CMake: set C++17 standard for benchmarks (required by newer re2).

This fixes https://github.com/skvadrik/re2c/issues/455 "Test for C++11
compiler at configure time".

.gitattributes

@@ -0,0 +1,2 @@
+* text=auto
+test/newlines/** -text

.github/workflows/benchmarks.yml

@@ -33,15 +33,15 @@ jobs:
 
     steps:
       - name: Checkout Code
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
 
       - name: Setup Python
-        uses: actions/setup-python@v2
+        uses: actions/setup-python@v3
         with:
           python-version: '3.8'
 
       - name: Setup Java
-        uses: actions/setup-python@v2
+        uses: actions/setup-java@v3
         with:
           distribution: 'adopt'
           java-version: '11'
@@ -52,6 +52,7 @@ jobs:
             Linux*)
               echo "CC=gcc" >> $GITHUB_ENV
               echo "CXX=g++" >> $GITHUB_ENV
+              sudo apt update
               sudo apt install libre2-dev libbenchmark-dev
               ;;
             macOS*)
@@ -71,6 +72,7 @@ jobs:
             -G "Unix Makefiles" \
             -DRE2C_BUILD_LIBS=ON \
             -DRE2C_BUILD_RE2GO=OFF \
+            -DRE2C_BUILD_RE2RUST=OFF \
             -DRE2C_BUILD_BENCHMARKS=ON \
             -DCMAKE_CXX_FLAGS="-O3"
 

.github/workflows/ci-autotools.yml

@@ -0,0 +1,158 @@
+name: CI autotools
+
+on:
+  push:
+  pull_request:
+
+defaults:
+  run:
+    shell: bash
+
+jobs:
+  build-and-test:
+    name: ${{ matrix.name }}
+    runs-on: ${{ matrix.os }}
+
+    strategy:
+      # When set to true, GitHub cancels
+      # all in-progress jobs if any matrix job fails.
+      fail-fast: false
+
+      matrix:
+
+        include:
+          - name: linux-gcc-intree
+            os: ubuntu-latest
+
+          - name: linux-gcc-ootree
+            os: ubuntu-latest
+
+          # macOS
+          - name: macos-clang-intree
+            os: macos-latest
+
+          - name: macos-clang-ootree
+            os: macos-latest
+
+    steps:
+      - name: Checkout Code
+        uses: actions/checkout@v3
+
+      - name: Setup apt cache (Linux)
+        if: startsWith(runner.os, 'Linux')
+        uses: actions/cache@v3
+        with:
+          path: /var/cache/apt
+          key: ${{ runner.os }}-apt-cache-${{ matrix.os }}-${{ github.sha }}
+          restore-keys: |
+            ${{ runner.os }}-apt-cache-${{ matrix.os }}-
+            ${{ runner.os }}-apt-cache-
+
+      - name: Setup Common Prerequisites
+        run: |
+          case ${{ runner.os }} in
+            Linux*)
+              sudo apt update
+              sudo apt install autoconf automake libtool bison
+              ;;
+            macOS*)
+              sudo xcode-select -switch /Applications/Xcode.app
+              echo "SDKROOT=$(xcodebuild -version -sdk macosx Path)" >> $GITHUB_ENV
+              echo "$(xcodebuild -version -sdk macosx Path)" >> $GITHUB_PATH
+
+              brew install autoconf automake libtool bison coreutils
+              # bison is not activated by default to avoid clobbering
+              # already present outdated system bison. Let's prepend
+              # brew's path.
+              echo "$(brew --prefix bison)/bin" >> $GITHUB_PATH
+              ;;
+          esac
+
+      - name: Setup Python
+        uses: actions/setup-python@v3
+        with:
+          python-version: 3.x
+          architecture: x64
+
+      - name: Install Python Dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install docutils pygments
+
+      - name: Generate ./configure
+        run: ./autogen.sh
+
+      - name: Fast Configure
+        run: |
+          ./configure \
+              --disable-golang \
+              --disable-rust \
+              --prefix=$PWD/install
+
+      - name: Fast Build
+        run: make -j$(nproc)
+
+      - name: Fast Install
+        run: make -j$(nproc) install
+
+      - name: Fast Clean
+        run: make -j$(nproc) distclean
+
+      - name: Minimal Install Test
+        working-directory: ./install/bin
+        run: ./re2c --version
+
+      - name: Get build dir
+        id: build-info
+        run: |
+          if echo "${{ matrix.name }}" | grep -q "ootree"; then
+            BUILD_DIR="$(pwd)/.build/${{ matrix.name }}-full"
+            mkdir -p "$BUILD_DIR"
+          else
+            BUILD_DIR="$(pwd)"
+          fi
+          echo "BUILD_DIR=$BUILD_DIR" >> $GITHUB_OUTPUT
+
+      - name: Full Configure
+        working-directory: ${{ steps.build-info.outputs.BUILD_DIR }}
+        run: |
+          $GITHUB_WORKSPACE/configure \
+              --prefix=$PWD/full-install \
+              --enable-libs \
+              --enable-parsers \
+              --enable-lexers \
+              --enable-docs \
+              --enable-debug \
+              RE2C_FOR_BUILD=$GITHUB_WORKSPACE/install/bin/re2c
+
+      - name: Full Build
+        working-directory: ${{ steps.build-info.outputs.BUILD_DIR }}
+        run: |
+          find $GITHUB_WORKSPACE/src -name '*.re' | xargs touch
+          make -j$(nproc)
+
+      - name: Run Main Test Suite
+        working-directory: ${{ steps.build-info.outputs.BUILD_DIR }}
+        run: bash -c "ulimit -s 256; make check -j$(nproc)"
+
+      - name: Run Skeleton Tests
+        working-directory: ${{ steps.build-info.outputs.BUILD_DIR }}
+        run: python run_tests.py --skeleton
+
+      - name: Full Install
+        working-directory: ${{ steps.build-info.outputs.BUILD_DIR }}
+        run: make -j$(nproc) install
+
+      - name: Archive test artifacts
+        # Run this step only if tests failed (otherwise there is nothing to upload).
+        if: failure()
+        uses: actions/upload-artifact@v3
+        with:
+          # To ensure that jobs don't overwrite existing artifacts,
+          # use a different name per job.
+          name: test-${{ matrix.name }}
+          # Tests run in a temporary directory that has a name test_<timestamp>.
+          path: ${{ steps.build-info.outputs.BUILD_DIR }}/test_[0-9]*
+          # Artifacts are retained for 90 days by default.
+          # In fact, we don't need such long period.
+          retention-days: 30

.github/workflows/ci-cmake.yml

@@ -0,0 +1,177 @@
+name: CI CMake
+
+on:
+  push:
+  pull_request:
+
+defaults:
+  run:
+    shell: bash
+
+jobs:
+  build-and-test:
+    name: ${{ matrix.name }}
+    runs-on: ${{ matrix.os }}
+
+    strategy:
+      # When set to true, GitHub cancels
+      # all in-progress jobs if any matrix job fails.
+      fail-fast: false
+
+      matrix:
+
+        include:
+          # Linux
+          - name: linux-gcc-ubsan-ootree
+            os: ubuntu-latest
+
+          - name: linux-gcc-asan-ootree
+            os: ubuntu-latest
+
+          - name: linux-gcc-valgrind-ootree
+            os: ubuntu-latest
+
+          - name: linux-gcc-debug-ootree-skeleton
+            os: ubuntu-latest
+
+          - name: linux-gcc-debug-intree-m32-skeleton
+            os: ubuntu-latest
+
+          - name: linux-gcc-release-ootree-skeleton
+            os: ubuntu-latest
+
+          # macOS
+          - name: macos-clang-debug-ootree-skeleton
+            os: macos-latest
+
+          - name: macos-clang-debug-intree-skeleton
+            os: macos-latest
+
+          - name: macos-clang-release-ootree-skeleton
+            os: macos-latest
+
+          # Windows
+          - name: windows-msvc-debug-ootree
+            os: windows-latest
+
+          - name: windows-msvc-debug-intree
+            os: windows-latest
+
+          - name: windows-msvc-release-ootree
+            os: windows-latest
+
+    steps:
+      - name: Checkout Code
+        uses: actions/checkout@v3
+
+      - name: Setup chocolatey Cache (Windows)
+        if: startsWith(runner.os, 'Windows')
+        uses: actions/cache@v3
+        with:
+          path: C:\Users\runneradmin\AppData\Local\Temp\chocolatey
+          key: ${{ runner.os }}-chocolatey-${{ matrix.os }}-${{ github.sha }}
+          restore-keys: |
+            ${{ runner.os }}-chocolatey-${{ matrix.os }}-
+            ${{ runner.os }}-chocolatey-
+
+      - name: Setup apt cache (Linux)
+        if: startsWith(runner.os, 'Linux')
+        uses: actions/cache@v3
+        with:
+          path: /var/cache/apt
+          key: ${{ runner.os }}-apt-cache-${{ matrix.os }}-${{ github.sha }}
+          restore-keys: |
+            ${{ runner.os }}-apt-cache-${{ matrix.os }}-
+            ${{ runner.os }}-apt-cache-
+
+      - name: Setup Common Prerequisites
+        run: |
+          case ${{ runner.os }} in
+            Linux*)
+              sudo apt update
+              sudo apt install gcc-multilib g++-multilib  # for -m32
+              sudo apt install valgrind
+              ;;
+            macOS*)
+              sudo xcode-select -switch /Applications/Xcode.app
+              echo "SDKROOT=$(xcodebuild -version -sdk macosx Path)" >> $GITHUB_ENV
+              echo "$(xcodebuild -version -sdk macosx Path)" >> $GITHUB_PATH
+              brew install coreutils  # for `nproc`
+              ;;
+            Windows*)
+              choco install --no-progress -y winflexbison3
+              choco install gnuwin32-coreutils.install  # for `nproc`
+              ;;
+          esac
+
+      - name: Setup Python
+        uses: actions/setup-python@v3
+        with:
+          python-version: 3.x
+          architecture: x64
+
+      - name: Install Python Dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install docutils pygments
+
+      - name: Check Available Configure Presets
+        run: cmake --list-presets
+
+      - name: Fast Configure
+        run: cmake --preset=${{ matrix.name }}-fast -DPython3_ROOT_DIR="${{ env.pythonLocation }}"
+
+      - name: Fast Build
+        run: cmake --build --preset=${{ matrix.name }}-fast -j$(nproc)
+
+      - name: Install
+        run: cmake --build --preset=${{ matrix.name }}-fast --target install
+
+      - name: Minimal Install Test
+        working-directory: ./install/bin
+        run: ./re2c --version
+
+      - name: Get build dir
+        id: build-info
+        run: |
+          if echo "${{ matrix.name }}" | grep -q "ootree"; then
+            BUILD_DIR="$(pwd)/.build/${{ matrix.name }}-full"
+          else
+            BUILD_DIR="$(pwd)"
+          fi
+          echo "BUILD_DIR=$BUILD_DIR" >> $GITHUB_OUTPUT
+
+      - name: Full Configure
+        run: cmake --preset=${{ matrix.name }}-full -DPython3_ROOT_DIR="${{ env.pythonLocation }}"
+
+      - name: Full Build
+        run: |
+          find src -name '*.re' | xargs touch
+          cmake --build --preset=${{ matrix.name }}-full -j$(nproc)
+
+      - name: Run Main Test Suite
+        run: bash -c "ulimit -s 256; cmake --build --preset=${{ matrix.name }}-full --target tests -j$(nproc)"
+
+      - name: Run Skeleton Tests
+        if: "contains(matrix.name, 'skeleton')"
+        working-directory: ${{ steps.build-info.outputs.BUILD_DIR }}
+        run: python run_tests.py --skeleton
+
+      - name: Run Valgrind Tests
+        if: "contains(matrix.name, 'valgrind')"
+        working-directory: ${{ steps.build-info.outputs.BUILD_DIR }}
+        run: python run_tests.py --valgrind
+
+      - name: Archive test artifacts
+        # Run this step only if tests failed (otherwise there is nothing to upload).
+        if: failure()
+        uses: actions/upload-artifact@v3
+        with:
+          # To ensure that jobs don't overwrite existing artifacts,
+          # use a different name per job.
+          name: test-${{ matrix.name }}
+          # Tests run in a temporary directory that has a name test_<timestamp>.
+          path: ${{ steps.build-info.outputs.BUILD_DIR }}/test_[0-9]*
+          # Artifacts are retained for 90 days by default.
+          # In fact, we don't need such long period.
+          retention-days: 30

.github/workflows/ci.yml

@@ -1,276 +0,0 @@
-name: CI
-
-on:
-  push:
-  pull_request:
-
-defaults:
-  run:
-    shell: bash
-
-jobs:
-  build-and-test:
-    name: ${{ matrix.name }}
-    runs-on: ${{ matrix.os }}
-
-    strategy:
-      # When set to true, GitHub cancels
-      # all in-progress jobs if any matrix job fails.
-      fail-fast: false
-
-      matrix:
-
-        include:
-          # Linux
-          - name: linux-gcc-ubsan-ootree
-            os: ubuntu-latest
-            generator: Unix Makefiles
-            build-type: Ubsan
-            build-dir: .build
-            cxx-flags: -O2
-            skeleton: ${{ false }}
-
-          - name: linux-gcc-asan-ootree
-            os: ubuntu-latest
-            generator: Unix Makefiles
-            build-type: Asan
-            build-dir: .build
-            cxx-flags: -O2
-            skeleton: ${{ false }}
-
-          - name: linux-gcc-debug-ootree-skeleton
-            os: ubuntu-latest
-            generator: Unix Makefiles
-            build-type: Debug
-            build-dir: .build
-            cxx-flags: -O2
-            skeleton: ${{ true }}
-
-          - name: linux-gcc-debug-intree-m32-skeleton
-            os: ubuntu-latest
-            generator: Unix Makefiles
-            build-type: Debug
-            build-dir: .
-            cxx-flags: -O2 -m32
-            ld-flags: -m32
-            skeleton: ${{ true }}
-
-          - name: linux-gcc-release-ootree-skeleton
-            os: ubuntu-latest
-            generator: Unix Makefiles
-            build-type: Release
-            build-dir: .build
-            skeleton: ${{ true }}
-
-          # macOS
-          - name: macos-clang-debug-ootree-skeleton
-            os: macos-latest
-            generator: Unix Makefiles
-            build-type: Debug
-            build-dir: .build
-            cxx-flags: -O2
-            skeleton: ${{ true }}
-
-          - name: macos-clang-debug-intree-skeleton
-            os: macos-latest
-            generator: Unix Makefiles
-            build-type: Debug
-            build-dir: .
-            cxx-flags: -O2
-            skeleton: ${{ true }}
-
-          - name: macos-clang-release-ootree-skeleton
-            os: macos-latest
-            generator: Unix Makefiles
-            build-type: Release
-            build-dir: .build
-            skeleton: ${{ true }}
-
-          # Windows
-          - name: windows-msvc-debug-ootree
-            os: windows-latest
-            generator: Visual Studio 16 2019
-            build-type: Debug
-            build-dir: .build
-            skeleton: ${{ false }}
-
-          - name: windows-msvc-debug-intree
-            os: windows-latest
-            generator: Visual Studio 16 2019
-            build-type: Debug
-            build-dir: .
-            skeleton: ${{ false }}
-
-          - name: windows-msvc-release-ootree
-            os: windows-latest
-            generator: Visual Studio 16 2019
-            build-type: Release
-            build-dir: .build
-            skeleton: ${{ false }}
-
-    steps:
-      - name: Checkout Code
-        uses: actions/checkout@v2
-
-      - name: Setup chocolatey Cache (Windows)
-        if: startsWith(runner.os, 'Windows')
-        uses: actions/cache@v2
-        with:
-          path: C:\Users\runneradmin\AppData\Local\Temp\chocolatey
-          key: ${{ runner.os }}-chocolatey-${{ matrix.os }}-${{ github.sha }}
-          restore-keys: |
-            ${{ runner.os }}-chocolatey-${{ matrix.os }}-
-            ${{ runner.os }}-chocolatey-
-
-      - name: Setup Common Prerequisites
-        run: |
-          case ${{ runner.os }} in
-            Linux*)
-              sudo apt install gcc-multilib g++-multilib  # for -m32
-              echo "CC=gcc" >> $GITHUB_ENV
-              echo "CXX=g++" >> $GITHUB_ENV
-              ;;
-            macOS*)
-              sudo xcode-select -switch /Applications/Xcode.app
-
-              echo "CC=$(xcrun -f clang)" >> $GITHUB_ENV
-              echo "CXX=$(xcrun -f clang++)" >> $GITHUB_ENV
-              echo "SDKROOT=$(xcodebuild -version -sdk macosx Path)" >> $GITHUB_ENV
-
-              echo "$(xcodebuild -version -sdk macosx Path)" >> $GITHUB_PATH
-              ;;
-            Windows*)
-              choco install --no-progress -y winflexbison3
-              ;;
-          esac
-
-      - name: Setup Python
-        uses: actions/setup-python@v2
-        with:
-          python-version: 3.x
-          architecture: x64
-
-      - name: Install Python Dependencies
-        # Disabled until this https://github.com/skvadrik/re2c/issues/327
-        # will be implemented
-        if: startsWith(runner.os, 'Windows') == false
-        run: |
-          python -m pip install --upgrade pip
-          python -m pip install docutils pygments
-
-      - name: Prepare Configure Command
-        id: configure-command
-        run: |
-          # Common CMake arguments
-          args=( -S . -B ${{ matrix.build-dir }} )
-
-          # Generator's specific stuff.
-          args+=( -G "${{ matrix.generator }}" )
-          case "${{ matrix.generator }}" in
-            'Visual Studio'*)
-              args+=( -A x64 )
-              ;;
-            *)
-              # CMAKE_BUILD_TYPE is not used by multi-config build
-              # types (e.g Visual Studio) at the configure time
-              # thus we add it only for single-config build types
-              args+=( -DCMAKE_BUILD_TYPE=${{ matrix.build-type }} )
-              ;;
-          esac
-
-          # CFLAGS, CXXFLAGS, LDFLAGS
-          args+=( -DCMAKE_C_FLAGS="${{ matrix.cxx-flags }}" )
-          args+=( -DCMAKE_CXX_FLAGS="${{ matrix.cxx-flags }}" )
-          args+=( -DCMAKE_EXE_LINKER_FLAGS="${{ matrix.ld-flags }}" )
-          args+=( -DCMAKE_SHARED_LINKER_FLAGS="${{ matrix.ld-flags }}" )
-
-          # Quote each argument to prevent word split.
-          echo "::set-output name=cmake::$(printf " '%s'" "${args[@]}")"
-
-      - name: Fast Configure
-        run: |
-          # Read prepared command into an array
-          args=( ${{ steps.configure-command.outputs.cmake }} )
-
-          # Amend Build Options
-          args+=( -DCMAKE_INSTALL_PREFIX="$(pwd)/install" )
-
-          # Disable some defaults for now to speed up the build as
-          # they are not necessary in a minimal installation
-          args+=( -DRE2C_BUILD_RE2GO=OFF )
-
-          # Print the resulting command line for debugging purposes
-          printf "cmake" && printf " '%s'" "${args[@]}" && printf "\n"
-
-          # Run CMake, pass quoted arguments to prevent word split.
-          printf " '%s'" "${args[@]}" | xargs cmake
-
-      - name: Fast Build
-        run: cmake --build ${{ matrix.build-dir }} --config ${{ matrix.build-type }}
-
-      - name: Install
-        run: |
-          cmake --build ${{ matrix.build-dir }} --config ${{ matrix.build-type }} --target install
-          echo -e "Install tree:\n$(find install | sed 's/^/-- /')"
-
-      - name: Minimal Install Test
-        working-directory: ./install/bin
-        run: ./re2c --version
-
-      - name: Full Configure
-        run: |
-          # Read prepared command into an array
-          args=( ${{ steps.configure-command.outputs.cmake }} )
-
-          # Amend Build Options
-          args+=( -DCMAKE_INSTALL_PREFIX="$(pwd)/install-2" -DRE2C_BUILD_RE2GO=ON -DRE2C_BUILD_LIBS=ON )
-          args+=( -DRE2C_REBUILD_LEXERS=ON -DRE2C_FOR_BUILD="$(pwd)/install/bin/re2c" )
-
-          # Windows does not have a native 'man' and 'sed' commands.
-          # Disabled until this https://github.com/skvadrik/re2c/issues/327
-          # will be implemented
-          case ${{ runner.os }} in
-            Windows*)
-              args+=( -DRE2C_REBUILD_DOCS=OFF )
-              ;;
-            *)
-              args+=( -DRE2C_REBUILD_DOCS=ON )
-              ;;
-          esac
-
-          # Print the resulting command line for debugging purposes
-          printf "cmake" && printf " '%s'" "${args[@]}" && printf "\n"
-
-          # Run CMake, pass quoted arguments to prevent word split.
-          printf " '%s'" "${args[@]}" | xargs cmake
-
-      - name: Full Build
-        run: |
-          find src -name '*.re' | xargs touch
-          cmake --build ${{ matrix.build-dir }} --config ${{ matrix.build-type }}
-
-      - name: Test
-        # Remove the following line after fix Windows tests
-        working-directory: ${{ matrix.build-dir }}
-        run: |
-          # main test suite
-          cmake --build . --target check
-
-          # skeleton tests (if enabled)
-          if [[ "x${{ matrix.skeleton }}" == "xtrue" ]] ; then
-            python run_tests.py --skeleton
-          fi
-
-      - name: Archive test artifacts
-        # Run this step only if tests failed (otherwise there is nothing to upload).
-        if: failure()
-        uses: actions/upload-artifact@v2
-        with:
-          # To ensure that jobs don't overwrite existing artifacts,
-          # use a different name per job.
-          name: test-${{ matrix.name }}
-          # Tests run in a temporary directory that has a name test_<timestamp>.
-          path: ${{ matrix.build-dir }}/test_[0-9]*
-          # Artifacts are retained for 90 days by default.
-          # In fact we don't need such long period.
-          retention-days: 30

.gitignore

@@ -11,6 +11,10 @@
 /m4
 /test_*
 /.build*
+/bazel-*
+
+# Ignore user defined CMake presets
+CMakeUserPresets.json
 
 # The files autogenerated by automake.
 Makefile.in

.travis.yml

@@ -1,9 +1,5 @@
 language: cpp
 
-os:
-  - linux
-  - freebsd
-
 # use Ubuntu 20.04 (Focal Fossa) that defaults to Python 3
 dist: focal
 
@@ -14,35 +10,25 @@ compiler:
   - gcc
   - clang
 
-before_install:
-  - if [ "$TRAVIS_OS_NAME" = "linux" ]; then apt install bison python3-docutils ; fi
-  - if [ "$TRAVIS_OS_NAME" = "freebsd" ]; then pkg install textproc/py-docutils; fi
+# preinstalled packages
+addons:
+  apt:
+    packages:
+    - bison
+    - python3-docutils
 
 # configurations: cmake/autotools, in-tree/out-of-tree, release/debug, etc.
 env:
-  - CMAKE=no  BUILD_DIR=.      CONFIG_DIR=.  COMMON_FLAGS="-O2 -g"
-  - CMAKE=no  BUILD_DIR=.build CONFIG_DIR=.. COMMON_FLAGS="-O2 -g"
-  - CMAKE=yes BUILD_DIR=.      CONFIG_DIR=.  BUILD_TYPE=Debug
+  - CMAKE=no  BUILD_DIR=.      CONFIG_DIR=.                                COMMON_FLAGS="-O2 -g"
+  - CMAKE=no  BUILD_DIR=.build CONFIG_DIR=.. CONFIG_FLAGS="--enable-debug" COMMON_FLAGS="-O2 -g"
+  - CMAKE=yes BUILD_DIR=.      CONFIG_DIR=.  BUILD_TYPE=Debug              COMMON_FLAGS="-O2 -g"
+  - CMAKE=yes BUILD_DIR=.build CONFIG_DIR=.. BUILD_TYPE=Debug              COMMON_FLAGS="-O2 -g" SKELETON=yes
   - CMAKE=yes BUILD_DIR=.build CONFIG_DIR=.. BUILD_TYPE=Release
-  - CMAKE=yes BUILD_DIR=.      CONFIG_DIR=.  BUILD_TYPE=Asan
-  - CMAKE=yes BUILD_DIR=.build CONFIG_DIR=.. BUILD_TYPE=Ubsan
-
-jobs:
-  exclude:
-  - os: freebsd
-    env: BUILD_TYPE=Asan
-  - os: freebsd
-    env: BUILD_TYPE=Ubsan
+  - CMAKE=yes BUILD_DIR=.build CONFIG_DIR=.. BUILD_TYPE=Debug              COMMON_FLAGS="-O2 -g -fsanitize=address"
+  - CMAKE=yes BUILD_DIR=.build CONFIG_DIR=.. BUILD_TYPE=Debug              COMMON_FLAGS="-O2 -g -fsanitize=undefined"
 
 script:
   - |
-    function ncpus() {
-        if [ "$TRAVIS_OS_NAME" = "freebsd" ]; then
-            sysctl -n hw.ncpu
-        else
-            nproc
-        fi
-    }
     function run_cmake() {
         cmake \
             -B . -S ${CONFIG_DIR} \
@@ -55,6 +41,7 @@ script:
     }
     function run_configure() {
         ${CONFIG_DIR}/configure \
+            ${CONFIG_FLAGS} \
             CFLAGS="${COMMON_FLAGS}" \
             CXXFLAGS="${COMMON_FLAGS}" \
             LDFLAGS="${COMMON_FLAGS}" \
@@ -69,7 +56,7 @@ script:
         mkdir -p ${BUILD_DIR} && cd ${BUILD_DIR}
         run_configure --prefix="$(pwd)/install"
     fi
-  - make -j$(ncpus)
+  - make -j$(nproc)
   - make install
   - |
     if [ x${CMAKE} = "xyes" ]; then
@@ -88,7 +75,10 @@ script:
             RE2C_FOR_BUILD="$(pwd)/install/bin/re2c"
     fi
   - find ${CONFIG_DIR}/src -name '*.re' | xargs touch
-  - make -j$(ncpus)
+  - make -j$(nproc)
   - |
-    make -j$(ncpus) check VERBOSE=1
-    ./run_tests.py --skeleton
+    if [ x${SKELETON} = "xyes" ]; then
+        ./run_tests.py --skeleton
+    else
+        make -j$(nproc) check VERBOSE=1
+    fi

BUILD.bazel

@@ -0,0 +1,45 @@
+package(default_visibility = ["//visibility:private"])
+
+genrule(
+    name = "re2c_config",
+    srcs = [
+        "CMakeLists.txt",
+        "//bazel:re2c_config.h",
+    ],
+    outs = ["config.h"],
+    cmd = """
+        cat $(location //:CMakeLists.txt) |\\
+            sed -rne 's/project\\(re2c VERSION ([^ ]+).*/#define PACKAGE_VERSION "\\1-bootstrap"/p' >> "$@"
+        cat $(location //bazel:re2c_config.h) >> "$@"
+    """,
+)
+
+cc_library(
+    name = "re2c_config_cc",
+    hdrs = ["config.h"],
+    defines = ["""RE2C_STDLIB_DIR='"$(BINDIR)/re2c.runfiles/re2c/include"'"""],
+    data = ["include/unicode_categories.re"],
+)
+
+cc_library(
+    name = "re2c_bootstrap_cc",
+    includes = ["bootstrap"],
+    hdrs = glob([
+        "bootstrap/src/**/*.h",
+        "src/**/*.h",
+    ]),
+    srcs = glob([
+        "bootstrap/src/**/*.cc",
+        "src/**/*.cc",
+    ]),
+    deps = [
+        "re2c_config_cc",  # Needed for 're2c_config_cc' data files
+    ],
+)
+
+cc_binary(
+    name = "re2c",
+    srcs = ["src/main.cc"],
+    deps = [":re2c_bootstrap_cc"],
+    visibility = ["//visibility:public"],
+)

BUILD.md

@@ -2,27 +2,24 @@ How to build
 ============
 
 
-
 Dependencies
 ------------
 
-To build re2c from a release tarball one needs only a C++ compiler and
-optionally Python 3 to run the tests.
+To build re2c from a release tarball one needs only a C++ compiler (supporting
+at least C++11) and Python 3 for build-time scripts and for testing. Building on
+Windows additionally requires either CMake or a Mingw build environment.
 
-To develop re2c one also needs CMake or Autotools. Re2c can be built with either
-of the two build systems, and they are both maintained and tested on CI. Other
-build dependencies include Bison (to rebuild parsers), rst2man (to rebuild
-documentation), Sphinx (to rebuild the website), and optionally Mingw and Wine
-(to run Windows tests on Linux). A few helper scripts are written in Haskell,
-but they are not essential for re2c development.
+To develop re2c one also needs CMake or Autotools (both build systems are
+maintained). Other dependencies include Bison (to rebuild parsers), Python 3
+Docutils (to rebuild documentation) and Sphinx (to rebuild the website).
+Benchmarks also use Google-benchmark, re2 and Java. A few helper scripts are
+written in Haskell, but they are not essential for re2c development.
 
-Re2c is a self-hosting lexer generator, meaning that parts of its source code
+re2c is a self-hosting lexer generator, meaning that parts of its source code
 are written in re2c (namely, all the source files that have a *.re* extension).
-By default re2c builds from pregenerated bootstrap files. To regenerate these
-files it is necessary to configure re2c in a special way and provide a path to
-an existing re2c binary (see the details below). Normally this is not necessary,
-and the pregenerated files can be used as long as the *.re* files are unchanged.
-
+By default re2c builds from pre-generated bootstrap files. To build from source
+*.re* files, it is necessary to reconfigure the build and provide a path to an
+existing re2c binary that will be used for bootstrap (see the details below).
 
 
 Build (Autotools)
@@ -51,20 +48,27 @@ The configure script has many options (to see them all, run
   * `--enable-golang`
     Build re2go (identical to `re2c --lang go`). This is on by default.
 
+  * `--enable-rust`
+    Build re2rust (identical to `re2c --lang rust`). This is on by default.
+
   * `--enable-lexers`
-    Enable regeneration of *.re* files (as opposed to using bootstrap files).
-    This requires setting `RE2C_FOR_BUILD` to an existing re2c executable
-    path.
+    Enable regeneration of lexers from *.re* files (as opposed to using
+    pre-generated bootstrap files). This requires setting `RE2C_FOR_BUILD` to
+    an existing re2c executable path.
+
+  * `--enable-parsers`
+    Enable regeneration of parsers from *.ypp* files with bison (as opposed to
+    using pre-generated bootstrap files).
 
   * `--enable-docs`
-    Enable regeneration of documentation (requires rst2man).
+    Enable regeneration of documentation (requires Python 3 Docutils).
 
   * `--enable-benchmarks`
-    Build benchmarks (requires google benchmarks library).
+    Build benchmarks (requires Google benchmarks library).
 
   * `--enable-benchmarks-regenerate`
     Regenerate C code for Ragel and Kleenex benchmarks (this will download and
-    build Ragel and Kleenex). Re2c benchmarks are always regenerated.
+    build Ragel and Kleenex). re2c benchmarks are always regenerated.
 
   * `--enable-libs`
     Build the experimental libre2c library that provides POSIX
@@ -82,10 +86,27 @@ There is a bunch of build scripts for specialized builds with Asan, Ubsan,
 GLIBCXX_DEBUG, etc. in the ``build`` subdirectory.
 
 
-
 Build (CMake)
 -------------
 
+re2c includes CMake presets for Linux, MacOS, and Windows. Use `cmake
+--list-presets` to see them.
+
+* The `-fast` presets have minimal dependencies and build re2c using
+  pre-generated manpages, lexers, and parsers.
+* The `-full` presets have additional prerequisites (Docutils, Bison, and a
+  pre-existing re2c binary) and re-generate the manpages, lexers, and parsers.
+
+If you just want to build re2c, use the `-release-ootree-fast` preset. The
+other presets are primarily of interest to developers.
+
+For example, to build the `windows-msvc-release-ootree-fast` preset:
+
+* CD to the root of the re2c folder (the folder containing this README).
+* Configure: `cmake --preset=windows-msvc-release-ootree-fast`
+* Build: `cmake --build --preset=windows-msvc-release-ootree-fast`
+* The binary will be in the `.build/<preset-name>` folder.
+
 CMake supports different build modes, which can be further customized with
 individual options. By default re2c builds with `-O2 -g` and does not set any
 particular mode to allow the user complete freedom in overriding the build
@@ -130,20 +151,28 @@ specific to re2c:
   * `-DRE2C_BUILD_RE2GO=yes`
     Build re2go executable (an alias to `re2c --lang go`). Enabled by default.
 
+  * `-DRE2C_BUILD_RE2RUST=yes`
+    Build re2rust executable (an alias to `re2c --lang rust`). Enabled by
+    default.
+
   * `-DRE2C_REBUILD_LEXERS=yes`
-    Enable regeneration of *.re* files (as opposed to using bootstrap files).
-    This requires setting `-DRE2C_FOR_BUILD` to an existing re2c executable
-    path.
+    Enable regeneration of lexers from *.re* files (as opposed to using
+    pre-generated bootstrap files). This requires setting `-DRE2C_FOR_BUILD` to
+    an existing re2c executable path.
+
+  * `-DRE2C_REBUILD_PARSERS=yes`
+    Enable regeneration of parsers from *.ypp* files with bison (as opposed to
+    using pre-generated bootstrap files).
 
   * `-DRE2C_REBUILD_DOCS=yes`
-    Enable regeneration of documentation (requires rst2man).
+    Enable regeneration of documentation (requires Python 3 Docutils).
 
   * `-DRE2C_BUILD_BENCHMARKS=yes`
-    Build benchmarks (requires google benchmarks library).
+    Build benchmarks (requires Google benchmarks library).
 
   * `-DRE2C_REGEN_BENCHMARKS=yes`
     Regenerate C code for Ragel and Kleenex benchmarks (this will download and
-    build Ragel and Kleenex). Re2c benchmarks are always regenerated.
+    build Ragel and Kleenex). re2c benchmarks are always regenerated.
 
   * `-DRE2C_BUILD_LIBS=yes`
     Build the experimental libre2c library that provides POSIX
@@ -161,11 +190,10 @@ There is a bunch of build scripts for specialized builds with Asan, Ubsan,
 GLIBCXX_DEBUG, etc. in the ``build`` subdirectory.
 
