +//! reference pictures, so rows can be coded in parallel. -use crate::mpeg2::headers::SCAN; +use crate::mpeg2::headers::{PictureType, SCAN}; use crate::mpeg2::mc::{Block, Frame, predict}; use crate::mpeg2::slice::NON_LINEAR_Q; -use crate::mpeg2::vlc::{MB_FWD, MB_INTRA, MB_PATTERN, MB_QUANT}; +use crate::mpeg2::vlc::{MB_BWD, MB_FWD, MB_INTRA, MB_PATTERN, MB_QUANT}; use super::bitwriter::BitWriter; use super::motion::Search; @@ -35,12 +35,17 @@ pub(crate) fn qcode_for(nonlinear: bool, qs: f64) -> u8 { /// What every slice of one picture shares (read-only). pub(crate) struct PicCtx<'a> { + /// I, P or B. + pub ptype: PictureType, /// The padded source picture. pub src: &'a Frame, - /// The reconstructed previous picture (P pictures); `None` codes an I picture. - pub reference: Option<&'a Frame>, - /// Vectors of the previous picture (motion search candidates), one per macroblock. - pub prev_mvs: &'a [[i32; 2]], + /// The reconstructed reference pictures: forward (P and B pictures) and backward (B + /// pictures). A B picture without a forward reference (the leading B pictures of a closed + /// GOP) predicts backward only. + pub fwd: Option<&'a Frame>, + pub bwd: Option<&'a Frame>, + /// Motion search candidates per macroblock, forward and backward (empty: none). + pub cand: [&'a [[i32; 2]]; 2], pub mb_w: usize, pub shx: u32, pub shy: u32, @@ -166,11 +171,22 @@ impl MbPix { struct SliceSt { qcode: u8, dc: [i32; 3], - pmv: [i32; 2], + /// Motion vector predictors, forward and backward. + pmv: [[i32; 2]; 2], /// Skipped macroblocks not yet covered by an address increment. skipped: usize, - /// The vector of the macroblock to the left (motion search candidate). - left_mv: [i32; 2], + /// The vectors of the macroblock to the left (motion search candidates). + left_mv: [[i32; 2]; 2], + /// Prediction directions (`MB_FWD` / `MB_BWD`) of the previous macroblock, or `MB_INTRA`: + /// a skipped macroblock in a B picture repeats them. + last_dirs: i16, +} + +/// A B macroblock's prediction: its directions and a vector per direction. +#[derive(Clone, Copy, Debug, PartialEq, Eq)] +struct BPred { + dirs: i16, + mvs: [[i32; 2]; 2], } /// One macroblock being coded. @@ -188,7 +204,7 @@ pub(crate) fn encode_row(ctx: &PicCtx, row: usize, out: &mut RowPlanes) -> RowOu let q0 = ctx.mb_qcode.get(row * ctx.mb_w).copied().unwrap_or(1); w.put(u32::from(q0), 5); w.bit(false); // extra_bit_slice - let mut st = SliceSt { qcode: q0, dc: [ctx.dc_reset(); 3], pmv: [0; 2], skipped: 0, left_mv: [0; 2] }; + let mut st = SliceSt { qcode: q0, dc: [ctx.dc_reset(); 3], pmv: [[0; 2]; 2], skipped: 0, left_mv: [[0; 2]; 2], last_dirs: MB_INTRA }; let mut ro = RowOut { bytes: Vec::new(), mvs: vec![[0; 2]; ctx.mb_w], stats: RowStats::default() }; for mbx in 0..ctx.mb_w { macroblock(ctx, &mut st, &mut w, (mbx, row), out, &mut ro); @@ -205,31 +221,105 @@ fn macroblock(ctx: &PicCtx, st: &mut SliceSt, w: &mut BitWriter, (mbx, mby): (us let qs = qscale(ctx.nonlinear, qcode); ro.stats.qsum += qs as u64; let planes = (ctx.src.w, ctx.src.cw); - let Some(reference) = ctx.reference else { - intra_mb(ctx, st, w, &mb).put(out, mbx, ctx.cmb(), planes); - ro.stats.intra += 1; - return; + let lambda = ((qs * 3 + 4) / 8).max(1) as u32; + let rec = match (ctx.ptype, ctx.fwd) { + (PictureType::P, Some(reference)) => { + let search = Search { reference, src: &src.y, mbx, mby, f_code: ctx.f_code, range: ctx.search, pmv: st.pmv[0], lambda }; + let found = search.run(&candidates(ctx, st, addr, 0)); + if intra_cost(&src.y).saturating_add(INTRA_BIAS) < found.sad { + None + } else { + let (rec, skipped) = inter_mb(ctx, st, w, &mb, reference, found.mv); + ro.stats.skipped += skipped as u32; + ro.mvs[mbx] = found.mv; + st.left_mv[0] = found.mv; + Some(rec) + } + } + (PictureType::B, _) if ctx.bwd.is_some() => b_decision(ctx, st, &mb, lambda).map(|pred| { + let (rec, skipped) = b_mb(ctx, st, w, &mb, pred); + ro.stats.skipped += skipped as u32; + st.left_mv = pred.mvs; + rec + }), + _ => None, }; - let mut candidates = vec![st.left_mv, st.pmv]; + let rec = rec.unwrap_or_else(|| { + ro.stats.intra += 1; + st.left_mv = [[0; 2]; 2]; + intra_mb(ctx, st, w, &mb) + }); + rec.put(out, mbx, ctx.cmb(), planes); +} + +/// Motion search candidates for direction `s` of macroblock `addr`: the left neighbour's vector, +/// the predictor, and the co-located and neighbouring candidates of the picture. +fn candidates(ctx: &PicCtx, st: &SliceSt, addr: usize, s: usize) -> Vec<[i32; 2]> { + let mut c = vec![st.left_mv[s], st.pmv[s]]; for a in [Some(addr), addr.checked_sub(ctx.mb_w), Some(addr + 1), Some(addr + ctx.mb_w)].into_iter().flatten() { - if let Some(&v) = ctx.prev_mvs.get(a) { - candidates.push(v); + if let Some(&v) = ctx.cand[s].get(a) { + c.push(v); } } - let lambda = ((qs * 3 + 4) / 8).max(1) as u32; - let search = Search { reference, src: &src.y, mbx, mby, f_code: ctx.f_code, range: ctx.search, pmv: st.pmv, lambda }; - let found = search.run(&candidates); - if intra_cost(&src.y).saturating_add(INTRA_BIAS) < found.sad { - intra_mb(ctx, st, w, &mb).put(out, mbx, ctx.cmb(), planes); - ro.stats.intra += 1; - st.left_mv = [0; 2]; - return; + c +} + +/// Luma SAD of a prediction against the source. +fn sad(a: &[u8; 256], b: &[u8; 256]) -> u32 { + a.iter().zip(b).map(|(x, y)| u32::from(x.abs_diff(*y))).sum() +} + +/// Choose a B macroblock's prediction: forward, backward or interpolated, from a motion search +/// per direction, or the skip prediction (the previous macroblock's directions with the +/// predictors). `None`: code it intra. +fn b_decision(ctx: &PicCtx, st: &SliceSt, mb: &MbIn, lambda: u32) -> Option { + let addr = mb.mby * ctx.mb_w + mb.mbx; + let mut found = [None, None]; + for (s, r) in [ctx.fwd, ctx.bwd].into_iter().enumerate() { + if let Some(reference) = r { + let search = + Search { reference, src: &mb.src.y, mbx: mb.mbx, mby: mb.mby, f_code: ctx.f_code, range: ctx.search, pmv: st.pmv[s], lambda }; + let f = search.run(&candidates(ctx, st, addr, s)); + found[s] = Some((f.mv, f.sad, search.mv_cost(f.mv))); + } } - let (rec, skipped) = inter_mb(ctx, st, w, &mb, reference, found.mv); - rec.put(out, mbx, ctx.cmb(), planes); - ro.stats.skipped += skipped as u32; - ro.mvs[mbx] = found.mv; - st.left_mv = found.mv; + let mut best: Option<(BPred, u32)> = None; + let mut consider = |pred: BPred, cost: u32| { + if best.is_none_or(|(_, c)| cost < c) { + best = Some((pred, cost)); + } + }; + for (s, dir) in [(0, MB_FWD), (1, MB_BWD)] { + if let Some((mv, sad, bits)) = found[s] { + let mut mvs = [[0; 2]; 2]; + mvs[s] = mv; + consider(BPred { dirs: dir, mvs }, sad.saturating_add(bits)); + } + } + if let (Some((f, _, bits_f)), Some((b, _, bits_b))) = (found[0], found[1]) { + let pred = BPred { dirs: MB_FWD | MB_BWD, mvs: [f, b] }; + let p = predict_b(ctx, pred, mb.mbx, mb.mby); + consider(pred, sad(&p.y, &mb.src.y).saturating_add(bits_f).saturating_add(bits_b)); + } + // the skip prediction costs no vector bits; the predictors came from the macroblock to the + // left, so they must be checked to stay inside the reference from here + let edge = mb.mbx == 0 || mb.mbx + 1 == ctx.mb_w; + if !edge && st.last_dirs & MB_INTRA == 0 && st.last_dirs != 0 { + let pred = BPred { dirs: st.last_dirs, mvs: st.pmv }; + let usable = [(0, MB_FWD, ctx.fwd), (1, MB_BWD, ctx.bwd)].into_iter().all(|(s, dir, r)| { + st.last_dirs & dir == 0 + || r.is_some_and(|reference| { + let search = + Search { reference, src: &mb.src.y, mbx: mb.mbx, mby: mb.mby, f_code: ctx.f_code, range: ctx.search, pmv: st.pmv[s], lambda }; + search.valid(st.pmv[s]) + }) + }); + if usable { + consider(pred, sad(&predict_b(ctx, pred, mb.mbx, mb.mby).y, &mb.src.y)); + } + } + let (pred, cost) = best?; + (intra_cost(&mb.src.y).saturating_add(INTRA_BIAS) >= cost).then_some(pred) } /// Σ |sample − block mean| over the four luma blocks: the cost of coding them intra. @@ -270,8 +360,12 @@ fn put_coefficients(w: &mut BitWriter, t: &EncTables, levels: &[i32; 64], table: t.dct_eob[table].put(w); } -fn put_mb_type(w: &mut BitWriter, t: &EncTables, predicted: bool, flags: i16) { - let table = if predicted { &t.mbtype_p } else { &t.mbtype_i }; +fn put_mb_type(w: &mut BitWriter, t: &EncTables, ptype: PictureType, flags: i16) { + let table = match ptype { + PictureType::P => &t.mbtype_p, + PictureType::B => &t.mbtype_b, + _ => &t.mbtype_i, + }; if let Some(c) = table.get(flags as usize).copied().flatten() { c.put(w); } @@ -283,12 +377,13 @@ fn intra_mb(ctx: &PicCtx, st: &mut SliceSt, w: &mut BitWriter, mb: &MbIn) -> MbP let quant = mb.qcode != st.qcode; put_increment(w, t, st.skipped + 1); st.skipped = 0; - put_mb_type(w, t, ctx.reference.is_some(), if quant { MB_INTRA | MB_QUANT } else { MB_INTRA }); + put_mb_type(w, t, ctx.ptype, if quant { MB_INTRA | MB_QUANT } else { MB_INTRA }); if quant { w.put(u32::from(mb.qcode), 5); st.qcode = mb.qcode; } - st.pmv = [0; 2]; + st.pmv = [[0; 2]; 2]; + st.last_dirs = MB_INTRA; let qs = qscale(ctx.nonlinear, st.qcode); let dc_mult = 8 >> ctx.dc_prec; let dc_max = (1 << (8 + ctx.dc_prec)) - 1; @@ -321,28 +416,16 @@ fn intra_mb(ctx: &PicCtx, st: &mut SliceSt, w: &mut BitWriter, mb: &MbIn) -> MbP /// whether it was skipped. fn inter_mb(ctx: &PicCtx, st: &mut SliceSt, w: &mut BitWriter, mb: &MbIn, reference: &Frame, mv: [i32; 2]) -> (MbPix, bool) { let t = enc_tables(); - let mut rec = predict_mb(ctx, reference, mv, mb.mbx, mb.mby); + let mut rec = predict_mb(ctx, Some(reference), mv, mb.mbx, mb.mby); let qs = qscale(ctx.nonlinear, mb.qcode); - let (cbw, _) = ctx.cmb(); - let n = ctx.blocks(); - let mut levels = [[0i32; 64]; 8]; - let mut cbp = 0u32; - for i in 0..n { - let (s, p) = (mb.src.block(i, cbw), rec.block(i, cbw)); - let f = fdct(&std::array::from_fn(|k| s[k] - p[k])); - for pos in 0..64 { - levels[i][pos] = quant_inter(f[pos], ctx.inter_matrix[pos], qs); - } - if levels[i].iter().any(|&l| l != 0) { - cbp |= 1 << (n - 1 - i); - } - } + let (levels, cbp) = residual(ctx, mb, &rec, qs); st.dc = [ctx.dc_reset(); 3]; + st.last_dirs = MB_FWD; // the first and last macroblock of a slice are always coded let edge = mb.mbx == 0 || mb.mbx + 1 == ctx.mb_w; if mv == [0, 0] && cbp == 0 && !edge { st.skipped += 1; - st.pmv = [0; 2]; + st.pmv = [[0; 2]; 2]; return (rec, true); } put_increment(w, t, st.skipped + 1); @@ -358,44 +441,140 @@ fn inter_mb(ctx: &PicCtx, st: &mut SliceSt, w: &mut BitWriter, mb: &MbIn, refere if quant { flags |= MB_QUANT; } - put_mb_type(w, t, true, flags); + put_mb_type(w, t, PictureType::P, flags); if quant { w.put(u32::from(mb.qcode), 5); st.qcode = mb.qcode; } if flags & MB_FWD != 0 { for c in 0..2 { - t.put_motion(w, wrap_delta(mv[c], st.pmv[c], ctx.f_code), ctx.f_code); + t.put_motion(w, wrap_delta(mv[c], st.pmv[0][c], ctx.f_code), ctx.f_code); } - st.pmv = mv; + st.pmv[0] = mv; } else { - st.pmv = [0; 2]; // "No MC" resets the predictors - } - if cbp != 0 { - if n == 8 { - t.cbp[(cbp >> 2) as usize].put(w); - w.put(cbp & 3, 2); - } else { - t.cbp[cbp as usize].put(w); - } - for i in 0..n { - if cbp & (1 << (n - 1 - i)) != 0 { - put_coefficients(w, t, &levels[i], 0, 0); - rec.store(i, cbw, &reconstruct(&levels[i], ctx.inter_matrix, qs, false, 0), true); - } - } + st.pmv = [[0; 2]; 2]; // "No MC" resets the predictors } + put_residual(ctx, w, &levels, cbp, &mut rec, qs); (rec, false) } -/// The frame prediction of a macroblock from `f` with vector `mv`, formed as the decoder forms -/// it (§7.6.4: chroma vectors are the luma ones halved towards zero). -fn predict_mb(ctx: &PicCtx, f: &Frame, mv: [i32; 2], mbx: usize, mby: usize) -> MbPix { - let (cbw, cbh) = ctx.cmb(); +/// Quantised residual of every block of `mb` against `pred`, and the coded_block_pattern. +fn residual(ctx: &PicCtx, mb: &MbIn, pred: &MbPix, qs: i32) -> ([[i32; 64]; 8], u32) { + let (cbw, _) = ctx.cmb(); + let n = ctx.blocks(); + let mut levels = [[0i32; 64]; 8]; + let mut cbp = 0u32; + for i in 0..n { + let (s, p) = (mb.src.block(i, cbw), pred.block(i, cbw)); + let f = fdct(&std::array::from_fn(|k| s[k] - p[k])); + for pos in 0..64 { + levels[i][pos] = quant_inter(f[pos], ctx.inter_matrix[pos], qs); + } + if levels[i].iter().any(|&l| l != 0) { + cbp |= 1 << (n - 1 - i); + } + } + (levels, cbp) +} + +/// Write coded_block_pattern and the coded blocks, adding their reconstruction to `rec`. +fn put_residual(ctx: &PicCtx, w: &mut BitWriter, levels: &[[i32; 64]; 8], cbp: u32, rec: &mut MbPix, qs: i32) { + if cbp == 0 { + return; + } + let t = enc_tables(); + let (cbw, _) = ctx.cmb(); + let n = ctx.blocks(); + if n == 8 { + t.cbp[(cbp >> 2) as usize].put(w); + w.put(cbp & 3, 2); + } else { + t.cbp[cbp as usize].put(w); + } + for i in 0..n { + if cbp & (1 << (n - 1 - i)) != 0 { + put_coefficients(w, t, &levels[i], 0, 0); + rec.store(i, cbw, &reconstruct(&levels[i], ctx.inter_matrix, qs, false, 0), true); + } + } +} + +/// Code a B macroblock with prediction `pred` (or skip it); returns its reconstruction and +/// whether it was skipped. +fn b_mb(ctx: &PicCtx, st: &mut SliceSt, w: &mut BitWriter, mb: &MbIn, pred: BPred) -> (MbPix, bool) { + let t = enc_tables(); + let mut rec = predict_b(ctx, pred, mb.mbx, mb.mby); + let qs = qscale(ctx.nonlinear, mb.qcode); + let (levels, cbp) = residual(ctx, mb, &rec, qs); + st.dc = [ctx.dc_reset(); 3]; + // a skipped B macroblock repeats the previous one's directions with the predictors as its + // vectors (§7.6.6.3); the first and last macroblock of a slice are always coded + let edge = mb.mbx == 0 || mb.mbx + 1 == ctx.mb_w; + let same = |s: usize, dir: i16| pred.dirs & dir == 0 || pred.mvs[s] == st.pmv[s]; + if cbp == 0 && !edge && st.last_dirs & MB_INTRA == 0 && pred.dirs == st.last_dirs && same(0, MB_FWD) && same(1, MB_BWD) { + st.skipped += 1; + return (rec, true); + } + put_increment(w, t, st.skipped + 1); + st.skipped = 0; + let quant = cbp != 0 && mb.qcode != st.qcode; + let mut flags = pred.dirs; + if cbp != 0 { + flags |= MB_PATTERN; + } + if quant { + flags |= MB_QUANT; + } + put_mb_type(w, t, PictureType::B, flags); + if quant { + w.put(u32::from(mb.qcode), 5); + st.qcode = mb.qcode; + } + for (s, dir) in [(0, MB_FWD), (1, MB_BWD)] { + if pred.dirs & dir != 0 { + for c in 0..2 { + t.put_motion(w, wrap_delta(pred.mvs[s][c], st.pmv[s][c], ctx.f_code), ctx.f_code); + } + st.pmv[s] = pred.mvs[s]; + } + } + st.last_dirs = pred.dirs; + put_residual(ctx, w, &levels, cbp, &mut rec, qs); + (rec, false) +} + +/// The prediction of a B macroblock: forward, backward, or the average of both, formed in the +/// decoder's order (forward first, then the backward prediction averaged in). +fn predict_b(ctx: &PicCtx, pred: BPred, mbx: usize, mby: usize) -> MbPix { let mut p = MbPix { y: [0; 256], c: [[0; 256]; 2] }; + let mut avg = false; + for (s, dir, r) in [(0, MB_FWD, ctx.fwd), (1, MB_BWD, ctx.bwd)] { + if pred.dirs & dir != 0 { + if let Some(f) = r { + predict_into(ctx, &mut p, f, pred.mvs[s], (mbx, mby), avg); + } + avg = true; + } + } + p +} + +/// The frame prediction of a macroblock from `f` with vector `mv` (gray without a reference). +fn predict_mb(ctx: &PicCtx, f: Option<&Frame>, mv: [i32; 2], mbx: usize, mby: usize) -> MbPix { + let mut p = MbPix { y: [128; 256], c: [[128; 256]; 2] }; + if let Some(f) = f { + predict_into(ctx, &mut p, f, mv, (mbx, mby), false); + } + p +} + +/// Form the frame prediction from `f` with vector `mv` into `p`, or average it with what `p` +/// holds, as the decoder forms it (§7.6.4: chroma vectors are the luma ones halved towards zero). +fn predict_into(ctx: &PicCtx, p: &mut MbPix, f: &Frame, mv: [i32; 2], (mbx, mby): (usize, usize), avg: bool) { + let (cbw, cbh) = ctx.cmb(); let blk = Block { x: (mbx * 16) as i32 + (mv[0] >> 1), y: (mby * 16) as i32 + (mv[1] >> 1), hx: mv[0] & 1 != 0, hy: mv[1] & 1 != 0, bw: 16, bh: 16 }; - predict(&mut p.y, 0, 16, f.plane(0, None), blk, false); + predict(&mut p.y, 0, 16, f.plane(0, None), blk, avg); let cmv = [if ctx.shx == 1 { mv[0] / 2 } else { mv[0] }, if ctx.shy == 1 { mv[1] / 2 } else { mv[1] }]; let cb = Block { x: (mbx * cbw) as i32 + (cmv[0] >> 1), @@ -406,7 +585,48 @@ fn predict_mb(ctx: &PicCtx, f: &Frame, mv: [i32; 2], mbx: usize, mby: usize) -> bh: cbh, }; for c in 0..2 { - predict(&mut p.c[c], 0, cbw, f.plane(1 + c, None), cb, false); + predict(&mut p.c[c], 0, cbw, f.plane(1 + c, None), cb, avg); + } +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn a_skip_prediction_never_leaves_the_reference() { + // three macroblocks across; the left neighbour's forward vector (+17 samples) fits its + // own position but points past the right edge from the middle macroblock + let mut f = Frame::new(48, 16, 24, 8); + for (i, v) in f.planes[0].iter_mut().enumerate() { + *v = ((i % 48) * 5 + (i / 48) * 11) as u8; + } + let ctx = PicCtx { + ptype: PictureType::B, + src: &f, + fwd: Some(&f), + bwd: Some(&f), + cand: [&[], &[]], + mb_w: 3, + shx: 1, + shy: 1, + intra_matrix: &crate::mpeg2::headers::DEFAULT_INTRA, + inter_matrix: &[16; 64], + nonlinear: true, + dc_prec: 1, + f_code: 3, + search: 31, + mb_qcode: &[4; 3], + }; + let bad = BPred { dirs: MB_FWD, mvs: [[34, 0], [0, 0]] }; + // the bitstream can carry the vector (f_code 3: -64..64): only the position rules it out + let at = |mbx| Search { reference: &f, src: &[0; 256], mbx, mby: 0, f_code: ctx.f_code, range: ctx.search, pmv: bad.mvs[0], lambda: 1 }; + assert!(at(0).valid(bad.mvs[0]) && !at(1).valid(bad.mvs[0])); + // a source that the out-of-range (edge-clamped) prediction would match exactly + let src = predict_b(&ctx, bad, 1, 0); + let st = SliceSt { qcode: 4, dc: [256; 3], pmv: bad.mvs, skipped: 0, left_mv: bad.mvs, last_dirs: MB_FWD }; + let mb = MbIn { src: &src, mbx: 1, mby: 0, qcode: 4 }; + let pred = b_decision(&ctx, &st, &mb, 1); + assert_ne!(pred, Some(bad)); } - p } diff --git a/crates/video/src/mpeg2/encode/rate.rs b/crates/video/src/mpeg2/encode/rate.rs index 65c23efc3..96c6f2aa1 100644 --- a/crates/video/src/mpeg2/encode/rate.rs +++ b/crates/video/src/mpeg2/encode/rate.rs @@ -1,18 +1,26 @@ //! Single-pass average bitrate control at picture level, after the target bit allocation of the //! MPEG-2 Test Model 5: each picture type keeps a complexity `X = bits · average quantiser`, the -//! bits left for the GOP are shared out by complexity, and a picture's quantiser is `X / target`. -//! There is no VBV model: the average converges on the target, single pictures may not. +//! bits left for the GOP are shared out by complexity (B pictures weighted down by `KB`), and a +//! picture's quantiser is `X / target`. There is no VBV model: the average converges on the +//! target, single pictures may not. + +use crate::mpeg2::headers::PictureType; + +/// TM5's quantiser ratio of B to P pictures (K_b / K_p). +const KB: f64 = 1.4; pub(crate) struct RateCtl { bps: f64, fps: f64, - /// Complexity of I and P pictures. + /// Complexity of I, P and B pictures. xi: f64, xp: f64, + xb: f64, /// Bits left for the current GOP (carries over between GOPs). remaining: f64, - /// P pictures left in the current GOP. + /// P and B pictures left in the current GOP. np: u32, + nb: u32, /// Range of quantiser_scale values. q_min: f64, q_max: f64, @@ -21,34 +29,60 @@ pub(crate) struct RateCtl { impl RateCtl { pub fn new(bps: f64, fps: f64, q_range: (f64, f64)) -> Self { // TM5's starting complexities - Self { bps, fps, xi: 160.0 * bps / 115.0, xp: 60.0 * bps / 115.0, remaining: 0.0, np: 0, q_min: q_range.0, q_max: q_range.1 } + Self { + bps, + fps, + xi: 160.0 * bps / 115.0, + xp: 60.0 * bps / 115.0, + xb: 42.0 * bps / 115.0, + remaining: 0.0, + np: 0, + nb: 0, + q_min: q_range.0, + q_max: q_range.1, + } } - /// A GOP of `pictures` pictures (one I, the rest P) starts. - pub fn start_gop(&mut self, pictures: u32) { - self.remaining += self.bps * f64::from(pictures) / self.fps; - self.np = pictures.saturating_sub(1); + /// A GOP of one I, `np` P and `nb` B pictures starts. + pub fn start_gop(&mut self, np: u32, nb: u32) { + self.remaining += self.bps * f64::from(1 + np + nb) / self.fps; + self.np = np; + self.nb = nb; } - /// quantiser_scale for the next picture. - pub fn qscale(&self, intra: bool) -> f64 { + /// The GOP ends `missing` pictures short of what `start_gop` budgeted (the encoder was + /// flushed): their bits are not carried into the next GOP. + pub fn end_gop(&mut self, missing: u32) { + self.remaining -= self.bps * f64::from(missing) / self.fps; + self.np = 0; + self.nb = 0; + } + + /// quantiser_scale for the next picture of type `ptype`. + pub fn qscale(&self, ptype: PictureType) -> f64 { let floor = self.bps / (8.0 * self.fps); - let (target, x) = if intra { - (self.remaining / (1.0 + f64::from(self.np) * self.xp / self.xi), self.xi) - } else { - (self.remaining / f64::from(self.np.max(1)), self.xp) + let (np, nb) = (f64::from(self.np), f64::from(self.nb)); + let (target, x) = match ptype { + PictureType::I => (self.remaining / (1.0 + np * self.xp / self.xi + nb * self.xb / (self.xi * KB)), self.xi), + PictureType::B => (self.remaining / (nb.max(1.0) + np * KB * self.xp / self.xb), self.xb), + _ => (self.remaining / (np.max(1.0) + nb * self.xb / (KB * self.xp)), self.xp), }; (x / target.max(floor)).clamp(self.q_min, self.q_max) } - /// The picture used `bits` at average quantiser_scale `avg_q`. - pub fn update(&mut self, intra: bool, bits: u64, avg_q: f64) { - let x = bits as f64 * avg_q; - if intra { - self.xi = x.max(1.0); - } else { - self.xp = x.max(1.0); - self.np = self.np.saturating_sub(1); + /// A picture of type `ptype` used `bits` at average quantiser_scale `avg_q`. + pub fn update(&mut self, ptype: PictureType, bits: u64, avg_q: f64) { + let x = (bits as f64 * avg_q).max(1.0); + match ptype { + PictureType::I => self.xi = x, + PictureType::B => { + self.xb = x; + self.nb = self.nb.saturating_sub(1); + } + _ => { + self.xp = x; + self.np = self.np.saturating_sub(1); + } } self.remaining -= bits as f64; } @@ -58,18 +92,26 @@ impl RateCtl { mod tests { use super::*; - /// A picture whose size is `complexity / q` (the model the controller assumes). - fn run(complexity_i: f64, complexity_p: f64, bps: f64, gops: u32) -> f64 { + /// A picture whose size is `complexity / q` (the model the controller assumes), in GOPs of + /// 12 pictures with `b` B pictures between anchors. + fn run(c: [f64; 3], b: usize, bps: f64, gops: u32) -> f64 { let fps = 25.0; + let types: Vec = (0..12) + .map(|k| match k { + 0 => PictureType::I, + k if k % (b + 1) == 0 => PictureType::P, + _ => PictureType::B, + }) + .collect(); + let np = types.iter().filter(|&&t| t == PictureType::P).count() as u32; let mut rc = RateCtl::new(bps, fps, (1.0, 112.0)); let mut total = 0.0; for _ in 0..gops { - rc.start_gop(12); - for k in 0..12 { - let intra = k == 0; - let q = rc.qscale(intra); - let bits = (if intra { complexity_i } else { complexity_p }) / q; - rc.update(intra, bits as u64, q); + rc.start_gop(np, 11 - np); + for &t in &types { + let q = rc.qscale(t); + let bits = c[t as usize] / q; + rc.update(t, bits as u64, q); total += bits; } } @@ -79,24 +121,40 @@ mod tests { #[test] fn converges_on_the_target() { // complexities whose target quantiser is inside the allowed range - for (ci, cp) in [(4e7, 1e7), (8e7, 2e6), (1e7, 9e6)] { - let got = run(ci, cp, 4e6, 20); - assert!((got / 4e6 - 1.0).abs() < 0.03, "{ci} {cp}: {got}"); + for (c, b) in [([4e7, 1e7, 7e6], 0), ([8e7, 2e6, 1e6], 0), ([1e7, 9e6, 8e6], 0), ([4e7, 1e7, 7e6], 2), ([6e7, 2e7, 5e6], 1)] { + let got = run(c, b, 4e6, 20); + assert!