Vendor Rust scraper library

components/web_discovery/browser/document_extractor/Cargo.toml

@@ -7,8 +7,7 @@ license = "MPL-2.0"
 
 [dependencies]
 cxx = "1"
-kuchikiki = "0.8.6-speedreader"
-html5ever = "0.29.1"
+scraper = "0.22"
 
 [lib]
 name = "document_extractor_cxx"

third_party/rust/chromium_crates_io/gnrt_config.toml

@@ -151,9 +151,13 @@ if (is_apple && target_cpu == "arm64") {
 build_script_outputs = [ "lib.rs" ]
 
 [crate.cssparser]
-build_script_outputs = [ "tokenizer.rs" ]
-extra_build_script_src_roots = [ "../build" ]
 license_files = ['LICENSE']
+extra_kv = { raw_gn = '''
+if (epoch == "0.28" || epoch == "0.27") {
+    build_script_inputs = [ "//brave/third_party/rust/chromium_crates_io/vendor/cssparser-" + cargo_pkg_version + "/build/match_byte.rs" ]
+    build_script_outputs = [ "tokenizer.rs" ]
+  }
+''' }
 
 [crate.cssparser-macros]
 license_files = ['LICENSE']
@@ -300,6 +304,13 @@ license_files = ['LICENSE-Apache']
 [crate.num-traits]
 extra_kv = { rustflags = [ "--cfg", "has_i128", "--cfg", "has_to_int_unchecked", "--cfg", "has_reverse_bits", "--cfg", "has_leading_trailing_ones", "--cfg", "has_int_assignop_ref", "--cfg", "has_div_euclid", "--cfg", "has_copysign" ] }
 
+[crate.phf_macros]
+extra_kv = { raw_gn = '''
+if (epoch == "0.11") {
+  output_name = "phf_macros_v0_11"
+}
+''' }
+
 [crate.powerfmt]
 license_files = ['LICENSE-Apache']
 
@@ -391,6 +402,9 @@ extra_kv = { rustenv = [ "TYPENUM_BUILD_CONSTS=\" + rebase_path(target_gen_dir)
 [crate.uint]
 license_files = ['../../../../../common/licenses/Apache-2.0']
 
+[crate.unicode-width]
+license_files = ['../../../../../common/licenses/Apache-2.0']
+
 [crate.wasi]
 license_files = ['../../../../../common/licenses/Apache-2.0']
 

third_party/rust/chromium_crates_io/vendor/cssparser-0.34.0/.cargo_vcs_info.json

@@ -0,0 +1,6 @@
+{
+  "git": {
+    "sha1": "5dc6d916a73159745e1fc1dd1f745e05d3130b0f"
+  },
+  "path_in_vcs": ""
+}

third_party/rust/chromium_crates_io/vendor/cssparser-0.34.0/.github/workflows/main.yml

@@ -0,0 +1,81 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+  workflow_dispatch:
+  merge_group:
+    types: [checks_requested]
+
+jobs:
+  linux-ci:
+    name: Linux
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        toolchain:
+          - nightly
+          - beta
+          - stable
+          - 1.63.0
+        features:
+          -
+          - --features dummy_match_byte
+        include:
+          - toolchain: nightly
+            features: --features bench
+          - toolchain: nightly
+            features: --features bench,dummy_match_byte
+    steps:
+      - uses: actions/checkout@v2
+
+      - name: Install toolchain
+        uses: actions-rs/toolchain@v1
+        with:
+          profile: minimal
+          toolchain: ${{ matrix.toolchain }}
+          override: true
+          components: rustfmt, clippy, ${{ matrix.toolchain == 'nightly' && 'miri,rust-src' || '' }}
+
+      - name: Cargo format & lint
+        run: cargo fmt --check && cargo clippy -- -Dwarnings
+
+      - name: Cargo build
+        run: cargo build ${{ matrix.features }}
+
+      - name: Cargo doc
+        run: cargo doc ${{ matrix.features }}
+
+      - name: Cargo test
+        run: cargo test ${{ matrix.features }}
+
+      - name: macros build
+        run: cargo build
+        working-directory: macros
+
+      - name: Color build
+        run: cargo build
+        working-directory: color
+
+      - name: Color test
+        run: cargo test
+        working-directory: color
+
+      - name: Cargo miri test
+        if: "matrix.toolchain == 'nightly'"
+        run: cargo miri test --features skip_long_tests ${{ matrix.features }}
+
+  build_result:
+    name: Result
+    runs-on: ubuntu-latest
+    needs:
+      - "linux-ci"
+
+    steps:
+      - name: Mark the job as successful
+        run: exit 0
+        if: success()
+      - name: Mark the job as unsuccessful
+        run: exit 1
+        if: "!success()"

third_party/rust/chromium_crates_io/vendor/cssparser-0.34.0/.gitignore

@@ -0,0 +1,3 @@
+target
+/Cargo.lock
+/.cargo/config

third_party/rust/chromium_crates_io/vendor/cssparser-0.34.0/Cargo.toml

@@ -0,0 +1,70 @@
+# THIS FILE IS AUTOMATICALLY GENERATED BY CARGO
+#
+# When uploading crates to the registry Cargo will automatically
+# "normalize" Cargo.toml files for maximal compatibility
+# with all versions of Cargo and also rewrite `path` dependencies
+# to registry (e.g., crates.io) dependencies.
+#
+# If you are reading this file be aware that the original Cargo.toml
+# will likely look very different (and much more reasonable).
+# See Cargo.toml.orig for the original contents.
+
+[package]
+edition = "2018"
+rust-version = "1.63"
+name = "cssparser"
+version = "0.34.0"
+authors = ["Simon Sapin <simon.sapin@exyr.org>"]
+exclude = [
+    "src/css-parsing-tests/**",
+    "src/big-data-url.css",
+]
+description = "Rust implementation of CSS Syntax Level 3"
+documentation = "https://docs.rs/cssparser/"
+readme = "README.md"
+keywords = [
+    "css",
+    "syntax",
+    "parser",
+]
+license = "MPL-2.0"
+repository = "https://github.com/servo/rust-cssparser"
+
+[profile.profiling]
+debug = 2
+inherits = "release"
+
+[dependencies.cssparser-macros]
+version = "0.6.1"
+
+[dependencies.dtoa-short]
+version = "0.3"
+
+[dependencies.itoa]
+version = "1.0"
+
+[dependencies.phf]
+version = "0.11.2"
+features = ["macros"]
+
+[dependencies.serde]
+version = "1.0"
+features = ["derive"]
+optional = true
+
+[dependencies.smallvec]
+version = "1.0"
+
+[dev-dependencies.difference]
+version = "2.0"
+
+[dev-dependencies.encoding_rs]
+version = "0.8"
+
+[dev-dependencies.serde_json]
+version = "1.0.25"
+
+[features]
+bench = []
+dummy_match_byte = []
+skip_long_tests = []