 
-
 Test
 ----
 
-Re2c has a main test suite and a number of unit tests. Run them all:
+re2c has a main test suite and a number of unit tests. Run them all:
   ```
   $ make check
   ```
@@ -190,9 +218,9 @@ Check the distribution (works with Autotools):
   $ make distcheck
   ```
 
-Re2c provides a helper script `build/__alltest.sh` that builds and tests
-various re2c build flavours with Asan, Ubsan, GLIBCXX_DEBUG, etc. There is a
-couple of fuzz-testing Haskell scripts in the *fuzz* subdirectory; they are
-based on the QuickCheck library and can be easily modified to fuzz-test various
-aspects of re2c by comparing current re2c version against older versions or
-against other regular expression libraries.
+re2c provides a helper script `build/__alltest.sh` that builds and tests various
+re2c build flavours with Asan, Ubsan, GLIBCXX_DEBUG, etc. There is a couple of
+fuzz-testing Haskell scripts in the *fuzz* subdirectory; they are based on the
+QuickCheck library and can be easily modified to fuzz-test various aspects of
+re2c by comparing current re2c version against older versions or against other
+regular expression libraries.

CHANGELOG

@@ -1,6 +1,255 @@
 Changelog
 =========
 
+
+----
+3.0x
+----
+
+3.1 (2023-07-19)
+~~~~~~~~~~~~~~~~
+
+- Added capturing groups with leftmost greedy semantics:
+
+  + Enabled with ``--leftmost-captures`` option or ``re2c:leftmost_captures``
+    configuration
+    (`55de79d8 <https://github.com/skvadrik/re2c/commit/55de79d8>`_,
+    `3a98b543 <https://github.com/skvadrik/re2c/commit/3a98b543>`_).
+
+- Added non-capturing groups:
+
+  + Added new syntax ``(! ...)`` for non-capturing groups
+    (`1edd25d3 <https://github.com/skvadrik/re2c/commit/1edd25d3>`_,
+    `b813c9b4 <https://github.com/skvadrik/re2c/commit/b813c9b4>`_,
+    `338806b9 <https://github.com/skvadrik/re2c/commit/338806b9>`_).
+
+  + Added the ability to flip defaults: make ``(...)`` capturing and ``(! ...)``
+    non-capturing with ``--invert-captures`` option or ``re2c:invert_captures``
+    configuration
+    (`20030ff1 <https://github.com/skvadrik/re2c/commit/20030ff1>`_,
+    `ce756195 <https://github.com/skvadrik/re2c/commit/ce756195>`_).
+
+- Regenerated Unicode include header to support a newer standard
+  (`e3ec2597 <https://github.com/skvadrik/re2c/commit/e3ec2597>`_).
+
+- Published TDFA paper: https://arxiv.org/abs/2206.01398, co-authored with
+  Angelo Borsotti
+  (`fa94d9c7 <https://github.com/skvadrik/re2c/commit/fa94d9c7>`_).
+
+- Removed experimental algorithms that are superseded by TDFA(1) and
+  generally less efficient:
+
+  + Removed staDFA algorithm and deprecated ``--stadfa`` option
+    (`ac5c06cc <https://github.com/skvadrik/re2c/commit/ac5c06cc>`_).
+
+  + Removed TDFA(0) algorithm and deprecated ``--no-lookahead`` option
+    (`dc8f264a <https://github.com/skvadrik/re2c/commit/dc8f264a>`_).
+
+  + (libre2c) Removed backward-matching algorithm
+    (`27256be1 <https://github.com/skvadrik/re2c/commit/27256be1>`_).
+
+  + (libre2c) Removed Kuklewicz POSIX disambiguation algorithm
+    (`aa97b014 <https://github.com/skvadrik/re2c/commit/aa97b014>`_).
+
+  + (libre2c) Removed GTOP shortest path finding algorithm
+    (`511a030c <https://github.com/skvadrik/re2c/commit/511a030c>`_).
+
+- Bug fixes:
+
+  + Fixed parsing of raw UTF-8 characters in Flex compatibility mode
+    (`d87f86ed <https://github.com/skvadrik/re2c/commit/d87f86ed>`_).
+
+  + Added header file to the dependencies generated with ``--depfile`` option
+    (`f807f763 <https://github.com/skvadrik/re2c/commit/f807f763>`_
+    and `2dda36aa <https://github.com/skvadrik/re2c/commit/2dda36aa>`_).
+
+  + Fixed stack overflow on large regular expressions by rewriting recursive
+    functions in iterative form
+    (`46a9b4c4 <https://github.com/skvadrik/re2c/commit/46a9b4c4>`_,
+    `aaf68292 <https://github.com/skvadrik/re2c/commit/aaf68292>`_,
+    `02e5d797 <https://github.com/skvadrik/re2c/commit/02e5d797>`_,
+    `5fffb187 <https://github.com/skvadrik/re2c/commit/5fffb187>`_)
+    and limited stack to 256K on GithubActions CI
+    (`111ee5da <https://github.com/skvadrik/re2c/commit/111ee5da>`_).
+
+- Build system:
+
+  + Added minimal http://bazel.build integration
+    (`3205c867 <https://github.com/skvadrik/re2c/commit/3205c867>`_).
+
+  + Added configure option ``--enable-parsers`` that regenerates bison parsers
+    (`9e0dbd3c <https://github.com/skvadrik/re2c/commit/9e0dbd3c>`_).
+
+  + Added CMake option ``RE2C_REBUILD_PARSERS``
+    (`6e91c22d <https://github.com/skvadrik/re2c/commit/6e91c22d>`_).
+
+  + With CMake, fixed documentation generation on Windows.
+
+- Codebase improvements:
+
+  + Moved the entire codebase to C++11.
+  + Added uniform error handling (return codes are now properly checked and
+    returned to the caller).
+  + Reorganized codegen subsystem in four well-defined phases (analyze,
+    generate, fixup, render) and separated codegen from parsing phase.
+  + Improved memory allocation by using slab allocators instead of global free
+    lists.
+  + Moved to pure API for bison parsers.
+  + Unified code style.
+
+- Testing:
+
+  + Added ``--verbose`` flag to run_tests.py and suppressed verbose output by
+    default.
+  + Multiple improvements of continuous testing with GithubActions.
+
+
+3.0 (2022-01-27)
+~~~~~~~~~~~~~~~~
+
+- Added code generation backend for Rust:
+
+  + Enabled with ``--lang rust`` option.
+  + A new ``re2rust`` binary (built by default, or configured with
+    ``--enable-rust`` Autoconf option and ``RE2C_BUILD_RE2RUST`` CMake option).
+
+- Added options:
+
+  + ``--loop-switch``
+  + ``--no-unsafe``
+
+- Added configurations;
+
+  + ``re2c:label:yyloop``
+  + ``re2c:unsafe``
+
+- Renamed options to use common naming scheme. The old names are supported as
+  aliases, so the change does not break existing code. Documentation has been
+  updated to use new names.
+
+  + ``--api`` is a new alias for ``--input``
+  + ``--ebcdic`` is a new alias for ``--ecb``
+  + ``--ucs2`` is a new alias for ``--wide-chars``
+  + ``--utf32`` is a new alias for ``--unicode``
+  + ``--utf16`` is a new alias for ``--utf-16``
+  + ``--utf8`` is a new alias for ``--utf-8``
+  + ``--header`` is a new alias for ``--type-header``
+
+- Renamed configurations to use common naming scheme and support proper scoping
+  under subcategories such as ``:define``, ``:label``, ``:variable``, etc. The
+  old names are supported as aliases, so the change does not break existing
+  code. Documentation has been updated to use new names.
+
+  + ``re2c:api`` is a new alias for ``re2c:flags:input``
+  + ``re2c:bit-vectors`` is a new alias for ``re2c:flags:bit-vectors``
+  + ``re2c:case-insensitive`` is a new alias for ``re2c:flags:case-insensitive``
+  + ``re2c:case-inverted`` is a new alias for ``re2c:flags:case-inverted``
+  + ``re2c:case-ranges`` is a new alias for ``re2c:flags:case-ranges``
+  + ``re2c:cond:prefix`` is a new alias for ``re2c:condprefix``
+  + ``re2c:cond:enumprefix`` is a new alias for ``re2c:condenumprefix``
+  + ``re2c:computed-gotos`` is a new alias for ``re2c:flags:computed-gotos``
+  + ``re2c:computed-gotos:threshold`` is a new alias for ``re2c:cgoto:threshold``
+  + ``re2c:debug-output`` is a new alias for ``re2c:flags:debug-output``
+  + ``re2c:encoding:ebcdic`` is a new alias for ``re2c:flags:ecb``
+  + ``re2c:encoding:utf32`` is a new alias for ``re2c:flags:unicode``
+  + ``re2c:encoding:ucs2`` is a new alias for ``re2c:flags:wide-chars``
+  + ``re2c:encoding:utf16`` is a new alias for ``re2c:flags:utf-16``
+  + ``re2c:encoding:utf8`` is a new alias for ``re2c:flags:utf-8``
+  + ``re2c:encoding-policy`` is a new alias for ``re2c:flags:encoding-policy``
+  + ``re2c:empty-class`` is a new alias for ``re2c:flags:empty-class``
+  + ``re2c:header`` is a new alias for ``re2c:flags:type-header``
+  + ``re2c:label:prefix`` is a new alias for ``re2c:labelprefix``
+  + ``re2c:label:yyfill`` is a new alias for ``re2c:label:yyFillLabel``
+  + ``re2c:label:start`` is a new alias for ``re2c:startlabel``
+  + ``re2c:nested-ifs`` is a new alias for ``re2c:flags:nested-ifs``
+  + ``re2c:posix-captures`` is a new alias for ``re2c:flags:posix-captures``
+  + ``re2c:tags`` is a new alias for ``re2c:flags:tags``
+  + ``re2c:variable:yych:conversion`` is a new alias for ``re2c:yych:conversion``
+  + ``re2c:variable:yych:emit`` is a new alias for ``re2c:yych:emit``
+  + ``re2c:variable:yybm:hex`` is a new alias for ``re2c:yybm:hex``
+  + ``re2c:unsafe`` is a new alias for ``re2c:flags:unsafe``
+
+- Added directive alias ``conditions:re2c`` for ``types:re2c``.
+
+- Multiple small changes in code generation, including some formatting changes
+  that result in large diffs in the generated code:
+
+  + Do not allocate indices for unused state labels (this results in a change in
+    state enumeration), commits
+    `919570c4 <https://github.com/skvadrik/re2c/commit/919570c4>`_ and
+    `82b704f6 <https://github.com/skvadrik/re2c/commit/82b704f6>`_.
+  + Do not generate redundant ``YYPEEK`` statements, commit
+    `cca31d22 <https://github.com/skvadrik/re2c/commit/cca31d22>`_.
+  + Do not generate ``YYDEBUG`` statements for unused states labels, commit
+    `a46f01e6 <https://github.com/skvadrik/re2c/commit/a46f01e6>`_.
+  + C backend: change formatting of switch statements, commit
+    `ed88e12e <https://github.com/skvadrik/re2c/commit/ed88e12e>`_.
+  + Go backend: render continuous character ranges in compact form, commit
+    `09161b14 <https://github.com/skvadrik/re2c/commit/09161b14>`_.
+  + Mark start and end of included .re files with line directives, commit
+    `48e83fca <https://github.com/skvadrik/re2c/commit/48e83fca>`_.
+
+- A fix to limit maximum allowed NFA and DFA size (to avoid out of memory
+  crashes and stack overflows), commit
+  `a3473fd7 <https://github.com/skvadrik/re2c/commit/a3473fd7>`_.
+
+- A fix to correctly compute fixed tags in trailing context, commit
+  `68e1ab71 <https://github.com/skvadrik/re2c/commit/68e1ab71>`_.
+
+- A fix to generate non-overlapping names for s-tag and m-tag variables, commit
+  `7c6b5c95 <https://github.com/skvadrik/re2c/commit/7c6b5c95>`_.
+
+- Infrastructural: added support for CMake presets.
+
+- Updated documentation.
+
+- Backwards-incompatible changes that are unlikely to affect any users:
+
+  + Restrict lexical contexts where ``%{`` is recognized as a block start,
+    commit `dba7d055 <https://github.com/skvadrik/re2c/commit/dba7d055>`_.
+  + Emit an error when repetition lower bound exceeds upper bound, commit
+    `039c1894 <https://github.com/skvadrik/re2c/commit/039c1894>`_.
+
+
+----
+2.2x
+----
+
+2.2 (2021-08-01)
+~~~~~~~~~~~~~~~~
+
+- Added named blocks and block lists in directives.
+
+- Added local blocks ``/*!local:re2c ... */``.
+
+- Added in-block ``!include`` directive.
+
+- Added in-block ``!use`` directive.
+
+- Allowed reusable blocks without ``-r --reusable`` option.
+
+- Allowed customizing the generated code with configurations for directives
+  ``max:re2c``, ``maxnmatch:re2c``, ``stags:re2c``, ``mtags:re2c`` and
+  ``types:re2c`` (see directive descriptions for details).
+
+- Forbid arbitrary text at the end of ``max:re2c`` directive. This may break
+  backwards compatibility, although it is unlikely that this was used by anyone.
+  The change was necessary in order to allow customization of the generated code
+  with configurations.
+
+- Deprecated configurations ``flags:i``, ``flags:no-debug-info`` in favour of
+  the global options ``-i``, ``--no-debug-info``.
+
+- Reimplemented re2c test runner in Python (thanks to
+  `Serghei Iakovlev <https://github.com/sergeyklay>`_). Improved integration
+  with GitHub Actions.
+
+- Changes in the experimental libre2c library: added new algorithms that
+  construct t-string or extract submatch on all repetitions; added TDFA
+  benchmark written in Java by Angelo Borsotti.
+
+- Updated documentation.
+
 ----
 2.1x
 ----

CMakeLists.txt

@@ -1,5 +1,5 @@
 cmake_minimum_required(VERSION 3.12)
-project(re2c VERSION 2.1.1 HOMEPAGE_URL "https://re2c.org/")
+project(re2c VERSION 3.1 HOMEPAGE_URL "https://re2c.org/")
 
 list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake")
 
@@ -17,46 +17,45 @@ if(RE2C_REBUILD_LEXERS AND NOT RE2C_FOR_BUILD)
     message(FATAL_ERROR "option RE2C_FOR_BUILD is required for RE2C_REBUILD_LEXERS")
 endif()
 
-option(RE2C_REBUILD_DOCS "Regenerate manpage" OFF)
-if(RE2C_REBUILD_DOCS)
-    find_program(RST2MAN NAMES rst2man rst2man.py)
-    if(NOT RST2MAN)
-        message(FATAL_ERROR "rst2man (or rst2man.py) program is required for RE2C_REBUILD_DOCS")
-    endif()
-
-    # Windows does not have a native 'man' and 'sed' commands.
-    # Therefore, we are looking for the Cygwin versions of 'man' and 'sed'.
-    if (WIN32)
-        find_program(MAN_EXECUTABLE man)
-        if(NOT MAN_EXECUTABLE)
-            message(FATAL_ERROR "man program is required for RE2C_REBUILD_DOCS")
-        endif()
-
-        find_program(SED_EXECUTABLE sed)
-        if(NOT SED_EXECUTABLE)
-            message(FATAL_ERROR "sed program is required for RE2C_REBUILD_DOCS")
-        endif()
-    endif()
+option(RE2C_REBUILD_PARSERS "Regenerate parsers with Bison" OFF)
+if(RE2C_REBUILD_PARSERS AND NOT RE2C_FOR_BUILD)
+    find_package(BISON REQUIRED)
 endif()
 
+option(RE2C_REBUILD_DOCS "Regenerate manpage" OFF)
 option(RE2C_BUILD_LIBS "Build libraries" OFF)
 
+
 option(RE2C_BUILD_RE2GO "Build re2go executable (an alias for `re2c --lang go`)" ON)
+option(RE2C_BUILD_RE2RUST "Build re2rust executable (an alias for `re2c --lang rust`)" ON)
 
 option(RE2C_BUILD_BENCHMARKS "Build benchmarks" OFF)
 option(RE2C_REGEN_BENCHMARKS "Regenerate C code for benchmarks" OFF)
 
 # checks for programs
-find_package(BISON)
 find_package(Python3 COMPONENTS Interpreter)
+# starting from cmake 3.19 find_package can do version check, but we are on 3.12
+if(Python3_VERSION VERSION_LESS 3.7)
+    message(FATAL_ERROR "python version 3.7 or higher is required")
+endif()
+if(RE2C_REBUILD_DOCS)
+    execute_process(
+        COMMAND "${Python3_EXECUTABLE}" -c "import docutils"
+        RESULT_VARIABLE EXIT_CODE
+        OUTPUT_QUIET
+    )
+    if (NOT ${EXIT_CODE} EQUAL 0)
+        message(FATAL_ERROR "python package docutils (needed for docs) not found")
+    endif()
+endif()
 
-# checks for C++ compiler flags
-set(CMAKE_CXX_STANDARD 98)
-set(CMAKE_CXX_EXTENSIONS OFF)
+# Use C++11 standard. Note that without `set(CMAKE_CXX_EXTENSIONS OFF)` this enables
+# -std=gnu++11, not -std=c++11. Extensions are needed to enable POSIX functions on
+# Cygwin and NetBSD.
+set(CMAKE_CXX_STANDARD 11)
+set(CMAKE_CXX_STANDARD_REQUIRED ON) # fail if the C++ compiler does not support this standard
 set(CMAKE_POSITION_INDEPENDENT_CODE ON) # make sure object libraries work with shared libraries
 
-# use system stdint.h if present, otherwise fallback to src/c99_stdint.h
-ac_check_headers("stdint.h")
 # needed for POSIX file API
 ac_check_headers("sys/types.h")
 ac_check_headers("sys/stat.h")
@@ -65,34 +64,25 @@ ac_check_headers("unistd.h")
 # windows POSIX-like API
 ac_check_headers("io.h")
 
-# list of possible types to use in typedefs
-ac_check_sizeof("char")
-ac_check_sizeof("short")
-ac_check_sizeof("int")
-ac_check_sizeof("long")
-ac_check_sizeof("long long")
-ac_check_sizeof("__int64")
-ac_check_sizeof("void *")
-# 64-bit integer constant suffix
-ac_check_sizeof("0l")
-ac_check_sizeof("0ll")
-ac_check_sizeof("0i8")
-
 # docs (manpages and help)
-set(re2c_manpage_source       "${CMAKE_CURRENT_BINARY_DIR}/doc/manpage.rst")
-set(re2c_help_source          "${CMAKE_CURRENT_BINARY_DIR}/doc/help.rst")
-set(re2c_manpage_bootstrap_c  "${CMAKE_CURRENT_SOURCE_DIR}/bootstrap/doc/re2c.1")
-set(re2c_manpage_bootstrap_go "${CMAKE_CURRENT_SOURCE_DIR}/bootstrap/doc/re2go.1")
-set(re2c_help_bootstrap       "${CMAKE_CURRENT_SOURCE_DIR}/bootstrap/src/msg/help.cc")
-set(re2c_manpage_c            "${CMAKE_CURRENT_BINARY_DIR}/doc/re2c.1")
-set(re2c_manpage_go           "${CMAKE_CURRENT_BINARY_DIR}/doc/re2go.1")
-set(re2c_help                 "${CMAKE_CURRENT_BINARY_DIR}/src/msg/help.cc")
-set(re2c_genhelp              "${CMAKE_CURRENT_SOURCE_DIR}/build/gen_help.sh")
-set(re2c_splitman             "${CMAKE_CURRENT_SOURCE_DIR}/build/split_man.sh")
+set(re2c_manpage_source         "${CMAKE_CURRENT_BINARY_DIR}/doc/manpage.rst")
+set(re2c_help_source            "${CMAKE_CURRENT_BINARY_DIR}/doc/help.rst")
+set(re2c_manpage_bootstrap_c    "${CMAKE_CURRENT_SOURCE_DIR}/bootstrap/doc/re2c.1")
+set(re2c_manpage_bootstrap_go   "${CMAKE_CURRENT_SOURCE_DIR}/bootstrap/doc/re2go.1")
+set(re2c_manpage_bootstrap_rust "${CMAKE_CURRENT_SOURCE_DIR}/bootstrap/doc/re2rust.1")
+set(re2c_help_bootstrap         "${CMAKE_CURRENT_SOURCE_DIR}/bootstrap/src/msg/help.cc")
+set(re2c_manpage_c              "${CMAKE_CURRENT_BINARY_DIR}/doc/re2c.1")
+set(re2c_manpage_go             "${CMAKE_CURRENT_BINARY_DIR}/doc/re2go.1")
+set(re2c_manpage_rust           "${CMAKE_CURRENT_BINARY_DIR}/doc/re2rust.1")
+set(re2c_help                   "${CMAKE_CURRENT_BINARY_DIR}/src/msg/help.cc")
+set(re2c_rst2man                "${CMAKE_CURRENT_SOURCE_DIR}/build/rst2man.py")
+set(re2c_rst2txt                "${CMAKE_CURRENT_SOURCE_DIR}/build/rst2txt.py")
+set(re2c_splitman               "${CMAKE_CURRENT_SOURCE_DIR}/build/split_man.py")
 set(re2c_docs
     "${re2c_help}"
     "${re2c_manpage_c}"
     "$<$<BOOL:${RE2C_BUILD_RE2GO}>:${re2c_manpage_go}>"
+    "$<$<BOOL:${RE2C_BUILD_RE2RUST}>:${re2c_manpage_rust}>"
 )
 
 set(top_srcdir "${CMAKE_CURRENT_SOURCE_DIR}")
@@ -136,19 +126,14 @@ add_library(re2c_objects_autogen_ver_to_vernum OBJECT
 )
 
 set(re2c_sources
-    src/codegen/code_goto.cc
-    src/codegen/combine.cc
-    src/codegen/gen_bitmaps.cc
-    src/codegen/gen_delayed.cc
-    src/codegen/gen_dfa.cc
-    src/codegen/gen_goto.cc
-    src/codegen/gen_program.cc
-    src/codegen/gen_state.cc
     src/codegen/helpers.cc
-    src/codegen/render.cc
+    src/codegen/output.cc
+    src/codegen/pass1_analyze.cc
+    src/codegen/pass2_generate.cc
+    src/codegen/pass3_fixup.cc
+    src/codegen/pass4_render.cc
     src/options/opt.cc
     src/options/symtab.cc
-    src/nfa/estimate_size.cc
     src/nfa/re_to_nfa.cc
     src/adfa/adfa.cc
     src/debug/dump_adfa.cc
@@ -174,27 +159,21 @@ set(re2c_sources
     src/dfa/fillpoints.cc
     src/dfa/find_state.cc
     src/dfa/minimization.cc
-    src/dfa/stacmd.cc
     src/dfa/tagver_table.cc
     src/dfa/tcmd.cc
-    src/encoding/ebcdic/ebcdic_regexp.cc
+    src/encoding/ebcdic.cc
     src/encoding/enc.cc
     src/encoding/range_suffix.cc
-    src/encoding/utf8/utf8_regexp.cc
-    src/encoding/utf8/utf8.cc
-    src/encoding/utf16/utf16_regexp.cc
-    src/encoding/utf16/utf16.cc
+    src/encoding/utf8.cc
+    src/encoding/utf16.cc
     src/msg/msg.cc
     src/msg/warn.cc
     src/regexp/ast_to_re.cc
     src/regexp/default_tags.cc
     src/regexp/fixed_tags.cc
     src/regexp/nullable.cc
-    src/regexp/re.cc
-    src/regexp/rule.cc
+    src/regexp/regexp.cc
     src/regexp/split_charset.cc
-    src/regexp/tag.cc
-    src/compile.cc
     src/skeleton/control_flow.cc
     src/skeleton/generate_code.cc
     src/skeleton/generate_data.cc
@@ -202,12 +181,8 @@ set(re2c_sources
     src/skeleton/skeleton.cc
     src/parse/ast.cc
     src/parse/input.cc
-    src/parse/parse.cc
-    src/parse/scanner.cc
-    src/parse/unescape.cc
     src/util/file_utils.cc
-    src/util/s_to_n32_unsafe.cc
-    src/util/temp_file.cc
+    src/util/string_utils.cc
     src/util/range.cc
     src/main.cc
     $<TARGET_OBJECTS:re2c_objects_autogen>
@@ -220,7 +195,13 @@ add_executable(re2c ${re2c_sources})
 # re2go
 if (RE2C_BUILD_RE2GO)
     add_executable(re2go ${re2c_sources})
-    target_compile_definitions(re2go PUBLIC "RE2C_LANG=LANG_GO")
+    target_compile_definitions(re2go PUBLIC "RE2C_LANG=Lang::GO")
+endif()
+
+# re2rust
+if (RE2C_BUILD_RE2RUST)
+    add_executable(re2rust ${re2c_sources})
+    target_compile_definitions(re2rust PUBLIC "RE2C_LANG=Lang::RUST")
 endif()
 
 re2c_bootstrap_lexer("src/parse/lex.re" "src/parse/lex.cc" "src/parse/lex.h")
@@ -232,8 +213,8 @@ re2c_bootstrap_parser("src/parse/parser.ypp" "src/parse/parser.cc" "src/parse/pa
 # docs
 set(re2c_docs_sources
     "${re2c_manpage_source}"
-    "doc/manual/api/api.rst_"
     "doc/manual/api/interface.rst_"
+    "doc/manual/conditions/blocks.rst_"
     "doc/manual/conditions/conditions.rst_"
     "doc/manual/configurations/configurations.rst_"
     "doc/manual/directives/directives.rst_"
@@ -258,13 +239,22 @@ set(re2c_docs_sources
     "doc/manual/state/state.rst_"
     "doc/manual/submatch/submatch_example_mtags.rst_"
     "doc/manual/submatch/submatch_example_posix.rst_"
-    "doc/manual/submatch/submatch_example_stags.rst_"
     "doc/manual/submatch/submatch_example_stags_fill.rst_"
+    "doc/manual/submatch/submatch_example_stags.rst_"
     "doc/manual/submatch/submatch.rst_"
+    "doc/manual/synopsis.rst_"
+    "doc/manual/syntax/api1.rst_"
+    "doc/manual/syntax/api2_c.rst_"
+    "doc/manual/syntax/api2_go.rst_"
+    "doc/manual/syntax/api2_rust.rst_"
+    "doc/manual/syntax/api3.rst_"
+    "doc/manual/syntax/intro.rst_"
     "doc/manual/syntax/syntax.rst_"
     "doc/manual/warnings/warnings_general.rst_"
     "doc/manual/warnings/warnings_list.rst_"
     "examples/c/01_basic.re"
+    "examples/c/01_basic.c"
+    "examples/c/conditions/parse_u32_blocks.re"
     "examples/c/conditions/parse_u32_conditions.re"
     "examples/c/encodings/unicode_identifier.re"
     "examples/c/eof/01_sentinel.re"
@@ -274,17 +264,19 @@ set(re2c_docs_sources
     "examples/c/fill/01_fill.re"
     "examples/c/fill/02_fill.re"
     "examples/c/headers/header.re"
-    "examples/c/headers/src/lexer/lexer.h"
+    "examples/c/headers/lexer/state.h"
     "examples/c/includes/include.re"
     "examples/c/includes/definitions.h"
     "examples/c/reuse/reuse.re"
     "examples/c/reuse/usedir.re"
     "examples/c/state/push.re"
-    "examples/c/submatch/01_stags.re"
     "examples/c/submatch/01_stags_fill.re"
+    "examples/c/submatch/01_stags.re"
     "examples/c/submatch/02_mtags.re"
     "examples/c/submatch/03_posix.re"
     "examples/go/01_basic.re"
+    "examples/go/01_basic.go"
+    "examples/go/conditions/parse_u32_blocks.re"
     "examples/go/conditions/parse_u32_conditions.re"
     "examples/go/encodings/unicode_identifier.re"
     "examples/go/eof/01_sentinel.re"
@@ -294,18 +286,42 @@ set(re2c_docs_sources
     "examples/go/fill/01_fill.re"
     "examples/go/fill/02_fill.re"
     "examples/go/headers/header.re"
-    "examples/go/headers/src/lexer/lexer.go"
+    "examples/go/headers/lexer/state.go"
     "examples/go/includes/include.re"
     "examples/go/includes/definitions.go"
     "examples/go/reuse/reuse.re"
     "examples/go/reuse/usedir.re"
     "examples/go/state/push.re"
+    "examples/go/submatch/01_stags_fill.re"
     "examples/go/submatch/01_stags.re"
     "examples/go/submatch/02_mtags.re"
     "examples/go/submatch/03_posix.re"
+    "examples/rust/01_basic.re"
+    "examples/rust/01_basic.rs"
+    "examples/rust/conditions/parse_u32_blocks.re"
+    "examples/rust/conditions/parse_u32_conditions.re"
+    "examples/rust/encodings/unicode_identifier.re"
+    "examples/rust/eof/01_sentinel.re"
+    "examples/rust/eof/02_bounds_checking.re"
+    "examples/rust/eof/03_eof_rule.re"
+    "examples/rust/eof/05_generic_api_eof_rule.re"
+    "examples/rust/fill/01_fill.re"
+    "examples/rust/fill/02_fill.re"
+    "examples/rust/headers/header.re"
+    "examples/rust/headers/lexer/state.rs"
+    "examples/rust/includes/include.re"
+    "examples/rust/includes/definitions.rs"
+    "examples/rust/reuse/reuse.re"
+    "examples/rust/reuse/usedir.re"
+    "examples/rust/state/push.re"
+    "examples/rust/submatch/01_stags_fill.re"
+    "examples/rust/submatch/01_stags.re"
+    "examples/rust/submatch/02_mtags.re"
+    "examples/rust/submatch/03_posix.re"
 )
-re2c_gen_manpage("${re2c_manpage_source}" "${re2c_manpage_c}"  "${re2c_manpage_bootstrap_c}"  "c")
-re2c_gen_manpage("${re2c_manpage_source}" "${re2c_manpage_go}" "${re2c_manpage_bootstrap_go}" "go")
+re2c_gen_manpage("${re2c_manpage_source}" "${re2c_manpage_c}"    "${re2c_manpage_bootstrap_c}"  "c")
+re2c_gen_manpage("${re2c_manpage_source}" "${re2c_manpage_go}"   "${re2c_manpage_bootstrap_go}" "go")
+re2c_gen_manpage("${re2c_manpage_source}" "${re2c_manpage_rust}" "${re2c_manpage_bootstrap_rust}" "rust")
 re2c_gen_help("${re2c_help_source}" "${re2c_help}" "${re2c_help_bootstrap}")
 add_custom_target(docs DEPENDS "${re2c_docs}")
 
@@ -318,6 +334,10 @@ if (CMAKE_SOURCE_DIR STREQUAL PROJECT_SOURCE_DIR)
         install(TARGETS re2go RUNTIME DESTINATION bin)
         install(FILES "${re2c_manpage_go}" DESTINATION "share/man/man1")
     endif()
+    if(RE2C_BUILD_RE2RUST)
+        install(TARGETS re2rust RUNTIME DESTINATION bin)
+        install(FILES "${re2c_manpage_rust}" DESTINATION "share/man/man1")
+    endif()
     install(FILES include/unicode_categories.re DESTINATION "${RE2C_STDLIB_DIR}")
 
     # rebuild all re2c sources using newly built re2c
@@ -352,11 +372,10 @@ if (CMAKE_SOURCE_DIR STREQUAL PROJECT_SOURCE_DIR)
         src/test/range/test.h
         src/util/range.cc
         src/util/range.h
-        src/util/static_assert.h
     )
     add_executable(re2c_test_s_to_n32_unsafe
         src/test/s_to_n32_unsafe/test.cc
-        src/util/s_to_n32_unsafe.cc
+        src/util/string_utils.cc
     )
     add_executable(re2c_test_ver_to_vernum
         src/test/ver_to_vernum/test.cc
@@ -398,21 +417,19 @@ if (RE2C_BUILD_LIBS)
     )
     re2c_bootstrap_lexer("lib/test_helper.re" "lib/test_helper.cc")
 
-    add_library(libre2c_objects OBJECT
+    set(libre2c_sources
         lib/regcomp.cc
         lib/regexec.cc
         lib/regexec_dfa.cc
-        lib/regexec_dfa_regless.cc
+        lib/regexec_dfa_multipass.cc
         lib/regexec_nfa_leftmost.cc
         lib/regexec_nfa_leftmost_trie.cc
         lib/regexec_nfa_posix.cc
         lib/regexec_nfa_posix_trie.cc
-        lib/regexec_nfa_posix_backward.cc
-        lib/regexec_nfa_posix_kuklewicz.cc
         lib/regfree.cc
         lib/stubs.cc
         src/parse/ast.cc
-        src/parse/unescape.cc
+        src/parse/input.cc
         src/options/opt.cc
         src/options/symtab.cc
         src/cfg/cfg.cc
@@ -438,36 +455,28 @@ if (RE2C_BUILD_LIBS)
         src/dfa/fillpoints.cc
         src/dfa/find_state.cc
         src/dfa/minimization.cc
-        src/dfa/stacmd.cc
         src/dfa/tagver_table.cc
         src/dfa/tcmd.cc
-        src/nfa/estimate_size.cc
         src/nfa/re_to_nfa.cc
         src/encoding/enc.cc
         src/encoding/range_suffix.cc
-        src/encoding/ebcdic/ebcdic_regexp.cc
-        src/encoding/utf16/utf16.cc
-        src/encoding/utf16/utf16_regexp.cc
-        src/encoding/utf8/utf8.cc
-        src/encoding/utf8/utf8_regexp.cc
+        src/encoding/ebcdic.cc
+        src/encoding/utf16.cc
+        src/encoding/utf8.cc
         src/msg/msg.cc
         src/msg/warn.cc
         src/regexp/ast_to_re.cc
         src/regexp/default_tags.cc
         src/regexp/fixed_tags.cc
         src/regexp/nullable.cc
-        src/regexp/re.cc
-        src/regexp/rule.cc
+        src/regexp/regexp.cc
         src/regexp/split_charset.cc
-        src/regexp/tag.cc
         src/skeleton/control_flow.cc
         src/skeleton/maxpath.cc
         src/skeleton/skeleton.cc
         src/util/range.cc
-        src/util/s_to_n32_unsafe.cc
-    )
-    set(libre2c_sources
-        $<TARGET_OBJECTS:libre2c_objects>
+        src/util/file_utils.cc
+        src/util/string_utils.cc
         $<TARGET_OBJECTS:libre2c_objects_autogen>
         $<TARGET_OBJECTS:re2c_objects_autogen_ver_to_vernum>
     )

CMakePresets.json

@@ -0,0 +1,498 @@
+{
+  "version": 3,
+  "cmakeMinimumRequired": {
+    "major": 3,
+    "minor": 21,
+    "patch": 1
+  },
+  "configurePresets": [
+    {
+      "name": "default-full-configuration",
+      "hidden": true,
+      "installDir": "${sourceDir}/install-2",
+      "cacheVariables": {
+        "RE2C_BUILD_RE2GO": true,
+        "RE2C_BUILD_RE2RUST": true,
+        "RE2C_BUILD_LIBS": true,
+        "RE2C_REBUILD_LEXERS": true,
+        "RE2C_REBUILD_DOCS": true,
+        "RE2C_FOR_BUILD": "${sourceDir}/install/bin/re2c"
+      }
+    },
+
+    {
+      "name": "linux-config-base",
+      "hidden": true,
+      "generator": "Unix Makefiles",
+      "condition": {
+        "type": "equals",
+        "lhs": "${hostSystemName}",
+        "rhs": "Linux"
+      },
+      "architecture": {
+        "value": "x64",
+        "strategy": "external"
+      },
+      "cacheVariables": {
+        "CMAKE_C_COMPILER": "gcc",
+        "CMAKE_CXX_COMPILER": "g++"
+      }
+    },
+    {
+      "name": "linux-gcc-ubsan-ootree-fast",
+      "displayName": "Linux x86_64: GCC / Ubsan / Out of tree (fast)",
+      "inherits": "linux-config-base",
+      "binaryDir": "${sourceDir}/.build/${presetName}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Ubsan",
+        "CMAKE_C_FLAGS": "-O2",
+        "CMAKE_CXX_FLAGS": "-O2",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "linux-gcc-ubsan-ootree-full",
+      "displayName": "Linux x86_64: GCC / Ubsan / Out of tree (full)",
+      "inherits": ["default-full-configuration", "linux-gcc-ubsan-ootree-fast"]
+    },
+    {
+      "name": "linux-gcc-asan-ootree-fast",
+      "displayName": "Linux x86_64: GCC / Asan / Out of tree (fast)",
+      "inherits": "linux-config-base",
+      "binaryDir": "${sourceDir}/.build/${presetName}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Asan",
+        "CMAKE_C_FLAGS": "-O2",
+        "CMAKE_CXX_FLAGS": "-O2",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "linux-gcc-asan-ootree-full",
+      "displayName": "Linux x86_64: GCC / Asan / Out of tree (full)",
+      "inherits": ["default-full-configuration", "linux-gcc-asan-ootree-fast"]
+    },
+    {
+      "name": "linux-gcc-debug-ootree-skeleton-fast",
+      "displayName": "Linux x86_64: GCC / Debug / Skeleton / Out of tree (fast)",
+      "inherits": "linux-config-base",
+      "binaryDir": "${sourceDir}/.build/${presetName}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Debug",
+        "CMAKE_C_FLAGS": "-O2",
+        "CMAKE_CXX_FLAGS": "-O2",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "linux-gcc-debug-ootree-skeleton-full",
+      "displayName": "Linux x86_64: GCC / Debug / Skeleton / Out of tree (full)",
+      "inherits": ["default-full-configuration", "linux-gcc-debug-ootree-skeleton-fast"]
+    },
+    {
+      "name": "linux-gcc-debug-intree-m32-skeleton-fast",
+      "displayName": "Linux x86: GCC / Debug / Skeleton / In tree (fast)",
+      "inherits": "linux-config-base",
+      "binaryDir": "${sourceDir}",
+      "installDir": "${sourceDir}/install",
+      "architecture": {
+        "value": "x86",
+        "strategy": "external"
+      },
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Debug",
+        "CMAKE_C_FLAGS": "-O2 -m32",
+        "CMAKE_CXX_FLAGS": "-O2 -m32",
+        "CMAKE_EXE_LINKER_FLAGS": "-m32",
+        "CMAKE_SHARED_LINKER_FLAGS": "-m32",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "linux-gcc-debug-intree-m32-skeleton-full",
+      "displayName": "Linux x86: GCC / Debug / Skeleton / In tree (full)",
+      "inherits": ["default-full-configuration", "linux-gcc-debug-intree-m32-skeleton-fast"]
+    },
+    {
+      "name": "linux-gcc-release-ootree-skeleton-fast",
+      "displayName": "Linux x86_64: GCC / Release / Skeleton / Out of tree (fast)",
+      "inherits": "linux-config-base",
+      "binaryDir": "${sourceDir}/.build/${presetName}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Release",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "linux-gcc-release-ootree-skeleton-full",
+      "displayName": "Linux x86_64: GCC / Release / Skeleton / Out of tree (full)",
+      "inherits": ["default-full-configuration", "linux-gcc-release-ootree-skeleton-fast"]
+    },
+    {
+      "name": "linux-gcc-valgrind-ootree-fast",
+      "displayName": "Linux x86_64: GCC / Valgrind / Out of tree (fast)",
+      "inherits": ["linux-gcc-release-ootree-skeleton-fast"]
+    },
+    {
+      "name": "linux-gcc-valgrind-ootree-full",
+      "displayName": "Linux x86_64: GCC / Valgrind / Out of tree (full)",
+      "inherits": ["linux-gcc-release-ootree-skeleton-full"]
+    },
+
+
+    {
+      "name": "macos-config-base",
+      "hidden": true,
+      "generator": "Unix Makefiles",
+      "condition": {
+        "type": "equals",
+        "lhs": "${hostSystemName}",
+        "rhs": "Darwin"
+      },
+      "architecture": {
+        "value": "x64",
+        "strategy": "external"
+      },
+      "cacheVariables": {
+        "CMAKE_C_COMPILER": "clang",
+        "CMAKE_CXX_COMPILER": "clang++"
+      }
+    },
+    {
+      "name": "macos-clang-debug-ootree-skeleton-fast",
+      "displayName": "macOS x86_64: Clang / Debug / Skeleton / Out of tree (fast)",
+      "inherits": "macos-config-base",
+      "binaryDir": "${sourceDir}/.build/${presetName}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Debug",
+        "CMAKE_C_FLAGS": "-O2",
+        "CMAKE_CXX_FLAGS": "-O2",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "macos-clang-debug-ootree-skeleton-full",
+      "displayName": "macOS x86_64: Clang / Debug / Skeleton / Out of tree (full)",
+      "inherits": ["default-full-configuration", "macos-clang-debug-ootree-skeleton-fast"]
+    },
+    {
+      "name": "macos-clang-debug-intree-skeleton-fast",
+      "displayName": "macOS x86_64: Clang / Debug / Skeleton / In tree (fast)",
+      "inherits": "macos-config-base",
+      "binaryDir": "${sourceDir}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Debug",
+        "CMAKE_C_FLAGS": "-O2",
+        "CMAKE_CXX_FLAGS": "-O2",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "macos-clang-debug-intree-skeleton-full",
+      "displayName": "macOS x86_64: Clang / Debug / Skeleton / In tree (full)",
+      "inherits": ["default-full-configuration", "macos-clang-debug-intree-skeleton-fast"]
+    },
+    {
+      "name": "macos-clang-release-ootree-skeleton-fast",
+      "displayName": "macOS x86_64: CLang / Release / Skeleton / Out of tree (fast)",
+      "inherits": "macos-config-base",
+      "binaryDir": "${sourceDir}/.build/${presetName}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Release",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "macos-clang-release-ootree-skeleton-full",
+      "displayName": "macOS x86_64: CLang / Release / Skeleton / Out of tree (full)",
+      "inherits": ["default-full-configuration", "macos-clang-release-ootree-skeleton-fast"]
+    },
+
+
+    {
+      "name": "windows-config-base",
+      "hidden": true,
+      "generator": "Visual Studio 17 2022",
+      "condition": {
+        "type": "equals",
+        "lhs": "${hostSystemName}",
+        "rhs": "Windows"
+      },
+      "architecture": {
+        "value": "x64",
+        "strategy": "external"
+      },
+      "cacheVariables": {
+        "CMAKE_C_COMPILER": "cl",
+        "CMAKE_CXX_COMPILER": "cl"
+      }