((got / 4e6 - 1.0).abs() < 0.03, "{c:?} b {b}: {got}"); } } + #[test] + fn a_gop_cut_short_gives_its_budget_back() { + let mut rc = RateCtl::new(4e6, 25.0, (1.0, 112.0)); + for _ in 0..20 { + // a GOP of 12 that ends after its I picture and two P pictures + rc.start_gop(11, 0); + for t in [PictureType::I, PictureType::P, PictureType::P] { + let q = rc.qscale(t); + rc.update(t, (1e7 / q) as u64, q); + } + rc.end_gop(9); + } + // what is left is the three pictures' rounding, not nine pictures' budget per GOP + assert!(rc.remaining.abs() < 3.0 * 4e6 / 25.0, "{}", rc.remaining); + } + #[test] fn the_quantiser_stays_in_range() { let mut rc = RateCtl::new(1e6, 25.0, (2.0, 62.0)); - rc.start_gop(2); + rc.start_gop(1, 0); // a picture far larger than its budget pushes the next quantiser to the top - rc.update(true, 50_000_000, 10.0); - rc.start_gop(2); - assert_eq!(rc.qscale(true), 62.0); + rc.update(PictureType::I, 50_000_000, 10.0); + rc.start_gop(1, 0); + assert_eq!(rc.qscale(PictureType::I), 62.0); // and a tiny one to the bottom let mut rc = RateCtl::new(1e6, 25.0, (2.0, 62.0)); - rc.start_gop(2); - rc.update(true, 10, 2.0); - assert_eq!(rc.qscale(true), 2.0); + rc.start_gop(1, 0); + rc.update(PictureType::I, 10, 2.0); + assert_eq!(rc.qscale(PictureType::I), 2.0); } } diff --git a/crates/video/src/mpeg2/encode/tables.rs b/crates/video/src/mpeg2/encode/tables.rs index b79dd92ea..e23929450 100644 --- a/crates/video/src/mpeg2/encode/tables.rs +++ b/crates/video/src/mpeg2/encode/tables.rs @@ -4,7 +4,7 @@ use std::sync::OnceLock; use crate::mpeg2::vlc::{ - CBP, DC_CHROMA, DC_LUMA, DCT_EOB, DCT_ESCAPE, DCT_ONE, DCT_ZERO, DCT_ZERO_LONG, MBA, MBA_ESCAPE, MBTYPE_I, MBTYPE_P, MOTION, dct_table, + CBP, DC_CHROMA, DC_LUMA, DCT_EOB, DCT_ESCAPE, DCT_ONE, DCT_ZERO, DCT_ZERO_LONG, MBA, MBA_ESCAPE, MBTYPE_B, MBTYPE_I, MBTYPE_P, MOTION, dct_table, parse_code, }; @@ -38,9 +38,10 @@ pub(crate) struct EncTables { /// macroblock_address_increment 1..=33 (index 0 unused) and the escape. pub mba: [Code; 34], pub mba_escape: Code, - /// macroblock_type by flags: I pictures, P pictures (index: flags, 0..32). + /// macroblock_type by flags: I, P and B pictures (index: flags, 0..32). pub mbtype_i: [Option; 32], pub mbtype_p: [Option; 32], + pub mbtype_b: [Option; 32], /// coded_block_pattern_420 (the first six blocks). pub cbp: [Code; 64], /// motion_code -16..=16 at index code + 16. @@ -105,6 +106,7 @@ pub(crate) fn enc_tables() -> &'static EncTables { mba_escape, mbtype_i: mb_types(MBTYPE_I), mbtype_p: mb_types(MBTYPE_P), + mbtype_b: mb_types(MBTYPE_B), cbp: by_value(CBP, 0), motion: by_value(MOTION, 16), dc_luma: by_value(DC_LUMA, 0), @@ -185,11 +187,25 @@ mod tests { use super::*; use crate::mpeg2::bits::Bits; use crate::mpeg2::slice::wrap_vector; - use crate::mpeg2::vlc::{MB_FWD, MB_INTRA, MB_PATTERN, MB_QUANT, rl, tables}; + use crate::mpeg2::vlc::{MB_BWD, MB_FWD, MB_INTRA, MB_PATTERN, MB_QUANT, rl, tables}; - /// Macroblock types the encoder sends (frame pictures, frame prediction, no B pictures). + /// Macroblock types the encoder sends in P pictures (frame pictures, frame prediction). const ENC_MB_TYPES: [i16; 7] = [MB_INTRA, MB_INTRA | MB_QUANT, MB_FWD | MB_PATTERN, MB_PATTERN, MB_FWD, MB_QUANT | MB_FWD | MB_PATTERN, MB_QUANT | MB_PATTERN]; + /// And in B pictures: every combination of directions, with and without a residual. + const ENC_MB_TYPES_B: [i16; 11] = [ + MB_INTRA, + MB_INTRA | MB_QUANT, + MB_FWD, + MB_BWD, + MB_FWD | MB_BWD, + MB_FWD | MB_PATTERN, + MB_BWD | MB_PATTERN, + MB_FWD | MB_BWD | MB_PATTERN, + MB_QUANT | MB_FWD | MB_PATTERN, + MB_QUANT | MB_BWD | MB_PATTERN, + MB_QUANT | MB_FWD | MB_BWD | MB_PATTERN, + ]; #[test] fn every_code_decodes_to_its_value() { @@ -223,6 +239,9 @@ mod tests { } check(t.mbtype_p[f as usize].unwrap_or_default(), &d.mbtype[1], f); } + for f in ENC_MB_TYPES_B { + check(t.mbtype_b[f as usize].unwrap_or_default(), &d.mbtype[2], f); + } for (k, vlc) in [&d.dct_zero, &d.dct_one].into_iter().enumerate() { for run in 0..=MAX_RUN { for level in 1..=MAX_LEVEL { diff --git a/crates/video/src/mpeg2/encode/tests.rs b/crates/video/src/mpeg2/encode/tests.rs index 5352161b3..01aac2b32 100644 --- a/crates/video/src/mpeg2/encode/tests.rs +++ b/crates/video/src/mpeg2/encode/tests.rs @@ -3,7 +3,7 @@ //! comparison is in `tests/mpeg2/encode.rs`. use super::*; -use crate::mpeg2::{Decoder, access_units, probe}; +use crate::mpeg2::{Decoder, probe}; /// A moving, textured test sequence: a gradient with a sharp pattern drifting by a fraction of a /// sample per picture, and a block moving across it. @@ -54,16 +54,21 @@ fn encode(cfg: EncoderConfig, frames: usize) -> (Vec, Vec, Vec

Vec { +/// Decode packets (each tagged with its presentation time); returns the pictures in display +/// order. +fn decode(packets: &[Packet]) -> Vec { let mut d = Decoder::new(); let mut out = Vec::new(); - for (i, r) in access_units(es).into_iter().enumerate() { - out.extend(d.decode(&es[r], i as i64).unwrap()); + for p in packets { + out.extend(d.decode(&p.data, p.pts).unwrap()); } out.extend(d.flush()); assert_eq!(d.errors(), 0, "decoder errors: {:?}", d.last_error()); @@ -71,9 +76,11 @@ fn decode(es: &[u8]) -> Vec { } fn assert_decodes_to_recon(name: &str, cfg: EncoderConfig, frames: usize) -> Vec { - let (es, recon, packets) = encode(cfg, frames); - let pics = decode(&es); + let (_, recon, packets) = encode(cfg, frames); + assert!(packets.windows(2).all(|w| w[0].dts < w[1].dts) && packets.iter().all(|p| p.dts <= p.pts), "{name}: decoding times"); + let pics = decode(&packets); assert_eq!(pics.len(), frames, "{name}: picture count"); + assert_eq!(recon.len(), frames, "{name}: reconstruction count"); for (i, (p, r)) in pics.iter().zip(&recon).enumerate() { assert_eq!(p.pts, i as i64, "{name}: display order"); for (plane, (a, b)) in [("Y", (&p.y, &r.y)), ("Cb", (&p.cb, &r.cb)), ("Cr", (&p.cr, &r.cr))] { @@ -101,6 +108,7 @@ fn intra_only_decodes_to_the_reconstruction() { fn long_p_gop_does_not_drift() { let mut cfg = EncoderConfig::new(96, 64, 25, 1); cfg.gop_size = 30; + cfg.b_pictures = 0; let packets = assert_decodes_to_recon("long gop", cfg, 30); assert_eq!(packets.iter().filter(|p| p.keyframe).count(), 1); assert!(packets[1..].iter().all(|p| p.picture_type == PictureType::P)); @@ -127,7 +135,201 @@ fn every_coding_option_decodes_to_the_reconstruction() { ("422 odd size", EncoderConfig { width: 70, height: 30, chroma: ChromaFormat::Yuv422, ..base.clone() }), ]; for (name, cfg) in cases { - assert_decodes_to_recon(name, cfg, 8); + assert_decodes_to_recon(name, cfg.clone(), 8); + assert_decodes_to_recon(name, EncoderConfig { b_pictures: 0, ..cfg }, 8); + } +} + +#[test] +fn b_pictures_decode_to_the_reconstruction() { + let base = EncoderConfig::new(96, 64, 25, 1); + for b in 1..=4 { + for closed in [true, false] { + for (gop, frames) in [(12, 30), (7, 17), (2, 6), (1, 4), (15, 16)] { + let name = format!("b {b} closed {closed} gop {gop} frames {frames}"); + let cfg = EncoderConfig { b_pictures: b, closed_gop: closed, gop_size: gop, ..base.clone() }; + let packets = assert_decodes_to_recon(&name, cfg, frames); + if gop > 1 { + assert!(packets.iter().any(|p| p.picture_type == PictureType::B), "{name}: no B pictures"); + } + } + } + } + let cases = [ + ("422", EncoderConfig { chroma: ChromaFormat::Yuv422, ..base.clone() }), + ("abr", EncoderConfig { rate: RateControl::Abr { kbps: 400 }, ..base.clone() }), + ("odd size", EncoderConfig { width: 50, height: 34, ..base.clone() }), + ("small range", EncoderConfig { search_range: 2, ..base.clone() }), + ("linear coarse", EncoderConfig { nonlinear_quant: false, rate: RateControl::Quantiser(25), ..base.clone() }), + ]; + for (name, cfg) in cases { + assert_decodes_to_recon(name, cfg, 14); + } +} + +#[test] +fn pictures_are_reordered_into_decoding_order() { + let cfg = EncoderConfig { gop_size: 12, b_pictures: 2, ..EncoderConfig::new(64, 48, 25, 1) }; + let (_, _, packets) = encode(cfg, 15); + let order: Vec<(PictureType, i64)> = packets.iter().map(|p| (p.picture_type, p.pts)).collect(); + use PictureType::{B, I, P}; + #[rustfmt::skip] + let want = [(I, 0), (P, 3), (B, 1), (B, 2), (P, 6), (B, 4), (B, 5), (P, 9), (B, 7), (B, 8), + // B10 and B11 lead the second GOP; at the end the last held picture becomes a P picture + (I, 12), (B, 10), (B, 11), (P, 14), (B, 13)]; + assert_eq!(order, want); + // decoding times: the presentation times in order, one interval early + let dts: Vec = packets.iter().map(|p| p.dts).collect(); + assert_eq!(dts, (-1..14).collect::>()); + // temporal_reference counts display order from each GOP's first picture + let pics = decode(&packets); + let tr: Vec = pics.iter().map(|p| p.temporal_reference).collect(); + assert_eq!(tr, [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4]); +} + +#[test] +fn a_closed_gop_decodes_on_its_own() { + for closed in [true, false] { + let cfg = EncoderConfig { gop_size: 6, b_pictures: 2, closed_gop: closed, ..EncoderConfig::new(64, 48, 25, 1) }; + let (_, recon, packets) = encode(cfg, 12); + // start decoding at the second GOP's I picture (display 6): B4 and B5 lead it + let at = packets.iter().position(|p| p.pts == 6).unwrap(); + assert!(packets[at].keyframe); + let mut d = Decoder::new(); + let mut pics = Vec::new(); + for p in &packets[at..] { + pics.extend(d.decode(&p.data, p.pts).unwrap()); + } + pics.extend(d.flush()); + let shown: Vec = pics.iter().map(|p| p.pts).collect(); + if closed { + assert_eq!(shown, (4..12).collect::>()); + for p in &pics { + assert_eq!(p.y, recon[p.pts as usize].y, "closed GOP: picture {}", p.pts); + } + } else { + // the leading B pictures need the previous GOP's last P picture + assert_eq!(shown, (6..12).collect::>()); + } + } +} + +#[test] +fn b_pictures_are_smaller_than_p_pictures() { + let cfg = EncoderConfig { gop_size: 12, b_pictures: 2, ..EncoderConfig::new(128, 96, 25, 1) }; + let (_, _, packets) = encode(cfg, 12); + let avg = |t: PictureType| { + let s: Vec = packets.iter().filter(|p| p.picture_type == t).map(|p| p.data.len()).collect(); + s.iter().sum::() as f64 / s.len().max(1) as f64 + }; + let (p, b) = (avg(PictureType::P), avg(PictureType::B)); + assert!(b * 1.5 < p, "P {p:.0} bytes, B {b:.0} bytes"); + // static content: the B pictures skip every macroblock but the first and last of a slice + let mut enc = Encoder::new(EncoderConfig { gop_size: 12, b_pictures: 2, ..EncoderConfig::new(128, 64, 25, 1) }).unwrap(); + let (y, cb, cr) = scene(128, 64, 64, 32, 0); + let f = YuvFrame { y: &y, cb: &cb, cr: &cr, y_stride: 128, c_stride: 64 }; + let mut packets = Vec::new(); + for t in 0..7 { + packets.extend(enc.encode(&f, t).unwrap()); + } + let i = packets[0].data.len(); + let bs: Vec<&Packet> = packets.iter().filter(|p| p.picture_type == PictureType::B).collect(); + assert_eq!(bs.len(), 4); + for p in bs { + // a picture header, four slice headers and eight edge macroblocks + assert!(p.data.len() < 64, "static B picture {} of {} bytes (I {i})", p.pts, p.data.len()); + } +} + +#[test] +fn presentation_times_must_step_evenly_with_b_pictures() { + let mut enc = Encoder::new(EncoderConfig::new(32, 32, 25, 1)).unwrap(); + let (y, c) = (vec![0u8; 32 * 32], vec![0u8; 16 * 16]); + let f = YuvFrame { y: &y, cb: &c, cr: &c, y_stride: 32, c_stride: 16 }; + // the I picture waits for the second time, which sets the step + assert!(enc.encode(&f, 10).unwrap().is_empty()); + let mut packets = enc.encode(&f, 12).unwrap(); + for bad in [12, 11, 15] { + assert!(matches!(enc.encode(&f, bad), Err(EncodeError::BadTimestamp(_))), "{bad}"); + } + packets.extend(enc.encode(&f, 14).unwrap()); + packets.extend(enc.flush()); + assert_eq!