Test parsing for external sources on CI

.github/workflows/ci.yml

@@ -0,0 +1,35 @@
+name: CI
+
+on:
+  push:
+    branches:
+      - main
+
+  pull_request:
+    branches:
+      - main
+    types:
+      - opened
+      - synchronize
+      - reopened
+      - unlocked
+      - ready_for_review
+
+jobs:
+  test-parsing:
+    if: github.event.pull_request.draft != true
+    concurrency:
+      group: ci-${{ github.event.pull_request.number || github.ref }}
+      cancel-in-progress: true
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout source code
+        uses: actions/checkout@v3
+
+      - name: Install the tools
+        run: npm install
+
+      - name: Run the parsing tests
+        run: |
+          npm run generate
+          npm run test-parsing

.gitignore

@@ -8,3 +8,4 @@ target
 /bindings/c/tree-sitter-*.pc
 /build
 /node_modules
+/test/sources

package.json

@@ -14,9 +14,10 @@
     "tree-sitter-cli": "^0.20.1"
   },
   "scripts": {
-    "test": "tree-sitter test"
-  }
-  ,
+    "test": "tree-sitter test",
+    "generate": "tree-sitter generate",
+    "test-parsing": "scripts/test-parsing"
+  },
   "tree-sitter": [
     {
       "scope": "source.zig",

scripts/known-parsing-failures.txt

@@ -0,0 +1,47 @@
+zig/test/cases/safety/slice with sentinel out of bounds - runtime len.zig
+zig/test/behavior/struct.zig
+zig/test/behavior/slice.zig
+zig/test/behavior/array.zig
+zig/test/behavior/pointers.zig
+zig/test/behavior/slice_sentinel_comptime.zig
+zig/test/behavior/bugs/12972.zig
+zig/test/behavior/packed_struct_explicit_backing_int.zig
+zig/lib/std/os.zig
+zig/lib/std/array_hash_map.zig
+zig/lib/std/hash/auto_hash.zig
+zig/lib/std/fs.zig
+zig/lib/std/mem.zig
+zig/lib/std/zig/c_translation.zig
+zig/lib/std/zig/system/windows.zig
+zig/lib/std/os/test.zig
+zig/lib/std/os/linux/bpf/btf.zig
+zig/lib/std/os/linux/bpf.zig
+zig/lib/std/os/windows.zig
+zig/lib/std/os/uefi/protocols/ip6_protocol.zig
+zig/lib/std/os/uefi/protocols/hii.zig
+zig/lib/std/os/uefi/protocols/simple_text_input_ex_protocol.zig
+zig/lib/std/os/uefi/protocols/simple_network_protocol.zig
+zig/lib/std/os/uefi/protocols/absolute_pointer_protocol.zig
+zig/lib/std/os/uefi/protocols/edid_override_protocol.zig
+zig/lib/std/os/uefi/tables/boot_services.zig
+zig/lib/std/unicode.zig
+zig/lib/std/meta/trait.zig
+zig/lib/std/Thread.zig
+zig/lib/std/mem/Allocator.zig
+zig/lib/std/json/test.zig
+zig/lib/std/child_process.zig
+zig/lib/std/macho.zig
+zig/lib/std/process.zig
+zig/lib/std/dwarf.zig
+zig/lib/std/json.zig
+zig/lib/std/c/openbsd.zig
+zig/lib/std/meta.zig
+zig/src/main.zig
+zig/src/link/Wasm/Object.zig
+zig/src/link/Wasm/Archive.zig
+zig/src/link/MachO/Archive.zig
+zig/src/Air.zig
+zig/src/codegen/llvm/bindings.zig
+zig/src/Autodoc.zig
+zig/src/Zir.zig
+zig/tools/process_headers.zig

scripts/test-parsing

@@ -0,0 +1,90 @@
+#!/bin/env bash
+# adapted from tree-sitter-typescript/script/parse-examples
+
+set -Eeu -o pipefail
+shopt -s inherit_errexit
+
+this_file="${BASH_SOURCE[0]}"
+if [ "${this_file::1}" != '/' ]; then
+  this_file="$PWD/$this_file"
+fi
+
+this_dir="${this_file%/*}"
+project_root="${this_dir%/*}"
+sources_dir="$project_root/test/sources"
+
+cd "$project_root"
+
+clone_repo() {
+  local owner=$1
+  local name=$2
+  local sha=$3
+  local path="$sources_dir/$name"
+
+  if [ -d "$path" ]; then
+    pushd "$path" > /dev/null
+    if [ "$(git rev-parse HEAD 2>/dev/null)" == "$sha"  ]; then
+      popd > /dev/null
+      return
+    else
+      popd > /dev/null
+      rm -rf "$path"
+      echo "Updating $owner/$name to $sha"
+    fi
+  else
+    echo "Cloning $owner/$name"
+  fi
+
+  mkdir -p "$path"
+  pushd "$path" > /dev/null
+  git init --quiet
+  git remote add origin "https://github.com/$owner/$name"
+  git pull --quiet --ff-only --depth 1 origin "$sha"
+  popd > /dev/null
+}
+
+# zig 0.10.1 -> https://github.com/ziglang/zig/releases/tag/0.10.1
+clone_repo ziglang zig b57081f039bd3f8f82210e8896e336e3c3a6869b
+
+known_failures=()
+while IFS= read -r line; do
+  if [[ "$line" =~ [^[:space:]] ]]; then
+    full_path="$sources_dir/$line"
+    if [ -e "$full_path" ]; then
+      if tree-sitter parse -q "$full_path" &>/dev/null; then
+        >&2 echo "File $full_path can be parsed without errors, but it's listed in $this_dir/known-parsing-failures.txt"
+        saw_error=true
+      else
+        known_failures+=("$sources_dir/$line")
+      fi
+    else
+      >&2 echo "File $full_path (listed in $this_dir/known-parsing-failures.txt) does not exist"
+      saw_error=true
+    fi
+  fi
+done < "$this_dir/known-parsing-failures.txt"
+if [ "${saw_error:-}" ]; then
+  exit 1
+fi
+
+while IFS= read -r line; do
+  for known_failure in "${known_failures[@]}"; do
+    if [ "$known_failure" == "$line" ]; then
+      continue 2
+    fi
+  done
+  parse_args+=("$line")
+done < <(
+  find "$sources_dir" \
+    -name "*.zig" \
+    -not -path "$sources_dir/zig/test/cases/compile_errors/**"
+)
+if [ ${#parse_args[*]} -eq 0 ]; then
+  >&2 echo "No files found to test"
+  exit 1
+fi
+
+echo "[STEP] The following files will be tested:"
+printf "%s\n" "${parse_args[@]}"
+
+tree-sitter parse -q "${parse_args[@]}"