+    },
+    {
+      "name": "windows-msvc-debug-ootree-fast",
+      "displayName": "Windows x86_64: MSVC / Debug / Out of tree (fast)",
+      "inherits": "windows-config-base",
+      "binaryDir": "${sourceDir}/.build/${presetName}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Debug",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "windows-msvc-debug-ootree-full",
+      "displayName": "Windows x86_64: MSVC / Debug / Out of tree (full)",
+      "inherits": ["default-full-configuration", "windows-msvc-debug-ootree-fast"]
+    },
+    {
+      "name": "windows-msvc-debug-intree-fast",
+      "displayName": "Windows x86_64: MSVC / Debug / In tree (fast)",
+      "inherits": "windows-config-base",
+      "binaryDir": "${sourceDir}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Debug",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "windows-msvc-debug-intree-full",
+      "displayName": "Windows x86_64: MSVC / Debug / In tree (full)",
+      "inherits": ["default-full-configuration", "windows-msvc-debug-intree-fast"]
+    },
+    {
+      "name": "windows-msvc-release-ootree-fast",
+      "displayName": "Windows x86_64: MSVC / Release / Out of tree (fast)",
+      "inherits": "windows-config-base",
+      "binaryDir": "${sourceDir}/.build/${presetName}",
+      "installDir": "${sourceDir}/install",
+      "cacheVariables": {
+        "CMAKE_BUILD_TYPE": "Release",
+        "RE2C_BUILD_RE2GO": false,
+        "RE2C_BUILD_RE2RUST": false
+      }
+    },
+    {
+      "name": "windows-msvc-release-ootree-full",
+      "displayName": "Windows x86_64: MSVC / Release / Out of tree (full)",
+      "inherits": ["default-full-configuration", "windows-msvc-release-ootree-fast"]
+    }
+  ],
+  "buildPresets": [
+    {
+      "name": "linux-build-base",
+      "hidden": true,
+      "condition": {
+        "type": "equals",
+        "lhs": "${hostSystemName}",
+        "rhs": "Linux"
+      }
+    },
+    {
+      "name": "linux-gcc-ubsan-ootree-fast",
+      "displayName": "Linux x86_64: GCC / Ubsan / Out of tree (fast)",
+      "configurePreset": "linux-gcc-ubsan-ootree-fast",
+      "inherits": "linux-build-base",
+      "configuration": "Ubsan"
+    },
+    {
+      "name": "linux-gcc-ubsan-ootree-full",
+      "displayName": "Linux x86_64: GCC / Ubsan / Out of tree (full)",
+      "configurePreset": "linux-gcc-ubsan-ootree-full",
+      "inherits": "linux-build-base",
+      "configuration": "Ubsan"
+    },
+    {
+      "name": "linux-gcc-asan-ootree-fast",
+      "displayName": "Linux x86_64: GCC / Asan / Out of tree (fast)",
+      "configurePreset": "linux-gcc-asan-ootree-fast",
+      "inherits": "linux-build-base",
+      "configuration": "Asan"
+    },
+    {
+      "name": "linux-gcc-asan-ootree-full",
+      "displayName": "Linux x86_64: GCC / Asan / Out of tree (full)",
+      "configurePreset": "linux-gcc-asan-ootree-full",
+      "inherits": "linux-build-base",
+      "configuration": "Asan"
+    },
+    {
+      "name": "linux-gcc-debug-ootree-skeleton-fast",
+      "displayName": "Linux x86_64: GCC / Debug / Skeleton / Out of tree (fast)",
+      "configurePreset": "linux-gcc-debug-ootree-skeleton-fast",
+      "inherits": "linux-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "linux-gcc-debug-ootree-skeleton-full",
+      "displayName": "Linux x86_64: GCC / Debug / Skeleton / Out of tree (full)",
+      "configurePreset": "linux-gcc-debug-ootree-skeleton-full",
+      "inherits": "linux-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "linux-gcc-debug-intree-m32-skeleton-fast",
+      "displayName": "Linux x86: GCC / Debug / Skeleton / In tree (fast)",
+      "configurePreset": "linux-gcc-debug-intree-m32-skeleton-fast",
+      "inherits": "linux-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "linux-gcc-debug-intree-m32-skeleton-full",
+      "displayName": "Linux x86: GCC / Debug / Skeleton / In tree (full)",
+      "configurePreset": "linux-gcc-debug-intree-m32-skeleton-full",
+      "inherits": "linux-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "linux-gcc-release-ootree-skeleton-fast",
+      "displayName": "Linux x86_64: GCC / Release / Skeleton / Out of tree (fast)",
+      "configurePreset": "linux-gcc-release-ootree-skeleton-fast",
+      "inherits": "linux-build-base",
+      "configuration": "Release"
+    },
+    {
+      "name": "linux-gcc-release-ootree-skeleton-full",
+      "displayName": "Linux x86_64: GCC / Release / Skeleton / Out of tree (full)",
+      "configurePreset": "linux-gcc-release-ootree-skeleton-full",
+      "inherits": "linux-build-base",
+      "configuration": "Release"
+    },
+    {
+      "name": "linux-gcc-valgrind-ootree-fast",
+      "displayName": "Linux x86_64: GCC / Valgrind / Out of tree (fast)",
+      "configurePreset": "linux-gcc-valgrind-ootree-fast",
+      "inherits": "linux-build-base",
+      "configuration": "Release"
+    },
+    {
+      "name": "linux-gcc-valgrind-ootree-full",
+      "displayName": "Linux x86_64: GCC / Valgrind / Out of tree (full)",
+      "configurePreset": "linux-gcc-valgrind-ootree-full",
+      "inherits": "linux-build-base",
+      "configuration": "Release"
+    },
+
+
+    {
+      "name": "macos-build-base",
+      "hidden": true,
+      "condition": {
+        "type": "equals",
+        "lhs": "${hostSystemName}",
+        "rhs": "Darwin"
+      }
+    },
+    {
+      "name": "macos-clang-debug-ootree-skeleton-fast",
+      "displayName": "macOS x86_64: Clang / Debug / Skeleton / Out of tree (fast)",
+      "configurePreset": "macos-clang-debug-ootree-skeleton-fast",
+      "inherits": "macos-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "macos-clang-debug-ootree-skeleton-full",
+      "displayName": "macOS x86_64: Clang / Debug / Skeleton / Out of tree (full)",
+      "configurePreset": "macos-clang-debug-ootree-skeleton-full",
+      "inherits": "macos-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "macos-clang-debug-intree-skeleton-fast",
+      "displayName": "macOS x86_64: Clang / Debug / Skeleton / In tree (fast)",
+      "configurePreset": "macos-clang-debug-intree-skeleton-fast",
+      "inherits": "macos-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "macos-clang-debug-intree-skeleton-full",
+      "displayName": "macOS x86_64: Clang / Debug / Skeleton / In tree (full)",
+      "configurePreset": "macos-clang-debug-intree-skeleton-full",
+      "inherits": "macos-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "macos-clang-release-ootree-skeleton-fast",
+      "displayName": "macOS x86_64: CLang / Release / Skeleton / Out of tree (fast)",
+      "configurePreset": "macos-clang-release-ootree-skeleton-fast",
+      "inherits": "macos-build-base",
+      "configuration": "Release"
+    },
+    {
+      "name": "macos-clang-release-ootree-skeleton-full",
+      "displayName": "macOS x86_64: CLang / Release / Skeleton / Out of tree (full)",
+      "configurePreset": "macos-clang-release-ootree-skeleton-full",
+      "inherits": "macos-build-base",
+      "configuration": "Release"
+    },
+
+
+    {
+      "name": "windows-build-base",
+      "hidden": true,
+      "condition": {
+        "type": "equals",
+        "lhs": "${hostSystemName}",
+        "rhs": "Windows"
+      }
+    },
+    {
+      "name": "windows-msvc-debug-ootree-fast",
+      "displayName": "Windows x86_64: MSVC / Debug / Out of tree (fast)",
+      "configurePreset": "windows-msvc-debug-ootree-fast",
+      "inherits": "windows-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "windows-msvc-debug-ootree-full",
+      "displayName": "Windows x86_64: MSVC / Debug / Out of tree (full)",
+      "configurePreset": "windows-msvc-debug-ootree-full",
+      "inherits": "windows-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "windows-msvc-debug-intree-fast",
+      "displayName": "Windows x86_64: MSVC / Debug / In tree (fast)",
+      "configurePreset": "windows-msvc-debug-intree-fast",
+      "inherits": "windows-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "windows-msvc-debug-intree-full",
+      "displayName": "Windows x86_64: MSVC / Debug / In tree (full)",
+      "configurePreset": "windows-msvc-debug-intree-full",
+      "inherits": "windows-build-base",
+      "configuration": "Debug"
+    },
+    {
+      "name": "windows-msvc-release-ootree-fast",
+      "displayName": "Windows x86_64: MSVC / Release / Out of tree (fast)",
+      "configurePreset": "windows-msvc-release-ootree-fast",
+      "inherits": "windows-build-base",
+      "configuration": "Release"
+    },
+    {
+      "name": "windows-msvc-release-ootree-full",
+      "displayName": "Windows x86_64: MSVC / Release / Out of tree (full)",
+      "configurePreset": "windows-msvc-release-ootree-full",
+      "inherits": "windows-build-base",
+      "configuration": "Release"
+    }
+  ]
+}

CONTRIBUTING.md

@@ -11,30 +11,45 @@ convenient).
 
 ## Language guidelines
 
-re2c is written in C++. The project uses a conservative language subset (for
-portability reasons the currently used standard is C++98). The code should stay
-simple and avoid complex object hierarchies and excessive use of template
-metaprogramming. It should allow the reader to concentrate on the underlying
-algorithms, without the language getting in the way.
+re2c is written in C++ (the current standard is C++11). The algorithms and
+mathematical concepts used in the code are quite complex, so the code itself
+should stay simple and let the developer focus on the underlying logic, without
+the language getting in the way. For example, try to avoid complex object
+hierarchies and excessive use of template metaprogramming.
 
-The line length limit is 90 characters, indentation is 4 spaces, and the
-indentation style is more or less K&R (egyptian braces).
+The line length limit is 100 characters, indentation is 4 spaces, and the
+indentation style is egyptian (the opening brace is on the same line). This is
+the approximate `astyle` command to auto-format re2c source code (it requires
+manual tweaking for argument/initializer lists and multiline comments):
 
-## Updating bootstrap files
+    $ # run in source directory
+    $ astyle --style=google \
+             --max-code-length=100 \
+             --align-pointer=type \
+             --align-reference=type \
+             --keep-one-line-statements \
+             --suffix=none \
+             --recursive '*'
 
-For changes that affect lexers (`.re` files), it is necessary to regenerate the
-corresponding bootstrap files. This can be done by reconfiguring and
-rebuilding re2c with options `--enable-lexers` (Autotools) or
-`-DRE2C_REBUILD_LEXERS=ON` (CMake). `RE2C_FOR_BUILD` should be defined as a
-path to an already existing re2c executable that will be used to regenerate
-the lexers.
+## Regenerating bootstrap files
 
-For changes that affect parser (`.ypp` files), it is necessary to regenerate
-bootstrap parser files. This happens automatically if bison is installed.
+By default re2c uses bootstrap files for autogenerated sources (lexers, parsers,
+docs) and does not attempt to regenerate them. Developers may need to enable
+"regeneration mode" if they modify the corresponding source files.
 
-For changes that update documentation, it is necessary to rebuild the
-bootstrap manpages by reconfiguring and rebuilding re2c with options
-`--enable-docs` (Autotools) or `-DRE2C_REBUILD_DOCS=ON` (CMake).
+For changes that affect lexers (`.re` files), regeneration mode is configured
+with options `--enable-lexers` (Autotools) or `-DRE2C_REBUILD_LEXERS=ON`
+(CMake). `RE2C_FOR_BUILD` should be defined as a path to an existing re2c
+executable that will be used to regenerate the lexers.
+
+For changes that affect parsers (`.ypp` files), regeneration mode is configured
+with options `--enable-parsers` (Autotools) or `-DRE2C_REBUILD_PARSERS=ON`
+(CMake). In this case `bison` must be installed on the system. Otherwise
+Autoconf or CMake will fail at configuration stage.
+
+For changes that update documentation, it is necessary to rebuild the bootstrap
+manpages by reconfiguring and rebuilding re2c with options `--enable-docs`
+(Autotools) or `-DRE2C_REBUILD_DOCS=ON` (CMake).
 
 Please do not commit trivial updates to the bootstrap files, such as changes of
 the generator version or generation date.

IDEAS.md

@@ -0,0 +1,46 @@
+If you run out of ideas of how to make re2c better...
+
+## Multibyte switch
+
+Reading multiple characters at a time is problematic due to the absence of
+alignment guarantees: even if unaligned reads are supported by the CPU, they are
+still likely to be slow.
+
+But it should be possible to read a few characters, combine them into one using
+shifts and bitwise operations, and then switch on the combined character. It can
+be done with existing API; no new primitives are needed.
+
+Of course it will only help in cases when the underlying DFA has long "chains"
+of transitions without branches that can be collapsed into one transition. But
+the optimization can be local (applied to some parts of the DFA).
+
+## User-defined language support
+
+The idea originated in a linux.org.ru thread (in Russian):
+    https://www.linux.org.ru/news/development/16759268?cid=16762746
+
+The original idea was to provide an API to render the generated AST into code
+and let the user implement it in the form of a plugin/library. This is
+problematic because 1) there is no stable AST, and 2) there are some language-
+agnostic optimizations and decisions that re2c takes way before the final
+rendering phase (so it is not as simple as providing a bunch or functions to
+generate code for different AST nodes).
+
+There may be another way to provide any-language support: let the user define a
+set of primitives/configurations that re2c will use for language-agnostic
+decisions and for code generation (some of them exist already, but not enough to
+fully redefine the target language). This is akin to the idea of Midnight
+Commander's skins or syntax files. re2c can provide default syntax files for
+some languages (it already provides "standard" Unicode categories).
+
+## Negation and complement
+
+It would be good to support these operations on regular expressions, but it
+requires some research. Does TDFA present any special difficulties?
+
+## Location support
+
+It would be good to have something like Bison locations (automatic tracking of
+line/column/filename inforamtion). This is primarily useful for tokenizers and
+large lexers.
+

MAINTAINERS.md

@@ -0,0 +1,41 @@
+If re2c is unmaintained for a long time and starts to bitrot, or if you have a
+fork and need to make it as simple as possible, here is a list of things that
+can be cut off rather safely vs. things that are essential.
+
+## Things to remove
+
+These are mostly experimental research projects that have never been used
+outside of lab (to the best of author's knowledge, that is):
+
+- Benchmarks: drop optional dependencies on re2, google-benchmark, Java (libre2c
+  has JNI bindings); simplify the build system(s).
+
+- libre2c: drop static/dynamic libraries support in the build system(s), e.g.
+  Libtool; remove a few internal options and a bunch of explicit template
+  instantiations in TDFA construction code that exist only for libre2c.
+
+- One of the build systems: Autotools or CMake, whichever you like less. Someone
+  will get unhappy in both cases (Windows is supported only by CMake, but distro
+  maintainers may have decades old recipes based on Autotools).
+
+## Things to remove only as a last resort
+
+- Skeleton: it is really useful for testing and codegen validation, but re2c can
+  live without it (and a few large tests can be dropped).
+
+- Fuzzers: same here, they are really useful to find bugs and test hypotheses,
+  but re2c can function without them.
+
+## Things to keep
+
+- Existing configurations / options / syntax: re2c must stay backwards
+  compatible or it will become unusable. One exception is internal/debug options
+  which may be removed with caution. For the rest, keep them as deprecated even
+  if they do nothing.
+
+- Language backends: they add little code complexity for a lot of value (and
+  discipline in the codegen subsystem).
+
+- Documentation (both on `master` branch and the website on `gh-pages` and
+  `gh-pages-gen` branches). Without docs no one will know how to use re2c.
+

Makefile.am

@@ -10,7 +10,7 @@ if DEBUG
 AM_CXXFLAGS += -DRE2C_DEBUG
 endif
 
-RE2CFLAGS = -br -W -Wno-match-empty-string
+RE2CFLAGS = -b -W -Wno-match-empty-string --no-generation-date
 
 # binary
 bin_PROGRAMS = re2c
@@ -20,6 +20,7 @@ noinst_PROGRAMS =
 re2c_HDR = \
 	src/constants.h \
 	src/codegen/code.h \
+	src/codegen/output.h \
 	src/codegen/helpers.h \
 	src/options/opt.h \
 	src/options/symtab.h \
@@ -30,70 +31,49 @@ re2c_HDR = \
 	src/dfa/determinization.h \
 	src/dfa/dfa.h \
 	src/dfa/posix_precedence.h \
-	src/dfa/stacmd.h \
 	src/dfa/tag_history.h \
 	src/dfa/tagver_table.h \
 	src/dfa/tcmd.h \
 	src/nfa/nfa.h \
-	src/encoding/case.h \
-	src/encoding/ebcdic/ebcdic.h \
-	src/encoding/ebcdic/ebcdic_regexp.h \
+	src/encoding/ebcdic.h \
 	src/encoding/enc.h \
 	src/encoding/range_suffix.h \
-	src/encoding/utf8/utf8.h \
-	src/encoding/utf8/utf8_regexp.h \
-	src/encoding/utf16/utf16_regexp.h \
-	src/encoding/utf16/utf16.h \
+	src/encoding/utf8.h \
+	src/encoding/utf16.h \
 	src/msg/location.h \
 	src/msg/msg.h \
 	src/msg/ver_to_vernum.h \
 	src/msg/warn.h \
-	src/regexp/re.h \
+	src/regexp/regexp.h \
 	src/regexp/rule.h \
 	src/regexp/tag.h \
-	src/compile.h \
 	src/skeleton/mtag_trie.h \
 	src/skeleton/path.h \
 	src/skeleton/skeleton.h \
 	src/parse/ast.h \
 	src/parse/input.h \
-	src/parse/parse.h \
-	src/parse/scanner.h \
-	src/parse/unescape.h \
 	src/debug/debug.h \
-	src/util/allocate.h \
+	src/util/allocator.h \
 	src/util/attribute.h \
-	src/util/c99_stdint.h \
+	src/util/containers.h \
+	src/util/check.h \
 	src/util/file_utils.h \
-	src/util/fixed_allocator.h \
 	src/util/forbid_copy.h \
-	src/util/free_list.h \
 	src/util/hash32.h \
-	src/util/local_increment.h \
-	src/util/lookup.h \
+	src/util/nowarn_in_bison.h \
 	src/util/range.h \
-	src/util/s_to_n32_unsafe.h \
-	src/util/slab_allocator.h \
-	src/util/smart_ptr.h \
-	src/util/static_assert.h \
 	src/util/string_utils.h \
-	src/util/temp_file.h \
-	src/util/u32lim.h \
-	src/util/uniq_vector.h
+	src/util/u32lim.h
+
 re2c_SRC = \
-	src/codegen/code_goto.cc \
-	src/codegen/combine.cc \
-	src/codegen/gen_bitmaps.cc \
-	src/codegen/gen_delayed.cc \
-	src/codegen/gen_dfa.cc \
-	src/codegen/gen_goto.cc \
-	src/codegen/gen_program.cc \
-	src/codegen/gen_state.cc \
 	src/codegen/helpers.cc \
-	src/codegen/render.cc \
+	src/codegen/output.cc \
+	src/codegen/pass1_analyze.cc \
+	src/codegen/pass2_generate.cc \
+	src/codegen/pass3_fixup.cc \
+	src/codegen/pass4_render.cc \
 	src/options/opt.cc \
 	src/options/symtab.cc \
-	src/nfa/estimate_size.cc \
 	src/nfa/re_to_nfa.cc \
 	src/adfa/adfa.cc \
 	src/debug/dump_adfa.cc \
@@ -119,27 +99,21 @@ re2c_SRC = \
 	src/dfa/fillpoints.cc \
 	src/dfa/find_state.cc \
 	src/dfa/minimization.cc \
-	src/dfa/stacmd.cc \
 	src/dfa/tagver_table.cc \
 	src/dfa/tcmd.cc \
-	src/encoding/ebcdic/ebcdic_regexp.cc \
+	src/encoding/ebcdic.cc \
 	src/encoding/enc.cc \
 	src/encoding/range_suffix.cc \
-	src/encoding/utf8/utf8_regexp.cc \
-	src/encoding/utf8/utf8.cc \
-	src/encoding/utf16/utf16_regexp.cc \
-	src/encoding/utf16/utf16.cc \
+	src/encoding/utf8.cc \
+	src/encoding/utf16.cc \
 	src/msg/msg.cc \
 	src/msg/warn.cc \
 	src/regexp/ast_to_re.cc \
 	src/regexp/default_tags.cc \
 	src/regexp/fixed_tags.cc \
 	src/regexp/nullable.cc \
-	src/regexp/re.cc \
-	src/regexp/rule.cc \
+	src/regexp/regexp.cc \
 	src/regexp/split_charset.cc \
-	src/regexp/tag.cc \
-	src/compile.cc \
 	src/skeleton/control_flow.cc \
 	src/skeleton/generate_code.cc \
 	src/skeleton/generate_data.cc \
@@ -147,12 +121,8 @@ re2c_SRC = \
 	src/skeleton/skeleton.cc \
 	src/parse/ast.cc \
 	src/parse/input.cc \
-	src/parse/parse.cc \
-	src/parse/scanner.cc \
-	src/parse/unescape.cc \
 	src/util/file_utils.cc \
-	src/util/s_to_n32_unsafe.cc \
-	src/util/temp_file.cc \
+	src/util/string_utils.cc \
 	src/util/range.cc
 re2c_SOURCES = \
 	src/main.cc \
@@ -188,10 +158,12 @@ re2c_BOOT_PARSER = bootstrap/src/parse/parser.cc
 re2c_BOOT_PARSER_HDR = bootstrap/src/parse/parser.h
 re2c_BOOT_DOC_C = bootstrap/doc/re2c.1
 re2c_BOOT_DOC_GO = bootstrap/doc/re2go.1
+re2c_BOOT_DOC_RUST = bootstrap/doc/re2rust.1
 re2c_BOOT_HELP = bootstrap/src/msg/help.cc
 re2c_BOOT = \
 	$(re2c_BOOT_DOC_C) \
 	$(re2c_BOOT_DOC_GO) \
+	$(re2c_BOOT_DOC_RUST) \
 	$(re2c_BOOT_HELP) \
 	$(re2c_BOOT_PARSER) \
 	$(re2c_BOOT_PARSER_HDR) \
@@ -215,8 +187,8 @@ re2c_CUSTOM = \
 # docs
 re2c_SRC_DOC = doc/manpage.rst
 re2c_SRC_DOC_EXT = \
-	doc/manual/api/api.rst_ \
 	doc/manual/api/interface.rst_ \
+	doc/manual/conditions/blocks.rst_ \
 	doc/manual/conditions/conditions.rst_ \
 	doc/manual/configurations/configurations.rst_ \
 	doc/manual/directives/directives.rst_ \
@@ -241,13 +213,22 @@ re2c_SRC_DOC_EXT = \
 	doc/manual/state/state.rst_ \
 	doc/manual/submatch/submatch_example_mtags.rst_ \
 	doc/manual/submatch/submatch_example_posix.rst_ \
-	doc/manual/submatch/submatch_example_stags.rst_ \
 	doc/manual/submatch/submatch_example_stags_fill.rst_ \
+	doc/manual/submatch/submatch_example_stags.rst_ \
 	doc/manual/submatch/submatch.rst_ \
+	doc/manual/synopsis.rst_ \
+	doc/manual/syntax/api1.rst_ \
+	doc/manual/syntax/api2_c.rst_ \
+	doc/manual/syntax/api2_go.rst_ \
+	doc/manual/syntax/api2_rust.rst_ \
+	doc/manual/syntax/api3.rst_ \
+	doc/manual/syntax/intro.rst_ \
 	doc/manual/syntax/syntax.rst_ \
 	doc/manual/warnings/warnings_general.rst_ \
 	doc/manual/warnings/warnings_list.rst_ \
 	examples/c/01_basic.re \
+	examples/c/01_basic.c \
+	examples/c/conditions/parse_u32_blocks.re \
 	examples/c/conditions/parse_u32_conditions.re \
 	examples/c/encodings/unicode_identifier.re \
 	examples/c/eof/01_sentinel.re \
@@ -257,17 +238,19 @@ re2c_SRC_DOC_EXT = \
 	examples/c/fill/01_fill.re \
 	examples/c/fill/02_fill.re \
 	examples/c/headers/header.re \
-	examples/c/headers/src/lexer/lexer.h \
+	examples/c/headers/lexer/state.h \
 	examples/c/includes/include.re \
 	examples/c/includes/definitions.h \
 	examples/c/reuse/reuse.re \
 	examples/c/reuse/usedir.re \
 	examples/c/state/push.re \
-	examples/c/submatch/01_stags.re \
 	examples/c/submatch/01_stags_fill.re \
+	examples/c/submatch/01_stags.re \
 	examples/c/submatch/02_mtags.re \
 	examples/c/submatch/03_posix.re \
 	examples/go/01_basic.re \
+	examples/go/01_basic.go \
+	examples/go/conditions/parse_u32_blocks.re \
 	examples/go/conditions/parse_u32_conditions.re \
 	examples/go/encodings/unicode_identifier.re \
 	examples/go/eof/01_sentinel.re \
@@ -277,38 +260,79 @@ re2c_SRC_DOC_EXT = \
 	examples/go/fill/01_fill.re \
 	examples/go/fill/02_fill.re \
 	examples/go/headers/header.re \
-	examples/go/headers/src/lexer/lexer.go \
+	examples/go/headers/lexer/state.go \
 	examples/go/includes/include.re \
 	examples/go/includes/definitions.go \
 	examples/go/reuse/reuse.re \
 	examples/go/reuse/usedir.re \
 	examples/go/state/push.re \
+	examples/go/submatch/01_stags_fill.re \
 	examples/go/submatch/01_stags.re \
 	examples/go/submatch/02_mtags.re \
-	examples/go/submatch/03_posix.re
+	examples/go/submatch/03_posix.re \
+	examples/rust/01_basic.re \
+	examples/rust/01_basic.rs \
+	examples/rust/conditions/parse_u32_blocks.re \
+	examples/rust/conditions/parse_u32_conditions.re \
+	examples/rust/encodings/unicode_identifier.re \
+	examples/rust/eof/01_sentinel.re \
+	examples/rust/eof/02_bounds_checking.re \
+	examples/rust/eof/03_eof_rule.re \
+	examples/rust/eof/05_generic_api_eof_rule.re \
+	examples/rust/fill/01_fill.re \
+	examples/rust/fill/02_fill.re \
+	examples/rust/headers/header.re \
+	examples/rust/headers/lexer/state.rs \
+	examples/rust/includes/include.re \
+	examples/rust/includes/definitions.rs \
+	examples/rust/reuse/reuse.re \
+	examples/rust/reuse/usedir.re \
+	examples/rust/state/push.re \
+	examples/rust/submatch/01_stags_fill.re \
+	examples/rust/submatch/01_stags.re \
+	examples/rust/submatch/02_mtags.re \
+	examples/rust/submatch/03_posix.re
 
 DOC_C = doc/re2c.1
 DOCS = $(DOC_C)
+
 if WITH_GOLANG
 DOC_GO = doc/re2go.1
 DOCS += $(DOC_GO)
 endif
+
+if WITH_RUST
+DOC_RUST = doc/re2rust.1
+DOCS += $(DOC_RUST)
+endif
+
 man_MANS = $(DOCS)
 
 # include files (stdlib)
 dist_stdlib_DATA = include/unicode_categories.re
 
+BAZELFILES := \
+	BUILD.bazel \
+	WORKSPACE \
+	bazel
+
 CMAKEFILES := \
 	CMakeLists.txt \
+	CMakePresets.json \
 	cmake
 
 EXTRA_DIST = \
 	$(re2c_BOOT) \
 	$(re2c_CUSTOM) \
 	$(re2c_SRC_DOC_EXT) \
+	$(BAZELFILES) \
 	$(CMAKEFILES) \
+	BUILD.md \
 	CHANGELOG \
+	CONTRIBUTING.md \
+	IDEAS.md \
 	LICENSE \
+	MAINTAINERS.md \
 	NO_WARRANTY \
 	README.md \
 	autogen.sh \
@@ -324,13 +348,8 @@ if REBUILD_LEXERS
 # generate lexer and update bootstrap files
 .re.cc:
 	$(AM_V_at)$(MKDIR_P) $(@D)
-	$(AM_V_GEN)if test -x $(RE2C_FOR_BUILD); \
-	then \
-		$(RE2C_FOR_BUILD) $(RE2CFLAGS) -o $@ $< && \
-		cp $(@:cc=[ch]*) $(top_srcdir)/bootstrap/$(@D); \
-	else \
-		echo "RE2C is not an executable file" && exit 1 ; \
-	fi
+	$(AM_V_GEN)$(RE2C_FOR_BUILD) $(RE2CFLAGS) -o $@ $<
+	$(AM_V_GEN)cp $(@:cc=[ch]*) $(top_srcdir)/bootstrap/$(@D);
 else
 # copy bootstrap files
 .re.cc:
@@ -339,19 +358,19 @@ else
 	$(AM_V_at)echo "Reconfigure with --enable-lexers to regenerate $@"
 endif
 
-# cannot update bootstrap with distcheck (it makes srcdir readonly),
-# so we update bootstrap only on incremental build (if re2c binary exists)
+if REBUILD_PARSERS
+# generate lexer and update bootstrap files
 .ypp.cc:
 	$(AM_V_at)$(MKDIR_P) $(@D)
-	$(AM_V_GEN)if test $(BISON) = "no"; \
-	then \
-		cp $(top_srcdir)/bootstrap/$(@:cc=*) $(@D); \
-	else \
-		$(BISON) --output=$@ --defines=$(@:cc=h) $< && \
-		if test -x $(top_builddir)/re2c$(EXEEXT); then \
-			cp $@ $(@:cc=h) $(top_srcdir)/bootstrap/$(@D); \
-		fi; \
-	fi
+	$(AM_V_GEN)$(BISON) --warnings --output=$@ --defines=$(@:cc=h) $<
+	$(AM_V_GEN)cp $@ $(@:cc=h) $(top_srcdir)/bootstrap/$(@D);
+else
+# copy bootstrap files
+.ypp.cc:
+	$(AM_V_at)$(MKDIR_P) $(@D)
+	$(AM_V_GEN)cp $(top_srcdir)/bootstrap/$(@:cc=*) $(@D);
+	$(AM_V_at)echo "Reconfigure with --enable-parsers to regenerate $@"
+endif
 
 # lexer depends on bison-generated header
 $(re2c_GEN_LEX): $(re2c_GEN_PARSER)
@@ -365,29 +384,35 @@ bootstrap: all
 .PHONY: docs
 if REBUILD_DOCS
 docs: $(DOCS) $(re2c_GEN_HELP)
-GENHELP = $(top_srcdir)/build/gen_help.sh
-SPLITMAN = $(top_srcdir)/build/split_man.sh
+RST2TXT = $(top_srcdir)/build/rst2txt.py
+RST2MAN = $(top_srcdir)/build/rst2man.py
+SPLITMAN = $(top_srcdir)/build/split_man.py
 # generate manpage for C
-$(DOC_C): $(re2c_SRC_DOC) $(re2c_SRC_DOC_EXT) $(SPLITMAN)
+$(DOC_C): $(re2c_SRC_DOC) $(re2c_SRC_DOC_EXT) $(SPLITMAN) $(RST2MAN)
 	$(AM_V_at)$(MKDIR_P) $(@D)
-	$(AM_V_GEN)$(SPLITMAN) $(top_builddir)/$(re2c_SRC_DOC) $(top_builddir)/$(re2c_SRC_DOC).c c \
-		&& $(RST2MAN) --tab-width=4 $(top_builddir)/$(re2c_SRC_DOC).c > $@ \
+	$(AM_V_GEN)$(PYTHON) $(SPLITMAN) $(top_builddir)/$(re2c_SRC_DOC) $(top_builddir)/$(re2c_SRC_DOC).c c \
+		&& $(PYTHON) $(RST2MAN) --tab-width=4 $(top_builddir)/$(re2c_SRC_DOC).c > $@ \
 		&& cp $@ $(top_srcdir)/$(re2c_BOOT_DOC_C) \
 		&& rm $(top_builddir)/$(re2c_SRC_DOC).c
 # generate manpage for Go
-$(DOC_GO): $(re2c_SRC_DOC) $(re2c_SRC_DOC_EXT) $(SPLITMAN)
+$(DOC_GO): $(re2c_SRC_DOC) $(re2c_SRC_DOC_EXT) $(SPLITMAN) $(RST2MAN)
 	$(AM_V_at)$(MKDIR_P) $(@D)
-	$(AM_V_GEN)$(SPLITMAN) $(top_builddir)/$(re2c_SRC_DOC) $(top_builddir)/$(re2c_SRC_DOC).go go \
-		&& $(RST2MAN) --tab-width=4 $(top_builddir)/$(re2c_SRC_DOC).go > $@ \
+	$(AM_V_GEN)$(PYTHON) $(SPLITMAN) $(top_builddir)/$(re2c_SRC_DOC) $(top_builddir)/$(re2c_SRC_DOC).go go \
+		&& $(PYTHON) $(RST2MAN) --tab-width=4 $(top_builddir)/$(re2c_SRC_DOC).go > $@ \
 		&& cp $@ $(top_srcdir)/$(re2c_BOOT_DOC_GO) \
 		&& rm $(top_builddir)/$(re2c_SRC_DOC).go
-# generate help
-$(re2c_GEN_HELP): $(re2c_CUSTOM_HELP) $(GENHELP)
+# generate manpage for Rust
+$(DOC_RUST): $(re2c_SRC_DOC) $(re2c_SRC_DOC_EXT) $(SPLITMAN) $(RST2MAN)
 	$(AM_V_at)$(MKDIR_P) $(@D)
-	$(AM_V_GEN)$(RST2MAN) $(top_builddir)/$(re2c_CUSTOM_HELP) > $@.1 \
-		&& $(GENHELP) $@.1 $@ \
-		&& cp $@ $(top_srcdir)/$(re2c_BOOT_HELP) \
-		&& rm $@.1
+	$(AM_V_GEN)$(PYTHON) $(SPLITMAN) $(top_builddir)/$(re2c_SRC_DOC) $(top_builddir)/$(re2c_SRC_DOC).rust rust \
+		&& $(PYTHON) $(RST2MAN) --tab-width=4 $(top_builddir)/$(re2c_SRC_DOC).rust > $@ \
+		&& cp $@ $(top_srcdir)/$(re2c_BOOT_DOC_RUST) \
+		&& rm $(top_builddir)/$(re2c_SRC_DOC).rust
+# generate help
+$(re2c_GEN_HELP): $(re2c_CUSTOM_HELP) $(re2c_SRC_DOC_EXT) $(RST2TXT)
+	$(AM_V_at)$(MKDIR_P) $(@D)
+	$(AM_V_GEN)$(PYTHON) $(RST2TXT) --variable-name=help $(top_builddir)/$(re2c_CUSTOM_HELP) $@ \
+		&& cp $@ $(top_srcdir)/$(re2c_BOOT_HELP)
 else
 docs: $(DOCS) $(re2c_GEN_HELP)
 	$(AM_V_at)echo "Reconfigure with --enable-docs to rebuild docs"
@@ -399,6 +424,10 @@ $(DOC_C): $(re2c_BOOT_DOC_C)
 $(DOC_GO): $(re2c_BOOT_DOC_GO)
 	$(AM_V_at)$(MKDIR_P) $(@D)
 	$(AM_V_GEN)cp $(top_srcdir)/$(re2c_BOOT_DOC_GO) $@
+# copy bootstrap manpage for Rust
+$(DOC_RUST): $(re2c_BOOT_DOC_RUST)
+	$(AM_V_at)$(MKDIR_P) $(@D)
+	$(AM_V_GEN)cp $(top_srcdir)/$(re2c_BOOT_DOC_RUST) $@
 # copy bootstrap help
 $(re2c_GEN_HELP): $(re2c_BOOT_HELP)
 	$(AM_V_at)$(MKDIR_P) $(@D)
@@ -411,25 +440,24 @@ all-local: docs $(re2c_BOOT_PARSER)
 re2c_TESTSUITE = run_tests.py
 .PHONY: tests
 tests: all $(re2c_TESTSUITE)
-	$(top_builddir)/$(re2c_TESTSUITE)
+	$(PYTHON) $(top_builddir)/$(re2c_TESTSUITE)
 .PHONY: vtests
 vtests: all $(re2c_TESTSUITE)
-	$(top_builddir)/$(re2c_TESTSUITE) --valgrind
+	$(PYTHON) $(top_builddir)/$(re2c_TESTSUITE) --valgrind
 .PHONY: wtests
 wtests: all $(re2c_TESTSUITE)
-	$(top_builddir)/$(re2c_TESTSUITE) --wine -j1
+	$(PYTHON) $(top_builddir)/$(re2c_TESTSUITE) --wine -j1
 
 re2c_test_range_SOURCES = \
 	src/test/range/test-impl.h \
 	src/test/range/test.cc \
 	src/test/range/test.h \
-	src/util/range.cc\
-	src/util/range.h \
-	src/util/static_assert.h
+	src/util/range.cc \
+	src/util/range.h
 
 re2c_test_s_to_n32_unsafe_SOURCES = \
 	src/test/s_to_n32_unsafe/test.cc \
-	src/util/s_to_n32_unsafe.cc
+	src/util/string_utils.cc
 
 re2c_test_ver_to_vernum_SOURCES = \
 	src/test/ver_to_vernum/test.cc \
@@ -464,7 +492,15 @@ endif
 # re2go
 if WITH_GOLANG
 bin_PROGRAMS += re2go
-re2go_CXXFLAGS = $(AM_CXXFLAGS) -DRE2C_LANG=LANG_GO
+re2go_CXXFLAGS = $(AM_CXXFLAGS) -DRE2C_LANG=Lang::GO
 re2go_SOURCES = $(re2c_SOURCES)
 nodist_re2go_SOURCES = $(nodist_re2c_SOURCES)
 endif
+
+# re2rust
+if WITH_RUST
+bin_PROGRAMS += re2rust
+re2rust_CXXFLAGS = $(AM_CXXFLAGS) -DRE2C_LANG=Lang::RUST
+re2rust_SOURCES = $(re2c_SOURCES)
+nodist_re2rust_SOURCES = $(nodist_re2c_SOURCES)
+endif

Makefile.lib.am

@@ -7,7 +7,7 @@ libre2c_la_LDFLAGS = -module -no-undefined
 
 libre2c_la_HDR = \
 	lib/lex.h \
-	lib/regcomp_dfa_regless.h \
+	lib/regcomp_dfa_multipass.h \
 	lib/regex.h \
 	lib/regex_impl.h \
 	lib/regoff_trie.h \
@@ -22,70 +22,53 @@ libre2c_la_HDR = \
 	src/dfa/determinization.h \
 	src/dfa/dfa.h \
 	src/dfa/posix_precedence.h \
-	src/dfa/stacmd.h \
 	src/dfa/tag_history.h \
 	src/dfa/tagver_table.h \
 	src/dfa/tcmd.h \
 	src/nfa/nfa.h \
-	src/encoding/case.h \
-	src/encoding/ebcdic/ebcdic.h \
-	src/encoding/ebcdic/ebcdic_regexp.h \
+	src/encoding/ebcdic.h \
 	src/encoding/enc.h \
 	src/encoding/range_suffix.h \
-	src/encoding/utf8/utf8.h \
-	src/encoding/utf8/utf8_regexp.h \
-	src/encoding/utf16/utf16_regexp.h \
-	src/encoding/utf16/utf16.h \
+	src/encoding/utf8.h \
+	src/encoding/utf16.h \
 	src/msg/location.h \
 	src/msg/msg.h \
 	src/msg/ver_to_vernum.h \
 	src/msg/warn.h \
-	src/regexp/re.h \
+	src/regexp/regexp.h \
 	src/regexp/rule.h \
 	src/regexp/tag.h \
-	src/compile.h \
 	src/skeleton/path.h \
 	src/skeleton/skeleton.h \
 	src/parse/ast.h \
-	src/parse/unescape.h \
 	src/parse/input.h \
 	src/parse/parser.h \
-	src/parse/scanner.h \
-	src/parse/unescape.h \
 	src/debug/debug.h \
-	src/util/allocate.h \
+	src/util/allocator.h \
 	src/util/attribute.h \
-	src/util/c99_stdint.h \
+	src/util/check.h \
+	src/util/containers.h \
 	src/util/file_utils.h \
 	src/util/forbid_copy.h \
-	src/util/free_list.h \
 	src/util/hash32.h \
-	src/util/local_increment.h \
-	src/util/lookup.h \
+	src/util/nowarn_in_bison.h \
 	src/util/range.h \
-	src/util/s_to_n32_unsafe.h \
-	src/util/slab_allocator.h \
-	src/util/smart_ptr.h \
-	src/util/static_assert.h \
 	src/util/string_utils.h \
-	src/util/u32lim.h \
-	src/util/uniq_vector.h
+	src/util/u32lim.h
 
 libre2c_la_SRC = \
 	lib/regcomp.cc \
 	lib/regexec.cc \
 	lib/regexec_dfa.cc \
-	lib/regexec_dfa_regless.cc \
+	lib/regexec_dfa_multipass.cc \
 	lib/regexec_nfa_leftmost.cc \
 	lib/regexec_nfa_leftmost_trie.cc \
 	lib/regexec_nfa_posix.cc \
 	lib/regexec_nfa_posix_trie.cc \
-	lib/regexec_nfa_posix_backward.cc \
-	lib/regexec_nfa_posix_kuklewicz.cc \
 	lib/regfree.cc \
 	lib/stubs.cc \
 	src/parse/ast.cc \
-	src/parse/unescape.cc \
+	src/parse/input.cc \
 	src/options/opt.cc \
 	src/options/symtab.cc \
 	src/cfg/cfg.cc \
@@ -111,33 +94,28 @@ libre2c_la_SRC = \
 	src/dfa/fillpoints.cc \
 	src/dfa/find_state.cc \
 	src/dfa/minimization.cc \
-	src/dfa/stacmd.cc \
 	src/dfa/tagver_table.cc \
 	src/dfa/tcmd.cc \
-	src/nfa/estimate_size.cc \
 	src/nfa/re_to_nfa.cc \
 	src/encoding/enc.cc \
 	src/encoding/range_suffix.cc \
-	src/encoding/ebcdic/ebcdic_regexp.cc \
-	src/encoding/utf16/utf16.cc \
-	src/encoding/utf16/utf16_regexp.cc \
-	src/encoding/utf8/utf8.cc \
-	src/encoding/utf8/utf8_regexp.cc \
+	src/encoding/ebcdic.cc \
+	src/encoding/utf16.cc \
+	src/encoding/utf8.cc \
 	src/msg/msg.cc \
 	src/msg/warn.cc \
 	src/regexp/ast_to_re.cc \
 	src/regexp/default_tags.cc \
 	src/regexp/fixed_tags.cc \
 	src/regexp/nullable.cc \
-	src/regexp/re.cc \
-	src/regexp/rule.cc \
+	src/regexp/regexp.cc \
 	src/regexp/split_charset.cc \
-	src/regexp/tag.cc \
 	src/skeleton/control_flow.cc \
 	src/skeleton/maxpath.cc \
 	src/skeleton/skeleton.cc \
 	src/util/range.cc \
-	src/util/s_to_n32_unsafe.cc
+	src/util/file_utils.cc \
+	src/util/string_utils.cc
 
 libre2c_la_SOURCES = \
 	$(libre2c_la_HDR) \

README.md

@@ -1,7 +1,7 @@
 DESCRIPTION
 -----------
 
-re2c is a free and open-source lexer generator for C, C++ and Go.
+re2c is a free and open-source lexer generator for C/C++, Go and Rust.
 
 Its main goal is generating fast lexers: at least as fast as their reasonably
 optimized hand-coded counterparts. Instead of using traditional table-driven
@@ -40,8 +40,8 @@ CONTRIBUTING
 ------------
 
 re2c is an Open Source, community-driven project. See
-[CONTRIBUTING.md](https://github.com/skvadrik/re2c/blob/master/CONTRIBUTING.md) for
-details about contributions to this repository.
+[CONTRIBUTING.md](https://github.com/skvadrik/re2c/blob/master/CONTRIBUTING.md)
+for details about contributions to this repository.
 
 
 BUILD
@@ -73,6 +73,7 @@ You are welcome to ask for help or share your thoughts and ideas.
 
 AUTHORS
 -------
-Re2c was originally written by Peter Bumbulis <peter@csg.uwaterloo.ca>.