(packets.iter().map(|p| (p.pts, p.dts)).collect::>(), [(10, 8), (14, 10), (12, 12)]); + // without B pictures any increasing (or not) times pass through + let mut enc = Encoder::new(EncoderConfig { b_pictures: 0, ..EncoderConfig::new(32, 32, 25, 1) }).unwrap(); + for pts in [5, 3, 9] { + let p = enc.encode(&f, pts).unwrap().remove(0); + assert_eq!((p.pts, p.dts), (pts, pts)); + } +} + +#[test] +fn decoding_times_increase_after_a_lone_picture_is_flushed() { + let mut enc = Encoder::new(EncoderConfig::new(32, 32, 25, 1)).unwrap(); + let (y, c) = (vec![0u8; 32 * 32], vec![0u8; 16 * 16]); + let f = YuvFrame { y: &y, cb: &c, cr: &c, y_stride: 32, c_stride: 16 }; + // one picture, flushed before a second sets the step: it decodes one unit early + assert!(enc.encode(&f, 100).unwrap().is_empty()); + let mut packets = enc.flush(); + assert_eq!(packets.iter().map(|p| (p.pts, p.dts)).collect::>(), [(100, 99)]); + // encoding goes on: the next picture decodes at the first one's presentation time + for pts in [105, 110] { + packets.extend(enc.encode(&f, pts).unwrap()); + } + packets.extend(enc.flush()); + assert_eq!(packets.iter().map(|p| (p.pts, p.dts)).collect::>(), [(100, 99), (105, 100), (110, 105)]); + assert!(packets[1].keyframe); +} + +#[test] +fn presentation_times_without_a_decoding_time_are_refused() { + let (y, c) = (vec![0u8; 32 * 32], vec![0u8; 16 * 16]); + let f = YuvFrame { y: &y, cb: &c, cr: &c, y_stride: 32, c_stride: 16 }; + // the decoding times are one step earlier: they must fit in an i64 + for (first, second) in [(i64::MIN + 1, i64::MIN + 3), (i64::MIN + 5, i64::MIN + 20)] { + let mut enc = Encoder::new(EncoderConfig::new(32, 32, 25, 1)).unwrap(); + assert!(matches!(enc.encode(&f, i64::MIN), Err(EncodeError::BadTimestamp(_)))); + assert!(enc.encode(&f, first).unwrap().is_empty()); + assert!(matches!(enc.encode(&f, second), Err(EncodeError::BadTimestamp(_))), "{second} after {first}"); + // a refused time leaves the encoder usable + assert_eq!(enc.flush().iter().map(|p| (p.pts, p.dts)).collect::>(), [(first, first - 1)]); + } + // the lowest times that have one + let mut enc = Encoder::new(EncoderConfig::new(32, 32, 25, 1)).unwrap(); + let mut packets = enc.encode(&f, i64::MIN + 1).unwrap(); + packets.extend(enc.encode(&f, i64::MIN + 2).unwrap()); + packets.extend(enc.flush()); + assert_eq!(packets.iter().map(|p| p.dts).collect::>(), [i64::MIN, i64::MIN + 1]); + // without B pictures every time passes through + let mut enc = Encoder::new(EncoderConfig { b_pictures: 0, ..EncoderConfig::new(32, 32, 25, 1) }).unwrap(); + assert_eq!(enc.encode(&f, i64::MIN).unwrap()[0].dts, i64::MIN); +} + +/// Average bitrate (bit/s) of the scene coded in groups of `group` pictures, flushed after each. +fn flushed_bitrate(cfg: EncoderConfig, group: usize, groups: usize) -> f64 { + let (w, h) = (cfg.width as usize, cfg.height as usize); + let fps = f64::from(cfg.fps_num) / f64::from(cfg.fps_den); + let mut enc = Encoder::new(cfg).unwrap(); + let mut bytes = 0; + for t in 0..group * groups { + let (y, cb, cr) = scene(w, h, w / 2, h / 2, t); + let f = YuvFrame { y: &y, cb: &cb, cr: &cr, y_stride: w, c_stride: w / 2 }; + bytes += enc.encode(&f, t as i64).unwrap().iter().map(|p| p.data.len()).sum::(); + if (t + 1) % group == 0 { + bytes += enc.flush().iter().map(|p| p.data.len()).sum::(); + } + } + bytes as f64 * 8.0 * fps / (group * groups) as f64 +} + +#[test] +fn flushing_short_gops_keeps_the_average_bitrate() { + // every flush ends a GOP of 12 after 4 pictures: the 8 that never came must not leave their + // bits to the next GOP + for b in [0, 2] { + let cfg = EncoderConfig { gop_size: 12, b_pictures: b, rate: RateControl::Abr { kbps: 600 }, ..EncoderConfig::new(176, 144, 25, 1) }; + let got = flushed_bitrate(cfg, 4, 15); + assert!((got / 600e3 - 1.0).abs() < 0.15, "b {b}: {got:.0} bit/s for 600000"); } } @@ -161,7 +363,7 @@ fn quality_follows_the_quantiser() { #[test] fn p_pictures_are_smaller_and_static_content_skips() { - let cfg = EncoderConfig { gop_size: 8, ..EncoderConfig::new(128, 64, 25, 1) }; + let cfg = EncoderConfig { gop_size: 8, b_pictures: 0, ..EncoderConfig::new(128, 64, 25, 1) }; let mut enc = Encoder::new(cfg).unwrap(); let (y, cb, cr) = scene(128, 64, 64, 32, 0); let f = YuvFrame { y: &y, cb: &cb, cr: &cr, y_stride: 128, c_stride: 64 }; @@ -191,7 +393,7 @@ fn the_sequence_header_describes_the_stream() { let enc = Encoder::new(cfg).unwrap(); assert_eq!(enc.profile_and_level(), pli, "{name}"); let info = probe(enc.sequence_header()).unwrap(); - assert_eq!((info.width, info.height, info.mpeg2, info.progressive_sequence, info.low_delay), (w, h, true, true, true)); + assert_eq!((info.width, info.height, info.mpeg2, info.progressive_sequence, info.low_delay), (w, h, true, true, false)); assert_eq!(info.codec_name(), format!("MPEG-2 Video ({name})")); assert_eq!(aspect, AspectRatio::Square); assert_eq!(info.sar, (1, 1)); @@ -199,6 +401,9 @@ fn the_sequence_header_describes_the_stream() { let enc = Encoder::new(EncoderConfig { aspect: AspectRatio::Dar16x9, ..EncoderConfig::new(720, 576, 30000, 1001) }).unwrap(); let info = probe(enc.sequence_header()).unwrap(); assert_eq!((info.dar, info.frame_rate), (Some((16, 9)), Some((30000, 1001)))); + // without B pictures the stream is low delay + let enc = Encoder::new(EncoderConfig { b_pictures: 0, ..EncoderConfig::new(720, 576, 25, 1) }).unwrap(); + assert!(probe(enc.sequence_header()).unwrap().low_delay); } #[test] @@ -214,6 +419,8 @@ fn bad_configurations_are_refused() { EncoderConfig { fps_den: 0, ..ok.clone() }, EncoderConfig { gop_size: 0, ..ok.clone() }, EncoderConfig { gop_size: 1024, ..ok.clone() }, + EncoderConfig { b_pictures: 5, ..ok.clone() }, + EncoderConfig { gop_size: 1023, b_pictures: 2, ..ok.clone() }, EncoderConfig { rate: RateControl::Quantiser(0), ..ok.clone() }, EncoderConfig { rate: RateControl::Quantiser(32), ..ok.clone() }, EncoderConfig { rate: RateControl::Abr { kbps: 0 }, ..ok.clone() }, @@ -231,7 +438,7 @@ fn bad_configurations_are_refused() { #[test] fn short_or_inconsistent_planes_are_refused() { - let mut enc = Encoder::new(EncoderConfig::new(32, 32, 25, 1)).unwrap(); + let mut enc = Encoder::new(EncoderConfig { b_pictures: 0, ..EncoderConfig::new(32, 32, 25, 1) }).unwrap(); let (y, c) = (vec![0u8; 32 * 32], vec![0u8; 16 * 16]); let ok = YuvFrame { y: &y, cb: &c, cr: &c, y_stride: 32, c_stride: 16 }; assert!(enc.encode(&ok, 0).is_ok()); diff --git a/crates/video/src/mpeg2/mod.rs b/crates/video/src/mpeg2/mod.rs index ae2bb1682..4fd59dced 100644 --- a/crates/video/src/mpeg2/mod.rs +++ b/crates/video/src/mpeg2/mod.rs @@ -1,5 +1,5 @@ //! Clean-room MPEG-2 video (ITU-T H.262 | ISO/IEC 13818-2) and MPEG-1 video (ISO/IEC 11172-2) -//! decoder, and the start of an MPEG-2 encoder ([`encode`], feature `mpeg2-encode`: I and P +//! decoder, and the start of an MPEG-2 encoder ([`encode`], feature `mpeg2-encode`: I, P and B //! pictures). //! //! - Main Profile at Low/Main/High-1440/High Level, 4:2:2 Profile (XDCAM HD422, IMX / D-10), diff --git a/crates/video/tests/mpeg2/encode.rs b/crates/video/tests/mpeg2/encode.rs index 5cb2377dc..122fe2d71 100644 --- a/crates/video/tests/mpeg2/encode.rs +++ b/crates/video/tests/mpeg2/encode.rs @@ -6,7 +6,7 @@ use std::path::{Path, PathBuf}; use std::process::{Command, Stdio}; -use craft_video::mpeg2::{ChromaFormat, Encoder, EncoderConfig, RateControl, ReconFrame, YuvFrame}; +use craft_video::mpeg2::{ChromaFormat, Encoder, EncoderConfig, Packet, RateControl, ReconFrame, YuvFrame}; use crate::common::{Stats, fixture_dir, pix_fmt_of}; use crate::oracle::{MAX_DIFFERING, MIN_PSNR, TOL_MAX}; @@ -39,24 +39,22 @@ fn sizes(cfg: &EncoderConfig) -> (usize, usize, usize) { (w * h, w / 2 * ch, w / 2) } -/// Encode `raw` frames; returns the elementary stream and the reconstruction. -fn encode(cfg: &EncoderConfig, raw: &[u8]) -> (Vec, Vec) { +/// Encode `raw` frames; returns the elementary stream, the reconstruction (display order) and +/// the packets (decoding order). +fn encode(cfg: &EncoderConfig, raw: &[u8]) -> (Vec, Vec, Vec) { let (ys, cs, cw) = sizes(cfg); let fsize = ys + 2 * cs; let mut enc = Encoder::new(cfg.clone()).expect("config"); enc.set_recon_capture(true); - let mut es = Vec::new(); + let mut packets = Vec::new(); for (i, f) in raw.chunks_exact(fsize).enumerate() { let frame = YuvFrame { y: &f[..ys], cb: &f[ys..ys + cs], cr: &f[ys + cs..], y_stride: cfg.width as usize, c_stride: cw }; - for p in enc.encode(&frame, i as i64).expect("encode") { - es.extend_from_slice(&p.data); - } - } - for p in enc.flush() { - es.extend_from_slice(&p.data); + packets.extend(enc.encode(&frame, i as i64).expect("encode")); } + packets.extend(enc.flush()); + let mut es: Vec = packets.iter().flat_map(|p| p.data.iter().copied()).collect(); es.extend_from_slice(enc.end_of_sequence()); - (es, enc.take_recon()) + (es, enc.take_recon(), packets) } /// ffmpeg's decode of `file`, failing on any decoding error. @@ -121,10 +119,11 @@ fn check(ff: &Path, c: &Case) -> (Vec, f64) { let raw = source(ff, c.cfg.width, c.cfg.height, c.frames, pix_fmt); let (ys, cs, _) = sizes(&c.cfg); let fsize = ys + 2 * cs; - let (es, recon) = encode(&c.cfg, &raw); + let (es, recon, packets) = encode(&c.cfg, &raw); let path = write(c.name, &es); let decoded = ffmpeg_decode(ff, &path, pix_fmt); assert_eq!(decoded.len(), c.frames * fsize, "{}: ffmpeg's frame count", c.name); + assert_eq!(recon.len(), c.frames, "{}: reconstruction count", c.name); let mut all = Stats::default(); for (i, (r, f)) in recon.iter().zip(decoded.chunks_exact(fsize)).enumerate() { let mut s = Stats::default(); @@ -139,15 +138,21 @@ fn check(ff: &Path, c: &Case) -> (Vec, f64) { assert!(quality >= c.min_psnr, "{}: luma PSNR {quality:.2} dB against the source", c.name); if let Some((stream, types)) = ffprobe(&path) { assert_eq!(stream, c.probe, "{}: ffprobe", c.name); - let gop = c.cfg.gop_size as usize; - let want: Vec<&str> = (0..c.frames).map(|i| if i % gop == 0 { "I" } else { "P" }).collect(); + // ffprobe lists the pictures in display order + let mut shown: Vec<&Packet> = packets.iter().collect(); + shown.sort_by_key(|p| p.pts); + let want: Vec = shown.iter().map(|p| format!("{:?}", p.picture_type)).collect(); assert_eq!(types, want, "{}: picture types", c.name); } + let count = |t: &str| packets.iter().filter(|p| format!("{:?}", p.picture_type) == t).count(); let kbps = es.len() as f64 * 8.0 * 25.0 / c.frames as f64 / 1000.0; println!( - "{}: {} frames, {kbps:.0} kbit/s, luma PSNR {quality:.2} dB, max diff to ffmpeg {}, {:.3}% differ", + "{}: {} frames ({} I, {} P, {} B), {kbps:.0} kbit/s, luma PSNR {quality:.2} dB, max diff to ffmpeg {}, {:.3}% differ", c.name, c.frames, + count("I"), + count("P"), + count("B"), all.max, 100.0 * all.diff as f64 / all.n as f64 ); @@ -160,6 +165,21 @@ fn ffmpeg_decodes_our_streams() { let cif = EncoderConfig { gop_size: 12, ..EncoderConfig::new(352, 288, 25, 1) }; let cases = [ Case { name: "cif_q4", cfg: cif.clone(), frames: 25, probe: "mpeg2video,Main,10,yuv420p", min_psnr: 36.0 }, + Case { name: "cif_ip", cfg: EncoderConfig { b_pictures: 0, ..cif.clone() }, frames: 25, probe: "mpeg2video,Main,10,yuv420p", min_psnr: 36.0 }, + Case { + name: "cif_open_gop", + cfg: EncoderConfig { closed_gop: false, gop_size: 15, ..cif.clone() }, + frames: 31, + probe: "mpeg2video,Main,10,yuv420p", + min_psnr: 35.0, + }, + Case { + name: "cif_b4", + cfg: EncoderConfig { b_pictures: 4, gop_size: 10, ..cif.clone() }, + frames: 23, + probe: "mpeg2video,Main,10,yuv420p", + min_psnr: 34.0, + }, Case { name: "cif_intra", cfg: EncoderConfig { gop_size: 1, rate: RateControl::Quantiser(8), ..cif.clone() }, @@ -191,10 +211,14 @@ fn ffmpeg_decodes_our_streams() { #[test] fn average_bitrate_reaches_its_target() { let Some(ff) = craft_media_testkit::ffmpeg_or_skip("mpeg2 encoder bitrate") else { return }; - for kbps in [800u32, 2500] { + for (kbps, b) in [(800u32, 2), (2500, 2), (800, 0)] { let c = Case { - name: if kbps == 800 { "cif_abr_800" } else { "cif_abr_2500" }, - cfg: EncoderConfig { rate: RateControl::Abr { kbps }, gop_size: 12, ..EncoderConfig::new(352, 288, 25, 1) }, + name: match (kbps, b) { + (800, 0) => "cif_abr_800_ip", + (800, _) => "cif_abr_800", + _ => "cif_abr_2500", + }, + cfg: EncoderConfig { rate: RateControl::Abr { kbps }, gop_size: 12, b_pictures: b, ..EncoderConfig::new(352, 288, 25, 1) }, frames: 72, probe: "mpeg2video,Main,10,yuv420p", min_psnr: 26.0, diff --git a/crates/xfl/Cargo.toml b/crates/xfl/Cargo.toml index ddf2b3a36..9c8a40e3c 100644 --- a/crates/xfl/Cargo.toml +++ b/crates/xfl/Cargo.toml @@ -11,5 +11,6 @@ repository.workspace = true workspace = true [dependencies] +craft-xml = { workspace = true } miniz_oxide = "0.8" thiserror = { workspace = true } diff --git a/crates/xfl/src/document.rs b/crates/xfl/src/document.rs new file mode 100644 index 000000000..8cde65513 --- /dev/null +++ b/crates/xfl/src/document.rs @@ -0,0 +1,524 @@ +//! Neutral views of the native XFL document schema. +//! +//! This layer recognizes the document, stage, scene, layer and keyframe fields defined by XFL. +//! It does not construct a consumer's animation model. More detailed payloads remain available +//! through [`Document::root`] for the application adapter. +use craft_xml::{Doctype, El, Error as XmlError, Limits, Local, Options, Pat, UnknownEntity}; +use std::borrow::Cow; + +/// Maximum bytes of XML in one XFL document or library item. +pub const MAX_XML_BYTES: usize = 256 * 1024 * 1024; +/// Maximum depth accepted by AniCraft's existing XFL XML adapter (root depth is one). +pub const MAX_XML_DEPTH: usize = 257; +/// Maximum XFL document elements accepted by the existing adapter. +pub const MAX_XML_ELEMENTS: usize = 4_000_000; + +/// An XFL `DOMDocument.xml` tree plus typed, zero-copy views over its native structure. +#[derive(Clone, Debug)] +pub struct Document { + root: El, +} + +/// Stage metadata stored on the root `DOMDocument` element. +#[derive(Clone, Copy, Debug)] +pub struct Stage<'a> { + pub width: Option, + pub height: Option, + pub frame_rate: Option, + pub background_color: Option<&'a str>, +} + +/// The script-language version and document class selected by an XFL `PublishSettings.xml`. +/// Values stay in their native string form so each consumer can apply its own domain rules. +#[derive(Clone, Debug, Default, PartialEq, Eq)] +pub struct PublishSettings<'a> { + pub action_script_version: Option>, + pub document_class: Option>, +} + +/// Bounds used when projecting native schema fields into a compact model. +#[derive(Clone, Copy, Debug)] +pub struct ModelLimits { + pub max_timelines: usize, + pub max_layers_per_timeline: usize, +} + +/// Compact XFL metadata that does not retain or duplicate the parsed XML tree. +#[derive(Clone, Debug, PartialEq)] +pub struct DocumentModel { + pub stage: StageModel, + pub timelines: Vec, + /// Present for a library `DOMSymbolItem`; absent for normal `DOMDocument` files. + pub symbol_timeline: Option, +} + +#[derive(Clone, Copy, Debug, Default, PartialEq)] +pub struct StageModel { + pub width: Option, + pub height: Option, + pub frame_rate: Option, + pub has_background_color: bool, +} + +#[derive(Clone, Debug, Default, PartialEq)] +pub struct TimelineModel { + pub layers: Vec, +} + +#[derive(Clone, Copy, Debug, Default, PartialEq, Eq)] +pub enum LayerType { + #[default] + Normal, + Folder, + Guide, + Mask, + Other, +} + +#[derive(Clone, Debug, Default, PartialEq)] +pub struct LayerModel { + pub layer_type: LayerType, + pub parent_layer_index: Option, + pub visible: Option, + pub locked: Option, + pub outline: Option, + pub height_multiplier: Option, + pub frames: Vec, +} + +#[derive(Clone, Copy, Debug, Default, PartialEq, Eq)] +pub enum TweenType { + Motion, + Shape, + MotionObject, + #[default] + Other, +} + +#[derive(Clone, Copy, Debug, Default, PartialEq, Eq)] +pub struct FrameModel { + pub index: Option, + pub duration: Option, + pub is_comment: bool, + pub tween_type: TweenType, +} + +/// A timeline (`DOMTimeline`) listed under `DOMDocument/timelines`. +#[derive(Clone, Copy, Debug)] +pub struct Timeline<'a> { + element: &'a El, +} + +/// A native `DOMLayer` element. +#[derive(Clone, Copy, Debug)] +pub struct Layer<'a> { + element: &'a El, +} + +/// A native `DOMFrame` element. +#[derive(Clone, Copy, Debug)] +pub struct Frame<'a> { + element: &'a El, +} + +fn attr_last<'a>(element: &'a El, name: &str) -> Option<&'a str> { + element.attrs.iter().rev().find(|(key, _)| key == name).map(|(_, value)| value.as_str()) +} + +fn number_f32(element: &El, name: &str) -> Option { + attr_last(element, name).and_then(|value| value.trim().parse::().ok()).filter(|value| value.is_finite()) +} + +fn number_u32(element: &El, name: &str) -> Option { + attr_last(element, name).and_then(|value| value.trim().parse::().ok()) +} + +fn boolean(element: &El, name: &str) -> Option { + attr_last(element, name).map(|value| value.trim() == "true") +} + +fn trimmed_text<'a>(element: &'a El) -> Option> { + match element.text() { + Cow::Borrowed(text) => { + let trimmed = text.trim(); + (!trimmed.is_empty()).then_some(Cow::Borrowed(trimmed)) + } + Cow::Owned(text) => { + let trimmed = text.trim(); + (!trimmed.is_empty()).then(|| Cow::Owned(trimmed.to_string())) + } + } +} + +impl Document { + /// Parse XFL XML with AniCraft's existing size, depth, element, attribute, entity, doctype + /// and malformed-end-tag behavior. The shared XML parser owns byte parsing; this crate owns + /// the XFL-specific limits and schema views. + pub fn parse(bytes: &[u8]) -> Result { + let defaults = Limits::DEFAULT; + let limits = Limits { + max_input: MAX_XML_BYTES, + max_depth: MAX_XML_DEPTH, + max_nodes: MAX_XML_ELEMENTS.saturating_mul(2).saturating_add(1), + max_elements: MAX_XML_ELEMENTS, + max_attrs: defaults.max_attrs, + max_attrs_total: defaults.max_attrs_total, + }; + let options = + Options { limits, strict_end_tags: true, unknown_entity: UnknownEntity::Literal, doctype: Doctype::Ignore, ..Options::default() }; + Ok(Self { root: options.parse(bytes)? }) + } + + /// The lossless XML tree used by consumer adapters for schema details this view does not + /// interpret. Text and child order come directly from `craft-xml`. + pub fn root(&self) -> &El { + &self.root + } + + /// The root's local element name (normally `DOMDocument`). + pub fn root_name(&self) -> &str { + &self.root.name + } + + /// Reads the current publish profile, falling back to the first profile and then the root. + /// This is also used for `PublishSettings.xml` files, whose schema is separate from the main + /// `DOMDocument` tree. + pub fn publish_settings(&self) -> PublishSettings<'_> { + let profile = self + .root + .elements() + .find(|element| Local("flash_profile").matches(&element.name) && boolean(element, "current") == Some(true)) + .or_else(|| self.root.elements().find(|element| Local("flash_profile").matches(&element.name))) + .unwrap_or(&self.root); + let properties = profile.child(Local("PublishFlashProperties")).unwrap_or(profile); + let value = |name: &str| properties.child(Local(name)).and_then(trimmed_text); + PublishSettings { action_script_version: value("ActionScriptVersion"), document_class: value("DocumentClass") } + } + + /// The XFL stage metadata, applying finite-number checks but leaving product defaults and + /// clamping to the consumer. + pub fn stage(&self) -> Stage<'_> { + Stage { + width: number_f32(&self.root, "width"), + height: number_f32(&self.root, "height"), + frame_rate: number_f32(&self.root, "frameRate"), + background_color: attr_last(&self.root, "backgroundColor"), + } + } + + /// Timelines in document order. Non-`DOMTimeline` extension nodes are skipped. + pub fn timelines(&self) -> impl Iterator> { + self.root + .child(Local("timelines")) + .into_iter() + .flat_map(El::elements) + .filter(|element| Local("DOMTimeline").matches(&element.name)) + .map(|element| Timeline { element }) + } + + /// The timeline nested under `DOMSymbolItem/timeline`, if this is library symbol XML. + pub fn symbol_timeline(&self) -> Option> { + self.root.child(Local("timeline"))?.child(Local("DOMTimeline")).map(|element| Timeline { element }) + } + + /// Projects bounded schema metadata and returns the XML tree by ownership. Consumers can + /// convert that tree into their compatibility representation without keeping two full trees + /// alive during import. + pub fn into_parts(self, limits: ModelLimits) -> (DocumentModel, El) { + let stage = StageModel { + width: number_f32(&self.root, "width"), + height: number_f32(&self.root, "height"), + frame_rate: number_f32(&self.root, "frameRate"), + has_background_color: attr_last(&self.root, "backgroundColor").is_some(), + }; + let timelines = self.timelines().take(limits.max_timelines).map(|timeline| timeline.into_model(limits.max_layers_per_timeline)).collect(); + let symbol_timeline = self.symbol_timeline().map(|timeline| timeline.into_model(limits.max_layers_per_timeline)); + (DocumentModel { stage, timelines, symbol_timeline }, self.root) + } +} + +impl<'a> Timeline<'a> { + pub fn name(self) -> Option<&'a str> { + attr_last(self.element, "name") + } + + /// Raw timeline XML for application-specific payload conversion. + pub fn element(self) -> &'a El { + self.element + } + + /// Layers in native order. Non-`DOMLayer` extension nodes are skipped. + pub fn layers(self) -> impl Iterator> { + self.element + .child(Local("layers")) + .into_iter() + .flat_map(El::elements) + .filter(|element| Local("DOMLayer").matches(&element.name)) + .map(|element| Layer { element }) + } + + fn into_model(self, max_layers: usize) -> TimelineModel { + TimelineModel { layers: self.layers().take(max_layers).map(Layer::into_model).collect() } + } +} + +impl<'a> Layer<'a> { + pub fn name(self) -> Option<&'a str> { + attr_last(self.element, "name") + } + + pub fn layer_type(self) -> Option<&'a str> { + attr_last(self.element, "layerType") + } + + pub fn parent_layer_index(self) -> Option { + number_u32(self.element, "parentLayerIndex") + } + + pub fn visible(self) -> Option { + boolean(self.element, "visible") + } + + pub fn locked(self) -> Option { + boolean(self.element, "locked") + } + + pub fn outline(self) -> Option { + boolean(self.element, "outline") + } + + pub fn color(self) -> Option<&'a str> { + attr_last(self.element, "color") + } + + pub fn height_multiplier(self) -> Option { + number_u32(self.element, "heightMultiplier") + } + + /// Frames in native order. Non-`DOMFrame` extension nodes are skipped. + pub fn frames(self) -> impl Iterator> { + self.element + .child(Local("frames")) + .into_iter() + .flat_map(El::elements) + .filter(|element| Local("DOMFrame").matches(&element.name)) + .map(|element| Frame { element }) + } + + fn into_model(self) -> LayerModel { + let layer_type = match attr_last(self.element, "layerType") { + Some("folder") => LayerType::Folder, + Some("guide") => LayerType::Guide, + Some("mask") => LayerType::Mask, + Some("normal") | None => LayerType::Normal, + Some(_) => LayerType::Other, + }; + LayerModel { + layer_type, + parent_layer_index: number_u32(self.element, "parentLayerIndex"), + visible: boolean(self.element, "visible"), + locked: boolean(self.element, "locked"), + outline: boolean(self.element, "outline"), + height_multiplier: number_u32(self.element, "heightMultiplier"), + frames: self.frames().map(Frame::into_model).collect(), + } + } +} + +impl<'a> Frame<'a> { + pub fn index(self) -> Option { + number_u32(self.element, "index") + } + + pub fn duration(self) -> Option { + number_u32(self.element, "duration") + } + + pub fn label_type(self) -> Option<&'a str> { + attr_last(self.element, "labelType") + } + + pub fn name(self) -> Option<&'a str> { + attr_last(self.element, "name") + } + + pub fn tween_type(self) -> Option<&'a str> { + attr_last(self.element, "tweenType") + } + + /// Raw frame XML for application-specific shape, action, sound and tween conversion. + pub fn element(self) -> &'a El { + self.element + } + + fn into_model(self) -> FrameModel { + let tween_type = match attr_last(self.element, "tweenType") { + Some("motion") => TweenType::Motion, + Some("shape") => TweenType::Shape, + Some("motion object") => TweenType::MotionObject, + _ => TweenType::Other, + }; + FrameModel { + index: number_u32(self.element, "index"), + duration: number_u32(self.element, "duration"), + is_comment: attr_last(self.element, "labelType").is_some_and(|value| value.trim() == "comment"), + tween_type, + } + } +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn reads_stage_scene_layer_and_frame_schema_without_building_an_app_model() { + let bytes = br##"]> + + + + + + + + + + +"##; + let document = Document::parse(bytes).unwrap(); + assert_eq!