-Since then many people contributed to the project. The current maintainer is
+
+re2c was originally written by Peter Bumbulis <peter@csg.uwaterloo.ca>. Since
+then many volunteers have contributed to the project. The current maintainer is
 Ulya Trofimovich <skvadrik@gmail.com>.

WORKSPACE

@@ -0,0 +1,14 @@
+workspace(name = "re2c")
+
+load("@bazel_tools//tools/build_defs/repo:http.bzl", "http_archive")
+
+http_archive(
+    name = "bazel_skylib",
+    urls = [
+        "https://mirror.bazel.build/github.com/bazelbuild/bazel-skylib/releases/download/1.3.0/bazel-skylib-1.3.0.tar.gz",
+        "https://github.com/bazelbuild/bazel-skylib/releases/download/1.3.0/bazel-skylib-1.3.0.tar.gz",
+    ],
+    sha256 = "74d544d96f4a5bb630d465ca8bbcfe231e3594e5aae57e1edbf17a6eb3ca2506",
+)
+load("@bazel_skylib//:workspace.bzl", "bazel_skylib_workspace")
+bazel_skylib_workspace()

autogen.sh

@@ -1,5 +1,3 @@
 #!/bin/sh
 
-git clean -fX
-mkdir -p m4
 autoreconf -i -W all

bazel/BUILD.bazel

@@ -0,0 +1,13 @@
+package(default_visibility = ["//visibility:private"])
+
+load("@bazel_skylib//:bzl_library.bzl", "bzl_library")
+
+exports_files([
+    "re2c_config.h",
+])
+
+bzl_library(
+    name = "re2c_bzl",
+    srcs = ["re2c.bzl"],
+    visibility = ["//visibility:public"],
+)

bazel/re2c.bzl

@@ -0,0 +1,70 @@
+# re2c generator rules - assuming that re2c is available as '@re2c//:re2c'.
+
+load("@bazel_skylib//lib:paths.bzl", "paths")
+
+visibility("public")
+
+# Define supported extensions as defined by:
+# . https://bazel.build/reference/be/c-cpp#cc_library_args
+# . https://github.com/bazelbuild/bazel/blob/master/src/main/java/com/google/devtools/build/lib/rules/cpp/CppFileTypes.java
+_C_SRC_EXT = [".c", ".cc", ".cpp", ".cxx", ".c++", ".C", ".cu", ".cl"]
+_C_HDR_EXT = [".h", ".hh", ".hpp", ".ipp", ".hxx", ".h++", ".inc", ".inl", ".tlh", ".tli", ".H", ".tcc"]
+_RE2C_EXT = [".re", ".re2c"]
+
+def _re2c_impl(ctx):
+    srcs = ctx.files.srcs
+    if len(srcs) != 1:
+        fail("Attribute `srcs` must contain exactly 1 re2c source file ({exts}).".format(exts=", ".join(_RE2C_EXT)))
+    outs = ctx.outputs.outs
+    if not outs or [o for o in outs if paths.split_extension(o.path)[1] not in _C_SRC_EXT + _C_HDR_EXT]:
+        fail("Attribute `outs` must contain at least 1 C/C++ shource or header file ({exts}).".format(exts=", ".join(_C_SRC_EXT + _C_HDR_EXT)))
+    ctx.actions.run_shell(
+        outputs = outs + [err_file],
+        tools = [ctx.executable._re2c_tool],
+        inputs = ctx.files.srcs + ctx.files.data,
+        command = """
+            "{re2c}" "{src}" -o "{out}" {re2c_args} {includes}
+        """.format(
+            re2c = ctx.executable._re2c_tool.path,
+            src = srcs[0].path,
+            out = outs[0].path,
+            includes = " ".join(list(dict([("-I" + f.dirname, 1) for f in ctx.files.data]))),
+            re2c_args = " ".join(ctx.attr.re2c_args),
+        ),
+        mnemonic = "RE2C",
+        progress_message = "Generating re2c %{label}",
+    )
+
+_re2c_attr = {
+    "srcs": attr.label_list(
+        doc = "The re2c source.",
+        allow_files = _RE2C_EXT,
+        mandatory = True,
+    ),
+    "data": attr.label_list(
+        doc = "Other input data files (e.g. include files).",
+        allow_files = _C_SRC_EXT + _C_HDR_EXT,
+        mandatory = False,
+    ),
+    "outs": attr.output_list(
+        doc = "The generated output source.",
+        mandatory = True,
+    ),
+    "re2c_args": attr.string_list(
+        doc = "Arguments to re2c invocation.",
+    ),
+    "_re2c_tool": attr.label(
+        doc = "The target of the 're2c' executable.",
+        default = Label("//:re2c"),
+        allow_single_file = True,
+        executable = True,
+        cfg = "exec",
+    ),
+}
+
+re2c_gen = rule(
+    attrs = _re2c_attr,
+    doc = """Use re2c tool to create an output file (e.g. '.cc') from a '.re' input.""",
+    output_to_genfiles = True,
+    implementation = _re2c_impl,
+)

bazel/re2c_config.h

@@ -0,0 +1,4 @@
+#define HAVE_SYS_TYPES_H 1
+#define HAVE_SYS_STAT_H 1
+#define HAVE_FCNTL_H 1
+#define HAVE_UNISTD_H 1

benchmarks/common/common.h

@@ -5,7 +5,7 @@
 #include <string>
 #include <vector>
 
-#include "src/util/c99_stdint.h"
+#include <stdint.h>
 
 
 enum engine_t {
@@ -23,7 +23,6 @@ struct bench_t {
     const char *name;
     const char *regexp;
     const char **strings;
-    int not_flags;
 };
 
 // Need to define benchmarks as functors for benchmark::RegisterBenchmark.

benchmarks/json2pgfplot.py

@@ -1,4 +1,4 @@
-#!/bin/env python3
+#!/usr/bin/env python3
 
 import argparse
 import json
@@ -12,7 +12,12 @@ plot_preamble = """
 
 \\usepackage[utf8]{inputenc}
 \\usepackage{amsmath, amssymb, amsfonts}
+\\usepackage{xcolor}
 \\usepackage{tikz, pgfplots, pgfplotstable}
+\\usepackage{pifont}
+
+\\definecolor{colorX}{HTML}{A1CAF1}
+\\definecolor{colorY}{HTML}{FBCEB1}
 """
 
 # Use Comic Neue font.
@@ -47,23 +52,23 @@ plot_begin = """
     axis y line*=none,
     axis x line=none,
     y=0.12in,
-    width=(\\plotwidth * 0.75),
+    width=\\plotwidth*1.15,
     bar width=\\barwidth,
     xshift=\\xplotshift,
     yshift=\\yplotshift,
     xmax=\\xmax,
-    xmin=-\\xmax/50,
+    xmin=0,
     ymax=0.7,
     ymin=0.3 - \\nrows,
     ytick=data,
-    yticklabels from table={\\table}{algo},
-    every axis title/.style={below right, at={(-0.5in,0)}},
-    title=\\title,
+    ytick style={draw=none},
+    yticklabels=\\empty,
+    title=\\empty,
     clip=false,
     visualization depends on={x > \\xmax \\as \\xoverflow},
     visualization depends on={x > \\xmax ? -x + \\xmax : 0 \\as \\xshift},
     nodes near coords={
-        \\ifdim \\xoverflow pt=1pt \\!\\Large{...}\\scriptsize\\! \\fi
+        \\ifdim \\xoverflow pt=1pt \\!\\ding{220}\\! \\scriptsize \\fi
         \\pgfmathprintnumber[assume math mode=true]{\\pgfplotspointmeta}
     },
     nodes near coords style={shift={(axis direction cs:\\xshift,0)}},
@@ -71,13 +76,21 @@ plot_begin = """
         fixed relative,
         precision = 10,
         1000 sep = {\\,}
-    },
-}}
+    },%s}}
+"""
+
+plot_labels_and_title = """
+    yticklabels from table={\\table}{algo},
+    every axis title/.style={below %s, at={(-0.08in,0)}},
+    title=\\title,
 """
 
 plot_begin_picture = """
 \\begin{tikzpicture}
-\\def\\xmax{%lf}
+
+\\begin{scope}[xshift=-0.2in, yshift=%sin * 0.75]
+%s
+\\end{scope}
 """
 
 plot_middle = """
@@ -85,12 +98,13 @@ plot_middle = """
 \\def\\yplotshift{%lfin}
 \\def\\nrows{%d},
 \\def\\title{%s},
+\\def\\xmax{%lf}
 
 \\pgfplotstableread {%s} \\table
 
-\\begin{axis}[styleX]
+\\begin{axis}[styleX,%s]
     \\clip(current axis.south west) rectangle(current axis.north east);
-    \\addplot[fill=lightgray] table[x=value, y expr=-\\coordindex] from \\table;
+    \\addplot[fill=%s] table[x=value, y expr=-\\coordindex] from \\table;
 \\end{axis}
 """
 
@@ -103,19 +117,35 @@ plot_end = """
 """
 
 
+VARIANTS = ['dfa_aot', 'dfa_jit', 'nfa']
+
+
 def parse_args():
+    variants = ', '.join(VARIANTS)
+
     parser = argparse.ArgumentParser()
     parser.add_argument("input", help="input JSON file")
     parser.add_argument("output", help="output TeX file")
+    parser.add_argument("--variant",
+        help="variant of the benchmark, one of %s" % variants)
     parser.add_argument("--relative-to",
         help="output CPU time relative to the specified algorithm")
     parser.add_argument("--font",
         help="use the specified font (supported fonts: cmbright, comicneue)")
-    return parser.parse_args()
+
+    args = parser.parse_args()
+    if args.variant not in VARIANTS:
+        sys.exit("please specify --variant (possible values: %s)" % variants)
+
+    return args
 
 
 # Generate benchmark title from benchmark name.
-def gen_title(oldname):
+def gen_title(oldname, remove_suffix):
+    if remove_suffix:
+        # cut off the last sub-word after the dash
+        oldname = oldname[:oldname.rfind("-")]
+
     # Name begins with a letter, so it must be a plain language word.
     if oldname[0].isalpha():
         return oldname
@@ -158,6 +188,68 @@ def time_format(time):
         return '%%.%dlf' % (-order + 1)
 
 
+class Params:
+    def __init__(self, args):
+        self.barwidth = 0.08
+
+        if args.variant == 'dfa_aot':
+            self.override_params_dfa_aot()
+        elif args.variant == 'dfa_jit':
+            self.override_params_dfa_jit()
+        elif args.variant == 'nfa':
+            self.override_params_nfa()
+
+        self.font = ''
+        if args.font == 'comicneue':
+            self.font = plot_font_comicneue
+        elif args.font == 'cmbright':
+            self.font = plot_font_cmbright
+        elif args.font:
+            sys.exit('error: unsupported font: "%s"' % font)
+
+        if args.relative_to: # override only if set
+            self.relative_to = args.relative_to
+
+    def override_params_dfa_aot(self):
+        self.plotwidth = 1.5
+        self.xshift = 2 * self.plotwidth
+        self.header = """
+            \\node[anchor=west] (x0) at (0.15in,0) {\\small\\bf{gcc (time)}};
+            \\node[anchor=west] (x1) at (1.65in,0) {\\small\\bf{gcc (size)}};
+            \\node[anchor=west] (x2) at (3.15in,0) {\\small\\bf{clang (time)}};
+            \\node[anchor=west] (x3) at (4.65in,0) {\\small\\bf{clang (size)}};
+        """
+        self.title_suffix = True
+        self.relative_to = None
+        self.title_annotate = False
+        self.global_styles = ""
+        self.local_styles = plot_labels_and_title % 'left'
+
+    def override_params_dfa_jit(self):
+        self.plotwidth = 2.7
+        self.xshift = self.plotwidth
+        self.header = """
+            \\node[anchor=west] (x0) at (0.15in,0) {\\small\\bf{regcomp (time)}};
+            \\node[anchor=west] (x1) at (2.85in,0) {\\small\\bf{regexec (time)}};
+        """
+        self.title_suffix = True
+        self.relative_to = ''
+        self.title_annotate = False
+        self.global_styles = ""
+        self.local_styles = plot_labels_and_title % 'left'
+
+    def override_params_nfa(self):
+        self.plotwidth = 2.5
+        self.xshift = self.plotwidth + 1
+        self.header = ""
+        self.only_1st_col_with_labels = False
+        self.title_suffix = False
+        self.relative_to = 'LG'
+        self.title_annotate = True
+        self.global_styles = plot_labels_and_title % 'right'
+        self.local_styles = ""
+
+
 # Split benchmarks into groups by regexp (the first component of the name).
 # If the relative mode is on, normalize CPU time by the time of the first
 # benchmark in each group.
@@ -175,7 +267,7 @@ def time_format(time):
 #   ], ...
 # }
 #
-def group_benchmarks(benchmarks, relative_to):
+def group_benchmarks(benchmarks, params):
     benchcount = len(benchmarks)
 
     groups = OrderedDict()
@@ -199,14 +291,14 @@ def group_benchmarks(benchmarks, relative_to):
     for name in groups:
         group = groups[name]
 
-        relto = index_of_algo(group, relative_to)
+        relto = index_of_algo(group, params.relative_to)
         (_, cputime0, binsize0, captures, regsize) = group[relto]
 
         timetbl = None
         if cputime0 != None:
             table = 'algo value\n'
             for (algo, time, _, _, _) in group:
-                time = time / (cputime0 if relative_to != None else 1)
+                time = time / (cputime0 if params.relative_to != None else 1)
                 maxtime = max(maxtime, time)
                 avgtime += time
                 table += ('{%s} ' + time_format(time) + '\n') % (algo, time)
@@ -221,74 +313,60 @@ def group_benchmarks(benchmarks, relative_to):
                 table += '{%s} %lf\n' % (algo, size)
             sizetbl = table
 
-        title = gen_title(name)
+        title = gen_title(name, params.title_suffix)
         nrows = len(group)
-        benchgroups[title] = (timetbl, sizetbl, nrows, regsize, captures)
+        benchgroups[name] = (title, timetbl, sizetbl, nrows, regsize, captures)
 
         nbench += nrows
         maxrows = max(maxrows, nrows)
 
     avgtime /= nbench
     avgsize /= nbench
-    maxtime = min(maxtime, 5 * avgtime)
-    maxsize = min(maxsize, 5 * avgsize)
+    factor = 10 if params.relative_to != None else 3
+    maxtime = min(maxtime, factor * avgtime)
+    maxsize = min(maxsize, factor * avgsize)
 
-    return benchgroups, maxtime, maxsize, maxrows
+    params.plotheight = 0.15 + maxrows * params.barwidth * 0.75
+    params.maxtime = maxtime
+    params.maxsize = maxsize
+
+    return benchgroups
 
 
 # Generate a Tikz picture with a PGF plot for each benchmark group. Crop
 # very long bars, so that they do not squash other bars to the left.
-def generate_plot(groups, font):
-    (benchgroups, maxtime, maxsize, maxrows) = groups
+def generate_plot(benchgroups, params):
+    plot = plot_preamble
+    plot += params.font
+    plot += plot_begin % (params.plotwidth, params.barwidth, params.global_styles)
+    plot += plot_begin_picture % (params.plotheight, params.header)
 
-    # Plot width is approximately half the usual page width.
-    plotwidth = 3.5
-    # Bar width is chosen impirically so that it looks good.
-    barwidth = 0.08
-    # Plot height is roughly bar width times maximum group size.
-    plotheight = 0.15 + maxrows * barwidth * 0.75
-
-    plot_time = ''
-    plot_size = ''
     i = 0
     for name, benchgroup in benchgroups.items():
         even = i % 2 == 0
         i += 1
-        (timetbl, sizetbl, nrows, regsize, captures) = benchgroup
+        (title, timetbl, sizetbl, nrows, regsize, captures) = benchgroup
 
-        xshift = 0 if even else plotwidth
-        yshift = -plotheight * (i if even else i - 1)
-        title = '\\textbf{%s}' % name
-        if regsize != None and captures != None:
+        xshift = 0 if even else params.xshift
+        yshift = -params.plotheight * (i if even else i - 1)
+        title = '\\textbf{%s}' % title
+        if params.title_annotate and regsize != None and captures != None:
             title += ' \\, (%d symbol%s, %d capture%s)' % (
                 regsize, 's' if regsize > 1 else '',
                 captures, 's' if captures > 1 else '')
 
         # Generate CPU time plots.
         if timetbl != None:
-            plot_time += plot_middle % (xshift, yshift, nrows, title, timetbl)
+            plot += plot_middle % (xshift, yshift, nrows, title, params.maxtime,
+                timetbl, params.local_styles if even else '', 'colorX')
 
         # Generate binary size plots.
         if sizetbl != None:
-            plot_size += plot_middle % (xshift, yshift, nrows, title, sizetbl)
-
-    # Glue plots together in one TeX document
-    plot = plot_preamble
-
-    if font == 'comicneue':
-        plot += plot_font_comicneue
-    elif font == 'cmbright':
-        plot += plot_font_cmbright
-    elif font != None:
-        sys.exit('error: unsupported font: "%s"' % font)
-
-    plot += plot_begin % (plotwidth, barwidth)
-
-    if plot_time != '':
-        plot += (plot_begin_picture % maxtime) + plot_time + plot_end_picture
-    if plot_size != '':
-        plot += (plot_begin_picture % maxsize) + plot_size + plot_end_picture
+            xshift += params.plotwidth
+            plot += plot_middle % (xshift, yshift, nrows, title, params.maxsize,
+                sizetbl, '', 'colorY')
 
+    plot += plot_end_picture
     plot += plot_end
 
     return plot
@@ -300,8 +378,9 @@ def main():
     with open(args.input) as f:
         input = json.load(f)
 
-    groups = group_benchmarks(input['benchmarks'], args.relative_to)
-    plot = generate_plot(groups, args.font)
+    params = Params(args)
+    groups = group_benchmarks(input['benchmarks'], params)
+    plot = generate_plot(groups, params)
 
     with open(args.output, 'w') as f:
         f.write(plot)

benchmarks/lookahead_tdfa/README

@@ -1,47 +0,0 @@
-
-To run benchmarks:
-
-1. Create symlink to the RE2C binary (version 1.0 or older) in the current
-directory.
-
-2. Generate input data: files ./http/gen/http.dat and ./uri/gen/uri.dat.
-    Random data can be generated with the following programs (written in
-    haskell):
-
-        http/gen/gen_http.hs
-        uri/gen/gen_uri.hs
-
-    Use the following scripts to build generators:
-
-        http/gen/mk_gen.sh
-        uri/gen/mk_gen.sh
-
-    The size of the generated data is hard-coded in the source code of
-    generators (and can be easily tweaked).
-
-    Alternatively, if you don't have haskell compiler or for any other reason
-    are unable to build generators, use small data samples provided for each
-    benchmark (one can copy-paste them many times with 'cat' to obtain a large
-    input file):
-
-        http/gen/sample.dat
-        uri/gen/sample.dat
-
-3.Run scripts:
-    ./http/simple/__bench.sh
-    ./http/simple/__bench_b.sh
-    ./http/simple/__bench_unopt.sh
-    ./http/rfc7230/__bench.sh
-    ./http/rfc7230/__bench_b.sh
-    ./http/rfc7230/__bench_unopt.sh
-    ./uri/simple/__bench.sh
-    ./uri/simple/__bench_b.sh
-    ./uri/simple/__bench_unopt.sh
-    ./uri/rfc3986/__bench.sh
-    ./uri/rfc3986/__bench_b.sh
-    ./uri/rfc3986/__bench_unopt.sh
-
-    Alternatively, run ./__run.sh (it will drop logs in each subdirectory).
-
-
-

benchmarks/lookahead_tdfa/__bench_utils.sh

@@ -1,116 +0,0 @@
-
-re2c="../../re2c"
-
-compile() {
-    name="${1}_${2}"
-    opts="$3"
-
-    [ -e ${re2c} ] || { echo "*** no re2c ***"; exit 1; }
-
-    ${re2c} ${opts}                -W ${name}.re        -o ${name}_lookahead.c    && ls -l --block-size=K ${name}_lookahead.c
-    ${re2c} ${opts} --no-lookahead -W ${name}.re        -o ${name}_no_lookahead.c && ls -l --block-size=K ${name}_no_lookahead.c
-    ${re2c} ${opts}                -W ${name}_notags.re -o ${name}_notags.c       && ls -l --block-size=K ${name}_notags.c
-}
-
-run() {
-    name1="$1"
-    name2="$2"
-    name="${name1}_${name2}"
-    input="../gen/${name1}.dat"
-
-    [ -e ${input} ] || { echo "*** no input file ***"; exit 1; }
-
-    strip ${name}
-    cmd="./${name} ${input}"
-    for i in `seq 1`; do $cmd >/dev/null 2>&1; done # warmup
-
-    npass=4
-    total=0
-    for i in `seq $npass`; do
-        t="`TIMEFORMAT='%E'; time ( $cmd 2>/dev/null ) 2>&1 1>/dev/null`"
-        echo "t:" $t
-        total=`echo $total + $t | bc -l`
-    done
-    average=`echo "$total / $npass" | bc -l`
-    echo "time: $average"
-
-    ls -l --block-size=K ${name}
-}
-
-#cflags="-Wall -O2"
-cflags="-O2"
-
-run_all() {
-    name1="$1"
-    name2="$2"
-    name="${name1}_${name2}"
-
-    echo "---------------- WARMUP (GCC -O2 lookahead) ----------------"
-    gcc     $cflags ${name}_lookahead.c    -o${name} && run ${name1} ${name2}
-    echo "---------------- END WARMUP ----------------"
-
-    echo "---------------- GCC -O2 lookahead ----------------"
-    gcc     $cflags ${name}_lookahead.c    -o${name} && run ${name1} ${name2}
-    echo "---------------- GCC -O2 no-lookahead ----------------"
-    gcc     $cflags ${name}_no_lookahead.c -o${name} && run ${name1} ${name2}
-    echo "---------------- GCC -O2 notags ----------------"
-    gcc     $cflags ${name}_notags.c       -o${name} && run ${name1} ${name2}
-
-    echo "---------------- CLANG -O2 lookahead ----------------"
-    clang   $cflags ${name}_lookahead.c    -o${name} && run ${name1} ${name2}
-    echo "---------------- CLANG -O2 no-lookahead ----------------"
-    clang   $cflags ${name}_no_lookahead.c -o${name} && run ${name1} ${name2}
-    echo "---------------- CLANG -O2 notags ----------------"
-    clang   $cflags ${name}_notags.c       -o${name} && run ${name1} ${name2}
-
-    echo "---------------- TCC -O2 lookahead ----------------"
-    tcc     $cflags ${name}_lookahead.c    -o${name} && run ${name1} ${name2}
-    echo "---------------- TCC -O2 no-lookahead ----------------"
-    tcc     $cflags ${name}_no_lookahead.c -o${name} && run ${name1} ${name2}
-    echo "---------------- TCC -O2 notags ----------------"
-    tcc     $cflags ${name}_notags.c       -o${name} && run ${name1} ${name2}
-
-    echo "---------------- PCC -O2 lookahead ----------------"
-    pcc     $cflags ${name}_lookahead.c    -o${name} && run ${name1} ${name2}
-    echo "---------------- PCC -O2 no-lookahead ----------------"
-    pcc     $cflags ${name}_no_lookahead.c -o${name} && run ${name1} ${name2}
-    echo "---------------- PCC -O2 notags ----------------"
-    pcc     $cflags ${name}_notags.c       -o${name} && run ${name1} ${name2}
-
-#    echo "---------------- NWCC -O2 lookahead ----------------"
-#    nwcc    $cflags ${name}_lookahead.c    -o${name} && run ${name1} ${name2}
-#    echo "---------------- NWCC -O2 no-lookahead ----------------"
-#    nwcc    $cflags ${name}_no_lookahead.c -o${name} && run ${name1} ${name2}
-#    echo "---------------- NWCC -O2 notags ----------------"
-#    nwcc    $cflags ${name}_notags.c       -o${name} && run ${name1} ${name2}
-
-    rm ${name}_lookahead.c
-    rm ${name}_no_lookahead.c
-    rm ${name}_notags.c
-    rm ${name}
-}
-
-verify() {
-    name1="$1"
-    name2="$2"
-    name=${name1}_${name2}
-    input="../gen/${name1}.dat.short"
-    output="${name1}.dat.short"
-
-    [ -e ${input} ] || { echo "*** no input file ***"; exit 1; }
-    [ -e ${re2c} ] || { echo "*** no re2c ***"; exit 1; }
-
-    ${re2c}                -W ${name}.re        -o ${name}_lookahead.c
-    ${re2c} --no-lookahead -W ${name}.re        -o ${name}_no_lookahead.c
-    ${re2c}                -W ${name}_notags.re -o ${name}_notags.c
-
-    g++ -Wall -O2 -DVERIFY ${name}_lookahead.c    -o${name} && ./${name} ${input} > ${output} && diff -q ${input} ${output} || echo "lookahead failed"
-    g++ -Wall -O2 -DVERIFY ${name}_no_lookahead.c -o${name} && ./${name} ${input} > ${output} && diff -q ${input} ${output} || echo "non-lookahead failed"
-    g++ -Wall -O2 -DVERIFY ${name}_notags.c       -o${name} && ./${name} ${input} > ${output} && diff -q ${input} ${output} || echo "notags failed"
-
-    rm ${name}_lookahead.c
-    rm ${name}_no_lookahead.c
-    rm ${name}_notags.c
-    rm ${name}
-    rm ${output}
-}

benchmarks/lookahead_tdfa/__run.sh

@@ -1,21 +0,0 @@
-#!/usr/bin/env bash
-
-set -e
-
-bench() {
-    cd "$1/$2/"
-
-#    ./__verify.sh
-
-    ./__bench.sh       1> log._ 2>&1
-    ./__bench_unopt.sh 1> log.unopt 2>&1
-    ./__bench_b.sh     1> log.b 2>&1
-
-    cd ../..
-}
-
-bench http rfc7230
-bench http simple
-bench uri  rfc3986
-bench uri  simple
-

benchmarks/lookahead_tdfa/http/gen/gen_http.hs

@@ -1,365 +0,0 @@
-import qualified Test.QuickCheck as Q
-import           Data.List (intercalate)
-import           Control.Monad (forM_, when, replicateM)
-
-    {-!re2c
-
-        crlf  = "\n";
-        sp    = " ";
-        htab  = "\t";
-        ows   = (sp | htab)*;
-        rws   = (sp | htab)+;
-        bws   = ows;
-        digit = [0-9];
-        alpha = [a-zA-Z];
-        hexdigit    = [0-9a-fA-F];
-        unreserved  = alpha | digit | [-._~];
-        pct_encoded = "%" hexdigit{2};
-        sub_delims  = [!$&'()*+,;=];
-        pchar       = unreserved | pct_encoded | sub_delims | [:@];
-        vchar = [\x1f-\x7e];
-        tchar = [-!#$%&'*+.^_`|~] | digit | alpha;
-        token = tchar+;
-        octet = [^\x00];
-
-
-        obs_fold       = crlf (sp | htab)+;
-        obs_text       = [\x80-\xff];
-        field_name     = token;
-        field_vchar    = vchar | obs_text;
-        field_content  = field_vchar ((sp | htab)+ field_vchar)?;
-        field_value    = (field_content | obs_fold)*;
-        header_field   = #h1 field_name #h2 ":" ows field_value ows #h3;
-        method         = token;
-        segment        = pchar*;
-        scheme         = alpha (alpha | digit | [-+.])*;
-        userinfo       = (unreserved | pct_encoded | sub_delims | ":")*;
-        dec_octet
-            = digit
-            | [\x31-\x39] digit
-            | "1" digit{2}
-            | "2" [\x30-\x34] digit
-            | "25" [\x30-\x35];
-        ipv4address    = dec_octet "." dec_octet "." dec_octet "." dec_octet;
-        h16            = hexdigit{1,4};
-        ls32           = h16 ":" h16 | ipv4address;
-        ipv6address
-            =                            (h16 ":"){6} ls32
-            |                       "::" (h16 ":"){5} ls32
-            | (               h16)? "::" (h16 ":"){4} ls32
-            | ((h16 ":"){0,1} h16)? "::" (h16 ":"){3} ls32
-            | ((h16 ":"){0,2} h16)? "::" (h16 ":"){2} ls32
-            | ((h16 ":"){0,3} h16)? "::"  h16 ":"     ls32
-            | ((h16 ":"){0,4} h16)? "::"              ls32
-            | ((h16 ":"){0,5} h16)? "::"              h16
-            | ((h16 ":"){0,6} h16)? "::";
-        ipvfuture      = "v" hexdigit+ "." (unreserved | sub_delims | ":" )+;
-        ip_literal     = "[" ( ipv6address | ipvfuture ) "]";
-        reg_name       = (unreserved | pct_encoded | sub_delims)*;
-        host           = ip_literal | ipv4address | reg_name;
-        port           = digit*;
-        authority      = (userinfo "@")? host (":" port)?;
-        query          = (pchar | [/?])*;
-        path_abempty   = ("/" pchar*)*;
-        path_absolute  = "/" (pchar+ ("/" pchar*)*)?;
-        path_rootless  = pchar+ ("/" pchar*)*;
-        path_empty     = "";
-        hier_part
-            = "//" authority path_abempty
-            | (path_absolute | path_rootless | path_empty);
-        absolute_uri   = scheme ":" hier_part ("?" query)?;
-        absolute_path  = ("/" segment)*;
-        origin_form    = absolute_path ("?" query)?;
-        http_name      = "HTTP";
-        http_version   = http_name "/" digit "." digit;
-        request_target = origin_form | absolute_uri | authority | "*";
-        request_line   = @m1 method @m2 sp @rt1 request_target @rt2 sp @v3 http_version @v4 crlf;
-        status_code    = digit{3};
-        reason_phrase  = (htab | sp | vchar | obs_text)*;
-        status_line    = @v1 http_version @v2 sp @s1 status_code @s2 sp @rp1 reason_phrase @rp2 crlf;
-        start_line     = (request_line | status_line);
-        message_body   = octet*;
-        message_head   = start_line (header_field crlf)* crlf;
-    -}
-
-type Scheme = String
-type User = String
-type Password = String
-type Host = String
-type Port = String
-type Path = String
-type Query = String
-data RequestTarget
-    = OriginForm Path Query
-    | AbsoluteURI Scheme User Password Host Port Path Query
-    | Authority User Password Host Port
-    | Asterisk
-type ReasonPhrase = String
-type HTTPVersion = String
-type Method = String
-type StatusCode = Int
-data StartLine
-    = RequestLine Method RequestTarget HTTPVersion
-    | StatusLine HTTPVersion StatusCode ReasonPhrase
-data HeaderField = HeaderField String
-data HTTPMsg = HTTPMsg StartLine [HeaderField]
-
-
-instance Show RequestTarget where
-    show (OriginForm p q) = p ++ (if q == "" then "" else "?" ++ q)
-    show (AbsoluteURI sc us pw ho pr pa q) = concat
-        [ sc
-        , ":"
-        , if ho == "" then "" else "//"
-        , us
-        , if us == "" then "" else ":"
-        , pw
-        , if pw == "" then "" else "@"
-        , ho
-        , if pr == "" then "" else ":"
-        , pr
-        , pa
-        , if q == "" then "" else "?"
-        , q
-        ]
-    show (Authority us pw ho pr) = concat
-        [ us
-        , if us == "" then "" else ":"
-        , pw
-        , if pw == "" then "" else "@"
-        , ho
-        , if pr == "" then "" else ":"
-        , pr
-        ]
-    show Asterisk = "*"
-instance Show StartLine where
-    show (RequestLine m rt v) = m ++ " " ++ show rt ++ " " ++ v ++ "\n"
-    show (StatusLine v sc rp) = v ++ " " ++ show sc ++ " " ++ rp ++ "\n"
-instance Show HeaderField where show (HeaderField hf) = hf
-instance Show HTTPMsg where
-    show (HTTPMsg sl hfs) = show sl ++ unlines (map show hfs) ++ "\n"
-
-
-instance Q.Arbitrary HTTPMsg where
-    arbitrary = do
-        scheme   <- Q.elements ["http", "https", "ftp", "file"]
-        user     <- lstr
-        password <- anstr
-        host     <- Q.frequency [(10, ipv4), (10, hostname), (1, ipv6)]
-        port     <- mayb $ show <$> (Q.choose (0, 10000) :: Q.Gen Int)
-        path     <- filepath
-        query    <- mayb anstr
-        request_target <- Q.frequency
-            [ (1, pure $ OriginForm path query)
-            , (1, pure $ OriginForm path "")
-
-            , (1, pure $ Authority user password host port)
-            , (1, pure $ Authority user password host "")
-            , (1, pure $ Authority ""   ""       host port)
-            , (1, pure $ Authority ""   ""       host "")
-
-            , (1, pure $ AbsoluteURI scheme user  password host port path query)
-            , (2, pure $ AbsoluteURI scheme ""    ""       host port path query)
-            , (1, pure $ AbsoluteURI scheme ""    ""       ""   ""   path query)
-
-            , (1, pure Asterisk)
-            ]
-        reason_phrase <- Q.listOf (Q.elements (alphanum ++ space)) :: Q.Gen [Char]
-        method <- Q.elements
-            [ "OPTIONS"
-            , "GET"
-            , "HEAD"
-            , "POST"
-            , "PUT"
-            , "DELETE"
-            , "TRACE"
-            , "CONNECT"
-            , "PATCH"
-            ]
-        status_code <- Q.choose (100, 999) :: Q.Gen Int
-        http_version <- do
-            n1 <- Q.choose (0, 9) :: Q.Gen Int
-            n2 <- Q.choose (0, 9) :: Q.Gen Int
-            return $ "HTTP/" ++ show n1 ++ "." ++ show n2
-        start_line <- Q.frequency
-            [ (1, pure $ RequestLine method request_target http_version)
-            , (1, pure $ StatusLine http_version status_code reason_phrase)
-            ]
-        -- copy-pasted from https://en.wikipedia.org/wiki/List_of_HTTP_header_fields
-        header_fields <- Q.listOf1 $ HeaderField <$> Q.elements
-            -- request examples
-            [ "Accept: text/plain"
-            , "Accept-Charset: utf-8"
-            , "Accept-Encoding: gzip, deflate"
-            , "Accept-Language: en-US"
-            , "Accept-Datetime: Thu, 31 May 2007 20:35:00 GMT"
-            , "Authorization: Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
-            , "Cache-Control: no-cache"
-            , "Connection: keep-alive"
-            , "Connection: Upgrade"
-            , "Cookie: $Version=1; Skin=new;"
-            , "Content-Length: 348"
-            , "Content-MD5: Q2hlY2sgSW50ZWdyaXR5IQ=="
-            , "Content-Type: application/x-www-form-urlencoded"
-            , "Date: Tue, 15 Nov 1994 08:12:31 GMT"
-            , "Expect: 100-continue"
-            , "Forwarded: for=192.0.2.60;proto=http;by=203.0.113.43"
-            , "Forwarded: for=192.0.2.43, for=198.51.100.17"
-            , "From: user@example.com"
-            , "Host: en.wikipedia.org:8080"
-            , "Host: en.wikipedia.org"
-            , "If-Match: \"737060cd8c284d8af7ad3082f209582d\""
-            , "If-Modified-Since: Sat, 29 Oct 1994 19:43:31 GMT"
-            , "If-None-Match: \"737060cd8c284d8af7ad3082f209582d\""
-            , "If-Range: \"737060cd8c284d8af7ad3082f209582d\""
-            , "If-Unmodified-Since: Sat, 29 Oct 1994 19:43:31 GMT"
-            , "Max-Forwards: 10"
-            , "Origin: http://www.example-social-network.com"
-            , "Pragma: no-cache"
-            , "Proxy-Authorization: Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ=="
-            , "Range: bytes=500-999"
-            , "Referer: http://en.wikipedia.org/wiki/Main_Page"
-            , "TE: trailers, deflate"
-            , "User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:12.0) Gecko/20100101 Firefox/12.0"
-            , "Upgrade: HTTPS/1.3, IRC/6.9, RTA/x11, websocket"
-            , "Via: 1.0 fred, 1.1 example.com (Apache/1.1)"
-            , "Warning: 199 Miscellaneous warning"
-            , "X-Requested-With: XMLHttpRequest"
-            , "DNT: 1 (Do Not Track Enabled)"
-            , "DNT: 0 (Do Not Track Disabled)"
-            , "X-Forwarded-For: client1, proxy1, proxy2"
-            , "X-Forwarded-For: 129.78.138.66, 129.78.64.103"
-            , "X-Forwarded-Host: en.wikipedia.org:8080"
-            , "X-Forwarded-Host: en.wikipedia.org"
-            , "X-Forwarded-Proto: https"
-            , "Front-End-Https: on"
-            , "X-HTTP-Method-Override: DELETE"
-            , "X-Att-Deviceid: GT-P7320/P7320XXLPG"
-            , "x-wap-profile: http://wap.samsungmobile.com/uaprof/SGH-I777.xml"
-            , "Proxy-Connection: keep-alive"
-            , "X-UIDH: ..."
-            , "X-Csrf-Token: i8XNjC4b8KVok4uw5RftR38Wgp2BFwql"
-            , "X-Request-ID: f058ebd6-02f7-4d3f-942e-904344e8cde5"
-            -- response examples
-            , "Access-Control-Allow-Origin: *"
-            , "Accept-Patch: text/example;charset=utf-8"
-            , "Accept-Ranges: bytes"
-            , "Age: 12"
-            , "Allow: GET, HEAD"
-            , "Alt-Svc: http/1.1=\"http2.example.com:8001\"; ma=7200"
-            , "Cache-Control: max-age=3600"
-            , "Connection: close"
-            , "Content-Disposition: attachment; filename=\"fname.ext\""
-            , "Content-Encoding: gzip"
-            , "Content-Language: da"
-            , "Content-Length: 51"
-            , "Content-Location: /index.htm"
-            , "Content-MD5: Q2hlY2sgSW50ZWdyaXR5IQ=="
-            , "Content-Range: bytes 21010-47021/47022"
-            , "Content-Type: text/html; charset=utf-8"
-            , "Date: Tue, 15 Nov 1994 08:12:31 GMT"
-            , "ETag: \"737060cd8c284d8af7ad3082f209582d\""
-            , "Expires: Thu, 01 Dec 1994 16:00:00 GMT"
-            , "Last-Modified: Tue, 15 Nov 1994 12:45:26 GMT"
-            , "Link: </feed>; rel=\"alternate\""
-            , "Location: http://www.w3.org/pub/WWW/People.html"
-            , "P3P: CP=\"This is not a P3P policy! See http://www.google.com/support/accounts/bin/answer.py?hl=en&answer=151657 for more info.\""
-            , "Pragma: no-cache"
-            , "Proxy-Authenticate: Basic"
-            , "Public-Key-Pins: max-age=2592000; pin-sha256=\"E9CZ9INDbd+2eRQozYqqbQ2yXLVKB9+xcprMF+44U1g=\";"
-            , "Refresh: 5; url=http://www.w3.org/pub/WWW/People.html"
-            , "Retry-After: 120"
-            , "Retry-After: Fri, 07 Nov 2014 23:59:59 GMT"
-            , "Server: Apache/2.4.1 (Unix)"
-            , "Set-Cookie: UserID=JohnDoe; Max-Age=3600; Version=1"
-            , "Strict-Transport-Security: max-age=16070400; includeSubDomains"
-            , "Trailer: Max-Forwards"
-            , "Transfer-Encoding: chunked"
-            , "Tk: ?"
-            , "Upgrade: HTTPS/1.3, IRC/6.9, RTA/x11, websocket"
-            , "Vary: *"
-            , "Vary: Accept-Language"
-            , "Via: 1.0 fred, 1.1 example.com (Apache/1.1)"
-            , "Warning: 199 Miscellaneous warning"
-            , "WWW-Authenticate: Basic"
-            , "X-Frame-Options: deny"
-            , "Status: 200 OK"
-            , "X-XSS-Protection: 1; mode=block"
-            , "X-WebKit-CSP: default-src 'self'"
-            , "X-Content-Type-Options: nosniff"
-            , "X-Powered-By: PHP/5.4.0"
-            , "X-UA-Compatible: IE=EmulateIE7"