(document.root_name(), "DOMDocument"); + let stage = document.stage(); + assert_eq!((stage.width, stage.height, stage.frame_rate), (Some(640.0), Some(480.0), Some(29.97))); + assert_eq!(stage.background_color, Some("#123456")); + let mut timelines = document.timelines(); + let timeline = timelines.next().unwrap(); + assert_eq!(timeline.name(), Some("Scene &custom;"), "declared entities stay literal"); + assert!(timelines.next().is_none()); + let mut layers = timeline.layers(); + let layer = layers.next().unwrap(); + assert_eq!(layer.name(), Some("Background")); + assert_eq!(layer.layer_type(), Some("normal")); + assert_eq!(layer.parent_layer_index(), Some(0)); + assert_eq!((layer.visible(), layer.locked(), layer.outline()), (Some(false), Some(true), Some(true))); + assert_eq!(layer.color(), Some("#abcdef")); + assert_eq!(layer.height_multiplier(), Some(2)); + let frame = layer.frames().next().unwrap(); + assert_eq!((frame.index(), frame.duration()), (Some(4), Some(3))); + assert_eq!((frame.name(), frame.label_type(), frame.tween_type()), (Some("Intro"), Some("name"), Some("motion"))); + assert_eq!(frame.element().child("elements").unwrap().child("Shape").unwrap().name, "Shape"); + } + + #[test] + fn uses_last_duplicate_attribute_like_the_legacy_xfl_adapter() { + let document = Document::parse(br##""##).unwrap(); + assert_eq!(document.stage().width, Some(2.0)); + } + + #[test] + fn preserves_lenient_doctype_entities_and_strict_end_tag_behavior() { + let document = Document::parse(br##"tail"##).unwrap(); + assert_eq!(document.stage().width, None); + assert!(Document::parse(b"").is_err()); + } + + #[test] + fn applies_the_existing_xfl_depth_and_attribute_budgets() { + let too_deep = format!("{}{}", "".repeat(MAX_XML_DEPTH + 1), "".repeat(MAX_XML_DEPTH + 1)); + assert!(matches!(Document::parse(too_deep.as_bytes()), Err(XmlError::TooDeep(_)))); + + let mut too_many_attrs = String::from(""); + assert!(matches!(Document::parse(too_many_attrs.as_bytes()), Err(XmlError::TooManyAttrs(_)))); + } + + #[test] + fn projects_neutral_schema_fields_and_returns_the_xml_tree_by_move() { + let document = Document::parse( + br##""##, + ) + .unwrap(); + let (model, root) = document.into_parts(ModelLimits { max_timelines: 1, max_layers_per_timeline: 1 }); + assert_eq!(model.stage, StageModel { width: Some(640.0), height: None, frame_rate: None, has_background_color: true }); + assert_eq!(model.timelines.len(), 1); + let layer = &model.timelines[0].layers[0]; + assert_eq!(layer.layer_type, LayerType::Mask); + assert_eq!(layer.visible, Some(false)); + assert_eq!(layer.frames[0], FrameModel { index: Some(7), duration: Some(2), is_comment: true, tween_type: TweenType::Shape }); + assert_eq!(root.attr("width"), Some("640")); + assert_eq!(root.child("timelines").unwrap().child("DOMTimeline").unwrap().attr("name"), Some("Scene")); + } + + #[test] + fn projects_symbol_timeline_under_the_native_symbol_wrapper() { + let document = Document::parse(b"").unwrap(); + let (model, _) = document.into_parts(ModelLimits { max_timelines: 10, max_layers_per_timeline: 10 }); + assert!(model.timelines.is_empty()); + assert_eq!(model.symbol_timeline.unwrap().layers[0].frames[0].index, Some(1)); + } + + #[test] + fn recognizes_prefixed_xfl_timeline_elements_by_local_name() { + let document = Document::parse( + b"", + ) + .unwrap(); + { + let mut timelines = document.timelines(); + let timeline = timelines.next().unwrap(); + assert!(timelines.next().is_none()); + let mut layers = timeline.layers(); + let layer = layers.next().unwrap(); + assert!(layers.next().is_none()); + let frame = layer.frames().next().unwrap(); + assert_eq!((frame.index(), frame.duration()), (Some(4), Some(2))); + } + let (model, _) = document.into_parts(ModelLimits { max_timelines: 1, max_layers_per_timeline: 1 }); + assert_eq!(model.timelines[0].layers[0].frames[0].index, Some(4)); + + let symbol = + Document::parse(b"").unwrap(); + assert!(symbol.symbol_timeline().is_some()); + } + + #[test] + fn reads_prefixed_publish_settings_children_by_local_name() { + let document = Document::parse( + b"3com.example.Main", + ) + .unwrap(); + let settings = document.publish_settings(); + assert_eq!(settings.action_script_version.as_deref(), Some("3")); + assert_eq!(settings.document_class.as_deref(), Some("com.example.Main")); + } + + #[test] + fn selects_current_publish_profile_and_uses_native_value_strings() { + let document = Document::parse( + br#"2OldClass 3 com.example.Main "#, + ) + .unwrap(); + let settings = document.publish_settings(); + assert_eq!(settings.action_script_version.as_deref(), Some("3")); + assert_eq!(settings.document_class.as_deref(), Some("com.example.Main")); + } + + #[test] + fn publish_settings_fall_back_to_first_profile_then_root() { + let first = Document::parse( + br#"23"#, + ) + .unwrap(); + assert_eq!(first.publish_settings().action_script_version.as_deref(), Some("2")); + + let root = Document::parse( + br#" 3 RootClass"#, + ) + .unwrap(); + let settings = root.publish_settings(); + assert_eq!(settings.action_script_version.as_deref(), Some("3")); + assert_eq!(settings.document_class.as_deref(), Some("RootClass")); + } +} diff --git a/crates/xfl/src/lib.rs b/crates/xfl/src/lib.rs index 83fdad40b..4a957111f 100644 --- a/crates/xfl/src/lib.rs +++ b/crates/xfl/src/lib.rs @@ -23,6 +23,8 @@ use std::cell::Cell; use std::collections::HashMap; +/// Neutral views of the XFL document, timeline, layer and frame schema. +pub mod document; /// XFL's edge-attribute grammar, independent of any consuming application's geometry types. pub mod edges; @@ -87,6 +89,19 @@ pub struct Archive<'a> { budget: Cell, } +/// The package root selected from an XFL archive's `DOMDocument.xml` paths. +#[derive(Clone, Debug, PartialEq, Eq)] +pub enum XflPackageRoot<'a> { + /// A document XML file is directly at the archive root. + Document, + /// The sole eligible top-level folder containing a document. + Folder(&'a str), + /// No document path was found. + Missing, + /// More than one eligible top-level folder contains a document, sorted by name. + Multiple(Vec<&'a str>), +} + const END: u32 = 0x0605_4b50; const END64: u32 = 0x0606_4b50; const LOCATOR64: u32 = 0x0706_4b50; @@ -527,6 +542,25 @@ impl<'a> Archive<'a> { self.entries.keys().map(String::as_str) } + /// Selects the package root from XFL document paths. A root-level document takes precedence; + /// otherwise only a single-level folder is eligible, with macOS metadata excluded. + pub fn xfl_package_root(&self) -> XflPackageRoot<'_> { + if self.entries.contains_key("DOMDocument.xml") { + return XflPackageRoot::Document; + } + let mut folders: Vec<&str> = self + .names() + .filter_map(|name| name.strip_suffix("/DOMDocument.xml")) + .filter(|folder| !folder.contains('/') && *folder != "__MACOSX") + .collect(); + folders.sort_unstable(); + match folders.as_slice() { + [] => XflPackageRoot::Missing, + [folder] => XflPackageRoot::Folder(folder), + _ => XflPackageRoot::Multiple(folders), + } + } + /// How many files and folder entries the archive lists. pub fn counts(&self) -> (usize, usize) { (self.entries.len(), self.folders) diff --git a/crates/xfl/src/tests.rs b/crates/xfl/src/tests.rs index ee4b2cf33..5ae0713ba 100644 --- a/crates/xfl/src/tests.rs +++ b/crates/xfl/src/tests.rs @@ -28,6 +28,32 @@ fn reads_stored_and_deflated_entries() { assert_eq!(a.read("missing").unwrap(), None); } +#[test] +fn xfl_package_root_prefers_root_then_selects_one_top_level_folder() { + let root = zip(&[("DOMDocument.xml", b"", true), ("Walk/DOMDocument.xml", b"", true)]); + let archive = Archive::open(&root).unwrap(); + assert_eq!(archive.xfl_package_root(), XflPackageRoot::Document); + + let folder = zip(&[("Walk/DOMDocument.xml", b"", true), ("Walk/LIBRARY/Clip.xml", b"", true)]); + let archive = Archive::open(&folder).unwrap(); + assert_eq!(archive.xfl_package_root(), XflPackageRoot::Folder("Walk")); +} + +#[test] +fn xfl_package_root_reports_sorted_candidates_and_ignores_metadata_and_nested_paths() { + let multiple = zip(&[ + ("Zebra/DOMDocument.xml", b"", true), + ("Apple/DOMDocument.xml", b"", true), + ("__MACOSX/DOMDocument.xml", b"not a project", true), + ]); + let archive = Archive::open(&multiple).unwrap(); + assert_eq!(archive.xfl_package_root(), XflPackageRoot::Multiple(vec!["Apple", "Zebra"])); + + let nested = zip(&[("One/Two/DOMDocument.xml", b"", true), ("__MACOSX/One/DOMDocument.xml", b"metadata", true)]); + let archive = Archive::open(&nested).unwrap(); + assert_eq!(archive.xfl_package_root(), XflPackageRoot::Missing); +} + #[test] fn the_checksum_matches_the_published_check_value() { assert_eq!(crc32(b"123456789"), 0xCBF4_3926); diff --git a/crates/xml/src/dtd.rs b/crates/xml/src/dtd.rs index 6bbca3554..9fa1dd497 100644 --- a/crates/xml/src/dtd.rs +++ b/crates/xml/src/dtd.rs @@ -3,11 +3,19 @@ //! Document Type Definitions: element declarations (content models) and required attributes, and //! validation of an XML document against them (Structure ▸ Validate). -use std::collections::HashMap; +use std::collections::{HashMap, HashSet}; /// How much text parameter entities may add during expansion (a pass count is not an output /// bound: each pass replaces whole strings, so references can inflate multiplicatively). const MAX_EXPANDED_BYTES: usize = 1024 * 1024; +const MAX_DTD_BYTES: usize = 1024 * 1024; +const MAX_VALIDATION_XML_BYTES: usize = 16 * 1024 * 1024; +const MAX_VALIDATION_NODES: usize = 100_000; +const MAX_VALIDATION_DEPTH: usize = 256; +const MAX_VALIDATION_ATTRIBUTES: usize = 1_000_000; +const MAX_MATCH_WORK: usize = 20_000_000; +const MAX_MATCH_POSITIONS: usize = 5_000_000; +const MAX_MATCH_STATES: usize = 100_000; /// A content particle: a name, a sequence, a choice, or one of those repeated (`?`, `*`, `+`). #[derive(Clone, Debug, PartialEq)] @@ -54,32 +62,38 @@ impl Dtd { } /// Markup declarations (``) with comments removed and quoted strings kept whole. -fn declarations(src: &str) -> Vec { +fn declarations(src: &str) -> Result, String> { let mut out = Vec::new(); let mut rest = src; while let Some(i) = rest.find("").map_or("", |j| &rest[j + 3..]); + let Some(j) = rest.find("-->") else { + return Err("unterminated DTD comment".into()); + }; + rest = &rest[j + 3..]; continue; } let mut quote = None; - let mut end = rest.len(); + let mut end = None; for (k, c) in rest.char_indices().skip(2) { match (quote, c) { (Some(q), c) if c == q => quote = None, (None, '"' | '\'') => quote = Some(c), (None, '>') => { - end = k; + end = Some(k); break; } _ => {} } } + let Some(end) = end else { + return Err("unterminated DTD declaration".into()); + }; out.push(rest[2..end].to_string()); - rest = rest.get(end + 1..).unwrap_or(""); + rest = &rest[end + 1..]; } - out + Ok(out) } fn tokens(s: &str) -> Vec { @@ -159,8 +173,48 @@ fn parse_model(s: &str) -> Result { _ => {} } if s.contains("#PCDATA") { - let t = tokens(&s.replace("#PCDATA", " #PCDATA ")); - return Ok(Model::Mixed(t.into_iter().filter(|x| !"()|*".contains(x.as_str()) && x != "#PCDATA").collect())); + let t = tokens(s); + if t.first().map(String::as_str) != Some("(") || t.get(1).map(String::as_str) != Some("#PCDATA") { + return Err("malformed mixed content model".into()); + } + let mut i = 2; + let mut allowed = Vec::new(); + if t.get(i).map(String::as_str) == Some(")") { + i += 1; + if t.get(i).map(String::as_str) == Some("*") { + i += 1; + } + if i != t.len() { + return Err("unexpected token after mixed content model".into()); + } + return Ok(Model::Mixed(allowed)); + } + loop { + if t.get(i).map(String::as_str) != Some("|") { + return Err("expected `|` in mixed content model".into()); + } + i += 1; + let Some(name) = t.get(i).filter(|name| !"(),|?*+".contains(name.as_str())) else { + return Err("expected element name in mixed content model".into()); + }; + if allowed.contains(name) { + return Err(format!("duplicate `{name}` in mixed content model")); + } + allowed.push(name.clone()); + i += 1; + match t.get(i).map(String::as_str) { + Some("|") => continue, + Some(")") => { + i += 1; + break; + } + _ => return Err("expected `|` or `)` in mixed content model".into()), + } + } + if t.get(i).map(String::as_str) != Some("*") || i + 1 != t.len() { + return Err("mixed content with element names must end in `*`".into()); + } + return Ok(Model::Mixed(allowed)); } let t = tokens(s); let mut i = 0; @@ -196,7 +250,33 @@ fn expand_parameter_entities(src: &str, ents: &[(String, String)]) -> Result") else { + return Err("unterminated DTD comment".into()); + }; + append(&mut out, &comment_tail[..end + 3])?; + rest = &comment_tail[end + 3..]; + continue; + } let (prefix, reference) = rest.split_at(i); append(&mut out, prefix)?; // `find` points to an ASCII `%`, so both split positions are UTF-8 boundaries. @@ -231,9 +311,12 @@ fn expand_parameter_entities(src: &str, ents: &[(String, String)]) -> Result Result { + if src.len() > MAX_DTD_BYTES { + return Err(format!("DTD exceeds the {MAX_DTD_BYTES} byte parsing limit")); + } // Parameter entities first, then expand references and read the declarations again. let mut ents: Vec<(String, String)> = Vec::new(); - for d in declarations(src) { + for d in declarations(src)? { if let Some(rest) = d.strip_prefix("ENTITY") { let rest = rest.trim_start(); if let Some(rest) = rest.strip_prefix('%') { @@ -249,7 +332,7 @@ pub fn parse(src: &str) -> Result { } let text = expand_parameter_entities(src, &ents)?; let mut dtd = Dtd::default(); - for d in declarations(&text) { + for d in declarations(&text)? { if let Some(rest) = d.strip_prefix("ELEMENT") { let rest = rest.trim_start(); let name: String = rest.chars().take_while(|c| !c.is_whitespace()).collect(); @@ -293,13 +376,31 @@ pub fn parse(src: &str) -> Result { } let Some((el, attrs)) = words.split_first() else { continue }; let mut k = 0; - while k + 2 < attrs.len() { + while k < attrs.len() { let name = &attrs[k]; - let default = attrs.get(k + 2).map(String::as_str).unwrap_or(""); + let Some(ty) = attrs.get(k + 1) else { + return Err(format!("incomplete ATTLIST for `{el}`")); + }; + k += 2; + // NOTATION is a two-token type: the keyword followed by its enumeration. + if ty == "NOTATION" { + if attrs.get(k).is_none_or(|group| !group.starts_with('(') || !group.ends_with(')')) { + return Err(format!("malformed NOTATION type for `{name}`")); + } + k += 1; + } + let Some(default) = attrs.get(k).map(String::as_str) else { + return Err(format!("missing default declaration for `{name}`")); + }; + k += 1; if default == "#REQUIRED" { dtd.required.entry(el.clone()).or_default().push(name.clone()); + } else if default == "#FIXED" { + if k >= attrs.len() { + return Err(format!("missing #FIXED value for `{name}`")); + } + k += 1; } - k += if default == "#FIXED" { 4 } else { 3 }; } } } @@ -313,18 +414,61 @@ pub fn parse(src: &str) -> Result { /// address of its node, which lives as long as the model) from a position. type EndsMemo = HashMap<(usize, usize), Vec>; +#[derive(Default)] +struct MatchBudget { + work: usize, + positions: usize, + states: usize, + exhausted: bool, + reported: bool, +} + +impl MatchBudget { + fn spend(&mut self, work: usize, positions: usize) -> Result<(), &'static str> { + if self.exhausted { + return Err("content model matching exceeded its resource budget"); + } + self.work = self.work.saturating_add(work); + self.positions = self.positions.saturating_add(positions); + if self.work > MAX_MATCH_WORK { + self.exhausted = true; + return Err("content model matching exceeded its work limit"); + } + if self.positions > MAX_MATCH_POSITIONS { + self.exhausted = true; + return Err("content model matching exceeded its memory limit"); + } + Ok(()) + } + + fn retain_state(&mut self) -> Result<(), &'static str> { + self.states = self.states.saturating_add(1); + if self.states > MAX_MATCH_STATES { + self.exhausted = true; + return Err("content model matching exceeded its state limit"); + } + Ok(()) + } +} + /// End positions after `cp` matches a prefix of `seq[i..]`. /// /// The positions of each group are worked out once per start position (`memo`): nested repetitions /// such as `((a*)*)*` ask the inner group for the same positions again and again, which is /// exponential in their depth without it. -fn ends(cp: &Cp, seq: &[&str], i: usize, memo: &mut EndsMemo) -> Vec { +fn ends(cp: &Cp, seq: &[&str], i: usize, memo: &mut EndsMemo, budget: &mut MatchBudget) -> Result, &'static str> { + budget.spend(1, 0)?; if let Cp::Name(n) = cp { - return if seq.get(i) == Some(&n.as_str()) { vec![i + 1] } else { vec![] }; + if seq.get(i) == Some(&n.as_str()) { + budget.spend(0, 1)?; + return Ok(vec![i + 1]); + } + return Ok(vec![]); } let key = (std::ptr::from_ref(cp) as usize, i); if let Some(done) = memo.get(&key) { - return done.clone(); + budget.spend(0, done.len())?; + return Ok(done.clone()); } let mut out: Vec = match cp { Cp::Name(_) => vec![], @@ -334,7 +478,9 @@ fn ends(cp: &Cp, seq: &[&str], i: usize, memo: &mut EndsMemo) -> Vec { for c in items { let mut next = Vec::new(); for p in at { - next.extend(ends(c, seq, p, memo)); + let found = ends(c, seq, p, memo, budget)?; + budget.spend(found.len(), found.len())?; + next.extend(found); } next.sort_unstable(); next.dedup(); @@ -345,22 +491,41 @@ fn ends(cp: &Cp, seq: &[&str], i: usize, memo: &mut EndsMemo) -> Vec { Cp::Choice(items) => { let mut all = Vec::new(); for c in items { - all.extend(ends(c, seq, i, memo)); + let found = ends(c, seq, i, memo, budget)?; + budget.spend(found.len(), found.len())?; + all.extend(found); } all } Cp::Rep(c, r) => { let mut all = if *r == '+' { vec![] } else { vec![i] }; - let mut frontier = ends(c, seq, i, memo); + let mut seen: HashSet = all.iter().copied().collect(); + let mut frontier = ends(c, seq, i, memo, budget)?; + budget.spend(frontier.len(), frontier.len())?; if *r == '?' { all.extend(frontier); } else { while !frontier.is_empty() { - let new: Vec = frontier.iter().copied().filter(|p| !all.contains(p)).collect(); + let mut new = Vec::new(); + for p in frontier.iter().copied() { + budget.spend(1, 0)?; + if seen.insert(p) { + budget.spend(0, 1)?; + new.push(p); + } + } + budget.spend(new.len(), new.len())?; all.extend(&new); let mut next = Vec::new(); + let mut next_seen = HashSet::new(); for p in new.into_iter().filter(|&p| p > i) { - next.extend(ends(c, seq, p, memo).into_iter().filter(|p| !all.contains(p))); + for found in ends(c, seq, p, memo, budget)? { + budget.spend(1, 0)?; + if !seen.contains(&found) && next_seen.insert(found) { + budget.spend(0, 1)?; + next.push(found); + } + } } // Two paths to one position are one position: kept apart they multiply per round. next.sort_unstable(); @@ -373,8 +538,10 @@ fn ends(cp: &Cp, seq: &[&str], i: usize, memo: &mut EndsMemo) -> Vec { }; out.sort_unstable(); out.dedup(); + budget.spend(out.len(), out.len())?; + budget.retain_state()?; memo.insert(key, out.clone()); - out + Ok(out) } fn show(cp: &Cp) -> String { @@ -392,60 +559,130 @@ struct Node { attrs: Vec, children: Vec, text: bool, + content: bool, +} + +fn bounded_node(e: &quick_xml::events::BytesStart<'_>, depth: usize, nodes: &mut usize, attributes: &mut usize) -> Result { + if depth >= MAX_VALIDATION_DEPTH { + return Err(format!("XML nesting exceeds the {MAX_VALIDATION_DEPTH} level validation limit")); + } + *nodes = nodes.saturating_add(1); + if *nodes > MAX_VALIDATION_NODES { + return Err(format!("XML exceeds the {MAX_VALIDATION_NODES} node validation limit")); + } + let mut attrs = Vec::new(); + for attr in e.attributes().with_checks(true) { + let attr = attr.map_err(|err| format!("invalid attribute: {err}"))?; + attrs.push(String::from_utf8_lossy(attr.key.as_ref()).to_string()); + } + *attributes = attributes.saturating_add(attrs.len()); + if *attributes > MAX_VALIDATION_ATTRIBUTES { + return Err(format!("XML exceeds the {MAX_VALIDATION_ATTRIBUTES} attribute validation limit")); + } + Ok(Node { name: String::from_utf8_lossy(e.name().as_ref()).to_string(), attrs, children: vec![], text: false, content: false }) } fn read_tree(xml: &str) -> Result { use quick_xml::events::Event; + if xml.len() > MAX_VALIDATION_XML_BYTES { + return Err(format!("XML exceeds the {} byte validation limit", MAX_VALIDATION_XML_BYTES)); + } let mut r = quick_xml::Reader::from_str(xml); let mut stack: Vec = Vec::new(); - let node = |e: &quick_xml::events::BytesStart| Node { - name: String::from_utf8_lossy(e.name().as_ref()).to_string(), - attrs: e.attributes().flatten().map(|a| String::from_utf8_lossy(a.key.as_ref()).to_string()).collect(), - children: vec![], - text: false, - }; + let mut root = None; + let mut nodes = 0usize; + let mut attributes = 0usize; loop { match r.read_event().map_err(|e| e.to_string())? { - Event::Start(e) => stack.push(node(&e)), + Event::Start(e) => { + if stack.is_empty() && root.is_some() { + return Err("multiple root elements".into()); + } + let n = bounded_node(&e, stack.len(), &mut nodes, &mut attributes)?; + if let Some(parent) = stack.last_mut() { + parent.content = true; + } + stack.push(n); + } Event::Empty(e) => { - let n = node(&e); + if stack.is_empty() && root.is_some() { + return Err("multiple root elements".into()); + } + let n = bounded_node(&e, stack.len(), &mut nodes, &mut attributes)?; match stack.last_mut() { - Some(p) => p.children.push(n), - None => return Ok(n), + Some(p) => { + p.content = true; + p.children.push(n); + } + None => root = Some(n), } } - Event::End(_) => { + Event::End(e) => { let n = stack.pop().ok_or("unbalanced end tag")?; + if n.name.as_bytes() != e.name().as_ref() { + return Err("mismatched end tag".into()); + } match stack.last_mut() { Some(p) => p.children.push(n), - None => return Ok(n), + None => root = Some(n), } } Event::Text(t) => { - if let Some(p) = stack.last_mut() - && !t.iter().all(u8::is_ascii_whitespace) - { - p.text = true; + if let Some(p) = stack.last_mut() { + p.content = true; + if !t.iter().all(u8::is_ascii_whitespace) { + p.text = true; + } + } else if !t.iter().all(u8::is_ascii_whitespace) { + return Err("text outside the root element".into()); } } - Event::CData(_) | Event::GeneralRef(_) => { + Event::CData(_) => { if let Some(p) = stack.last_mut() { p.text = true; + p.content = true; } } - Event::Eof => return Err("no root element".into()), + Event::GeneralRef(reference) => { + let name = reference.decode().map_err(|e| e.to_string())?; + let accepted = if let Some(character) = reference.resolve_char_ref().map_err(|e| e.to_string())? { + crate::is_xml_char(character) + } else { + matches!