-            , "X-UA-Compatible: IE=edge"
-            , "X-UA-Compatible: Chrome=1"
-            , "X-Content-Duration: 42.666"
-            , "Upgrade-Insecure-Requests: 1"
-            , "X-Request-ID: f058ebd6-02f7-4d3f-942e-904344e8cde5"
-            ]
-        return $ HTTPMsg start_line header_fields
-
-
-lalpha :: [Char]
-lalpha = ['a'..'z']
-
-ualpha :: [Char]
-ualpha = ['A'..'Z']
-
-alpha :: [Char]
-alpha = lalpha ++ ualpha
-
-digit :: [Char]
-digit = ['0'..'9']
-
-hexdigit :: [Char]
-hexdigit = ['a'..'f'] ++ ['A'..'F'] ++ digit
-
-alphanum :: [Char]
-alphanum = alpha ++ digit
-
-space :: [Char]
-space = [' ', '\t']
-
-lstr :: Q.Gen String
-lstr = Q.listOf1 (Q.elements lalpha) :: Q.Gen [Char]
-
-anstr :: Q.Gen String
-anstr = Q.listOf1 (Q.elements alphanum) :: Q.Gen [Char]
-
-ipv4 :: Q.Gen String
-ipv4 = do
-    let octet = show <$> (Q.choose (0,255) :: Q.Gen Int)
-    intercalate "." <$> replicateM 4 octet
-
-ipv6 :: Q.Gen String
-ipv6 = do
-    let hexgroup = do
-            k <- Q.choose (1, 4) :: Q.Gen Int
-            replicateM k (Q.elements hexdigit)
-    n <- Q.choose (0, 8) :: Q.Gen Int
-    m <- Q.choose (1, n) :: Q.Gen Int
-    hh <- replicateM n hexgroup
-    return $ (++ "]") . ("[" ++) $ case n of
-        8 -> intercalate ":" hh
-        _ -> concat [intercalate ":" (take m hh), "::", intercalate ":" (drop m hh)]
-
-hostname :: Q.Gen String
-hostname = do
-    subdomains <- Q.listOf1 (take 10 <$> lstr)
-    return $ intercalate "." $ take 5 subdomains
-
-filepath :: Q.Gen String
-filepath = do
-    subdirs <- Q.listOf1 lstr
-    return $ concat $ map ("/" ++) $ take 10 subdirs
-
-mayb :: Q.Gen String -> Q.Gen String
-mayb gen = Q.oneof [gen, pure ""]
-
-main :: IO ()
-main = do
-    let f = "http.dat"
-    writeFile f ""
-    forM_ [0 .. 170000 :: Int] $ \i -> do
-        when (i `mod` 1000 == 0) $ print i
-        msg <- Q.generate (Q.arbitrary :: Q.Gen HTTPMsg)
-        appendFile f $ show msg

benchmarks/lookahead_tdfa/http/gen/mk_gen.sh

@@ -1 +0,0 @@
-ghc -O2 -Wall gen_http.hs

benchmarks/lookahead_tdfa/http/gen/sample.dat

@@ -1,101 +0,0 @@
-HEAD file://cwidoaajjb.pqzeztqtg.iagnuhmdq.npliaihmqn.yuptnwn/hlhkojvjvzmsgccwcpwhh/qfliekillmmdz/aypkwffmdojvpdodmrouzxuupqp/gjqnmbk/mnbtpocjbe/ttvqsflbsqirgahmutddqucg/nkycantossaikwulmgemuqttwx/foqeceoyiowyeftxpr/syjcvolygaguylfkx/gwgx?7B9Jq5CNzjzSFzJl7cCL4oUX HTTP/6.2
-Content-Disposition: attachment; filename="fname.ext"
-Content-Language: da
-If-Match: "737060cd8c284d8af7ad3082f209582d"
-If-None-Match: "737060cd8c284d8af7ad3082f209582d"
-P3P: CP="This is not a P3P policy! See http://www.google.com/support/accounts/bin/answer.py?hl=en&answer=151657 for more info."
-Connection: Upgrade
-X-Powered-By: PHP/5.4.0
-Proxy-Authenticate: Basic
-Warning: 199 Miscellaneous warning
-Expires: Thu, 01 Dec 1994 16:00:00 GMT
-Retry-After: Fri, 07 Nov 2014 23:59:59 GMT
-Upgrade: HTTPS/1.3, IRC/6.9, RTA/x11, websocket
-X-Att-Deviceid: GT-P7320/P7320XXLPG
-Alt-Svc: http/1.1="http2.example.com:8001"; ma=7200
-Accept-Ranges: bytes
-Content-Language: da
-Cache-Control: no-cache
-Age: 12
-Accept-Encoding: gzip, deflate
-Content-Length: 51
-WWW-Authenticate: Basic
-Vary: *
-Upgrade-Insecure-Requests: 1
-Strict-Transport-Security: max-age=16070400; includeSubDomains
-Public-Key-Pins: max-age=2592000; pin-sha256="E9CZ9INDbd+2eRQozYqqbQ2yXLVKB9+xcprMF+44U1g=";
-Pragma: no-cache
-Content-Length: 51
-Host: en.wikipedia.org
-X-Frame-Options: deny
-
-POST https://hubgkuzz.qqwqvsfccy.psvyjbuva.afjqkipasn.zrdeeznzgo:7150/yfztaoqwaqlyuxfdkosgkegikcn/xme HTTP/6.6
-Cache-Control: max-age=3600
-Content-Length: 51
-Content-Type: application/x-www-form-urlencoded
-Upgrade: HTTPS/1.3, IRC/6.9, RTA/x11, websocket
-Alt-Svc: http/1.1="http2.example.com:8001"; ma=7200
-Date: Tue, 15 Nov 1994 08:12:31 GMT
-Accept-Charset: utf-8
-Cookie: $Version=1; Skin=new;
-X-Request-ID: f058ebd6-02f7-4d3f-942e-904344e8cde5
-
-CONNECT zwusyxlehg.qxqjpqqjzq.rnrnmjbspj.ugbkvsriyw.gpchzrlqkk:8638 HTTP/7.4
-Pragma: no-cache
-Transfer-Encoding: chunked
-Content-Location: /index.htm
-X-Att-Deviceid: GT-P7320/P7320XXLPG
-Public-Key-Pins: max-age=2592000; pin-sha256="E9CZ9INDbd+2eRQozYqqbQ2yXLVKB9+xcprMF+44U1g=";
-Accept-Patch: text/example;charset=utf-8
-Host: en.wikipedia.org:8080
-X-Forwarded-For: client1, proxy1, proxy2
-Accept-Datetime: Thu, 31 May 2007 20:35:00 GMT
-Proxy-Authorization: Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ==
-DNT: 1 (Do Not Track Enabled)
-X-UA-Compatible: Chrome=1
-Tk: ?
-Refresh: 5; url=http://www.w3.org/pub/WWW/People.html
-X-UA-Compatible: IE=edge
-Expect: 100-continue
-DNT: 0 (Do Not Track Disabled)
-WWW-Authenticate: Basic
-Content-Type: application/x-www-form-urlencoded
-Public-Key-Pins: max-age=2592000; pin-sha256="E9CZ9INDbd+2eRQozYqqbQ2yXLVKB9+xcprMF+44U1g=";
-Authorization: Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ==
-X-Content-Type-Options: nosniff
-X-Csrf-Token: i8XNjC4b8KVok4uw5RftR38Wgp2BFwql
-X-Forwarded-Host: en.wikipedia.org
-X-Att-Deviceid: GT-P7320/P7320XXLPG
-Content-Location: /index.htm
-Accept-Datetime: Thu, 31 May 2007 20:35:00 GMT
-
-HTTP/9.5 559 u
-Date: Tue, 15 Nov 1994 08:12:31 GMT
-Content-Length: 348
-X-Forwarded-Proto: https
-Proxy-Authorization: Basic QWxhZGRpbjpvcGVuIHNlc2FtZQ==
-Expires: Thu, 01 Dec 1994 16:00:00 GMT
-WWW-Authenticate: Basic
-Allow: GET, HEAD
-Warning: 199 Miscellaneous warning
-
-POST wllvzjwvpp.mhagladifk.joaq.rnprjmcpum.opadaxtssr HTTP/6.4
-Warning: 199 Miscellaneous warning
-Proxy-Authenticate: Basic
-Content-Encoding: gzip
-X-Forwarded-Host: en.wikipedia.org
-Strict-Transport-Security: max-age=16070400; includeSubDomains
-Content-Language: da
-Allow: GET, HEAD
-Retry-After: Fri, 07 Nov 2014 23:59:59 GMT
-If-Unmodified-Since: Sat, 29 Oct 1994 19:43:31 GMT
-X-Forwarded-For: client1, proxy1, proxy2
-
-HTTP/0.6 463 2bs
-From: user@example.com
-X-Forwarded-Host: en.wikipedia.org
-X-Forwarded-Host: en.wikipedia.org
-X-UA-Compatible: IE=edge
-X-UA-Compatible: IE=edge
-Date: Tue, 15 Nov 1994 08:12:31 GMT
-

benchmarks/lookahead_tdfa/http/rfc7230/__bench.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile http rfc7230 ""
-run_all http rfc7230
-

benchmarks/lookahead_tdfa/http/rfc7230/__bench_b.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile http rfc7230 "-b"
-run_all http rfc7230
-

benchmarks/lookahead_tdfa/http/rfc7230/__bench_unopt.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile http rfc7230 "--no-optimize-tags"
-run_all http rfc7230
-

benchmarks/lookahead_tdfa/http/rfc7230/__verify.sh

@@ -1,4 +0,0 @@
-
-. ../../__bench_utils.sh
-
-verify http rfc7230

benchmarks/lookahead_tdfa/http/rfc7230/http_rfc7230.re

@@ -1,318 +0,0 @@
-#include <assert.h>
-#include <stddef.h>
-#include <stdlib.h>
-#include <stdio.h>
-#include <string.h>
-
-typedef struct taglist_t {
-    struct taglist_t *pred;
-    long dist;
-} taglist_t;
-
-typedef struct taglistpool_t {
-    taglist_t *head;
-    taglist_t *next;
-    taglist_t *last;
-} taglistpool_t;
-
-typedef struct {
-    FILE *file;
-    char *buf;
-    char *lim;
-    char *cur;
-    char *mar;
-    char *tok;
-    /*!stags:re2c format = "char *@@;\n"; */
-    /*!mtags:re2c format = "taglist_t *@@;\n"; */
-    taglistpool_t tlp;
-    int eof;
-} input_t;
-
-static void taglistpool_clear(taglistpool_t *tlp, input_t *in)
-{
-    tlp->next = tlp->head;
-    /*!mtags:re2c format = "in->@@ = 0;\n"; */
-}
-
-static void taglistpool_init(taglistpool_t *tlp)
-{
-    static const unsigned size = 1024 * 1024;
-    tlp->head = (taglist_t*)malloc(size * sizeof(taglist_t));
-    tlp->next = tlp->head;
-    tlp->last = tlp->head + size;
-}
-
-static void taglistpool_free(taglistpool_t *tlp)
-{
-    free(tlp->head);
-    tlp->head = tlp->next = tlp->last = NULL;
-}
-
-static taglist_t *taglistpool_next(taglistpool_t *tlp)
-{
-    if (tlp->next < tlp->last) {
-        return tlp->next++;
-    }
-
-    const unsigned size = tlp->last - tlp->head;
-    taglist_t *head = (taglist_t*)malloc(2 * size * sizeof(taglist_t));
-    memcpy(head, tlp->head, size * sizeof(taglist_t));
-    free(tlp->head);
-    tlp->head = head;
-    tlp->next = head + size;
-    tlp->last = head + size * 2;
-    return tlp->next++;
-}
-
-static void taglist(taglist_t **ptl, const char *b, const char *t, taglistpool_t *tlp)
-{
-    taglist_t *tl = taglistpool_next(tlp);
-    tl->pred = *ptl;
-    tl->dist = t - b;
-    *ptl = tl;
-}
-
-/*!max:re2c*/
-static const size_t SIZE = 4096;
-
-static void init_input(input_t *in, const char *fname)
-{
-    in->file = fopen(fname, "r");
-    in->buf = (char*) malloc(SIZE + YYMAXFILL);
-    in->lim = in->buf + SIZE;
-    in->cur = in->lim;
-    in->mar = in->lim;
-    in->tok = in->lim;
-    /*!stags:re2c format = "in->@@ = 0;\n"; */
-    /*!mtags:re2c format = "in->@@ = 0;\n"; */
-    taglistpool_init(&in->tlp);
-    in->eof = 0;
-}
-
-static void free_input(input_t *in)
-{
-    fclose(in->file);
-    free(in->buf);
-    taglistpool_free(&in->tlp);
-}
-
-static int fill(input_t *in, size_t need)
-{
-    size_t free;
-
-    if (in->eof) return 1;
-
-    free = in->tok - in->buf;
-    if (free < need) return 2;
-
-    memmove(in->buf, in->tok, in->lim - in->tok);
-    in->lim -= free;
-    in->cur -= free;
-    in->mar -= free;
-    in->tok -= free;
-    /*!stags:re2c format = "if (in->@@) in->@@ -= free;\n"; */
-    in->lim += fread(in->lim, 1, free, in->file);
-    if (in->lim < in->buf + SIZE) {
-        in->eof = 1;
-        memset(in->lim, 0, YYMAXFILL);
-        in->lim += YYMAXFILL;
-    }
-    return 0;
-}
-
-static void print_headers(const char *tok,
-    const taglist_t *h1, const taglist_t *h2,
-    const taglist_t *h3, const taglist_t *h4,
-    const taglist_t *h5)
-{
-    if (!h1 || !h2 || !h3 || !h4 || !h5) {
-        assert(!h1 && !h2 && !h3 && !h4 && !h5);
-        return;
-    }
-    print_headers(tok, h1->pred, h2->pred, h3->pred, h4->pred, h5->pred);
-    printf("%.*s%.*s%.*s%.*s\n",
-        (int)(h2->dist - h1->dist), tok + h1->dist,
-        (int)(h3->dist - h2->dist), tok + h2->dist,
-        (int)(h4->dist - h3->dist), tok + h3->dist,
-        (int)(h5->dist - h4->dist), tok + h4->dist);
-}
-
-#define YYCTYPE        char
-#define YYCURSOR       in->cur
-#define YYMARKER       in->mar
-#define YYLIMIT        in->lim
-#define YYMTAGP(tl)    taglist(&tl, in->tok, in->cur, &in->tlp)
-#define YYMTAGN(tl)    taglist(&tl, in->tok, NULL, &in->tlp)
-#define YYFILL(n)      if (fill(in, n) != 0) return 2;
-
-static int lex(input_t *in, long *count, long *total)
-{
-    const char *of, *au, *at,
-        *hs1, *hs3, *m1, *p1, *p3, *p5, *q1, *q3,
-        *hs2, *hs4, *m2, *p2, *p4, *p6, *q2, *q4,
-        *r1, *r3, *rp1, *s1, *st1, *u1, *u3, *v1, *v3,
-        *r2, *r4, *rp2, *s2, *st2, *u2, *u4, *v2, *v4;
-    taglist_t *h1, *h2, *h3, *h4, *h5;
-    long c, t;
-
-    c = 0; t = 0;
-    of = au = at
-        = hs1 = hs3 = m1 = p1 = p3 = p5 = q1 = q3
-        = hs2 = hs4 = m2 = p2 = p4 = p6 = q2 = q4
-        = r1 = r3 = rp1 = s1 = st1 = u1 = u3 = v1 = v3
-        = r2 = r4 = rp2 = s2 = st2 = u2 = u4 = v2 = v4 = NULL;
-loop:
-    in->tok = in->cur;
-/*!re2c
-    re2c:flags:tags = 1;
-    re2c:tags:expression = "in->@@";
-
-    end = "\x00";
-    eol = "\n";
-
-    crlf        = eol;
-    sp          = " ";
-    htab        = "\t";
-    ows         = (sp | htab)*;
-    digit       = [0-9];
-    alpha       = [a-zA-Z];
-    hexdigit    = [0-9a-fA-F];
-    unreserved  = alpha | digit | [-._~];
-    pct_encoded = "%" hexdigit{2};
-    sub_delims  = [!$&'()*+,;=];
-    pchar       = unreserved | pct_encoded | sub_delims | [:@];
-    vchar       = [\x1f-\x7e];
-    tchar       = [-!#$%&'*+.^_`|~] | digit | alpha;
-
-    obs_fold       = crlf (sp | htab)+;
-    obs_text       = [\x80-\xff];
-    field_name     = tchar+;
-    field_vchar    = vchar | obs_text;
-    field_content  = field_vchar ((sp | htab)+ field_vchar)?;
-    field_value    = (field_content | obs_fold)*;
-    header_field   = #h1 field_name #h2 ":" ows #h3 field_value #h4 ows #h5;
-    scheme         = alpha (alpha | digit | [-+.])*;
-    userinfo       = (unreserved | pct_encoded | sub_delims | ":")*;
-    dec_octet
-        = digit
-        | [\x31-\x39] digit
-        | "1" digit{2}
-        | "2" [\x30-\x34] digit
-        | "25" [\x30-\x35];
-    ipv4address    = dec_octet "." dec_octet "." dec_octet "." dec_octet;
-    h16            = hexdigit{1,4};
-    ls32           = h16 ":" h16 | ipv4address;
-    ipv6address
-        =                            (h16 ":"){6} ls32
-        |                       "::" (h16 ":"){5} ls32
-        | (               h16)? "::" (h16 ":"){4} ls32
-        | ((h16 ":"){0,1} h16)? "::" (h16 ":"){3} ls32
-        | ((h16 ":"){0,2} h16)? "::" (h16 ":"){2} ls32
-        | ((h16 ":"){0,3} h16)? "::"  h16 ":"     ls32
-        | ((h16 ":"){0,4} h16)? "::"              ls32
-        | ((h16 ":"){0,5} h16)? "::"              h16
-        | ((h16 ":"){0,6} h16)? "::";
-    ipvfuture      = "v" hexdigit+ "." (unreserved | sub_delims | ":" )+;
-    ip_literal     = "[" ( ipv6address | ipvfuture ) "]";
-    reg_name       = (unreserved | pct_encoded | sub_delims)*;
-    path_abempty   = ("/" pchar*)*;
-    path_absolute  = "/" (pchar+ ("/" pchar*)*)?;
-    path_rootless  = pchar+ ("/" pchar*)*;
-    path_empty     = "";
-    host           = ip_literal | ipv4address | reg_name;
-    port           = digit*;
-    query          = (pchar | [/?])*;
-    absolute_uri   = @s1 scheme @s2 ":"
-        ( "//" (@u1 userinfo @u2 "@")? @hs1 host @hs2 (":" @r1 port @r2)? @p1 path_abempty @p2
-        | @p3 (path_absolute | path_rootless | path_empty) @p4
-        ) ("?" @q1 query @q2)?;
-    authority      = (@u3 userinfo @u4 "@")? @hs3 host @hs4 (":" @r3 port @r4)?;
-    origin_form    = @p5 path_abempty @p6 ("?" @q3 query @q4)?;
-    http_name      = "HTTP";
-    http_version   = http_name "/" digit "." digit;
-    request_target
-        = @at authority
-        | @au absolute_uri
-        | @of origin_form
-        | "*";
-    method         = tchar+;
-    request_line   = @m1 method @m2 sp request_target sp @v3 http_version @v4 crlf;
-    status_code    = digit{3};
-    reason_phrase  = (htab | sp | vchar | obs_text)*;
-    status_line    = @v1 http_version @v2 sp @st1 status_code @st2 sp @rp1 reason_phrase @rp2 crlf;
-    start_line     = (request_line | status_line);
-    message_head   = start_line (header_field crlf)* crlf;
-
-    *   { return 1; }
-    end { *count = c; *total = t; return 0; }
-    eol { goto loop; }
-    message_head {
-#ifndef VERIFY
-        c += 1;
-        if (st1) t += (v2 - v1) + (st2 - st1) + (rp2 - rp1);
-        if (m1) {
-            if (of) t += (p6 - p5) + (q4 - q3);
-            if (au) t += (s2 - s1) + (u2 - u1) + (hs2 - hs1)
-                + (r2 - r1) + (p2 - p1) + (p4 - p3) + (q2 - q1);
-            if (at) t += (u4 - u3) + (hs4 - hs3) + (r4 - r3);
-            t += (v4 - v3);
-        }
-        for (; h1 != 0; h1 = h1->pred, h2 = h2->pred,
-            h3 = h3->pred, h4 = h4->pred, h5 = h5->pred) {
-            t += (h2->dist - h1->dist) + (h3->dist - h2->dist)
-                + (h4->dist - h3->dist) + (h5->dist - h4->dist);
-        }
-#else
-        if (st1) {
-            printf("%.*s %.*s %.*s\n",
-                (int)(v2 - v1), v1,
-                (int)(st2 - st1), st1,
-                (int)(rp2 - rp1), rp1);
-        } else if (m1) {
-            printf("%.*s ", (int)(m2 - m1), m1);
-            if (of) {
-                printf("%.*s", (int)(p6 - p5), p5);
-                if (q3) printf("?%.*s", (int)(q4 - q3), q3);
-            } else if (au) {
-                printf("%.*s:", (int)(s2 - s1), s1);
-                if (p1) printf("//");
-                if (u1) printf("%.*s@", (int)(u2 - u1), u1);
-                printf("%.*s", (int)(hs2 - hs1), hs1);
-                if (r1) printf(":%.*s", (int)(r2 - r1), r1);
-                if (p1) printf("%.*s",  (int)(p2 - p1), p1);
-                if (p3) printf("%.*s",  (int)(p4 - p3), p3);
-                if (q1) printf("?%.*s", (int)(q2 - q1), q1);
-            } else if (at) {
-                if (u3) printf("%.*s@", (int)(u4 - u3), u3);
-                printf("%.*s", (int)(hs4 - hs3), hs3);
-                if (r3) printf(":%.*s", (int)(r4 - r3), r3);
-            } else {
-                printf("*");
-            }
-            printf(" %.*s\n", (int)(v4 - v3), v3);
-        }
-        print_headers(in->tok, h1, h2, h3, h4, h5);
-        printf("\n");
-#endif
-        taglistpool_clear(&in->tlp, in);
-        goto loop;
-    }
-*/
-}
-
-int main(int argc, char **argv)
-{
-    long count, total;
-    input_t in;
-    init_input(&in, argv[1]);
-
-    switch (lex(&in, &count, &total)) {
-        case 0: fprintf(stderr, "ok, parsed %ld HTTPs, peeked %ld chars\n", count, total); break;
-        case 1: fprintf(stderr, "syntax error: %s\n", in.cur); break;
-        case 2: fprintf(stderr, "fill failed\n"); break;
-        default: fprintf(stderr, "panic!\n"); break;
-    }
-
-    free_input(&in);
-    return 0;
-}

benchmarks/lookahead_tdfa/http/rfc7230/http_rfc7230_notags.re

@@ -1,180 +0,0 @@
-#include <stddef.h>
-#include <stdlib.h>
-#include <stdio.h>
-#include <string.h>
-
-typedef struct {
-    FILE *file;
-    char *buf;
-    char *lim;
-    char *cur;
-    char *mar;
-    char *tok;
-    int eof;
-} input_t;
-
-/*!max:re2c*/
-static const size_t SIZE = 4096;
-
-static void init_input(input_t *in, const char *fname)
-{
-    in->file = fopen(fname, "r");
-    in->buf = (char*) malloc(SIZE + YYMAXFILL);
-    in->lim = in->buf + SIZE;
-    in->cur = in->lim;
-    in->mar = in->lim;
-    in->tok = in->lim;
-    in->eof = 0;
-}
-
-static void free_input(input_t *in)
-{
-    fclose(in->file);
-    free(in->buf);
-}
-
-static int fill(input_t *in, size_t need)
-{
-    size_t free;
-
-    if (in->eof) return 1;
-
-    free = in->tok - in->buf;
-    if (free < need) return 2;
-
-    memmove(in->buf, in->tok, in->lim - in->tok);
-    in->lim -= free;
-    in->cur -= free;
-    in->mar -= free;
-    in->tok -= free;
-    in->lim += fread(in->lim, 1, free, in->file);
-    if (in->lim < in->buf + SIZE) {
-        in->eof = 1;
-        memset(in->lim, 0, YYMAXFILL);
-        in->lim += YYMAXFILL;
-    }
-    return 0;
-}
-
-#define YYCTYPE   char
-#define YYCURSOR  in->cur
-#define YYMARKER  in->mar
-#define YYLIMIT   in->lim
-#define YYFILL(n) if (fill(in, n) != 0) return 2;
-
-static int lex(input_t *in, long *count, long *total)
-{
-    long c, t;
-
-    c = 0; t = 0;
-loop:
-    in->tok = in->cur;
-/*!re2c
-
-    end = "\x00";
-    eol = "\n";
-
-    crlf        = eol;
-    sp          = " ";
-    htab        = "\t";
-    ows         = (sp | htab)*;
-    digit       = [0-9];
-    alpha       = [a-zA-Z];
-    hexdigit    = [0-9a-fA-F];
-    unreserved  = alpha | digit | [-._~];
-    pct_encoded = "%" hexdigit{2};
-    sub_delims  = [!$&'()*+,;=];
-    pchar       = unreserved | pct_encoded | sub_delims | [:@];
-    vchar       = [\x1f-\x7e];
-    tchar       = [-!#$%&'*+.^_`|~] | digit | alpha;
-
-    obs_fold       = crlf (sp | htab)+;
-    obs_text       = [\x80-\xff];
-    field_name     = tchar+;
-    field_vchar    = vchar | obs_text;
-    field_content  = field_vchar ((sp | htab)+ field_vchar)?;
-    field_value    = (field_content | obs_fold)*;
-    header_field   = field_name ":" ows field_value ows;
-    scheme         = alpha (alpha | digit | [-+.])*;
-    userinfo       = (unreserved | pct_encoded | sub_delims | ":")*;
-    dec_octet
-        = digit
-        | [\x31-\x39] digit
-        | "1" digit{2}
-        | "2" [\x30-\x34] digit
-        | "25" [\x30-\x35];
-    ipv4address    = dec_octet "." dec_octet "." dec_octet "." dec_octet;
-    h16            = hexdigit{1,4};
-    ls32           = h16 ":" h16 | ipv4address;
-    ipv6address
-        =                            (h16 ":"){6} ls32
-        |                       "::" (h16 ":"){5} ls32
-        | (               h16)? "::" (h16 ":"){4} ls32
-        | ((h16 ":"){0,1} h16)? "::" (h16 ":"){3} ls32
-        | ((h16 ":"){0,2} h16)? "::" (h16 ":"){2} ls32
-        | ((h16 ":"){0,3} h16)? "::"  h16 ":"     ls32
-        | ((h16 ":"){0,4} h16)? "::"              ls32
-        | ((h16 ":"){0,5} h16)? "::"              h16
-        | ((h16 ":"){0,6} h16)? "::";
-    ipvfuture      = "v" hexdigit+ "." (unreserved | sub_delims | ":" )+;
-    ip_literal     = "[" ( ipv6address | ipvfuture ) "]";
-    reg_name       = (unreserved | pct_encoded | sub_delims)*;
-    path_abempty   = ("/" pchar*)*;
-    path_absolute  = "/" (pchar+ ("/" pchar*)*)?;
-    path_rootless  = pchar+ ("/" pchar*)*;
-    path_empty     = "";
-    host           = ip_literal | ipv4address | reg_name;
-    port           = digit*;
-    query          = (pchar | [/?])*;
-    absolute_uri   =  scheme  ":"
-        ( "//" (userinfo "@")? host (":" port)? path_abempty
-        | (path_absolute | path_rootless | path_empty)
-        ) ("?" query)?;
-    authority      = (userinfo "@")? host (":" port)?;
-    origin_form    = path_abempty ("?" query)?;
-    http_name      = "HTTP";
-    http_version   = http_name "/" digit "." digit;
-    request_target
-        = authority
-        | absolute_uri
-        | origin_form
-        | "*";
-    method         = tchar+;
-    request_line   = method sp request_target sp http_version crlf;
-    status_code    = digit{3};
-    reason_phrase  = (htab | sp | vchar | obs_text)*;
-    status_line    = http_version sp status_code sp reason_phrase crlf;
-    start_line     = (request_line | status_line);
-    message_head   = start_line (header_field crlf)* crlf;
-
-    *   { return 1; }
-    end { *count = c; return 0; }
-    eol { goto loop; }
-    message_head {
-#ifndef VERIFY
-        c += 1;
-        t += in->cur - in->tok;
-#else
-        printf("%.*s", (int)(in->cur - in->tok), in->tok);
-#endif
-        goto loop;
-    }
-*/
-}
-
-int main(int argc, char **argv)
-{
-    long count, total;
-    input_t in;
-    init_input(&in, argv[1]);
-
-    switch (lex(&in, &count, &total)) {
-        case 0: fprintf(stderr, "ok, parsed %ld HTTPs, peeked %ld chars\n", count, total); break;
-        case 1: fprintf(stderr, "syntax error: %s\n", in.cur); break;
-        case 2: fprintf(stderr, "fill failed\n"); break;
-        default: fprintf(stderr, "panic!\n"); break;
-    }
-
-    free_input(&in);
-    return 0;
-}

benchmarks/lookahead_tdfa/http/simple/__bench.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile http simple ""
-run_all http simple
-

benchmarks/lookahead_tdfa/http/simple/__bench_b.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile http simple "-b"
-run_all http simple
-

benchmarks/lookahead_tdfa/http/simple/__bench_unopt.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile http simple "--no-optimize-tags"
-run_all http simple
-

benchmarks/lookahead_tdfa/http/simple/__verify.sh

@@ -1,4 +0,0 @@
-
-. ../../__bench_utils.sh
-
-verify http simple

benchmarks/lookahead_tdfa/http/simple/http_simple.re

@@ -1,231 +0,0 @@
-#include <assert.h>
-#include <stddef.h>
-#include <stdlib.h>
-#include <stdio.h>
-#include <string.h>
-
-typedef struct taglist_t {
-    struct taglist_t *pred;
-    long dist;
-} taglist_t;
-
-typedef struct taglistpool_t {
-    taglist_t *head;
-    taglist_t *next;
-    taglist_t *last;
-} taglistpool_t;
-
-typedef struct {
-    FILE *file;
-    char *buf;
-    char *lim;
-    char *cur;
-    char *mar;
-    char *tok;
-    /*!stags:re2c format = "char *@@;\n"; */
-    /*!mtags:re2c format = "taglist_t *@@;\n"; */
-    taglistpool_t tlp;
-    int eof;
-} input_t;
-
-static void taglistpool_clear(taglistpool_t *tlp, input_t *in)
-{
-    tlp->next = tlp->head;
-    /*!mtags:re2c format = "in->@@ = 0;\n"; */
-}
-
-static void taglistpool_init(taglistpool_t *tlp)
-{
-    static const unsigned size = 1024 * 1024;
-    tlp->head = (taglist_t*)malloc(size * sizeof(taglist_t));
-    tlp->next = tlp->head;
-    tlp->last = tlp->head + size;
-}
-
-static void taglistpool_free(taglistpool_t *tlp)
-{
-    free(tlp->head);
-    tlp->head = tlp->next = tlp->last = NULL;
-}
-
-static taglist_t *taglistpool_next(taglistpool_t *tlp)
-{
-    if (tlp->next < tlp->last) {
-        return tlp->next++;
-    }
-
-    const unsigned size = tlp->last - tlp->head;
-    taglist_t *head = (taglist_t*)malloc(2 * size * sizeof(taglist_t));
-    memcpy(head, tlp->head, size * sizeof(taglist_t));
-    free(tlp->head);
-    tlp->head = head;
-    tlp->next = head + size;
-    tlp->last = head + size * 2;
-    return tlp->next++;
-}
-
-static void taglist(taglist_t **ptl, const char *b, const char *t, taglistpool_t *tlp)
-{
-    taglist_t *tl = taglistpool_next(tlp);
-    tl->pred = *ptl;
-    tl->dist = t - b;
-    *ptl = tl;
-}
-
-/*!max:re2c*/
-static const size_t SIZE = 4096;
-
-static void init_input(input_t *in, const char *fname)
-{
-    in->file = fopen(fname, "r");
-    in->buf = (char*) malloc(SIZE + YYMAXFILL);
-    in->lim = in->buf + SIZE;
-    in->cur = in->lim;
-    in->mar = in->lim;
-    in->tok = in->lim;
-    /*!stags:re2c format = "in->@@ = 0;\n"; */
-    /*!mtags:re2c format = "in->@@ = 0;\n"; */
-    taglistpool_init(&in->tlp);
-    in->eof = 0;
-}
-
-static void free_input(input_t *in)
-{
-    fclose(in->file);
-    free(in->buf);
-    taglistpool_free(&in->tlp);
-}
-
-static int fill(input_t *in, size_t need)
-{
-    size_t free;
-
-    if (in->eof) return 1;
-
-    free = in->tok - in->buf;
-    if (free < need) return 2;
-
-    memmove(in->buf, in->tok, in->lim - in->tok);
-    in->lim -= free;
-    in->cur -= free;
-    in->mar -= free;
-    in->tok -= free;
-    /*!stags:re2c format = "if (in->@@) in->@@ -= free;\n"; */
-    in->lim += fread(in->lim, 1, free, in->file);
-    if (in->lim < in->buf + SIZE) {
-        in->eof = 1;
-        memset(in->lim, 0, YYMAXFILL);
-        in->lim += YYMAXFILL;
-    }
-    return 0;
-}
-
-static void print_headers(const char *tok,
-    const taglist_t *h1, const taglist_t *h2, const taglist_t *h3)
-{
-    if (!h1 || !h2 || !h3) {
-        assert(!h1 && !h2 && !h3);
-        return;
-    }
-    print_headers(tok, h1->pred, h2->pred, h3->pred);
-    printf("%.*s%.*s\n",
-        (int)(h2->dist - h1->dist), tok + h1->dist,
-        (int)(h3->dist - h2->dist), tok + h2->dist);
-}
-
-#define YYCTYPE        char
-#define YYCURSOR       in->cur
-#define YYMARKER       in->mar
-#define YYLIMIT        in->lim
-#define YYMTAGP(tl)    taglist(&tl, in->tok, in->cur, &in->tlp)
-#define YYMTAGN(tl)    taglist(&tl, in->tok, NULL, &in->tlp)
-#define YYFILL(n)      if (fill(in, n) != 0) return 2;
-
-static int lex(input_t *in, long *count, long *total)
-{
-    const char *s1, *s2, *v1, *v2, *v3, *v4, *m1, *m2, *rp1, *rp2, *rt1, *rt2;
-    taglist_t *h1, *h2, *h3;
-    long c, t;
-
-    c = 0; t = 0;
-loop:
-    in->tok = in->cur;
-/*!re2c
-    re2c:flags:tags = 1;
-    re2c:tags:expression = "in->@@";
-
-    end            = "\x00";
-    eol            = "\n";
-    sp             = " ";
-    htab           = "\t";
-    ows            = (sp | htab)*;
-    char           = [-._~%!$&'()*+,;=a-zA-Z0-9];
-    tchar          = [-._~%!$&'*+#^`|a-zA-Z0-9];
-    vchar          = [\x1f-\x7e\x80-\xff];
-    scheme         = [-+.a-zA-Z0-9]+;
-    userinfo       = (char | [:])+;
-    host           = (char | [:[\]])+;
-    port           = [0-9]*;
-    path           = (char | [:@/])*;
-    query          = (char | [:@?/])*;
-    obs_fold       = eol (sp | htab)+;
-    field_content  = vchar ((sp | htab)+ vchar)?;
-    header_field   = #h1 tchar+ ":" #h2 ows (field_content | obs_fold)* ows #h3;
-    authority      = (userinfo "@")? host (":" port)?;
-    absolute_uri   = scheme ":" ("//" (userinfo "@")? host (":" port)?)? path ("?" query)?;
-    origin_form    = "/" path ("?" query)?;
-    http_version   = "HTTP/" [0-9] "." [0-9];
-    request_target = authority | absolute_uri | origin_form | "*";
-    method         = tchar+;
-    request_line   = @m1 method @m2 sp @rt1 request_target @rt2 sp @v3 http_version @v4 eol;
-    status_code    = [0-9]{3};
-    reason_phrase  = (htab | sp | vchar)*;
-    status_line    = @v1 http_version @v2 sp @s1 status_code @s2 sp @rp1 reason_phrase @rp2 eol;
-    start_line     = (request_line | status_line);
-    message_head   = start_line (header_field eol)* eol;
-
-    *   { return 1; }
-    end { *count = c; *total = t; return 0; }
-    eol { goto loop; }
-    message_head {
-#ifndef VERIFY
-        c += 1;
-        t += (v2 - v1) + (s2 - s1) + (rp2 - rp1)
-            + (m2 - m1) + (rt2 - rt1) + (v4 - v3);
-        for (; h1 != 0; h1 = h1->pred, h2 = h2->pred, h3 = h3->pred) {
-            t += (h2->dist - h1->dist) + (h3->dist - h2->dist);
-        }
-#else
-        if (s1) printf("%.*s %.*s %.*s\n",
-            (int)(v2 - v1), v1,
-            (int)(s2 - s1), s1,
-            (int)(rp2 - rp1), rp1);
-        if (m1) printf("%.*s %.*s %.*s\n",
-            (int)(m2 - m1), m1,
-            (int)(rt2 - rt1), rt1,
-            (int)(v4 - v3), v3);
-        print_headers(in->tok, h1, h2, h3);
-        printf("\n");
-#endif
-        taglistpool_clear(&in->tlp, in);
-        goto loop;
-    }
-*/
-}
-
-int main(int argc, char **argv)
-{
-    long count, total;
-    input_t in;
-    init_input(&in, argv[1]);
-
-    switch (lex(&in, &count, &total)) {
-        case 0: fprintf(stderr, "ok, parsed %ld HTTPs, peeked %ld chars\n", count, total); break;
-        case 1: fprintf(stderr, "syntax error: %s\n", in.cur); break;
-        case 2: fprintf(stderr, "fill failed\n"); break;
-        default: fprintf(stderr, "panic!\n"); break;
-    }
-
-    free_input(&in);
-    return 0;
-}

benchmarks/lookahead_tdfa/http/simple/http_simple_notags.re

@@ -1,134 +0,0 @@
-#include <assert.h>
-#include <stddef.h>
-#include <stdlib.h>
-#include <stdio.h>
-#include <string.h>
-
-typedef struct {
-    FILE *file;
-    char *buf;
-    char *lim;
-    char *cur;
-    char *mar;
-    char *tok;
-    int eof;
-} input_t;
-
-/*!max:re2c*/
-static const size_t SIZE = 4096;
-
-static void init_input(input_t *in, const char *fname)
-{
-    in->file = fopen(fname, "r");
-    in->buf = (char*) malloc(SIZE + YYMAXFILL);
-    in->lim = in->buf + SIZE;
-    in->cur = in->lim;
-    in->mar = in->lim;
-    in->tok = in->lim;
-    in->eof = 0;
-}
-
-static void free_input(input_t *in)
-{
-    fclose(in->file);
-    free(in->buf);
-}
-
-static int fill(input_t *in, size_t need)
-{
-    size_t free;
-
-    if (in->eof) return 1;
-
-    free = in->tok - in->buf;
-    if (free < need) return 2;
-
-    memmove(in->buf, in->tok, in->lim - in->tok);
-    in->lim -= free;
-    in->cur -= free;
-    in->mar -= free;
-    in->tok -= free;
-    in->lim += fread(in->lim, 1, free, in->file);
-    if (in->lim < in->buf + SIZE) {
-        in->eof = 1;
-        memset(in->lim, 0, YYMAXFILL);
-        in->lim += YYMAXFILL;
-    }
-    return 0;
-}
-
-#define YYCTYPE            char
-#define YYCURSOR           in->cur
-#define YYMARKER           in->mar
-#define YYLIMIT            in->lim
-#define YYFILL(n)          if (fill(in, n) != 0) return 2;
-
-static int lex(input_t *in, long *count, long *total)
-{
-    long c, t;
-
-    c = 0; t = 0;
-loop:
-    in->tok = in->cur;
-/*!re2c
-    end            = "\x00";
-    eol            = "\n";
-    sp             = " ";
-    htab           = "\t";
-    ows            = (sp | htab)*;
-    char           = [-._~%!$&'()*+,;=a-zA-Z0-9];
-    tchar          = [-._~%!$&'*+#^`|a-zA-Z0-9];
-    vchar          = [\x1f-\x7e\x80-\xff];
-    scheme         = [-+.a-zA-Z0-9]+;
-    userinfo       = (char | [:])+;
-    host           = (char | [:[\]])+;
-    port           = [0-9]*;
-    path           = (char | [:@/])*;
-    query          = (char | [:@?/])*;
-    obs_fold       = eol (sp | htab)+;
-    field_content  = vchar ((sp | htab)+ vchar)?;
-    header_field   = tchar+ ":" ows (field_content | obs_fold)* ows;
-    authority      = (userinfo "@")? host (":" port)?;
-    absolute_uri   = scheme ":" ("//" (userinfo "@")? host (":" port)?)? path ("?" query)?;
-    origin_form    = "/" path ("?" query)?;
-    http_version   = "HTTP/" [0-9] "." [0-9];
-    request_target = authority | absolute_uri | origin_form | "*";
-    method         = tchar+;
-    request_line   = method sp request_target sp http_version eol;
-    status_code    = [0-9]{3};
-    reason_phrase  = (htab | sp | vchar)*;
-    status_line    = http_version sp status_code sp reason_phrase eol;
-    start_line     = (request_line | status_line);
-    message_head   = start_line (header_field eol)* eol;
-
-    *   { return 1; }
-    end { *count = c; *total = t; return 0; }
-    eol { goto loop; }
-    message_head {
-#ifndef VERIFY
-        c += 1;
-        t += in->cur - in->tok;
-#else
-        printf("%.*s", (int)(in->cur - in->tok), in->tok);
-#endif
-        goto loop;
-    }
-*/
-}
-
-int main(int argc, char **argv)
-{
-    long count, total;
-    input_t in;
-    init_input(&in, argv[1]);
-
-    switch (lex(&in, &count, &total)) {
-        case 0: fprintf(stderr, "ok, parsed %ld HTTPs, peeked %ld chars\n", count, total); break;
-        case 1: fprintf(stderr, "syntax error: %s\n", in.cur); break;
-        case 2: fprintf(stderr, "fill failed\n"); break;
-        default: fprintf(stderr, "panic!\n"); break;
-    }
-
-    free_input(&in);
-    return 0;
-}

benchmarks/lookahead_tdfa/uri/gen/gen_uri.hs

@@ -1,111 +0,0 @@
-{-# LANGUAGE LambdaCase #-}
-
-import qualified Test.QuickCheck as Q
-import qualified Test.QuickCheck.Monadic as QM
-import qualified System.Random as SR
-import qualified System.Process as SP
-import qualified System.Exit as SE
-import qualified Data.ByteString.Char8 as BS
-import           Data.Char (chr)
-import           Data.List (intercalate)
-import           Control.Monad (forM_, when, replicateM)
-import           Data.Foldable (foldlM)
-
-type Scheme = String
-type User = String
-type Password = String
-type Host = String
-type Port = String
-type Path = String
-type Query = String
-type Fragment = String
-data URI = URI Scheme User Password Host Port Path Query Fragment
-
-instance Show URI where
-    show (URI sc us pw ho po pa qu fr) = concat
-        [ sc
-        , ":"
-        , if ho == "" then "" else "//"
-        , us
-        , if us == "" then "" else ":"
-        , pw
-        , if pw == "" then "" else "@"
-        , ho
-        , if po == "" then "" else ":"
-        , po
-        , pa
-        , if qu == "" then "" else "?"
-        , qu
-        , if fr == "" then "" else "#"
-        , fr
-        ]
-
-lalpha = ['a'..'z']
-ualpha = ['A'..'Z']
-alpha = lalpha ++ ualpha
-digit = ['0'..'9']
-hexdigit = ['a'..'f'] ++ ['A'..'F'] ++ digit
-alphanum = alpha ++ digit
-
-lstr :: Q.Gen String
-lstr = Q.listOf1 (Q.elements lalpha) :: Q.Gen [Char]
-
-anstr :: Q.Gen String
-anstr = Q.listOf1 (Q.elements alphanum) :: Q.Gen [Char]
-
-ipv4 :: Q.Gen String
-ipv4 = do
-    let octet = show <$> (Q.choose (0,255) :: Q.Gen Int)
-    intercalate "." <$> replicateM 4 octet
-
-ipv6 :: Q.Gen String
-ipv6 = do
-    let hexgroup = do
-            k <- Q.choose (1, 4) :: Q.Gen Int
-            replicateM k (Q.elements hexdigit)
-    n <- Q.choose (0, 8) :: Q.Gen Int
-    m <- Q.choose (1, n) :: Q.Gen Int
-    hh <- replicateM n hexgroup
-    return $ (++ "]") . ("[" ++) $ case n of
-        8 -> intercalate ":" hh
-        _ -> concat [intercalate ":" (take m hh), "::", intercalate ":" (drop m hh)]
-
-hostname :: Q.Gen String
-hostname = do
-    subdomains <- Q.listOf1 (take 10 <$> lstr)
-    return $ intercalate "." $ take 5 subdomains
-
-filepath :: Q.Gen String
-filepath = do
-    subdirs <- Q.listOf1 lstr
-    return $ concat $ map ("/" ++) $ take 10 subdirs
-
-mayb :: Q.Gen String -> Q.Gen String
-mayb gen = Q.oneof [gen, pure ""]
-
-instance Q.Arbitrary URI where
-    arbitrary = do
-        scheme   <- Q.elements ["http", "https", "ftp", "file"]
-        user     <- lstr
-        password <- anstr
-        host     <- Q.frequency [(10, ipv4), (10, hostname), (1, ipv6)]
-        port     <- mayb $ show <$> (Q.choose (0, 10000) :: Q.Gen Int)
-        path     <- filepath
-        query    <- mayb anstr
-        fragment <- mayb anstr
-
-        Q.frequency
-            [ (1, pure $ URI scheme user  password host port path query fragment)
-            , (2, pure $ URI scheme ""    ""       host port path query fragment)
-            , (1, pure $ URI scheme ""    ""       ""   ""   path query fragment)
-            ]
-
-main = do
-    let f = "uri.dat"
-    writeFile f ""
-
-    forM_ [0..600000] $ \i -> do
-        when (i `mod` 100 == 0) $ print i
-
-        uri <- Q.generate (Q.arbitrary :: Q.Gen URI)
-        appendFile f $ show uri ++ "\n"

benchmarks/lookahead_tdfa/uri/gen/mk_gen.sh

@@ -1 +0,0 @@
-ghc -O2 -Wall gen_uri.hs

benchmarks/lookahead_tdfa/uri/gen/sample.dat

@@ -1,20 +0,0 @@
-https://130.226.207.254/jbtvkjzrjthmbnypntksydqelq/jjdsj/udvjilrctugdnuolrrzkr/cwkvkxjhehiodwfytl/ufgjg/nfgpg/iu#mE39x1S2vXd1o
-https:/ohgudpayfmlfhcoyvgjhyqfbua/linrqhrhlaiixdbxonbmn/atrgtgdphcz/rr/dtdhuqshwxhewfn/fyuawfip/qugtmsmwxuywignctpbtabge/acijwctkdqawozgns/eipjsaiiurudzqrviveu/bijkdhwtmlneypz#EhkI5ylLokJNKaBsd
-http://cjop.xhsqcbwtp.vrfzoaodga.nnkbwbmq.cfvgulvcfj/geuhnhetwsqjfuojmvaziemd/eodavsnwgauluaclpjqb/phmdaviaz#AlBK76vA
-http:/ijnjyrzldk/jyapegxurikjmip/olhhrswcvepexhshr/whlsdorizwdovszxcuwnnbexfy/ohsfdqoryjotnqrzagziyg/ajxvofxromjwirayehuw/swubrfzbwtmtvomstbaumtvksghg#D
-https://220.234.52.233/peiygqiuiwezsr/qnymfpaurkzifllbwtkzqgpkcdp/hbeschxipcpf/soakqi/bxbhzothilaxznoiprg/gaqsbgqobypjyzwdyvcmr/fszwjmnixfc/t/vlumvtpktjjkuzxix/nqqwnxwkyinkjngixkcgrobuwe#L7cwKfIxLBv
-http://190.102.237.163/jfymqrdxatqxxosr/eodoflhmefpbqunvzkwkxfgxj/ife/bbijkiueazcfwodgsddft/xpegxlfbdzafenjaqgjbnqrsb/poizlwooajtnejxpycftrcsfjkruln?DEVzXYliuaVThP2db4oE32q#fXoCDA1BJyjOUBJwpMOOC36Q7o6n
-https://prlmszykmotajvhfingpy:5gJpVNFdVUCBumKIV0ciYat@tyekbzfzyn.kirctezfig.gvefopbqlf.onykhnytqd.uofiltbvdv/jiqe/odskecdswmouqaunyfuerpefcusuxa/fkqdmaueh/vzkskyxrpqkfzzjgujqmkt/qfnzplmxvunhvjq/kskwzjeueepwokxguatetegnjzgu/pysxiyoyssbfiibuonhk/wmzxsguoqoeutnkaanqqd/pkpbwvoepeutjbedkdimko/vogpbm#fBzuGdIV22C
-https://uqaqxbiszxuwfd:mIai0iqs9Ps3UEvpmu@qwnsormfld.uikzf.rajsatakhr:2634/pcervqbppzxjydwilqwoxa/byyvgqvnqjmqgvnjmbio/sifgcfwfhlxrnzpjlo/zhdejqnlcgmmknd/bpqtpqsxyrlvfohbkwavetd/lcsxvflqocclqvilbuvgcggupysof/wzuwqttblllpdjxxqzjnv/uqzxnhuobrldciwbnrfqmv/jpbashfkljtfciiqivtdpeqrfhz/hdjtjdgfyja
-file://ngngiosyeynlnyegckpwp:QoW1iBFx@171.234.221.207/ltdgogjynxrrnjygsltmcfihfsqfvk/mixerqujwbektqyzimadzdx/urb/vrvjsofccprkaavyyp/sycatuazj/vlmmmvozycmvpicrpcipfxywp/zddjtiahb
-https://58.187.10.250:8256/ialbromwxodapxebh/lwddbb/fghhkaztbruuibdfwjym/vrwkmqrnqumknb/kytsuocatyouppiec/wxdwpeudwfghhdwiagjrgllqhoaxhn/tnugkxbiimo/ydrzmbihqoibkcvm/rqyczmveyz/nefixtpefrkuxntolkwgiamp
-ftp://hukiuvtpzvmwmnxgomsujflyk:b1MSVh7pUWPjK24nnJoOOwKIET36s@6.119.196.62:5869/cyhsyhjykv#NlE8g3hz1Ebta8kk7LUvTm
-ftp://fhvvtgyuua.ogahfrqbvg.cughxcqaqu.nyfdeg.mzbkmrnfib:4544/vwqfmulbaoxczkm/clixvtifydkj/shrflpkrq/vi/sruymrqxodjrbwvhtbwzvmlsys/uv/ybywv#ADdXUBIIgy
-http:/laxn/rugglbrcb/znzlrhyoxapkznmedipktxhqzeohlw/rfmg/uyntiipnzdwjlfy/fktcgxrgcsfkr/rbbcxzybieozevucqxpmeq/apqyszyxgpq/zbknrolqhmtwjshockr/kicrlzsbtx?VbJp1VLkYC#mYPZplrVnb7UHSatrpGAwi
-https:/ztpluoa/jphvoweijirbvnmek/pbvimnscpb/catfuxpwqihpqbnrjvkhmbdagvqzsu/czjmvoa/btgpsmqroaxkqt/kcvpypw/eqbifrwpexhsvhmrdtfj/y/gqmjqiyqfkjatby#xDfmbp6mzuOS73yV
-https://jqhvw:ssV@[4::a8:5ab]:7064/tto/xjdjuxhkqhuvmdbwn/kejaxevbfon/nxrvxlozfd/nefd/drhf/ya/meebpkpao/go/sv?XZtzUKrb4Qq19CsMmL
-https:/nrmkjtzsirjeybbznmbxefn/tbunkcjyoksunhdiiec/iuguxifycwqd/cjaqmevkjkwf/lysgdnbidkjtulmtdsxuiw/ykmcmukxjjmzdutq/nndpaxzkknxawibpc/bsuqrecyypwqxsbbslv/hzfitcvz?zVwBAXzGU6
-https://179.106.111.234:4602/kaojoxgqtrtqjqbxojhplkgo/sakwfopartajvd/jdjmzkjzxsutx/jvorvytygowpcugwcpdp/gf/skqbpbmaxznokzio/jmgdwytfu/zwronpnfltlqsdkm/bjkpnbqymcebiccdvdh/gelbqajbdysgljczflyocuxqehpvrn#iePJE2SwM2H2z7pGWE3MZQu79
-https://sgcoeyexpydtvuvrlncfv:cCVO9XO6W6lY58itexKFMeputSs7@199.131.253.98/cxoyvdvmlofvxhlfdoosxyrzhei/dqnifd/kuweqsnj/wrcmglfaorwemv/oyotkyibrraabiwatl/mw/fmyxtdztnyqmciijifflmb/kyfprlhxhcemmr/ds/ubuqzto
-ftp://226.47.189.219/uttpjalgzirumdxcqstmop/rsavfunauwrnzbrbccv/qvnzgkypemnulagkcnxgmhg/vdzboggzuttu/wimrcceehdaznqnvguhi/hcmbkir/vxcincavxbheiuuqoukjsnd/rnkr/repqv/cdncqpiqhrughnlhbu?04HBSdquIv5jRN5POmtHr#t8TxSWH16brGRC79JCiBCp8bEyF
-http://lkjzakdejf.kwbbhfnwhj.cypikjr.wogqvpwhbz.dwyohxizdk:4775/iwnmuufehazvnzgsrqm/gdjpojnwrlwikzpoevnpnkvw/wrbctzxujufpuyeue/dlzajtkhkbuj/zptqonydlp/qatdpxbpwvndzwrmynkfrbrpyix/ehavmzuuptjcv/bbosue/xulpwugitisuaagcmlvwtfykvr/izafvzsb?dY4#eIfVkGptVdWsH8D

benchmarks/lookahead_tdfa/uri/rfc3986/__bench.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile uri rfc3986 ""
-run_all uri rfc3986
-

benchmarks/lookahead_tdfa/uri/rfc3986/__bench_b.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile uri rfc3986 "-b"
-run_all uri rfc3986
-

benchmarks/lookahead_tdfa/uri/rfc3986/__bench_unopt.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile uri rfc3986 "--no-optimize-tags"
-run_all uri rfc3986
-

benchmarks/lookahead_tdfa/uri/rfc3986/__verify.sh

@@ -1,4 +0,0 @@
-
-. ../../__bench_utils.sh
-
-verify uri rfc3986

benchmarks/lookahead_tdfa/uri/rfc3986/uri_rfc3986.re

@@ -1,178 +0,0 @@
-#include <stddef.h>
-#include <stdlib.h>
-#include <stdio.h>
-#include <string.h>
-
-/*!max:re2c*/
-static const size_t SIZE = 4096;
-
-typedef struct {
-    FILE *file;
-    char *buf;
-    char *lim;
-    char *cur;
-    char *mar;
-    char *tok;
-    /*!stags:re2c format = "char *@@;\n"; */
-    int eof;
-} input_t;
-
-static void init_input(input_t *in, const char *fname)
-{
-    in->file = fopen(fname, "r");
-    in->buf = (char*) malloc(SIZE + YYMAXFILL);
-    in->lim = in->buf + SIZE;
-    in->cur = in->lim;
-    in->mar = in->lim;
-    in->tok = in->lim;
-    /*!stags:re2c format = "in->@@ = 0;\n"; */
-    in->eof = 0;
-}
-
-static void free_input(input_t *in)
-{
-    free(in->buf);
-    fclose(in->file);
-}
-
-static int fill(input_t *in, size_t need)
-{
-    size_t free;
-
-    if (in->eof) return 1;
-
-    free = in->tok - in->buf;
-    if (free < need) return 2;
-
-    memmove(in->buf, in->tok, in->lim - in->tok);
-    in->lim -= free;
-    in->cur -= free;
-    in->mar -= free;
-    in->tok -= free;
-    /*!stags:re2c format = "if (in->@@) in->@@ -= free;\n"; */
-    in->lim += fread(in->lim, 1, free, in->file);
-    if (in->lim < in->buf + SIZE) {
-        in->eof = 1;
-        memset(in->lim, 0, YYMAXFILL);
-        in->lim += YYMAXFILL;
-    }
-    return 0;
-}
-
-static int lex(input_t *in, long *count, long *total)
-{
-    const char
-        *s1, *u1, *h1, *h3, *h5, *r1, *p1, *p3, *q1, *f1,
-        *s2, *u2, *h2, *h4, *h6, *r2, *p2, *p4, *q2, *f2;
-    long c, t;
-
-    c = 0; t = 0;
-loop:
-    in->tok = in->cur;
-/*!re2c
-
-    re2c:define:YYCTYPE = char;
-    re2c:define:YYCURSOR = in->cur;
-    re2c:define:YYMARKER = in->mar;
-    re2c:define:YYLIMIT = in->lim;
-    re2c:define:YYFILL = "if (fill(in, @@) != 0) return 2;";
-    re2c:define:YYFILL:naked = 1;
-    re2c:flags:tags = 1;
-    re2c:tags:expression = "in->@@";
-
-    end = "\x00";
-    eol = "\n";
-
-    alpha       = [a-zA-Z];
-    digit       = [0-9];
-    hexdigit    = [0-9a-fA-F];
-    unreserved  = alpha | digit | [-._~];
-    pct_encoded = "%" hexdigit{2};
-    sub_delims  = [!$&'()*+,;=];
-    pchar       = unreserved | pct_encoded | sub_delims | [:@];
-
-    scheme = @s1 alpha (alpha | digit | [-+.])* @s2;
-    userinfo = @u1 (unreserved | pct_encoded | sub_delims | ":")* @u2;
-    dec_octet
-        = digit
-        | [\x31-\x39] digit
-        | "1" digit{2}
-        | "2" [\x30-\x34] digit
-        | "25" [\x30-\x35];
-    ipv4address = dec_octet "." dec_octet "." dec_octet "." dec_octet;
-    h16         = hexdigit{1,4};
-    ls32        = h16 ":" h16 | ipv4address;
-    ipv6address
-        =                            (h16 ":"){6} ls32
-        |                       "::" (h16 ":"){5} ls32
-        | (               h16)? "::" (h16 ":"){4} ls32
-        | ((h16 ":"){0,1} h16)? "::" (h16 ":"){3} ls32
-        | ((h16 ":"){0,2} h16)? "::" (h16 ":"){2} ls32
-        | ((h16 ":"){0,3} h16)? "::"  h16 ":"     ls32
-        | ((h16 ":"){0,4} h16)? "::"              ls32
-        | ((h16 ":"){0,5} h16)? "::"              h16
-        | ((h16 ":"){0,6} h16)? "::";
-    ipvfuture   = "v" hexdigit+ "." (unreserved | sub_delims | ":" )+;
-    ip_literal  = "[" ( ipv6address | ipvfuture ) "]";
-    reg_name    = (unreserved | pct_encoded | sub_delims)*;
-    host
-        = @h1 ip_literal  @h2
-        | @h3 ipv4address @h4
-        | @h5 reg_name    @h6;
-    port      = @r1 digit* @r2;
-    authority = (userinfo "@")? host (":" port)?;
-    path_abempty  = ("/" pchar*)*;
-    path_absolute = "/" (pchar+ ("/" pchar*)*)?;
-    path_rootless = pchar+ ("/" pchar*)*;
-    path_empty    = "";
-    hier_part
-        = "//" authority @p1 path_abempty @p2
-        | @p3 (path_absolute | path_rootless | path_empty) @p4;
-    query    = @q1 (pchar | [/?])* @q2;
-    fragment = @f1 (pchar | [/?])* @f2;
-    uri = scheme ":" hier_part ("?" query)? ("#" fragment)?;
-
-    *   { return 1; }
-    end { *count = c; *total = t; return 0; }
-    eol { goto loop; }
-    uri {
-#ifndef VERIFY
-        c += 1;
-        t += (s2 - s1) + (u2 - u1) + (h2 - h1) + (h4 - h3)
-            + (h6 - h5) + (r2 - r1) + (p2 - p1) + (p4 - p3)
-            + (q2 - q1) + (f2 - f1);
-#else
-                printf("%.*s:", (int)(s2 - s1), s1);
-        if (p1) printf("//");
-        if (u1) printf("%.*s@", (int)(u2 - u1), u1);
-        if (h1) printf("%.*s",  (int)(h2 - h1), h1);
-        if (h3) printf("%.*s",  (int)(h4 - h3), h3);
-        if (h5) printf("%.*s",  (int)(h6 - h5), h5);
-        if (r1) printf(":%.*s", (int)(r2 - r1), r1);
-        if (p1) printf("%.*s",  (int)(p2 - p1), p1);
-        if (p3) printf("%.*s",  (int)(p4 - p3), p3);
-        if (q1) printf("?%.*s", (int)(q2 - q1), q1);
-        if (f1) printf("#%.*s", (int)(f2 - f1), f1);
-        printf("\n");
-#endif
-        goto loop;
-    }
-*/
-}
-
-int main(int argc, char **argv)
-{
-    long count, total;
-    input_t in;
-    init_input(&in, argv[1]);
-
-    switch (lex(&in, &count, &total)) {
-        case 0: fprintf(stderr, "ok, parsed %ld URIs, peeked %ld chars\n", count, total); break;
-        case 1: fprintf(stderr, "syntax error\n"); break;
-        case 2: fprintf(stderr, "yyfill failed\n"); break;
-        default: fprintf(stderr, "panic!\n"); break;
-    }
-
-    free_input(&in);
-    return 0;
-}

benchmarks/lookahead_tdfa/uri/rfc3986/uri_rfc3986_notags.re

@@ -1,153 +0,0 @@
-#include <stddef.h>
-#include <stdlib.h>
-#include <stdio.h>
-#include <string.h>
-
-/*!max:re2c*/
-static const size_t SIZE = 4096;
-
-typedef struct {
-    FILE *file;
-    char *buf;
-    char *lim;
-    char *cur;
-    char *mar;
-    char *tok;
-    int eof;
-} input_t;
-
-static void init_input(input_t *in, const char *fname)
-{
-    in->file = fopen(fname, "r");
-    in->buf = (char*) malloc(SIZE + YYMAXFILL);
-    in->lim = in->buf + SIZE;
-    in->cur = in->lim;
-    in->mar = in->lim;
-    in->tok = in->lim;
-    in->eof = 0;
-}
-
-static void free_input(input_t *in)
-{
-    free(in->buf);
-    fclose(in->file);
-}
-
-static int fill(input_t *in, size_t need)
-{
-    size_t free;
-
-    if (in->eof) return 1;
-
-    free = in->tok - in->buf;
-    if (free < need) return 2;
-
-    memmove(in->buf, in->tok, in->lim - in->tok);
-    in->lim -= free;
-    in->cur -= free;
-    in->mar -= free;
-    in->tok -= free;
-    in->lim += fread(in->lim, 1, free, in->file);
-    if (in->lim < in->buf + SIZE) {
-        in->eof = 1;
-        memset(in->lim, 0, YYMAXFILL);
-        in->lim += YYMAXFILL;
-    }
-    return 0;
-}
-
-static int lex(input_t *in, long *count, long *total)
-{
-    long c, t;
-
-    c = 0; t = 0;
-loop:
-    in->tok = in->cur;
-/*!re2c
-    re2c:define:YYCTYPE = char;
-    re2c:define:YYCURSOR = in->cur;
-    re2c:define:YYMARKER = in->mar;
-    re2c:define:YYLIMIT = in->lim;
-    re2c:define:YYFILL = "if (fill(in, @@) != 0) return 2;";
-    re2c:define:YYFILL:naked = 1;
-
-    end = "\x00";
-    eol = "\n";
-
-    alpha         = [a-zA-Z];
-    digit         = [0-9];
-    hexdigit      = [0-9a-fA-F];
-    unreserved    = alpha | digit | [-._~];
-    pct_encoded   = "%" hexdigit{2};
-    sub_delims    = [!$&'()*+,;=];
-    pchar         = unreserved | pct_encoded | sub_delims | [:@];
-
-    scheme        = alpha (alpha | digit | [-+.])*;
-    userinfo      = (unreserved | pct_encoded | sub_delims | ":")*;
-    dec_octet
-        = digit
-        | [\x31-\x39] digit
-        | "1" digit{2}
-        | "2" [\x30-\x34] digit
-        | "25" [\x30-\x35];
-    ipv4address   = dec_octet "." dec_octet "." dec_octet "." dec_octet;
-    h16           = hexdigit{1,4};
-    ls32          = h16 ":" h16 | ipv4address;
-    ipv6address
-        =                            (h16 ":"){6} ls32
-        |                       "::" (h16 ":"){5} ls32
-        | (               h16)? "::" (h16 ":"){4} ls32
-        | ((h16 ":"){0,1} h16)? "::" (h16 ":"){3} ls32
-        | ((h16 ":"){0,2} h16)? "::" (h16 ":"){2} ls32
-        | ((h16 ":"){0,3} h16)? "::"  h16 ":"     ls32
-        | ((h16 ":"){0,4} h16)? "::"              ls32
-        | ((h16 ":"){0,5} h16)? "::"              h16
-        | ((h16 ":"){0,6} h16)? "::";
-    ipvfuture     = "v" hexdigit+ "." (unreserved | sub_delims | ":" )+;
-    ip_literal    = "[" ( ipv6address | ipvfuture ) "]";
-    reg_name      = (unreserved | pct_encoded | sub_delims)*;
-    host          = ip_literal | ipv4address | reg_name;
-    port          = digit*;
-    authority     = (userinfo "@")? host (":" port)?;
-    path_abempty  = ("/" pchar*)*;
-    path_absolute = "/" (pchar+ ("/" pchar*)*)?;
-    path_rootless = pchar+ ("/" pchar*)*;
-    path_empty    = "";
-    hier_part
-        = "//" authority path_abempty
-        | (path_absolute | path_rootless | path_empty);
-    query         = (pchar | [/?])*;
-    fragment      = (pchar | [/?])*;
-    uri           = scheme ":" hier_part ("?" query)? ("#" fragment)?;
-
-    *   { return 1; }
-    end { *count = c; *total = t; return 0; }
-    eol { goto loop; }
-    uri {
-#ifndef VERIFY
-        c += 1;
-        t += in->cur - in->tok;
-#else
-        printf("%.*s\n", (int)(in->cur - in->tok), in->tok);
-#endif
-        goto loop;
-    }
-*/
-}
-
-int main(int argc, char **argv)
-{
-    long count, total;
-    input_t in;
-    init_input(&in, argv[1]);
-
-    switch (lex(&in, &count, &total)) {
-        case 0: fprintf(stderr, "ok, parsed %ld URIs, peeked %ld chars\n", count, total); break;
-        case 1: fprintf(stderr, "syntax error\n"); break;
-        case 2: fprintf(stderr, "yyfill failed\n"); break;
-        default: fprintf(stderr, "panic!\n"); break;
-    }
-
-    free_input(&in);
-    return 0;
-}

benchmarks/lookahead_tdfa/uri/simple/__bench.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile uri simple ""
-run_all uri simple
-

benchmarks/lookahead_tdfa/uri/simple/__bench_b.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile uri simple "-b"
-run_all uri simple
-

benchmarks/lookahead_tdfa/uri/simple/__bench_unopt.sh

@@ -1,6 +0,0 @@
-
-. ../../__bench_utils.sh
-
-compile uri simple "--no-optimize-tags"
-run_all uri simple
-

benchmarks/lookahead_tdfa/uri/simple/__verify.sh

@@ -1,4 +0,0 @@
-
-. ../../__bench_utils.sh
-
-verify uri simple

benchmarks/lookahead_tdfa/uri/simple/uri_simple.re

@@ -1,134 +0,0 @@
-#include <stddef.h>
-#include <stdlib.h>
-#include <stdio.h>
-#include <string.h>
-
-/*!max:re2c*/
-static const size_t SIZE = 4096;
-
-typedef struct {
-    FILE *file;
-    char *buf;
-    char *lim;
-    char *cur;
-    char *mar;
-    char *tok;
-    /*!stags:re2c format = "char *@@;\n"; */
-    int eof;
-} input_t;
-
-static void init_input(input_t *in, const char *fname)
-{
-    in->file = fopen(fname, "r");
-    in->buf = (char*) malloc(SIZE + YYMAXFILL);
-    in->lim = in->buf + SIZE;
-    in->cur = in->lim;
-    in->mar = in->lim;
-    in->tok = in->lim;
-    /*!stags:re2c format = "in->@@ = 0;\n"; */
-    in->eof = 0;
-}
-
-static void free_input(input_t *in)
-{
-    free(in->buf);
-    fclose(in->file);
-}
-
-static int fill(input_t *in, size_t need)
-{
-    size_t free;
-
-    if (in->eof) return 1;
-
-    free = in->tok - in->buf;
-    if (free < need) return 2;
-
-    memmove(in->buf, in->tok, in->lim - in->tok);
-    in->lim -= free;
-    in->cur -= free;
-    in->mar -= free;
-    in->tok -= free;
-    /*!stags:re2c format = "if (in->@@) in->@@ -= free;\n"; */
-    in->lim += fread(in->lim, 1, free, in->file);
-    if (in->lim < in->buf + SIZE) {
-        in->eof = 1;
-        memset(in->lim, 0, YYMAXFILL);
-        in->lim += YYMAXFILL;
-    }
-    return 0;
-}
-
-static int lex(input_t *in, long *count, long *total)
-{
-    const char
-        *s1, *u1, *h1, *r1, *p1, *q1, *f1,
-        *s2, *u2, *h2, *r2, *p2, *q2, *f2;
-    long c, t;
-
-    c = 0; t = 0;
-loop:
-    in->tok = in->cur;
-/*!re2c
-
-    re2c:define:YYCTYPE = char;
-    re2c:define:YYCURSOR = in->cur;
-    re2c:define:YYMARKER = in->mar;
-    re2c:define:YYLIMIT = in->lim;
-    re2c:define:YYFILL = "if (fill(in, @@) != 0) return 2;";
-    re2c:define:YYFILL:naked = 1;
-    re2c:flags:tags = 1;
-    re2c:tags:expression = "in->@@";
-
-    end       = "\x00";
-    eol       = "\n";
-    char      = [-._~%!$&'()*+,;=a-zA-Z0-9];
-    scheme    = @s1 [-+.a-zA-Z0-9]+ @s2;
-    userinfo  = @u1 (char | [:])+ @u2;
-    host      = @h1 (char | [:[\]])+ @h2;
-    port      = @r1 [0-9]* @r2;
-    path      = @p1 (char | [:@/])* @p2;
-    query     = @q1 (char | [:@?/])* @q2;
-    fragment  = @f1 (char | [:@?/])* @f2;
-    uri       = scheme ":" ("//" (userinfo "@")? host (":" port)?)? path ("?" query)? ("#" fragment)?;
-
-    *   { return 1; }
-    end { *count = c; *total = t; return 0; }
-    eol { goto loop; }
-    uri {
-#ifndef VERIFY
-        c += 1;
-        t += (s2 - s1) + (u2 - u1) + (h2 - h1)
-            + (r2 - r1) + (p2 - p1) + (q2 - q1) + (f2 - f1);
-#else
-                printf("%.*s:", (int)(s2 - s1), s1);
-        if (h1) printf("//");
-        if (u1) printf("%.*s@", (int)(u2 - u1), u1);
-        if (h1) printf("%.*s",  (int)(h2 - h1), h1);
-        if (r1) printf(":%.*s", (int)(r2 - r1), r1);
-                printf("%.*s",  (int)(p2 - p1), p1);
-        if (q1) printf("?%.*s", (int)(q2 - q1), q1);
-        if (f1) printf("#%.*s", (int)(f2 - f1), f1);
-        printf("\n");
-#endif
-        goto loop;
-    }
-*/
-}
-
-int main(int argc, char **argv)
-{
-    long count, total;
-    input_t in;
-    init_input(&in, argv[1]);
-
-    switch (lex(&in, &count, &total)) {
-        case 0: fprintf(stderr, "ok, parsed %ld URIs, peeked %ld chars\n", count, total); break;
-        case 1: fprintf(stderr, "syntax error\n"); break;
-        case 2: fprintf(stderr, "yyfill failed\n"); break;
-        default: fprintf(stderr, "panic!\n"); break;
-    }
-
-    free_input(&in);
-    return 0;
-}

benchmarks/lookahead_tdfa/uri/simple/uri_simple_notags.re

@@ -1,116 +0,0 @@
-#include <stddef.h>
-#include <stdlib.h>
-#include <stdio.h>
-#include <string.h>
-
-/*!max:re2c*/
-static const size_t SIZE = 4096;
-
-typedef struct {
-    FILE *file;
-    char *buf;
-    char *lim;
-    char *cur;
-    char *mar;
-    char *tok;
-    int eof;
-} input_t;
-
-static void init_input(input_t *in, const char *fname)
-{
-    in->file = fopen(fname, "r");
-    in->buf = (char*) malloc(SIZE + YYMAXFILL);
-    in->lim = in->buf + SIZE;
-    in->cur = in->lim;
-    in->mar = in->lim;
-    in->tok = in->lim;
-    in->eof = 0;
-}
-
-static void free_input(input_t *in)
-{
-    free(in->buf);
-    fclose(in->file);
-}
-
-static int fill(input_t *in, size_t need)
-{
-    size_t free;
-
-    if (in->eof) return 1;
-
-    free = in->tok - in->buf;
-    if (free < need) return 2;
-
-    memmove(in->buf, in->tok, in->lim - in->tok);
-    in->lim -= free;
-    in->cur -= free;
-    in->mar -= free;
-    in->tok -= free;
-    in->lim += fread(in->lim, 1, free, in->file);
-    if (in->lim < in->buf + SIZE) {
-        in->eof = 1;
-        memset(in->lim, 0, YYMAXFILL);
-        in->lim += YYMAXFILL;
-    }
-    return 0;
-}
-
-static int lex(input_t *in, long *count, long *total)
-{
-    long c, t;
-
-    c = 0; t = 0;
-loop:
-    in->tok = in->cur;
-/*!re2c
-    re2c:define:YYCTYPE = char;
-    re2c:define:YYCURSOR = in->cur;
-    re2c:define:YYMARKER = in->mar;
-    re2c:define:YYLIMIT = in->lim;
-    re2c:define:YYFILL = "if (fill(in, @@) != 0) return 2;";
-    re2c:define:YYFILL:naked = 1;
-
-    end       = "\x00";
-    eol       = "\n";
-    char      = [-._~%!$&'()*+,;=a-zA-Z0-9];
-    scheme    = [-+.a-zA-Z0-9]+;
-    userinfo  = (char | [:])+;
-    host      = (char | [:[\]])+;
-    port      = [0-9]*;
-    path      = (char | [:@/])*;
-    query     = (char | [:@?/])*;
-    fragment  = (char | [:@?/])*;
-    uri       = scheme ":" ("//" (userinfo "@")? host (":" port)?)? path ("?" query)? ("#" fragment)?;
-
-    *   { return 1; }
-    end { *count = c; *total = t; return 0; }
-    eol { goto loop; }
-    uri {
-#ifndef VERIFY
-        c += 1;
-        t += in->cur - in->tok;
-#else
-        printf("%.*s\n", (int)(in->cur - in->tok), in->tok);
-#endif
-        goto loop;
-    }
-*/
-}
-
-int main(int argc, char **argv)
-{
-    long count, total;
-    input_t in;
-    init_input(&in, argv[1]);
-
-    switch (lex(&in, &count, &total)) {
-        case 0: fprintf(stderr, "ok, parsed %ld URIs, peeked %ld chars\n", count, total); break;
-        case 1: fprintf(stderr, "syntax error\n"); break;
-        case 2: fprintf(stderr, "yyfill failed\n"); break;
-        default: fprintf(stderr, "panic!\n"); break;
-    }
-
-    free_input(&in);
-    return 0;
-}

benchmarks/submatch_dfa_aot/CMakeLists.txt

@@ -125,8 +125,15 @@ foreach(bench ${BENCHMARKS})
 endforeach()
 
 # re2c
+file(MAKE_DIRECTORY "${ENG_DIR}/re2c/")
 file(MAKE_DIRECTORY "${GEN_DIR}/re2c/")
 set(RE2C "${CMAKE_BINARY_DIR}/re2c")
+set(RE2C3 "${ENG_DIR}/re2c/re2c3")
+add_custom_command(
+    OUTPUT "${RE2C3}"
+    COMMAND "${CMAKE_CURRENT_SOURCE_DIR}/engines/re2c/getre2c3.sh"
+    WORKING_DIRECTORY "${ENG_DIR}/re2c/"
+)
 set(RE2C_FLAGS "--reusable" "--tags" "--no-generation-date" "--no-version")
 set(COMMON_RE2C
     "${SRC_DIR}/re2c/common.re"
@@ -135,28 +142,44 @@ set(COMMON_RE2C
     "${SRC_DIR}/re2c/include-eof/fill.re"
     "${SRC_DIR}/re2c/include-eof/fill_email.re"
 )
+# deprecated algorithms were removed after re2c-3.0
+set(DEPRECATED_ALGS "tdfa0" "stadfa")
 foreach(bench ${BENCHMARKS})
 foreach(eof "" "-eof")
 foreach(alg "tdfa1" "tdfa0" "stadfa")
     set(src_file "${SRC_DIR}/re2c/${bench}.re")
     set(gen_file "${GEN_DIR}/re2c/${bench}${eof}-${alg}.c")
     set(pregen_file "${PREGEN_DIR}/re2c/${bench}${eof}-${alg}.c")
-    # always regenerate re2c benchmarks (regardless of RE2C_REGEN_BENCHMARKS)
-    file(RELATIVE_PATH rel_src_file "${CMAKE_CURRENT_BINARY_DIR}" "${src_file}")
-    file(RELATIVE_PATH rel_gen_file "${CMAKE_CURRENT_BINARY_DIR}" "${gen_file}")
-    file(RELATIVE_PATH rel_inc_path "${CMAKE_CURRENT_BINARY_DIR}" "${SRC_DIR}/re2c/include${eof}")
-    set(re2c_flags ${RE2C_FLAGS} "-I" "${rel_inc_path}")
-    if("${alg}" STREQUAL "tdfa0")
-        set(re2c_flags ${re2c_flags} "--no-lookahead")
-    elseif("${alg}" STREQUAL "stadfa")
-        set(re2c_flags ${re2c_flags} "--stadfa")
+    # always regenerate re2c benchmarks, except for deprecated algorithms
+    if(RE2C_REGEN_BENCHMARKS OR NOT (alg IN_LIST DEPRECATED_ALGS))
+        # for deprecated algorithms use re2c-3.0
+        if(alg IN_LIST DEPRECATED_ALGS)
+            set(re2c_for_gen "${RE2C3}")
+        else()
+            set(re2c_for_gen "${RE2C}")
+        endif()
+        file(RELATIVE_PATH rel_src_file "${CMAKE_CURRENT_BINARY_DIR}" "${src_file}")
+        file(RELATIVE_PATH rel_gen_file "${CMAKE_CURRENT_BINARY_DIR}" "${gen_file}")
+        file(RELATIVE_PATH rel_inc_path "${CMAKE_CURRENT_BINARY_DIR}" "${SRC_DIR}/re2c/include${eof}")
+        set(re2c_flags ${RE2C_FLAGS} "-I" "${rel_inc_path}")
+        if("${alg}" STREQUAL "tdfa0")
+            set(re2c_flags ${re2c_flags} "--no-lookahead")
+        elseif("${alg}" STREQUAL "stadfa")
+            set(re2c_flags ${re2c_flags} "--stadfa")
+        endif()
+        add_custom_command(
+            OUTPUT "${gen_file}"
+            COMMAND "${re2c_for_gen}" ${re2c_flags} "${rel_src_file}" -o "${rel_gen_file}"
+            COMMAND "${CMAKE_COMMAND}" -E copy_if_different "${gen_file}" "${pregen_file}"
+            DEPENDS "${src_file}" ${COMMON_RE2C} "${re2c_for_gen}"
+        )
+    else()
+        add_custom_command(
+            OUTPUT "${gen_file}"
+            COMMAND "${CMAKE_COMMAND}" -E copy "${pregen_file}" "${gen_file}"
+            DEPENDS "${pregen_file}"
+        )
     endif()
-    add_custom_command(
-        OUTPUT "${gen_file}"
-        COMMAND "${RE2C}" ${re2c_flags} "${rel_src_file}" -o "${rel_gen_file}"
-        COMMAND "${CMAKE_COMMAND}" -E copy_if_different "${gen_file}" "${pregen_file}"
-        DEPENDS "${src_file}" ${COMMON_RE2C} "${RE2C}"
-    )
     list(APPEND GEN "${gen_file}")
 endforeach()
 endforeach()

benchmarks/submatch_dfa_aot/Makefile.am

@@ -7,6 +7,7 @@ DAT_DIR = data
 CFLAGS = -O3 -I $(SRC_DIR)
 RAGEL = $(ENG_DIR)/ragel/ragel7
 KLEENEX = $(ENG_DIR)/kleenex/kexc
+RE2C3 = $(ENG_DIR)/re2c/re2c3
 RE2C = $(top_builddir)/re2c
 RE2C_FLAGS_COMMON = --reusable --tags --no-generation-date --no-version
 RE2C_FLAGS     = $(RE2C_FLAGS_COMMON) -I $(SRC_DIR)/re2c/include
@@ -60,21 +61,23 @@ BIN_RAGEL = $(patsubst $(GEN_DIR)%.c, $(BIN_DIR)%, $(GEN_RAGEL))
 
 COMMON_RAGEL = $(COMMON_SRC) $(SRC_DIR)/ragel/common.c
 
-GEN_RE2C_TDFA1      = $(patsubst %, $(GEN_DIR)/re2c/%-tdfa1.c,      $(BENCHMARKS))
-GEN_RE2C_TDFA0      = $(patsubst %, $(GEN_DIR)/re2c/%-tdfa0.c,      $(BENCHMARKS))
-GEN_RE2C_STADFA     = $(patsubst %, $(GEN_DIR)/re2c/%-stadfa.c,     $(BENCHMARKS))
-GEN_RE2C_EOF_TDFA1  = $(patsubst %, $(GEN_DIR)/re2c/%-eof-tdfa1.c,  $(BENCHMARKS))
-GEN_RE2C_EOF_TDFA0  = $(patsubst %, $(GEN_DIR)/re2c/%-eof-tdfa0.c,  $(BENCHMARKS))
-GEN_RE2C_EOF_STADFA = $(patsubst %, $(GEN_DIR)/re2c/%-eof-stadfa.c, $(BENCHMARKS))
+GEN_RE2C_TDFA1     = $(patsubst %, $(GEN_DIR)/re2c/%-tdfa1.c,     $(BENCHMARKS))
+GEN_RE2C_EOF_TDFA1 = $(patsubst %, $(GEN_DIR)/re2c/%-eof-tdfa1.c, $(BENCHMARKS))
 GEN_RE2C = \
     $(GEN_RE2C_TDFA1) \
-    $(GEN_RE2C_TDFA0) \
-    $(GEN_RE2C_STADFA) \
-    $(GEN_RE2C_EOF_TDFA1) \
-    $(GEN_RE2C_EOF_TDFA0) \
-    $(GEN_RE2C_EOF_STADFA)
+    $(GEN_RE2C_EOF_TDFA1)
 
-BIN_RE2C = $(patsubst $(GEN_DIR)%.c, $(BIN_DIR)%, $(GEN_RE2C))
+GEN_RE2C3_TDFA0      = $(patsubst %, $(GEN_DIR)/re2c/%-tdfa0.c,      $(BENCHMARKS))
+GEN_RE2C3_EOF_TDFA0  = $(patsubst %, $(GEN_DIR)/re2c/%-eof-tdfa0.c,  $(BENCHMARKS))
+GEN_RE2C3_STADFA     = $(patsubst %, $(GEN_DIR)/re2c/%-stadfa.c,     $(BENCHMARKS))
+GEN_RE2C3_EOF_STADFA = $(patsubst %, $(GEN_DIR)/re2c/%-eof-stadfa.c, $(BENCHMARKS))
+GEN_RE2C3 = \
+    $(GEN_RE2C_TDFA0) \
+    $(GEN_RE2C_EOF_TDFA0)  \
+    $(GEN_RE2C3_STADFA) \
+    $(GEN_RE2C3_EOF_STADFA)
+
+BIN_RE2C = $(patsubst $(GEN_DIR)%.c, $(BIN_DIR)%, $(GEN_RE2C) $(GEN_RE2C3))
 
 COMMON_RE2C = \
     $(COMMON_SRC) \
@@ -97,7 +100,7 @@ GEN_KLEENEX = $(patsubst %, $(GEN_DIR)/kleenex/%.c, \
 
 BIN_KLEENEX = $(patsubst $(GEN_DIR)%.c, $(BIN_DIR)%, $(GEN_KLEENEX))
 
-GEN = $(GEN_RAGEL) $(GEN_RE2C) $(GEN_KLEENEX)
+GEN = $(GEN_RAGEL) $(GEN_RE2C) $(GEN_RE2C3) $(GEN_KLEENEX)
 
 BIN_ = $(BIN_RAGEL) $(BIN_RE2C) $(BIN_KLEENEX)
 BIN_GCC   = $(patsubst %, %-gcc, $(BIN_))
@@ -113,7 +116,7 @@ DAT = $(patsubst %, $(DAT_DIR)/%/big, \
 
 all-local: $(BIN) $(DAT)
 
-# always regenerate re2c benchmarks
+# always regenerate re2c benchmarks (except for deprecated algorithms)
 
 $(GEN_RE2C_TDFA1): $(GEN_DIR)/%-tdfa1.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C)
 	$(AM_V_at)mkdir -p $(@D)
@@ -125,29 +128,31 @@ $(GEN_RE2C_EOF_TDFA1): $(GEN_DIR)/%-eof-tdfa1.c: $(SRC_DIR)/%.re $(COMMON_RE2C)
 	$(AM_V_GEN)$(RE2C) $(RE2C_FLAGS_EOF) $< -o $@
 	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
 
-$(GEN_RE2C_TDFA0): $(GEN_DIR)/%-tdfa0.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C)
-	$(AM_V_at)mkdir -p $(@D)
-	$(AM_V_GEN)$(RE2C) $(RE2C_FLAGS) --no-lookahead $< -o $@
-	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
+# optionally regenerate ragel, kleenex and re2c-3.0 benchmarks
+# (TDFA(0) and staDFA algorithms were removed in re2c versions after 3.0)
 
-$(GEN_RE2C_EOF_TDFA0): $(GEN_DIR)/%-eof-tdfa0.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C)
-	$(AM_V_at)mkdir -p $(@D)
-	$(AM_V_GEN)$(RE2C) $(RE2C_FLAGS_EOF) --no-lookahead $< -o $@
-	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
-
-$(GEN_RE2C_STADFA): $(GEN_DIR)/%-stadfa.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C)
-	$(AM_V_at)mkdir -p $(@D)
-	$(AM_V_GEN)$(RE2C) $(RE2C_FLAGS) --stadfa $< -o $@
-	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
-
-$(GEN_RE2C_EOF_STADFA): $(GEN_DIR)/%-eof-stadfa.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C)
-	$(AM_V_at)mkdir -p $(@D)
-	$(AM_V_GEN)$(RE2C) $(RE2C_FLAGS_EOF) --stadfa $< -o $@
-	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
-
-# optionally regenerate ragel and kleenex benchmarks
 if REGEN_BENCHMARKS
 
+$(GEN_RE2C3_TDFA0): $(GEN_DIR)/%-tdfa0.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C3)
+	$(AM_V_at)mkdir -p $(@D)
+	$(AM_V_GEN)$(RE2C3) $(RE2C_FLAGS) --no-lookahead $< -o $@
+	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
+
+$(GEN_RE2C3_EOF_TDFA0): $(GEN_DIR)/%-eof-tdfa0.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C3)
+	$(AM_V_at)mkdir -p $(@D)
+	$(AM_V_GEN)$(RE2C3) $(RE2C_FLAGS_EOF) --no-lookahead $< -o $@
+	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
+
+$(GEN_RE2C3_STADFA): $(GEN_DIR)/%-stadfa.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C3)
+	$(AM_V_at)mkdir -p $(@D)
+	$(AM_V_GEN)$(RE2C3) $(RE2C_FLAGS) --stadfa $< -o $@
+	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
+
+$(GEN_RE2C3_EOF_STADFA): $(GEN_DIR)/%-eof-stadfa.c: $(SRC_DIR)/%.re $(COMMON_RE2C) $(RE2C3)
+	$(AM_V_at)mkdir -p $(@D)
+	$(AM_V_GEN)$(RE2C3) $(RE2C_FLAGS_EOF) --stadfa $< -o $@
+	$(AM_V_at)if ! cmp -s $@ $(PREGEN_DIR)/re2c/$(@F) ; then cp -f $@ $(PREGEN_DIR)/re2c/$(@F) ; fi
+
 $(GEN_RAGEL): $(GEN_DIR)/%.c: $(SRC_DIR)/%.rl $(COMMON_RAGEL) $(RAGEL)
 	$(AM_V_at)mkdir -p $(@D)
 	$(AM_V_GEN)$(RAGEL) -G2 $< -o $@
@@ -160,7 +165,7 @@ $(GEN_KLEENEX): $(GEN_DIR)/%.c: $(SRC_DIR)/%.kex $(KLEENEX)
 
 else
 
-$(GEN_RAGEL) $(GEN_KLEENEX): $(GEN_DIR)/%: $(PREGEN_DIR)/%
+$(GEN_RAGEL) $(GEN_RE2C3) $(GEN_KLEENEX): $(GEN_DIR)/%: $(PREGEN_DIR)/%
 	$(AM_V_at)mkdir -p $(@D)
 	$(AM_V_GEN)cp -f $< $@
 
@@ -188,5 +193,9 @@ $(RAGEL):
 	$(AM_V_at)mkdir -p $(@D) && cp $(srcdir)/engines/ragel/getragel7.sh $(@D)
 	$(AM_V_GEN)( cd $(@D) && ./getragel7.sh )
 
+$(RE2C3):
+	$(AM_V_at)mkdir -p $(@D) && cp $(srcdir)/engines/re2c/getre2c3.sh $(@D)
+	$(AM_V_GEN)( cd $(@D) && ./getre2c3.sh )
+
 clean-local:
 	$(AM_V_at)rm -f $(GEN) $(OBJ) $(BIN) $(DAT)

benchmarks/submatch_dfa_aot/engines/kleenex/getkleenex.sh

@@ -1,11 +1,11 @@
-set -e
+set -ex
 
 rm -rf kleenexlang
 git clone --recursive https://github.com/diku-kmc/kleenexlang.git
 cd kleenexlang
-cabal v1-sandbox init && cabal v1-sandbox add-source regexps-syntax
-cabal v1-install --dependencies-only
-cabal v1-configure && cabal v1-build
+mkdir "install"
+cabal build
+cabal install --install-method=copy --installdir="$(pwd)/install"
 cd ..
 rm -f kexc
-ln -s kleenexlang/dist/build/kexc/kexc kexc
+ln -s "kleenexlang/install/kexc" kexc

benchmarks/submatch_dfa_aot/engines/ragel/getragel7.sh

@@ -1,4 +1,4 @@
-set -e
+set -ex
 
 if [ -e ragel7 ] ; then
     echo "ragel7 is already installed, quitting"
@@ -10,7 +10,10 @@ RAGEL_DIR=ragel-7.0.4
 
 wget http://www.colm.net/files/colm/"$COLM_DIR".tar.gz -O - | tar zx \
     && cd "$COLM_DIR" \
-    && ./configure --prefix=$(pwd)/install \
+    && sed -i 's|^#!/bin/bash|#!/usr/bin/env bash|' src/colm-wrap.sh \
+    && ./configure \
+        --prefix=$(pwd)/install \
+        --disable-manual \
     && make -j$(nproc) \
     && make install \
     && cd ..

benchmarks/submatch_dfa_aot/engines/re2c/getre2c3.sh

@@ -0,0 +1,22 @@
+set -ex
+
+if [ -e re2c3 ] ; then
+    echo "re2c3 is already installed, quitting"
+    exit 0
+fi
+
+RE2C_DIR=re2c-3.0
+
+wget https://github.com/skvadrik/re2c/releases/download/3.0/re2c-3.0.tar.xz \
+    && tar xf re2c-3.0.tar.xz \
+    && rm re2c-3.0.tar.xz \
+    && cd "$RE2C_DIR" \
+    && ./configure \
+        --disable-golang \
+        --disable-rust \
+        --prefix=$(pwd)/install \
+    && make -j$(nproc) \
+    && make install \
+    && cd ..
+
+ln -s "$RE2C_DIR"/install/bin/re2c re2c3

benchmarks/submatch_dfa_aot/pregen/re2c/04__apache_log-eof-tdfa1.c

@@ -1,5 +1,6 @@
 /* Generated by re2c */
 #line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 #include <assert.h>
 #include <stdlib.h>
 #include "common.h"
@@ -24,7 +25,9 @@ typedef struct {
     char *cur;
     char *mar;
     char *tok;
-    char *yyt1;
+    
+#line 30 "gen/re2c/04__apache_log-eof-tdfa1.c"
+char *yyt1;
 char *yyt10;
 char *yyt11;
 char *yyt12;
@@ -39,8 +42,12 @@ char *yyt6;
 char *yyt7;
 char *yyt8;
 char *yyt9;
+#line 34 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
     
+#line 49 "gen/re2c/04__apache_log-eof-tdfa1.c"
+#line 35 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
     taglistpool_t tlp;
     int eof;
 } input_t;
@@ -49,6 +56,9 @@ static inline void taglistpool_clear(taglistpool_t *tlp, input_t *in)
 {
     tlp->next = tlp->head;
     
+#line 60 "gen/re2c/04__apache_log-eof-tdfa1.c"
+#line 43 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
 }
 
 static inline void taglistpool_init(taglistpool_t *tlp)
@@ -86,6 +96,7 @@ static inline void taglist(taglist_t **ptl, const char *b, const char *t, taglis
     *ptl = tl;
 }
 
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/include-eof/fill.re"
 #line 6 "../../../benchmarks/submatch_dfa_aot/src/re2c/include-eof/fill.re"
 
 
@@ -102,7 +113,9 @@ static inline int fill(input_t *in)
     in->cur -= free;