(name.as_ref(), "amp" | "lt" | "gt" | "apos" | "quot") + }; + if !accepted { + return Err(format!("unsupported general entity reference `&{name};`")); + } + if let Some(p) = stack.last_mut() { + p.text = true; + p.content = true; + } else { + return Err("reference outside the root element".into()); + } + } + Event::Comment(_) | Event::PI(_) => { + if let Some(parent) = stack.last_mut() { + parent.content = true; + } + } + Event::Eof => { + if !stack.is_empty() { + return Err("unclosed element".into()); + } + return root.ok_or_else(|| "no root element".into()); + } _ => {} } } } -fn check(dtd: &Dtd, n: &Node, path: &str, out: &mut Vec) { +fn check(dtd: &Dtd, n: &Node, path: &str, out: &mut Vec, budget: &mut MatchBudget) { let mut problem = |m: String| out.push(Problem { path: path.to_string(), message: m }); match dtd.model(&n.name) { None => problem(format!("`{}` isn't declared in the DTD", n.name)), Some(Model::Any) => {} Some(Model::Empty) => { - if n.text || !n.children.is_empty() { + if n.content { problem(format!("`{}` should be empty", n.name)); } } @@ -461,9 +698,19 @@ fn check(dtd: &Dtd, n: &Node, path: &str, out: &mut Vec) { problem(format!("`{}` can't contain text", n.name)); } let seq: Vec<&str> = n.children.iter().map(|c| c.name.as_str()).collect(); - if !ends(cp, &seq, 0, &mut EndsMemo::new()).contains(&seq.len()) { - let got = if seq.is_empty() { "nothing".to_string() } else { seq.join(", ") }; - problem(format!("`{}` contains {got}; the DTD asks for {}", n.name, show(cp))); + let matched = ends(cp, &seq, 0, &mut EndsMemo::new(), budget); + match matched { + Ok(positions) if positions.contains(&seq.len()) => {} + Ok(_) => { + let got = if seq.is_empty() { "nothing".to_string() } else { seq.join(", ") }; + problem(format!("`{}` contains {got}; the DTD asks for {}", n.name, show(cp))); + } + Err(error) => { + if !budget.reported { + problem(format!("`{}`: {error}", n.name)); + budget.reported = true; + } + } } } } @@ -476,17 +723,21 @@ fn check(dtd: &Dtd, n: &Node, path: &str, out: &mut Vec) { for c in &n.children { let k = seen.entry(c.name.as_str()).or_default(); *k += 1; - check(dtd, c, &format!("{path}/{}[{k}]", c.name), out); + check(dtd, c, &format!("{path}/{}[{k}]", c.name), out, budget); } } /// Validate `xml` against `dtd`; an empty list when it's valid. +/// +/// Validation rejects input above the internal byte, depth, node, attribute and content-model +/// matching budgets. Only predefined entities and numeric character references are resolved; a +/// general entity reference is reported as unsupported rather than expanded or fetched. pub fn validate(dtd: &Dtd, xml: &str) -> Vec { match read_tree(xml) { Ok(root) => { let mut out = Vec::new(); let path = root.name.clone(); - check(dtd, &root, &path, &mut out); + check(dtd, &root, &path, &mut out, &mut MatchBudget::default()); out } Err(e) => vec![Problem { path: String::new(), message: format!("not well-formed XML: {e}") }], @@ -745,4 +996,68 @@ mod tests { assert_eq!(problems.len(), 1, "{problems:?}"); assert!(problems[0].message.starts_with("`r` contains a, a, a"), "{problems:?}"); } + + #[test] + fn incomplete_declarations_and_malformed_mixed_models_are_rejected() { + assert!(parse("")).is_err(), "{model}"); + } + assert!(parse("").is_ok()); + assert!(parse("").is_ok()); + assert!(parse("").is_ok()); + } + + #[test] + fn parameter_entities_are_not_expanded_inside_comments() { + let dtd = parse( + r#" + +"#, + ) + .unwrap(); + assert_eq!(dtd.model("r"), Some(&Model::Empty)); + assert!(!validate(&dtd, "").is_empty()); + } + + #[test] + fn notation_attribute_type_keeps_following_required_attribute_aligned() { + let dtd = parse("").unwrap(); + assert!(validate(&dtd, "").iter().any(|p| p.message.contains("required attribute `id`"))); + assert!(validate(&dtd, "").is_empty()); + } + + #[test] + fn malformed_or_unbounded_xml_is_rejected_before_validation() { + let dtd = parse("").unwrap(); + for xml in ["", ""] { + assert!(!validate(&dtd, xml).is_empty(), "{xml}"); + } + let deep = format!("{}{}", "".repeat(MAX_VALIDATION_DEPTH + 1), "".repeat(MAX_VALIDATION_DEPTH + 1)); + let problems = validate(&dtd, &deep); + assert!(problems.iter().any(|p| p.message.contains("nesting exceeds")), "{problems:?}"); + } + + #[test] + fn references_must_be_declared_and_empty_elements_must_have_no_content() { + let dtd = parse("").unwrap(); + assert!(validate(&dtd, "&missing;").iter().any(|p| p.message.contains("unsupported general entity"))); + assert!(validate(&dtd, "&known;").iter().any(|p| p.message.contains("unsupported general entity"))); + assert!(validate(&dtd, "A&").is_empty()); + assert!(!validate(&dtd, "�").is_empty()); + for xml in [" ", "", "", "&"] { + assert!(validate(&dtd, xml).iter().any(|p| p.message.contains("should be empty")), "{xml}"); + } + assert!(validate(&dtd, "").is_empty()); + } + + #[test] + fn ambiguous_content_matching_stops_at_its_budget() { + let dtd = parse("").unwrap(); + let xml = format!("{}", "".repeat(10_000)); + let problems = validate(&dtd, &xml); + assert!(problems.iter().any(|p| p.message.contains("matching exceeded its ")), "{}", problems.len()); + } } diff --git a/sonar-project.properties b/sonar-project.properties index d08c71698..f30f3ae5f 100644 --- a/sonar-project.properties +++ b/sonar-project.properties @@ -15,6 +15,29 @@ sonar.cpd.exclusions=**/*tables.rs # S4790 (weak hash): the ICC specification defines the profile ID as the MD5 of the profile, so # no other algorithm produces an ID other software accepts. It identifies, it does not protect. -sonar.issue.ignore.multicriteria=icc_profile_id +# +# The encrypted packages below are the same case: each format derives its keys from MD5 or SHA-1, +# so substituting a stronger hash stops the files the module reads and writes from opening. +sonar.issue.ignore.multicriteria=icc_profile_id,xls_filepass_hash,pdf_standard_security_hash,odf_package_hash,gha-lockfile sonar.issue.ignore.multicriteria.icc_profile_id.ruleKey=rust:S4790 sonar.issue.ignore.multicriteria.icc_profile_id.resourceKey=crates/cms/src/write.rs +# xls: [MS-OFFCRYPTO] 2.3.6.1-2.3.6.2, where RC4 1.1 takes its key and verifier from MD5, and +# 2.3.5.1-2.3.5.2, where RC4 CryptoAPI takes them from SHA-1. +sonar.issue.ignore.multicriteria.xls_filepass_hash.ruleKey=rust:S4790 +sonar.issue.ignore.multicriteria.xls_filepass_hash.resourceKey=crates/xls/src/crypt.rs +# pdf: ISO 32000-1 7.6.3.3, Algorithms 2-7. Revisions 2-4 derive the file key and the /O and /U +# values from MD5; revisions 5 and 6 use SHA-256 and SHA-512 (Algorithm 2.B). +sonar.issue.ignore.multicriteria.pdf_standard_security_hash.ruleKey=rust:S4790 +sonar.issue.ignore.multicriteria.pdf_standard_security_hash.resourceKey=crates/pdf/src/crypt.rs +# odf: OpenDocument 4.16.6 and 4.16.3, a start key and a file checksum made with SHA-1, each used +# only where the package's manifest names it. +sonar.issue.ignore.multicriteria.odf_package_hash.ruleKey=rust:S4790 +sonar.issue.ignore.multicriteria.odf_package_hash.resourceKey=crates/odf/src/crypt.rs + +# githubactions:S7637 asks for every action to be pinned to a commit SHA. A workflow covered by +# .github/workflows/actions.lock runs the commit the lockfile records, so its tag refs are pinned +# and the lockfile covers the actions those actions use too. Sonar does not read lockfiles yet, so +# the rule flags the tag refs anyway: +# https://community.sonarsource.com/t/githubactions-s7637-false-positives-for-actions-pinned-through-native-actions-lock/190814 +sonar.issue.ignore.multicriteria.gha-lockfile.ruleKey=githubactions:S7637 +sonar.issue.ignore.multicriteria.gha-lockfile.resourceKey=.github/workflows/** diff --git a/vendor/README.craft-reader.md b/vendor/README.craft-reader.md index 3439657b5..8fc625fb9 100644 --- a/vendor/README.craft-reader.md +++ b/vendor/README.craft-reader.md @@ -444,3 +444,18 @@ a decoder's limit from damage). Not ported from those commits: the page-wide all refused-mask handling in `hayro-syntax` and `hayro-interpret` (0ed33990, 4aa732ea, e2464e14), which rest on the decoded-image cache of 558ffd7f (a performance change). Tests: those that came with the patches in `hayro-jbig2`, and PdfCraft's raster tests adapted in `hayro-jbig2/src/pdfcraft_tests.rs`. + +## Craft review note: the one direct `unsafe` + +`hayro-syntax/src/page.rs` holds the snapshot's only direct `unsafe` block: a +`core::mem::transmute` of `&XRef` to `&'static XRef` in the page cache. The vendored crates are +outside the workspace's `forbid(unsafe_code)`, so it compiles. Its safety argument is the field +order of the owning struct together with the `Arc` outliving the borrow. Treat it as an +audited exception, not a pattern to copy: if the caching code is re-synced, re-read that +invariant and re-run the PDF tests that drop pages while a cache entry is live. + +`hayro-syntax` also has an `unsafe` cargo feature that pulls in `flate2`, `memchr` and +SIMD-capable decoders. `craft-pdf/read` depends on it with `default-features = false, +features = ["std"]`, so the feature stays off; a review of the resolved graph should keep it +that way. The CI check for that resolved feature set is a CI change under AGENTS.md rule 11 and +is proposed separately rather than in a fix pull request. diff --git a/vendor/hayro-syntax/src/pdf.rs b/vendor/hayro-syntax/src/pdf.rs index a13e3f669..88b881b7f 100644 --- a/vendor/hayro-syntax/src/pdf.rs +++ b/vendor/hayro-syntax/src/pdf.rs @@ -186,6 +186,16 @@ impl Pdf { pub fn metadata(&self) -> &Metadata { self.xref.metadata() } + + /// Return the document information dictionary resolved through the final trailer and xref. + pub fn info(&self) -> Option> { + self.xref.info_id().and_then(|id| self.xref.get(id)) + } + + /// Whether the resolved trailer declares the document encrypted. + pub fn is_encrypted(&self) -> bool { + self.xref.is_encrypted() + } } fn find_version(data: &[u8]) -> Option { diff --git a/vendor/hayro-syntax/src/xref.rs b/vendor/hayro-syntax/src/xref.rs index 5fced8a88..0d2dadb64 100644 --- a/vendor/hayro-syntax/src/xref.rs +++ b/vendor/hayro-syntax/src/xref.rs @@ -449,6 +449,8 @@ impl XRef { let td = TrailerData { pages_ref: pages_ref.into(), root_ref: root_ref.into(), + info_ref: trailer_dict.get_ref(INFO).map(Into::into), + encrypted: trailer_dict.contains_key(ENCRYPT), version, }; @@ -461,6 +463,8 @@ impl XRef { let td = TrailerData { pages_ref: pages_ref.into(), root_ref, + info_ref: None, + encrypted: false, version: None, }; @@ -526,6 +530,16 @@ impl XRef { self.trailer_data().root_ref } + /// Return the document information dictionary reference from the resolved trailer, if any. + pub fn info_id(&self) -> Option { + self.trailer_data().info_ref + } + + /// Whether the resolved trailer has an `/Encrypt` entry. + pub fn is_encrypted(&self) -> bool { + self.trailer_data().encrypted + } + /// Whether the PDF has optional content groups. pub fn has_optional_content_groups(&self) -> bool { match &self.0 { @@ -885,6 +899,8 @@ struct MapRepr { pub(crate) struct TrailerData { pub(crate) pages_ref: ObjectIdentifier, pub(crate) root_ref: ObjectIdentifier, + pub(crate) info_ref: Option, + pub(crate) encrypted: bool, pub(crate) version: Option, } @@ -893,6 +909,8 @@ impl TrailerData { Self { pages_ref: ObjectIdentifier::new(0, 0), root_ref: ObjectIdentifier::new(0, 0), + info_ref: None, + encrypted: false, version: None, } }