     in->mar -= free;
     in->tok -= free;
-    if (in->yyt1) in->yyt1 -= free;
+    
+#line 118 "gen/re2c/04__apache_log-eof-tdfa1.c"
+if (in->yyt1) in->yyt1 -= free;
 if (in->yyt10) in->yyt10 -= free;
 if (in->yyt11) in->yyt11 -= free;
 if (in->yyt12) in->yyt12 -= free;
@@ -117,6 +130,7 @@ if (in->yyt6) in->yyt6 -= free;
 if (in->yyt7) in->yyt7 -= free;
 if (in->yyt8) in->yyt8 -= free;
 if (in->yyt9) in->yyt9 -= free;
+#line 21 "../../../benchmarks/submatch_dfa_aot/src/re2c/include-eof/fill.re"
 
 
     in->lim += fread(in->lim, 1, free, stdin);
@@ -136,7 +150,9 @@ static inline void init_input(input_t *in)
     in->cur = in->lim;
     in->mar = in->lim;
     in->tok = in->lim;
-    in->yyt1 = 0;
+    
+#line 155 "gen/re2c/04__apache_log-eof-tdfa1.c"
+in->yyt1 = 0;
 in->yyt10 = 0;
 in->yyt11 = 0;
 in->yyt12 = 0;
@@ -151,12 +167,17 @@ in->yyt6 = 0;
 in->yyt7 = 0;
 in->yyt8 = 0;
 in->yyt9 = 0;
+#line 40 "../../../benchmarks/submatch_dfa_aot/src/re2c/include-eof/fill.re"
 
     
+#line 174 "gen/re2c/04__apache_log-eof-tdfa1.c"
+#line 41 "../../../benchmarks/submatch_dfa_aot/src/re2c/include-eof/fill.re"
+
     taglistpool_init(&in->tlp);
     in->eof = 0;
     fill(in);
 }
+#line 81 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
 
 static inline void free_input(input_t *in)
@@ -190,6 +211,7 @@ int main(int argc, char **argv)
     POST;
     return 0;
 }
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 
 
 static int lex(input_t *in, Output *out)
@@ -201,477 +223,477 @@ static int lex(input_t *in, Output *out)
 loop:
     in->tok = in->cur;
 
-#line 205 "gen/re2c/04__apache_log-eof-tdfa1.c"
+#line 227 "gen/re2c/04__apache_log-eof-tdfa1.c"
 {
 	char yych;
-yyFillLabel3_0:
+yyFillLabel0:
 	yych = *in->cur;
 	switch (yych) {
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		in->yyt1 = in->cur;
-		goto yy4;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_0;
-			goto yy47;
-		}
-		goto yy2;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			in->yyt1 = in->cur;
+			goto yy3;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel0;
+				goto yy32;
+			}
+			goto yy1;
 	}
-yy2:
+yy1:
 	++in->cur;
-yy3:
+yy2:
 #line 33 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 	{ return 1; }
-#line 236 "gen/re2c/04__apache_log-eof-tdfa1.c"
-yy4:
+#line 258 "gen/re2c/04__apache_log-eof-tdfa1.c"
+yy3:
 	in->mar = ++in->cur;
-yyFillLabel3_1:
+yyFillLabel1:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt2 = in->cur;
-		goto yy5;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy8;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_1;
-		}
-		goto yy3;
+		case '\t':
+		case ' ':
+			in->yyt2 = in->cur;
+			goto yy4;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy6;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel1;
+			}
+			goto yy2;
+	}
+yy4:
+	++in->cur;
+yyFillLabel2:
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy4;
+		case '-': goto yy7;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel2;
+			}
+			goto yy5;
 	}
 yy5:
+	in->cur = in->mar;
+	goto yy2;
+yy6:
 	++in->cur;
-yyFillLabel3_2:
+yyFillLabel3:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy5;
-	case '-':	goto yy10;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_2;
-		}
-		goto yy7;
+		case '\t':
+		case ' ':
+			in->yyt2 = in->cur;
+			goto yy4;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy6;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel3;
+			}
+			goto yy5;
 	}
 yy7:
-	in->cur = in->mar;
-	goto yy3;
-yy8:
 	++in->cur;
-yyFillLabel3_3:
+yyFillLabel4:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt2 = in->cur;
-		goto yy5;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy8;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_3;
-		}
-		goto yy7;
+		case '\t':
+		case ' ':
+			in->yyt3 = in->cur;
+			goto yy8;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel4;
+			}
+			goto yy5;
+	}
+yy8:
+	++in->cur;
+yyFillLabel5:
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy8;
+		case '-': goto yy9;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel5;
+			}
+			goto yy5;
+	}
+yy9:
+	++in->cur;
+yyFillLabel6:
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt4 = in->cur;
+			goto yy10;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel6;
+			}
+			goto yy5;
 	}
 yy10:
 	++in->cur;
-yyFillLabel3_4:
+yyFillLabel7:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt3 = in->cur;
-		goto yy11;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_4;
-		}
-		goto yy7;
+		case '\t':
+		case ' ': goto yy10;
+		case '[':
+			in->yyt5 = in->cur;
+			goto yy11;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel7;
+			}
+			goto yy5;
 	}
 yy11:
 	++in->cur;
-yyFillLabel3_5:
+yyFillLabel8:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy11;
-	case '-':	goto yy13;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_5;
-		}
-		goto yy7;
+		case 0x00:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel8;
+				goto yy5;
+			}
+			goto yy12;
+		case ']': goto yy5;
+		default: goto yy13;
 	}
-yy13:
+yy12:
 	++in->cur;
-yyFillLabel3_6:
+yyFillLabel9:
 	yych = *in->cur;
+yy13:
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt4 = in->cur;
-		goto yy14;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_6;
-		}
-		goto yy7;
+		case '\n': goto yy5;
+		case ']': goto yy14;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel9;
+				goto yy5;
+			}
+			goto yy12;
 	}
 yy14:
 	++in->cur;
-yyFillLabel3_7:
+yyFillLabel10:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy14;
-	case '[':
-		in->yyt5 = in->cur;
-		goto yy16;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_7;
-		}
-		goto yy7;
+		case '\t':
+		case ' ':
+			in->yyt6 = in->cur;
+			goto yy15;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel10;
+			}
+			goto yy5;
+	}
+yy15:
+	++in->cur;
+yyFillLabel11:
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy15;
+		case '"':
+			in->yyt7 = in->cur;
+			goto yy16;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel11;
+			}
+			goto yy5;
 	}
 yy16:
 	++in->cur;
-yyFillLabel3_8:
+yyFillLabel12:
 	yych = *in->cur;
 	switch (yych) {
-	case 0x00:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_8;
-			goto yy7;
-		}
-		goto yy17;
-	case ']':	goto yy7;
-	default:	goto yy18;
+		case 0x00:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel12;
+				goto yy5;
+			}
+			goto yy17;
+		case '"': goto yy5;
+		default: goto yy18;
 	}
 yy17:
 	++in->cur;
-yyFillLabel3_9:
+yyFillLabel13:
 	yych = *in->cur;
 yy18:
 	switch (yych) {
-	case '\n':	goto yy7;
-	case ']':	goto yy19;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_9;
-			goto yy7;
-		}
-		goto yy17;
+		case '\n': goto yy5;
+		case '"': goto yy19;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel13;
+				goto yy5;
+			}
+			goto yy17;
 	}
 yy19:
 	++in->cur;
-yyFillLabel3_10:
+yyFillLabel14:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt6 = in->cur;
-		goto yy20;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_10;
-		}
-		goto yy7;
+		case '\t':
+		case ' ':
+			in->yyt8 = in->cur;
+			goto yy20;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel14;
+			}
+			goto yy5;
 	}
 yy20:
 	++in->cur;
-yyFillLabel3_11:
+yyFillLabel15:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy20;
-	case '"':
-		in->yyt7 = in->cur;
-		goto yy22;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_11;
-		}
-		goto yy7;
+		case '\t':
+		case ' ': goto yy20;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			in->yyt9 = in->cur;
+			goto yy21;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel15;
+			}
+			goto yy5;
+	}
+yy21:
+	++in->cur;
+yyFillLabel16:
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt10 = in->cur;
+			goto yy22;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy21;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel16;
+			}
+			goto yy5;
 	}
 yy22:
 	++in->cur;
-yyFillLabel3_12:
+yyFillLabel17:
 	yych = *in->cur;
 	switch (yych) {
-	case 0x00:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_12;
-			goto yy7;
-		}
-		goto yy23;
-	case '"':	goto yy7;
-	default:	goto yy24;
+		case '\t':
+		case ' ': goto yy22;
+		case '-':
+			in->yyt11 = in->cur;
+			goto yy23;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			in->yyt11 = in->cur;
+			goto yy24;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel17;
+			}
+			goto yy5;
 	}
 yy23:
 	++in->cur;
-yyFillLabel3_13:
+yyFillLabel18:
 	yych = *in->cur;
-yy24:
 	switch (yych) {
-	case '\n':	goto yy7;
-	case '"':	goto yy25;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_13;
-			goto yy7;
-		}
-		goto yy23;
+		case '\t':
+		case ' ':
+			in->yyt12 = in->cur;
+			goto yy25;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel18;
+			}
+			goto yy5;
+	}
+yy24:
+	++in->cur;
+yyFillLabel19:
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt12 = in->cur;
+			goto yy25;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy24;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel19;
+			}
+			goto yy5;
 	}
 yy25:
 	++in->cur;
-yyFillLabel3_14:
+yyFillLabel20:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt8 = in->cur;
-		goto yy26;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_14;
-		}
-		goto yy7;
+		case '\t':
+		case ' ': goto yy25;
+		case '"':
+			in->yyt13 = in->cur;
+			goto yy26;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel20;
+			}
+			goto yy5;
 	}
 yy26:
 	++in->cur;
-yyFillLabel3_15:
+yyFillLabel21:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy26;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		in->yyt9 = in->cur;
-		goto yy28;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_15;
-		}
-		goto yy7;
+		case '\n': goto yy5;
+		case '"': goto yy27;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel21;
+				goto yy5;
+			}
+			goto yy26;
+	}
+yy27:
+	++in->cur;
+yyFillLabel22:
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt14 = in->cur;
+			goto yy28;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel22;
+			}
+			goto yy5;
 	}
 yy28:
 	++in->cur;
-yyFillLabel3_16:
+yyFillLabel23:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt10 = in->cur;
-		goto yy30;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy28;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_16;
-		}
-		goto yy7;
+		case '\t':
+		case ' ': goto yy28;
+		case '"':
+			in->yyt15 = in->cur;
+			goto yy29;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel23;
+			}
+			goto yy5;
+	}
+yy29:
+	++in->cur;
+yyFillLabel24:
+	yych = *in->cur;
+	switch (yych) {
+		case '\n': goto yy5;
+		case '"': goto yy30;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel24;
+				goto yy5;
+			}
+			goto yy29;
 	}
 yy30:
 	++in->cur;
-yyFillLabel3_17:
+yyFillLabel25:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy30;
-	case '-':
-		in->yyt11 = in->cur;
-		goto yy32;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		in->yyt11 = in->cur;
-		goto yy33;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_17;
-		}
-		goto yy7;
+		case '\n': goto yy31;
+		default:
+			if (in->lim <= in->cur) {
+				if (fill(in) == 0) goto yyFillLabel25;
+			}
+			goto yy5;
 	}
-yy32:
-	++in->cur;
-yyFillLabel3_18:
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt12 = in->cur;
-		goto yy35;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_18;
-		}
-		goto yy7;
-	}
-yy33:
-	++in->cur;
-yyFillLabel3_19:
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt12 = in->cur;
-		goto yy35;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy33;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_19;
-		}
-		goto yy7;
-	}
-yy35:
-	++in->cur;
-yyFillLabel3_20:
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':	goto yy35;
-	case '"':
-		in->yyt13 = in->cur;
-		goto yy37;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_20;
-		}
-		goto yy7;
-	}
-yy37:
-	++in->cur;
-yyFillLabel3_21:
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':	goto yy7;
-	case '"':	goto yy39;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_21;
-			goto yy7;
-		}
-		goto yy37;
-	}
-yy39:
-	++in->cur;
-yyFillLabel3_22:
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt14 = in->cur;
-		goto yy40;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_22;
-		}
-		goto yy7;
-	}
-yy40:
-	++in->cur;
-yyFillLabel3_23:
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':	goto yy40;
-	case '"':
-		in->yyt15 = in->cur;
-		goto yy42;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_23;
-		}
-		goto yy7;
-	}
-yy42:
-	++in->cur;
-yyFillLabel3_24:
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':	goto yy7;
-	case '"':	goto yy44;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_24;
-			goto yy7;
-		}
-		goto yy42;
-	}
-yy44:
-	++in->cur;
-yyFillLabel3_25:
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':	goto yy45;
-	default:
-		if (in->lim <= in->cur) {
-			if (fill(in) == 0) goto yyFillLabel3_25;
-		}
-		goto yy7;
-	}
-yy45:
+yy31:
 	++in->cur;
 	h1 = in->yyt1;
 	h2 = in->yyt2;
@@ -705,11 +727,11 @@ yy45:
         outc(out, '\n');
         goto loop;
     }
-#line 709 "gen/re2c/04__apache_log-eof-tdfa1.c"
-yy47:
+#line 731 "gen/re2c/04__apache_log-eof-tdfa1.c"
+yy32:
 #line 5 "../../../benchmarks/submatch_dfa_aot/src/re2c/include-eof/fill.re"
 	{ return 0; }
-#line 713 "gen/re2c/04__apache_log-eof-tdfa1.c"
+#line 735 "gen/re2c/04__apache_log-eof-tdfa1.c"
 }
 #line 47 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 

benchmarks/submatch_dfa_aot/pregen/re2c/04__apache_log-stadfa.c

@@ -1,5 +1,6 @@
 /* Generated by re2c */
 #line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 #include <assert.h>
 #include <stdlib.h>
 #include "common.h"
@@ -24,7 +25,9 @@ typedef struct {
     char *cur;
     char *mar;
     char *tok;
-    char *yyt1;
+    
+#line 30 "gen/re2c/04__apache_log-stadfa.c"
+char *yyt1;
 char *yyt10;
 char *yyt11;
 char *yyt12;
@@ -39,8 +42,12 @@ char *yyt6;
 char *yyt7;
 char *yyt8;
 char *yyt9;
+#line 34 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
     
+#line 49 "gen/re2c/04__apache_log-stadfa.c"
+#line 35 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
     taglistpool_t tlp;
     int eof;
 } input_t;
@@ -49,6 +56,9 @@ static inline void taglistpool_clear(taglistpool_t *tlp, input_t *in)
 {
     tlp->next = tlp->head;
     
+#line 60 "gen/re2c/04__apache_log-stadfa.c"
+#line 43 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
 }
 
 static inline void taglistpool_init(taglistpool_t *tlp)
@@ -86,10 +96,13 @@ static inline void taglist(taglist_t **ptl, const char *b, const char *t, taglis
     *ptl = tl;
 }
 
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 #line 4 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
+#line 104 "gen/re2c/04__apache_log-stadfa.c"
 #define YYMAXFILL 12
+#line 6 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
 static inline int fill(input_t *in, size_t need)
@@ -105,7 +118,9 @@ static inline int fill(input_t *in, size_t need)
     in->cur -= free;
     in->mar -= free;
     in->tok -= free;
-    if (in->yyt1) in->yyt1 -= free;
+    
+#line 123 "gen/re2c/04__apache_log-stadfa.c"
+if (in->yyt1) in->yyt1 -= free;
 if (in->yyt10) in->yyt10 -= free;
 if (in->yyt11) in->yyt11 -= free;
 if (in->yyt12) in->yyt12 -= free;
@@ -120,6 +135,7 @@ if (in->yyt6) in->yyt6 -= free;
 if (in->yyt7) in->yyt7 -= free;
 if (in->yyt8) in->yyt8 -= free;
 if (in->yyt9) in->yyt9 -= free;
+#line 21 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
     in->lim += fread(in->lim, 1, free, stdin);
@@ -140,7 +156,9 @@ static inline void init_input(input_t *in)
     in->cur = in->lim;
     in->mar = in->lim;
     in->tok = in->lim;
-    in->yyt1 = 0;
+    
+#line 161 "gen/re2c/04__apache_log-stadfa.c"
+in->yyt1 = 0;
 in->yyt10 = 0;
 in->yyt11 = 0;
 in->yyt12 = 0;
@@ -155,11 +173,16 @@ in->yyt6 = 0;
 in->yyt7 = 0;
 in->yyt8 = 0;
 in->yyt9 = 0;
+#line 41 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
     
+#line 180 "gen/re2c/04__apache_log-stadfa.c"
+#line 42 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
+
     taglistpool_init(&in->tlp);
     in->eof = 0;
 }
+#line 81 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
 
 static inline void free_input(input_t *in)
@@ -193,6 +216,7 @@ int main(int argc, char **argv)
     POST;
     return 0;
 }
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 
 
 static int lex(input_t *in, Output *out)
@@ -204,461 +228,461 @@ static int lex(input_t *in, Output *out)
 loop:
     in->tok = in->cur;
 
-#line 208 "gen/re2c/04__apache_log-stadfa.c"
+#line 232 "gen/re2c/04__apache_log-stadfa.c"
 {
 	char yych;
 	if ((in->lim - in->cur) < 8) if (fill(in, 8) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case 0x00:	goto yy2;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy6;
-	default:	goto yy4;
+		case 0x00: goto yy1;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy4;
+		default: goto yy2;
 	}
-yy2:
+yy1:
 	++in->cur;
 #line 3 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 	{ return 0; }
-#line 232 "gen/re2c/04__apache_log-stadfa.c"
-yy4:
+#line 256 "gen/re2c/04__apache_log-stadfa.c"
+yy2:
 	++in->cur;
-yy5:
+yy3:
 #line 33 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 	{ return 1; }
-#line 238 "gen/re2c/04__apache_log-stadfa.c"
-yy6:
+#line 262 "gen/re2c/04__apache_log-stadfa.c"
+yy4:
 	yych = *(in->mar = ++in->cur);
 	in->yyt1 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy7;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy9;
-	default:	goto yy5;
+		case '\t':
+		case ' ': goto yy5;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy7;
+		default: goto yy3;
 	}
-yy7:
+yy5:
 	yych = *++in->cur;
 	in->yyt2 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy11;
-	case '-':	goto yy13;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy8;
+		case '-': goto yy9;
+		default: goto yy6;
 	}
-yy8:
+yy6:
 	in->cur = in->mar;
-	goto yy5;
-yy9:
+	goto yy3;
+yy7:
 	++in->cur;
 	if ((in->lim - in->cur) < 7) if (fill(in, 7) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy7;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy9;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy5;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy7;
+		default: goto yy6;
 	}
-yy11:
+yy8:
 	++in->cur;
 	if ((in->lim - in->cur) < 6) if (fill(in, 6) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy11;
-	case '-':	goto yy13;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy8;
+		case '-': goto yy9;
+		default: goto yy6;
 	}
-yy13:
+yy9:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy14;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy10;
+		default: goto yy6;
 	}
-yy14:
+yy10:
 	yych = *++in->cur;
 	in->yyt3 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy15;
-	case '-':	goto yy17;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy11;
+		case '-': goto yy12;
+		default: goto yy6;
 	}
-yy15:
+yy11:
 	++in->cur;
 	if ((in->lim - in->cur) < 4) if (fill(in, 4) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy15;
-	case '-':	goto yy17;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy11;
+		case '-': goto yy12;
+		default: goto yy6;
+	}
+yy12:
+	yych = *++in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy13;
+		default: goto yy6;
+	}
+yy13:
+	yych = *++in->cur;
+	in->yyt4 = in->cur - 1;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy14;
+		case '[': goto yy15;
+		default: goto yy6;
+	}
+yy14:
+	++in->cur;
+	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy14;
+		case '[': goto yy15;
+		default: goto yy6;
+	}
+yy15:
+	yych = *++in->cur;
+	in->yyt5 = in->cur - 1;
+	switch (yych) {
+		case '\n':
+		case ']': goto yy6;
+		default: goto yy16;
+	}
+yy16:
+	++in->cur;
+	if ((in->lim - in->cur) < 4) if (fill(in, 4) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\n': goto yy6;
+		case ']': goto yy17;
+		default: goto yy16;
 	}
 yy17:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy18;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy18;
+		default: goto yy6;
 	}
 yy18:
 	yych = *++in->cur;
-	in->yyt4 = in->cur - 1;
+	in->yyt6 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy19;
-	case '[':	goto yy21;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy19;
+		case '"': goto yy20;
+		default: goto yy6;
 	}
 yy19:
 	++in->cur;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy19;
-	case '[':	goto yy21;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy19;
+		case '"': goto yy20;
+		default: goto yy6;
 	}
-yy21:
-	yych = *++in->cur;
-	in->yyt5 = in->cur - 1;
-	switch (yych) {
-	case '\n':
-	case ']':	goto yy8;
-	default:	goto yy22;
-	}
-yy22:
-	++in->cur;
-	if ((in->lim - in->cur) < 4) if (fill(in, 4) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':	goto yy8;
-	case ']':	goto yy24;
-	default:	goto yy22;
-	}
-yy24:
-	yych = *++in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':	goto yy25;
-	default:	goto yy8;
-	}
-yy25:
-	yych = *++in->cur;
-	in->yyt6 = in->cur - 1;
-	switch (yych) {
-	case '\t':
-	case ' ':	goto yy26;
-	case '"':	goto yy28;
-	default:	goto yy8;
-	}
-yy26:
-	++in->cur;
-	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':	goto yy26;
-	case '"':	goto yy28;
-	default:	goto yy8;
-	}
-yy28:
+yy20:
 	yych = *++in->cur;
 	in->yyt7 = in->cur - 1;
 	switch (yych) {
-	case '\n':
-	case '"':	goto yy8;
-	default:	goto yy29;
+		case '\n':
+		case '"': goto yy6;
+		default: goto yy21;
 	}
-yy29:
+yy21:
 	++in->cur;
 	if ((in->lim - in->cur) < 12) if (fill(in, 12) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\n':	goto yy8;
-	case '"':	goto yy31;
-	default:	goto yy29;
+		case '\n': goto yy6;
+		case '"': goto yy22;
+		default: goto yy21;
 	}
-yy31:
+yy22:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy32;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy23;
+		default: goto yy6;
 	}
-yy32:
+yy23:
 	yych = *++in->cur;
 	in->yyt8 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy33;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy35;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy24;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy25;
+		default: goto yy6;
 	}
-yy33:
+yy24:
 	++in->cur;
 	if ((in->lim - in->cur) < 10) if (fill(in, 10) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy33;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy35;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy24;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy25;
+		default: goto yy6;
 	}
-yy35:
+yy25:
 	yych = *++in->cur;
 	in->yyt9 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy36;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy37;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy26;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy27;
+		default: goto yy6;
 	}
-yy36:
+yy26:
 	yych = *++in->cur;
 	in->yyt10 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy39;
-	case '-':	goto yy41;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy42;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy28;
+		case '-': goto yy29;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy30;
+		default: goto yy6;
 	}
-yy37:
+yy27:
 	++in->cur;
 	if ((in->lim - in->cur) < 9) if (fill(in, 9) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy36;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy37;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy26;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy27;
+		default: goto yy6;
 	}
-yy39:
+yy28:
 	++in->cur;
 	if ((in->lim - in->cur) < 8) if (fill(in, 8) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy39;
-	case '-':	goto yy41;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy42;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy28;
+		case '-': goto yy29;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy30;
+		default: goto yy6;
 	}
-yy41:
+yy29:
 	yych = *++in->cur;
 	in->yyt11 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy43;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy31;
+		default: goto yy6;
 	}
-yy42:
+yy30:
 	yych = *++in->cur;
 	in->yyt11 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy43;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy44;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy31;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy32;
+		default: goto yy6;
 	}
-yy43:
+yy31:
 	yych = *++in->cur;
 	in->yyt12 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy46;
-	case '"':	goto yy48;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy33;
+		case '"': goto yy34;
+		default: goto yy6;
 	}
-yy44:
+yy32:
 	++in->cur;
 	if ((in->lim - in->cur) < 7) if (fill(in, 7) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy43;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy44;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy31;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy32;
+		default: goto yy6;
 	}
-yy46:
+yy33:
 	++in->cur;
 	if ((in->lim - in->cur) < 6) if (fill(in, 6) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy46;
-	case '"':	goto yy48;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy33;
+		case '"': goto yy34;
+		default: goto yy6;
 	}
-yy48:
+yy34:
 	yych = *++in->cur;
 	in->yyt13 = in->cur - 1;
 	switch (yych) {
-	case '\n':	goto yy8;
-	case '"':	goto yy51;
-	default:	goto yy49;
+		case '\n': goto yy6;
+		case '"': goto yy36;
+		default: goto yy35;
 	}
-yy49:
+yy35:
 	++in->cur;
 	if ((in->lim - in->cur) < 5) if (fill(in, 5) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\n':	goto yy8;
-	case '"':	goto yy51;
-	default:	goto yy49;
+		case '\n': goto yy6;
+		case '"': goto yy36;
+		default: goto yy35;
 	}
-yy51:
+yy36:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy52;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy37;
+		default: goto yy6;
 	}
-yy52:
+yy37:
 	yych = *++in->cur;
 	in->yyt14 = in->cur - 1;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy53;
-	case '"':	goto yy55;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy38;
+		case '"': goto yy39;
+		default: goto yy6;
 	}
-yy53:
+yy38:
 	++in->cur;
 	if ((in->lim - in->cur) < 3) if (fill(in, 3) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy53;
-	case '"':	goto yy55;
-	default:	goto yy8;
+		case '\t':
+		case ' ': goto yy38;
+		case '"': goto yy39;
+		default: goto yy6;
 	}
-yy55:
+yy39:
 	yych = *++in->cur;
 	in->yyt15 = in->cur - 1;
 	switch (yych) {
-	case '\n':	goto yy8;
-	case '"':	goto yy58;
-	default:	goto yy56;
+		case '\n': goto yy6;
+		case '"': goto yy41;
+		default: goto yy40;
 	}
-yy56:
+yy40:
 	++in->cur;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\n':	goto yy8;
-	case '"':	goto yy58;
-	default:	goto yy56;
+		case '\n': goto yy6;
+		case '"': goto yy41;
+		default: goto yy40;
 	}
-yy58:
+yy41:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\n':	goto yy59;
-	default:	goto yy8;
+		case '\n': goto yy42;
+		default: goto yy6;
 	}
-yy59:
+yy42:
 	++in->cur;
 	h1 = in->yyt1;
 	h2 = in->yyt2;
@@ -692,7 +716,7 @@ yy59:
         outc(out, '\n');
         goto loop;
     }
-#line 696 "gen/re2c/04__apache_log-stadfa.c"
+#line 720 "gen/re2c/04__apache_log-stadfa.c"
 }
 #line 47 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 

benchmarks/submatch_dfa_aot/pregen/re2c/04__apache_log-tdfa0.c

@@ -1,5 +1,6 @@
 /* Generated by re2c */
 #line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 #include <assert.h>
 #include <stdlib.h>
 #include "common.h"
@@ -24,7 +25,9 @@ typedef struct {
     char *cur;
     char *mar;
     char *tok;
-    char *yyt1;
+    
+#line 30 "gen/re2c/04__apache_log-tdfa0.c"
+char *yyt1;
 char *yyt10;
 char *yyt11;
 char *yyt12;
@@ -39,8 +42,12 @@ char *yyt6;
 char *yyt7;
 char *yyt8;
 char *yyt9;
+#line 34 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
     
+#line 49 "gen/re2c/04__apache_log-tdfa0.c"
+#line 35 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
     taglistpool_t tlp;
     int eof;
 } input_t;
@@ -49,6 +56,9 @@ static inline void taglistpool_clear(taglistpool_t *tlp, input_t *in)
 {
     tlp->next = tlp->head;
     
+#line 60 "gen/re2c/04__apache_log-tdfa0.c"
+#line 43 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
 }
 
 static inline void taglistpool_init(taglistpool_t *tlp)
@@ -86,10 +96,13 @@ static inline void taglist(taglist_t **ptl, const char *b, const char *t, taglis
     *ptl = tl;
 }
 
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 #line 4 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
+#line 104 "gen/re2c/04__apache_log-tdfa0.c"
 #define YYMAXFILL 2
+#line 6 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
 static inline int fill(input_t *in, size_t need)
@@ -105,7 +118,9 @@ static inline int fill(input_t *in, size_t need)
     in->cur -= free;
     in->mar -= free;
     in->tok -= free;
-    if (in->yyt1) in->yyt1 -= free;
+    
+#line 123 "gen/re2c/04__apache_log-tdfa0.c"
+if (in->yyt1) in->yyt1 -= free;
 if (in->yyt10) in->yyt10 -= free;
 if (in->yyt11) in->yyt11 -= free;
 if (in->yyt12) in->yyt12 -= free;
@@ -120,6 +135,7 @@ if (in->yyt6) in->yyt6 -= free;
 if (in->yyt7) in->yyt7 -= free;
 if (in->yyt8) in->yyt8 -= free;
 if (in->yyt9) in->yyt9 -= free;
+#line 21 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
     in->lim += fread(in->lim, 1, free, stdin);
@@ -140,7 +156,9 @@ static inline void init_input(input_t *in)
     in->cur = in->lim;
     in->mar = in->lim;
     in->tok = in->lim;
-    in->yyt1 = 0;
+    
+#line 161 "gen/re2c/04__apache_log-tdfa0.c"
+in->yyt1 = 0;
 in->yyt10 = 0;
 in->yyt11 = 0;
 in->yyt12 = 0;
@@ -155,11 +173,16 @@ in->yyt6 = 0;
 in->yyt7 = 0;
 in->yyt8 = 0;
 in->yyt9 = 0;
+#line 41 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
     
+#line 180 "gen/re2c/04__apache_log-tdfa0.c"
+#line 42 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
+
     taglistpool_init(&in->tlp);
     in->eof = 0;
 }
+#line 81 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
 
 static inline void free_input(input_t *in)
@@ -193,6 +216,7 @@ int main(int argc, char **argv)
     POST;
     return 0;
 }
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 
 
 static int lex(input_t *in, Output *out)
@@ -204,397 +228,397 @@ static int lex(input_t *in, Output *out)
 loop:
     in->tok = in->cur;
 
-#line 208 "gen/re2c/04__apache_log-tdfa0.c"
+#line 232 "gen/re2c/04__apache_log-tdfa0.c"
 {
 	char yych;
 	in->yyt1 = in->cur;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur++;
 	switch (yych) {
-	case 0x00:	goto yy2;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		in->yyt15 = in->cur;
-		goto yy6;
-	default:	goto yy4;
+		case 0x00: goto yy1;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			in->yyt15 = in->cur;
+			goto yy4;
+		default: goto yy2;
 	}
-yy2:
+yy1:
 #line 3 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 	{ return 0; }
-#line 234 "gen/re2c/04__apache_log-tdfa0.c"
-yy4:
-yy5:
+#line 258 "gen/re2c/04__apache_log-tdfa0.c"
+yy2:
+yy3:
 #line 33 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 	{ return 1; }
-#line 239 "gen/re2c/04__apache_log-tdfa0.c"
-yy6:
+#line 263 "gen/re2c/04__apache_log-tdfa0.c"
+yy4:
 	yych = *(in->mar = in->cur);
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		goto yy7;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		in->yyt15 = in->cur;
-		goto yy10;
-	default:	goto yy5;
+		case '\t':
+		case ' ':
+			++in->cur;
+			goto yy5;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			in->yyt15 = in->cur;
+			goto yy7;
+		default: goto yy3;
 	}
-yy7:
+yy5:
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		goto yy7;
-	case '-':
-		++in->cur;
-		in->yyt2 = in->cur;
-		goto yy12;
-	default:	goto yy9;
+		case '\t':
+		case ' ':
+			++in->cur;
+			goto yy5;
+		case '-':
+			++in->cur;
+			in->yyt2 = in->cur;
+			goto yy8;
+		default: goto yy6;
+	}
+yy6:
+	in->cur = in->mar;
+	goto yy3;
+yy7:
+	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			++in->cur;
+			goto yy5;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			in->yyt15 = in->cur;
+			goto yy7;
+		default: goto yy6;
+	}
+yy8:
+	yych = *in->cur;
+	switch (yych) {
+		case '-': goto yy6;
+		default: goto yy10;
 	}
 yy9:
-	in->cur = in->mar;
-	goto yy5;
+	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
+	yych = *in->cur;
 yy10:
-	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			++in->cur;
+			goto yy9;
+		case '-':
+			++in->cur;
+			in->yyt3 = in->cur;
+			goto yy11;
+		default: goto yy6;
+	}
+yy11:
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		goto yy7;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		in->yyt15 = in->cur;
-		goto yy10;
-	default:	goto yy9;
+		case '[': goto yy6;
+		default: goto yy13;
 	}
 yy12:
-	yych = *in->cur;
-	switch (yych) {
-	case '-':	goto yy9;
-	default:	goto yy14;
-	}
-yy13:
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
-yy14:
+yy13:
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		goto yy13;
-	case '-':
-		++in->cur;
-		in->yyt3 = in->cur;
-		goto yy15;
-	default:	goto yy9;
+		case '\t':
+		case ' ':
+			++in->cur;
+			in->yyt4 = in->cur;
+			goto yy12;
+		case '[':
+			++in->cur;
+			goto yy14;
+		default: goto yy6;
+	}
+yy14:
+	yych = *in->cur;
+	switch (yych) {
+		case ']': goto yy6;
+		default: goto yy16;
 	}
 yy15:
-	yych = *in->cur;
-	switch (yych) {
-	case '[':	goto yy9;
-	default:	goto yy17;
-	}
-yy16:
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
-yy17:
+yy16:
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		in->yyt4 = in->cur;
-		goto yy16;
-	case '[':
-		++in->cur;
-		goto yy18;
-	default:	goto yy9;
+		case '\n': goto yy6;
+		case ']':
+			++in->cur;
+			in->yyt5 = in->cur;
+			goto yy17;
+		default:
+			++in->cur;
+			goto yy15;
+	}
+yy17:
+	yych = *in->cur;
+	switch (yych) {
+		case '"': goto yy6;
+		default: goto yy19;
 	}
 yy18:
-	yych = *in->cur;
-	switch (yych) {
-	case ']':	goto yy9;
-	default:	goto yy20;
-	}
-yy19:
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
-yy20:
+yy19:
 	switch (yych) {
-	case '\n':	goto yy9;
-	case ']':
-		++in->cur;
-		in->yyt5 = in->cur;
-		goto yy21;
-	default:
-		++in->cur;
-		goto yy19;
+		case '\t':
+		case ' ':
+			++in->cur;
+			in->yyt6 = in->cur;
+			goto yy18;
+		case '"':
+			++in->cur;
+			goto yy20;
+		default: goto yy6;
+	}
+yy20:
+	yych = *in->cur;
+	switch (yych) {
+		case '"': goto yy6;
+		default: goto yy22;
 	}
 yy21:
-	yych = *in->cur;
-	switch (yych) {
-	case '"':	goto yy9;
-	default:	goto yy23;
-	}
-yy22:
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
-yy23:
+yy22:
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		in->yyt6 = in->cur;
-		goto yy22;
-	case '"':
-		++in->cur;
-		goto yy24;
-	default:	goto yy9;
+		case '\n': goto yy6;
+		case '"':
+			++in->cur;
+			in->yyt7 = in->cur;
+			goto yy23;
+		default:
+			++in->cur;
+			goto yy21;
+	}
+yy23:
+	yych = *in->cur;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy6;
+		default: goto yy25;
 	}
 yy24:
-	yych = *in->cur;
-	switch (yych) {
-	case '"':	goto yy9;
-	default:	goto yy26;
-	}
-yy25:
-	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
-	yych = *in->cur;
-yy26:
-	switch (yych) {
-	case '\n':	goto yy9;
-	case '"':
-		++in->cur;
-		in->yyt7 = in->cur;
-		goto yy27;
-	default:
-		++in->cur;
-		goto yy25;
-	}
-yy27:
-	yych = *in->cur;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy9;
-	default:	goto yy29;
-	}
-yy28:
 	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
 	yych = *in->cur;
-yy29:
+yy25:
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		in->yyt8 = in->cur;
-		goto yy28;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		in->yyt9 = in->cur;
-		goto yy30;
-	default:	goto yy9;
+		case '\t':
+		case ' ':
+			++in->cur;
+			in->yyt8 = in->cur;
+			goto yy24;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			in->yyt9 = in->cur;
+			goto yy26;
+		default: goto yy6;
+	}
+yy26:
+	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			++in->cur;
+			in->yyt10 = in->cur;
+			goto yy27;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			in->yyt9 = in->cur;
+			goto yy26;
+		default: goto yy6;
+	}
+yy27:
+	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			++in->cur;
+			in->yyt10 = in->cur;
+			goto yy27;
+		case '-':
+			++in->cur;
+			in->yyt11 = in->cur;
+			goto yy28;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			in->yyt11 = in->cur;
+			goto yy29;
+		default: goto yy6;
+	}
+yy28:
+	yych = *in->cur;
+	switch (yych) {
+		case '"': goto yy6;
+		default: goto yy31;
+	}
+yy29:
+	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			++in->cur;
+			in->yyt12 = in->cur;
+			goto yy30;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			in->yyt11 = in->cur;
+			goto yy29;
+		default: goto yy6;
 	}
 yy30:
 	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
 	yych = *in->cur;
+yy31:
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		in->yyt10 = in->cur;
-		goto yy32;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		in->yyt9 = in->cur;
-		goto yy30;
-	default:	goto yy9;
+		case '\t':
+		case ' ':
+			++in->cur;
+			in->yyt12 = in->cur;
+			goto yy30;
+		case '"':
+			++in->cur;
+			goto yy32;
+		default: goto yy6;
 	}
 yy32:
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		in->yyt10 = in->cur;
-		goto yy32;
-	case '-':
-		++in->cur;
-		in->yyt11 = in->cur;
-		goto yy34;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		in->yyt11 = in->cur;
-		goto yy35;
-	default:	goto yy9;
+		case '\n': goto yy6;
+		case '"':
+			++in->cur;
+			in->yyt13 = in->cur;
+			goto yy33;
+		default:
+			++in->cur;
+			goto yy32;
+	}
+yy33:
+	yych = *in->cur;
+	switch (yych) {
+		case '"': goto yy6;
+		default: goto yy35;
 	}
 yy34:
-	yych = *in->cur;
-	switch (yych) {
-	case '"':	goto yy9;
-	default:	goto yy38;
-	}
-yy35:
 	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
 	yych = *in->cur;
+yy35:
 	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		in->yyt12 = in->cur;
-		goto yy37;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		in->yyt11 = in->cur;
-		goto yy35;
-	default:	goto yy9;
+		case '\t':
+		case ' ':
+			++in->cur;
+			in->yyt14 = in->cur;
+			goto yy34;
+		case '"':
+			++in->cur;
+			goto yy36;
+		default: goto yy6;
+	}
+yy36:
+	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\n': goto yy6;
+		case '"':
+			++in->cur;
+			goto yy37;
+		default:
+			++in->cur;
+			goto yy36;
 	}
 yy37:
-	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
 	yych = *in->cur;
+	switch (yych) {
+		case '\n':
+			++in->cur;
+			goto yy38;
+		default: goto yy6;
+	}
 yy38:
-	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		in->yyt12 = in->cur;
-		goto yy37;
-	case '"':
-		++in->cur;
-		goto yy39;
-	default:	goto yy9;
-	}
-yy39:
-	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':	goto yy9;
-	case '"':
-		++in->cur;
-		in->yyt13 = in->cur;
-		goto yy41;
-	default:
-		++in->cur;
-		goto yy39;
-	}
-yy41:
-	yych = *in->cur;
-	switch (yych) {
-	case '"':	goto yy9;
-	default:	goto yy43;
-	}
-yy42:
-	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
-	yych = *in->cur;
-yy43:
-	switch (yych) {
-	case '\t':
-	case ' ':
-		++in->cur;
-		in->yyt14 = in->cur;
-		goto yy42;
-	case '"':
-		++in->cur;
-		goto yy44;
-	default:	goto yy9;
-	}
-yy44:
-	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':	goto yy9;
-	case '"':
-		++in->cur;
-		goto yy46;
-	default:
-		++in->cur;
-		goto yy44;
-	}
-yy46:
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':
-		++in->cur;
-		goto yy47;
-	default:	goto yy9;
-	}
-yy47:
 	h1 = in->yyt1;
 	h2 = in->yyt15;
 	i2 = in->yyt2;
@@ -627,7 +651,7 @@ yy47:
         outc(out, '\n');
         goto loop;
     }
-#line 631 "gen/re2c/04__apache_log-tdfa0.c"
+#line 655 "gen/re2c/04__apache_log-tdfa0.c"
 }
 #line 47 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 

benchmarks/submatch_dfa_aot/pregen/re2c/04__apache_log-tdfa1.c

@@ -1,5 +1,6 @@
 /* Generated by re2c */
 #line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 #include <assert.h>
 #include <stdlib.h>
 #include "common.h"
@@ -24,7 +25,9 @@ typedef struct {
     char *cur;
     char *mar;
     char *tok;
-    char *yyt1;
+    
+#line 30 "gen/re2c/04__apache_log-tdfa1.c"
+char *yyt1;
 char *yyt10;
 char *yyt11;
 char *yyt12;
@@ -39,8 +42,12 @@ char *yyt6;
 char *yyt7;
 char *yyt8;
 char *yyt9;
+#line 34 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
     
+#line 49 "gen/re2c/04__apache_log-tdfa1.c"
+#line 35 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
     taglistpool_t tlp;
     int eof;
 } input_t;
@@ -49,6 +56,9 @@ static inline void taglistpool_clear(taglistpool_t *tlp, input_t *in)
 {
     tlp->next = tlp->head;
     
+#line 60 "gen/re2c/04__apache_log-tdfa1.c"
+#line 43 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
 }
 
 static inline void taglistpool_init(taglistpool_t *tlp)
@@ -86,10 +96,13 @@ static inline void taglist(taglist_t **ptl, const char *b, const char *t, taglis
     *ptl = tl;
 }
 
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 #line 4 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
+#line 104 "gen/re2c/04__apache_log-tdfa1.c"
 #define YYMAXFILL 2
+#line 6 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
 static inline int fill(input_t *in, size_t need)
@@ -105,7 +118,9 @@ static inline int fill(input_t *in, size_t need)
     in->cur -= free;
     in->mar -= free;
     in->tok -= free;
-    if (in->yyt1) in->yyt1 -= free;
+    
+#line 123 "gen/re2c/04__apache_log-tdfa1.c"
+if (in->yyt1) in->yyt1 -= free;
 if (in->yyt10) in->yyt10 -= free;
 if (in->yyt11) in->yyt11 -= free;
 if (in->yyt12) in->yyt12 -= free;
@@ -120,6 +135,7 @@ if (in->yyt6) in->yyt6 -= free;
 if (in->yyt7) in->yyt7 -= free;
 if (in->yyt8) in->yyt8 -= free;
 if (in->yyt9) in->yyt9 -= free;
+#line 21 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
     in->lim += fread(in->lim, 1, free, stdin);
@@ -140,7 +156,9 @@ static inline void init_input(input_t *in)
     in->cur = in->lim;
     in->mar = in->lim;
     in->tok = in->lim;
-    in->yyt1 = 0;
+    
+#line 161 "gen/re2c/04__apache_log-tdfa1.c"
+in->yyt1 = 0;
 in->yyt10 = 0;
 in->yyt11 = 0;
 in->yyt12 = 0;
@@ -155,11 +173,16 @@ in->yyt6 = 0;
 in->yyt7 = 0;
 in->yyt8 = 0;
 in->yyt9 = 0;
+#line 41 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
     
+#line 180 "gen/re2c/04__apache_log-tdfa1.c"
+#line 42 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
+
     taglistpool_init(&in->tlp);
     in->eof = 0;
 }
+#line 81 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
 
 static inline void free_input(input_t *in)
@@ -193,6 +216,7 @@ int main(int argc, char **argv)
     POST;
     return 0;
 }
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 
 
 static int lex(input_t *in, Output *out)
@@ -204,350 +228,350 @@ static int lex(input_t *in, Output *out)
 loop:
     in->tok = in->cur;
 
-#line 208 "gen/re2c/04__apache_log-tdfa1.c"
+#line 232 "gen/re2c/04__apache_log-tdfa1.c"
 {
 	char yych;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case 0x00:	goto yy2;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		in->yyt1 = in->cur;
-		goto yy6;
-	default:	goto yy4;
+		case 0x00: goto yy1;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			in->yyt1 = in->cur;
+			goto yy4;
+		default: goto yy2;
 	}
-yy2:
+yy1:
 	++in->cur;
 #line 3 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 	{ return 0; }
-#line 234 "gen/re2c/04__apache_log-tdfa1.c"
-yy4:
+#line 258 "gen/re2c/04__apache_log-tdfa1.c"
+yy2:
 	++in->cur;
-yy5:
+yy3:
 #line 33 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 	{ return 1; }
-#line 240 "gen/re2c/04__apache_log-tdfa1.c"
-yy6:
+#line 264 "gen/re2c/04__apache_log-tdfa1.c"
+yy4:
 	yych = *(in->mar = ++in->cur);
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt2 = in->cur;
-		goto yy7;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy10;
-	default:	goto yy5;
+		case '\t':
+		case ' ':
+			in->yyt2 = in->cur;
+			goto yy5;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy7;
+		default: goto yy3;
 	}
-yy7:
+yy5:
 	++in->cur;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy7;
-	case '-':	goto yy12;
-	default:	goto yy9;
+		case '\t':
+		case ' ': goto yy5;
+		case '-': goto yy8;
+		default: goto yy6;
 	}
-yy9:
+yy6:
 	in->cur = in->mar;
-	goto yy5;
-yy10:
+	goto yy3;
+yy7:
 	++in->cur;
 	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt2 = in->cur;
-		goto yy7;
-	case '.':
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy10;
-	default:	goto yy9;
+		case '\t':
+		case ' ':
+			in->yyt2 = in->cur;
+			goto yy5;
+		case '.':
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy7;
+		default: goto yy6;
+	}
+yy8:
+	yych = *++in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt3 = in->cur;
+			goto yy9;
+		default: goto yy6;
+	}
+yy9:
+	++in->cur;
+	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy9;
+		case '-': goto yy10;
+		default: goto yy6;
+	}
+yy10:
+	yych = *++in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt4 = in->cur;
+			goto yy11;
+		default: goto yy6;
+	}
+yy11:
+	++in->cur;
+	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy11;
+		case '[':
+			in->yyt5 = in->cur;
+			goto yy12;
+		default: goto yy6;
 	}
 yy12:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt3 = in->cur;
-		goto yy13;
-	default:	goto yy9;
+		case ']': goto yy6;
+		default: goto yy14;
 	}
 yy13:
 	++in->cur;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
+yy14:
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy13;
-	case '-':	goto yy15;
-	default:	goto yy9;
+		case '\n': goto yy6;
+		case ']': goto yy15;
+		default: goto yy13;
 	}
 yy15:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt4 = in->cur;
-		goto yy16;
-	default:	goto yy9;
+		case '\t':
+		case ' ':
+			in->yyt6 = in->cur;
+			goto yy16;
+		default: goto yy6;
 	}
 yy16:
 	++in->cur;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy16;
-	case '[':
-		in->yyt5 = in->cur;
-		goto yy18;
-	default:	goto yy9;
+		case '\t':
+		case ' ': goto yy16;
+		case '"':
+			in->yyt7 = in->cur;
+			goto yy17;
+		default: goto yy6;
 	}
-yy18:
+yy17:
 	yych = *++in->cur;
 	switch (yych) {
-	case ']':	goto yy9;
-	default:	goto yy20;
+		case '"': goto yy6;
+		default: goto yy19;
 	}
-yy19:
+yy18:
 	++in->cur;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
-yy20:
+yy19:
 	switch (yych) {
-	case '\n':	goto yy9;
-	case ']':	goto yy21;
-	default:	goto yy19;
+		case '\n': goto yy6;
+		case '"': goto yy20;
+		default: goto yy18;
 	}
-yy21:
+yy20:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt6 = in->cur;
-		goto yy22;
-	default:	goto yy9;
+		case '\t':
+		case ' ':
+			in->yyt8 = in->cur;
+			goto yy21;
+		default: goto yy6;
+	}
+yy21:
+	++in->cur;
+	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy21;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			in->yyt9 = in->cur;
+			goto yy22;
+		default: goto yy6;
 	}
 yy22:
+	++in->cur;
+	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt10 = in->cur;
+			goto yy23;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy22;
+		default: goto yy6;
+	}
+yy23:
 	++in->cur;
 	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy22;
-	case '"':
-		in->yyt7 = in->cur;
-		goto yy24;
-	default:	goto yy9;
+		case '\t':
+		case ' ': goto yy23;
+		case '-':
+			in->yyt11 = in->cur;
+			goto yy24;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			in->yyt11 = in->cur;
+			goto yy25;
+		default: goto yy6;
 	}
 yy24:
 	yych = *++in->cur;
 	switch (yych) {
-	case '"':	goto yy9;
-	default:	goto yy26;
+		case '\t':
+		case ' ':
+			in->yyt12 = in->cur;
+			goto yy26;
+		default: goto yy6;
 	}
 yy25:
 	++in->cur;
-	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
+	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
 	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt12 = in->cur;
+			goto yy26;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy25;
+		default: goto yy6;
+	}
 yy26:
-	switch (yych) {
-	case '\n':	goto yy9;
-	case '"':	goto yy27;
-	default:	goto yy25;
-	}
-yy27:
-	yych = *++in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt8 = in->cur;
-		goto yy28;
-	default:	goto yy9;
-	}
-yy28:
 	++in->cur;
 	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':	goto yy28;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		in->yyt9 = in->cur;
-		goto yy30;
-	default:	goto yy9;
+		case '\t':
+		case ' ': goto yy26;
+		case '"':
+			in->yyt13 = in->cur;
+			goto yy27;
+		default: goto yy6;
+	}
+yy27:
+	++in->cur;
+	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\n': goto yy6;
+		case '"': goto yy28;
+		default: goto yy27;
+	}
+yy28:
+	yych = *++in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ':
+			in->yyt14 = in->cur;
+			goto yy29;
+		default: goto yy6;
+	}
+yy29:
+	++in->cur;
+	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
+	yych = *in->cur;
+	switch (yych) {
+		case '\t':
+		case ' ': goto yy29;
+		case '"':
+			in->yyt15 = in->cur;
+			goto yy30;
+		default: goto yy6;
 	}
 yy30:
 	++in->cur;
-	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
+	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt10 = in->cur;
-		goto yy32;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy30;
-	default:	goto yy9;
+		case '\n': goto yy6;
+		case '"': goto yy31;
+		default: goto yy30;
+	}
+yy31:
+	yych = *++in->cur;
+	switch (yych) {
+		case '\n': goto yy32;
+		default: goto yy6;
 	}
 yy32:
-	++in->cur;
-	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':	goto yy32;
-	case '-':
-		in->yyt11 = in->cur;
-		goto yy34;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		in->yyt11 = in->cur;
-		goto yy35;
-	default:	goto yy9;
-	}
-yy34:
-	yych = *++in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt12 = in->cur;
-		goto yy37;
-	default:	goto yy9;
-	}
-yy35:
-	++in->cur;
-	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt12 = in->cur;
-		goto yy37;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy35;
-	default:	goto yy9;
-	}
-yy37:
-	++in->cur;
-	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':	goto yy37;
-	case '"':
-		in->yyt13 = in->cur;
-		goto yy39;
-	default:	goto yy9;
-	}
-yy39:
-	++in->cur;
-	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':	goto yy9;
-	case '"':	goto yy41;
-	default:	goto yy39;
-	}
-yy41:
-	yych = *++in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':
-		in->yyt14 = in->cur;
-		goto yy42;
-	default:	goto yy9;
-	}
-yy42:
-	++in->cur;
-	if (in->lim <= in->cur) if (fill(in, 1) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\t':
-	case ' ':	goto yy42;
-	case '"':
-		in->yyt15 = in->cur;
-		goto yy44;
-	default:	goto yy9;
-	}
-yy44:
-	++in->cur;
-	if ((in->lim - in->cur) < 2) if (fill(in, 2) != 0) return 1;
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':	goto yy9;
-	case '"':	goto yy46;
-	default:	goto yy44;
-	}
-yy46:
-	yych = *++in->cur;
-	switch (yych) {
-	case '\n':	goto yy47;
-	default:	goto yy9;
-	}
-yy47:
 	++in->cur;
 	h1 = in->yyt1;
 	h2 = in->yyt2;
@@ -581,7 +605,7 @@ yy47:
         outc(out, '\n');
         goto loop;
     }
-#line 585 "gen/re2c/04__apache_log-tdfa1.c"
+#line 609 "gen/re2c/04__apache_log-tdfa1.c"
 }
 #line 47 "../../../benchmarks/submatch_dfa_aot/src/re2c/04__apache_log.re"
 

benchmarks/submatch_dfa_aot/pregen/re2c/05__datetime-stadfa.c

@@ -1,5 +1,6 @@
 /* Generated by re2c */
 #line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 #include <assert.h>
 #include <stdlib.h>
 #include "common.h"
@@ -24,12 +25,18 @@ typedef struct {
     char *cur;
     char *mar;
     char *tok;
-    char *yyt1;
+    
+#line 30 "gen/re2c/05__datetime-stadfa.c"
+char *yyt1;
 char *yyt2;
 char *yyt3;
 char *yyt4;
+#line 34 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
     
+#line 38 "gen/re2c/05__datetime-stadfa.c"
+#line 35 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
     taglistpool_t tlp;
     int eof;
 } input_t;
@@ -38,6 +45,9 @@ static inline void taglistpool_clear(taglistpool_t *tlp, input_t *in)
 {
     tlp->next = tlp->head;
     
+#line 49 "gen/re2c/05__datetime-stadfa.c"
+#line 43 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
 }
 
 static inline void taglistpool_init(taglistpool_t *tlp)
@@ -75,10 +85,13 @@ static inline void taglist(taglist_t **ptl, const char *b, const char *t, taglis
     *ptl = tl;
 }
 
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 #line 4 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
+#line 93 "gen/re2c/05__datetime-stadfa.c"
 #define YYMAXFILL 26
+#line 6 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
 static inline int fill(input_t *in, size_t need)
@@ -94,10 +107,13 @@ static inline int fill(input_t *in, size_t need)
     in->cur -= free;
     in->mar -= free;
     in->tok -= free;
-    if (in->yyt1) in->yyt1 -= free;
+    
+#line 112 "gen/re2c/05__datetime-stadfa.c"
+if (in->yyt1) in->yyt1 -= free;
 if (in->yyt2) in->yyt2 -= free;
 if (in->yyt3) in->yyt3 -= free;
 if (in->yyt4) in->yyt4 -= free;
+#line 21 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
     in->lim += fread(in->lim, 1, free, stdin);
@@ -118,15 +134,22 @@ static inline void init_input(input_t *in)
     in->cur = in->lim;
     in->mar = in->lim;
     in->tok = in->lim;
-    in->yyt1 = 0;
+    
+#line 139 "gen/re2c/05__datetime-stadfa.c"
+in->yyt1 = 0;
 in->yyt2 = 0;
 in->yyt3 = 0;
 in->yyt4 = 0;
+#line 41 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
     
+#line 147 "gen/re2c/05__datetime-stadfa.c"
+#line 42 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
+
     taglistpool_init(&in->tlp);
     in->eof = 0;
 }
+#line 81 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
 
 static inline void free_input(input_t *in)
@@ -160,6 +183,7 @@ int main(int argc, char **argv)
     POST;
     return 0;
 }
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
 
 
 static int lex(input_t *in, Output *out)
@@ -170,144 +194,144 @@ static int lex(input_t *in, Output *out)
 loop:
     in->tok = in->cur;
 
-#line 174 "gen/re2c/05__datetime-stadfa.c"
+#line 198 "gen/re2c/05__datetime-stadfa.c"
 {
 	char yych;
 	if ((in->lim - in->cur) < 26) if (fill(in, 26) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case 0x00:	goto yy2;
-	case '0':	goto yy6;
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy7;
-	default:	goto yy4;
+		case 0x00: goto yy1;
+		case '0': goto yy4;
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy5;
+		default: goto yy2;
 	}
-yy2:
+yy1:
 	++in->cur;
 #line 3 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 	{ return 0; }
-#line 197 "gen/re2c/05__datetime-stadfa.c"
-yy4:
+#line 221 "gen/re2c/05__datetime-stadfa.c"
+yy2:
 	++in->cur;
-yy5:
+yy3:
 #line 20 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
 	{ return 1; }
-#line 203 "gen/re2c/05__datetime-stadfa.c"
-yy6:
+#line 227 "gen/re2c/05__datetime-stadfa.c"
+yy4:
 	yych = *(in->mar = ++in->cur);
 	in->yyt2 = in->cur - 1;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy8;
-	default:	goto yy5;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy6;
+		default: goto yy3;
+	}
+yy5:
+	yych = *(in->mar = ++in->cur);
+	in->yyt2 = in->cur - 1;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy8;
+		default: goto yy3;
+	}
+yy6:
+	yych = *++in->cur;
+	in->yyt3 = in->yyt2;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy9;
+		default: goto yy7;
 	}
 yy7:
-	yych = *(in->mar = ++in->cur);
-	in->yyt2 = in->cur - 1;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy10;
-	default:	goto yy5;
-	}
+	in->cur = in->mar;
+	goto yy3;
 yy8:
 	yych = *++in->cur;
 	in->yyt3 = in->yyt2;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy11;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy10;
+		default: goto yy7;
 	}
 yy9:
-	in->cur = in->mar;
-	goto yy5;
+	yych = *++in->cur;
+	in->yyt4 = in->yyt3;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy11;
+		default: goto yy7;
+	}
 yy10:
 	yych = *++in->cur;
+	in->yyt4 = in->yyt3;
 	in->yyt3 = in->yyt2;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy12;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy12;
+		default: goto yy7;
 	}
 yy11:
-	yych = *++in->cur;
-	in->yyt4 = in->yyt3;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy13;
-	default:	goto yy9;
-	}
-yy12:
-	yych = *++in->cur;
-	in->yyt4 = in->yyt3;
-	in->yyt3 = in->yyt2;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy14;
-	default:	goto yy9;
-	}
-yy13:
 	yych = *++in->cur;
 	in->yyt1 = in->yyt4;
 	switch (yych) {
-	case '-':	goto yy16;
-	default:	goto yy9;
+		case '-': goto yy13;
+		default: goto yy7;
 	}
-yy14:
+yy12:
 	++in->cur;
 	if ((in->lim - in->cur) < 22) if (fill(in, 22) != 0) return 1;
 	yych = *in->cur;
@@ -315,250 +339,250 @@ yy14:
 	in->yyt4 = in->yyt3;
 	in->yyt3 = in->yyt2;
 	switch (yych) {
-	case '-':	goto yy16;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy14;
-	default:	goto yy9;
+		case '-': goto yy13;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy12;
+		default: goto yy7;
+	}
+yy13:
+	yych = *++in->cur;
+	switch (yych) {
+		case '0': goto yy14;
+		case '1': goto yy15;
+		default: goto yy7;
+	}
+yy14:
+	yych = *++in->cur;
+	switch (yych) {
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy16;
+		default: goto yy7;
+	}
+yy15:
+	yych = *++in->cur;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2': goto yy16;
+		default: goto yy7;
 	}
 yy16:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':	goto yy17;
-	case '1':	goto yy18;
-	default:	goto yy9;
+		case '-': goto yy17;
+		default: goto yy7;
 	}
 yy17:
 	yych = *++in->cur;
 	switch (yych) {
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy19;
-	default:	goto yy9;
+		case '0': goto yy18;
+		case '1':
+		case '2': goto yy19;
+		case '3': goto yy20;
+		default: goto yy7;
 	}
 yy18:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':	goto yy19;
-	default:	goto yy9;
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy21;
+		default: goto yy7;
 	}
 yy19:
 	yych = *++in->cur;
 	switch (yych) {
-	case '-':	goto yy20;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy21;
+		default: goto yy7;
 	}
 yy20:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':	goto yy21;
-	case '1':
-	case '2':	goto yy22;
-	case '3':	goto yy23;
-	default:	goto yy9;
+		case '0':
+		case '1': goto yy21;
+		default: goto yy7;
 	}
 yy21:
 	yych = *++in->cur;
 	switch (yych) {
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy24;
-	default:	goto yy9;
+		case 'T': goto yy22;
+		default: goto yy7;
 	}
 yy22:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy24;
-	default:	goto yy9;
+		case '0':
+		case '1': goto yy23;
+		case '2': goto yy24;
+		default: goto yy7;
 	}
 yy23:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':	goto yy24;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy25;
+		default: goto yy7;
 	}
 yy24:
 	yych = *++in->cur;
 	switch (yych) {
-	case 'T':	goto yy25;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3': goto yy25;
+		default: goto yy7;
 	}
 yy25:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':	goto yy26;
-	case '2':	goto yy27;
-	default:	goto yy9;
+		case ':': goto yy26;
+		default: goto yy7;
 	}
 yy26:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy28;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5': goto yy27;
+		default: goto yy7;
 	}
 yy27:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':	goto yy28;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy28;
+		default: goto yy7;
 	}
 yy28:
 	yych = *++in->cur;
 	switch (yych) {
-	case ':':	goto yy29;
-	default:	goto yy9;
+		case ':': goto yy29;
+		default: goto yy7;
 	}
 yy29:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':	goto yy30;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5': goto yy30;
+		default: goto yy7;
 	}
 yy30:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy31;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy31;
+		default: goto yy7;
 	}
 yy31:
 	yych = *++in->cur;
 	switch (yych) {
-	case ':':	goto yy32;
-	default:	goto yy9;
+		case '+':
+		case '-': goto yy32;
+		case 'Z': goto yy33;
+		default: goto yy7;
 	}
 yy32:
 	yych = *++in->cur;
+	in->yyt2 = in->cur - 1;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':	goto yy33;
-	default:	goto yy9;
+		case '0':
+		case '1': goto yy34;
+		case '2': goto yy35;
+		default: goto yy7;
 	}
 yy33:
 	yych = *++in->cur;
+	in->yyt2 = in->cur - 1;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy34;
-	default:	goto yy9;
+		case '\n': goto yy36;
+		default: goto yy7;
 	}
 yy34:
 	yych = *++in->cur;
 	switch (yych) {
-	case '+':
-	case '-':	goto yy35;
-	case 'Z':	goto yy36;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy37;
+		default: goto yy7;
 	}
 yy35:
 	yych = *++in->cur;
-	in->yyt2 = in->cur - 1;
 	switch (yych) {
-	case '0':
-	case '1':	goto yy37;
-	case '2':	goto yy38;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3': goto yy37;
+		default: goto yy7;
 	}
 yy36:
-	yych = *++in->cur;
-	in->yyt2 = in->cur - 1;
-	switch (yych) {
-	case '\n':	goto yy39;
-	default:	goto yy9;
-	}
-yy37:
-	yych = *++in->cur;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy41;
-	default:	goto yy9;
-	}
-yy38:
-	yych = *++in->cur;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':	goto yy41;
-	default:	goto yy9;
-	}
-yy39:
 	++in->cur;
 	y1 = in->yyt1;
 	z1 = in->yyt2;
@@ -586,44 +610,44 @@ yy39:
         outc(out, '\n');
         goto loop;
     }
-#line 590 "gen/re2c/05__datetime-stadfa.c"
-yy41:
+#line 614 "gen/re2c/05__datetime-stadfa.c"
+yy37:
 	yych = *++in->cur;
 	switch (yych) {
-	case ':':	goto yy42;
-	default:	goto yy9;
+		case ':': goto yy38;
+		default: goto yy7;
 	}
-yy42:
+yy38:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':	goto yy43;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5': goto yy39;
+		default: goto yy7;
 	}
-yy43:
+yy39:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy44;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy40;
+		default: goto yy7;
 	}
-yy44:
+yy40:
 	yych = *++in->cur;
 	switch (yych) {
-	case '\n':	goto yy39;
-	default:	goto yy9;
+		case '\n': goto yy36;
+		default: goto yy7;
 	}
 }
 #line 32 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"

benchmarks/submatch_dfa_aot/pregen/re2c/05__datetime-tdfa0.c

@@ -1,5 +1,6 @@
 /* Generated by re2c */
 #line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 #include <assert.h>
 #include <stdlib.h>
 #include "common.h"
@@ -24,10 +25,16 @@ typedef struct {
     char *cur;
     char *mar;
     char *tok;
-    char *yyt1;
+    
+#line 30 "gen/re2c/05__datetime-tdfa0.c"
+char *yyt1;
 char *yyt2;
+#line 34 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
     
+#line 36 "gen/re2c/05__datetime-tdfa0.c"
+#line 35 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
     taglistpool_t tlp;
     int eof;
 } input_t;
@@ -36,6 +43,9 @@ static inline void taglistpool_clear(taglistpool_t *tlp, input_t *in)
 {
     tlp->next = tlp->head;
     
+#line 47 "gen/re2c/05__datetime-tdfa0.c"
+#line 43 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
 }
 
 static inline void taglistpool_init(taglistpool_t *tlp)
@@ -73,10 +83,13 @@ static inline void taglist(taglist_t **ptl, const char *b, const char *t, taglis
     *ptl = tl;
 }
 
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 #line 4 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
+#line 91 "gen/re2c/05__datetime-tdfa0.c"
 #define YYMAXFILL 26
+#line 6 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
 static inline int fill(input_t *in, size_t need)
@@ -92,8 +105,11 @@ static inline int fill(input_t *in, size_t need)
     in->cur -= free;
     in->mar -= free;
     in->tok -= free;
-    if (in->yyt1) in->yyt1 -= free;
+    
+#line 110 "gen/re2c/05__datetime-tdfa0.c"
+if (in->yyt1) in->yyt1 -= free;
 if (in->yyt2) in->yyt2 -= free;
+#line 21 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
     in->lim += fread(in->lim, 1, free, stdin);
@@ -114,13 +130,20 @@ static inline void init_input(input_t *in)
     in->cur = in->lim;
     in->mar = in->lim;
     in->tok = in->lim;
-    in->yyt1 = 0;
+    
+#line 135 "gen/re2c/05__datetime-tdfa0.c"
+in->yyt1 = 0;
 in->yyt2 = 0;
+#line 41 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
     
+#line 141 "gen/re2c/05__datetime-tdfa0.c"
+#line 42 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
+
     taglistpool_init(&in->tlp);
     in->eof = 0;
 }
+#line 81 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
 
 static inline void free_input(input_t *in)
@@ -154,6 +177,7 @@ int main(int argc, char **argv)
     POST;
     return 0;
 }
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
 
 
 static int lex(input_t *in, Output *out)
@@ -164,457 +188,457 @@ static int lex(input_t *in, Output *out)
 loop:
     in->tok = in->cur;
 
-#line 168 "gen/re2c/05__datetime-tdfa0.c"
+#line 192 "gen/re2c/05__datetime-tdfa0.c"
 {
 	char yych;
 	in->yyt1 = in->cur;
 	if ((in->lim - in->cur) < 26) if (fill(in, 26) != 0) return 1;
 	yych = *in->cur++;
 	switch (yych) {
-	case 0x00:	goto yy2;
-	case '0':	goto yy6;
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy7;
-	default:	goto yy4;
+		case 0x00: goto yy1;
+		case '0': goto yy4;
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy5;
+		default: goto yy2;
 	}
-yy2:
+yy1:
 #line 3 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 	{ return 0; }
-#line 191 "gen/re2c/05__datetime-tdfa0.c"
-yy4:
-yy5:
+#line 215 "gen/re2c/05__datetime-tdfa0.c"
+yy2:
+yy3:
 #line 20 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
 	{ return 1; }
-#line 196 "gen/re2c/05__datetime-tdfa0.c"
-yy6:
+#line 220 "gen/re2c/05__datetime-tdfa0.c"
+yy4:
 	yych = *(in->mar = in->cur);
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy8;
-	default:	goto yy5;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy6;
+		default: goto yy3;
+	}
+yy5:
+	yych = *(in->mar = in->cur);
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy8;
+		default: goto yy3;
+	}
+yy6:
+	yych = *in->cur;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy9;
+		default: goto yy7;
 	}
 yy7:
-	yych = *(in->mar = in->cur);
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy10;
-	default:	goto yy5;
-	}
+	in->cur = in->mar;
+	goto yy3;
 yy8:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy11;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy10;
+		default: goto yy7;
 	}
 yy9:
-	in->cur = in->mar;
-	goto yy5;
+	yych = *in->cur;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy11;
+		default: goto yy7;
+	}
 yy10:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy12;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy12;
+		default: goto yy7;
 	}
 yy11:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy13;
-	default:	goto yy9;
+		case '-':
+			++in->cur;
+			goto yy13;
+		default: goto yy7;
 	}
 yy12:
+	if ((in->lim - in->cur) < 22) if (fill(in, 22) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy14;
-	default:	goto yy9;
+		case '-':
+			++in->cur;
+			goto yy13;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy12;
+		default: goto yy7;
 	}
 yy13:
 	yych = *in->cur;
 	switch (yych) {
-	case '-':
-		++in->cur;
-		goto yy16;
-	default:	goto yy9;
+		case '0':
+			++in->cur;
+			goto yy14;
+		case '1':
+			++in->cur;
+			goto yy15;
+		default: goto yy7;
 	}
 yy14:
-	if ((in->lim - in->cur) < 22) if (fill(in, 22) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '-':
-		++in->cur;
-		goto yy16;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy14;
-	default:	goto yy9;
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy16;
+		default: goto yy7;
+	}
+yy15:
+	yych = *in->cur;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+			++in->cur;
+			goto yy16;
+		default: goto yy7;
 	}
 yy16:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-		++in->cur;
-		goto yy17;
-	case '1':
-		++in->cur;
-		goto yy18;
-	default:	goto yy9;
+		case '-':
+			++in->cur;
+			goto yy17;
+		default: goto yy7;
 	}
 yy17:
 	yych = *in->cur;
 	switch (yych) {
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy19;
-	default:	goto yy9;
+		case '0':
+			++in->cur;
+			goto yy18;
+		case '1':
+		case '2':
+			++in->cur;
+			goto yy19;
+		case '3':
+			++in->cur;
+			goto yy20;
+		default: goto yy7;
 	}
 yy18:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-		++in->cur;
-		goto yy19;
-	default:	goto yy9;
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy21;
+		default: goto yy7;
 	}
 yy19:
 	yych = *in->cur;
 	switch (yych) {
-	case '-':
-		++in->cur;
-		goto yy20;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy21;
+		default: goto yy7;
 	}
 yy20:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-		++in->cur;
-		goto yy21;
-	case '1':
-	case '2':
-		++in->cur;
-		goto yy22;
-	case '3':
-		++in->cur;
-		goto yy23;
-	default:	goto yy9;
+		case '0':
+		case '1':
+			++in->cur;
+			goto yy21;
+		default: goto yy7;
 	}
 yy21:
 	yych = *in->cur;
 	switch (yych) {
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy24;
-	default:	goto yy9;
+		case 'T':
+			++in->cur;
+			goto yy22;
+		default: goto yy7;
 	}
 yy22:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy24;
-	default:	goto yy9;
+		case '0':
+		case '1':
+			++in->cur;
+			goto yy23;
+		case '2':
+			++in->cur;
+			goto yy24;
+		default: goto yy7;
 	}
 yy23:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-		++in->cur;
-		goto yy24;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy25;
+		default: goto yy7;
 	}
 yy24:
 	yych = *in->cur;
 	switch (yych) {
-	case 'T':
-		++in->cur;
-		goto yy25;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+			++in->cur;
+			goto yy25;
+		default: goto yy7;
 	}
 yy25:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-		++in->cur;
-		goto yy26;
-	case '2':
-		++in->cur;
-		goto yy27;
-	default:	goto yy9;
+		case ':':
+			++in->cur;
+			goto yy26;
+		default: goto yy7;
 	}
 yy26:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy28;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+			++in->cur;
+			goto yy27;
+		default: goto yy7;
 	}
 yy27:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-		++in->cur;
-		goto yy28;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy28;
+		default: goto yy7;
 	}
 yy28:
 	yych = *in->cur;
 	switch (yych) {
-	case ':':
-		++in->cur;
-		goto yy29;
-	default:	goto yy9;
+		case ':':
+			++in->cur;
+			goto yy29;
+		default: goto yy7;
 	}
 yy29:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-		++in->cur;
-		goto yy30;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+			++in->cur;
+			goto yy30;
+		default: goto yy7;
 	}
 yy30:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy31;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			in->yyt2 = in->cur;
+			goto yy31;
+		default: goto yy7;
 	}
 yy31:
 	yych = *in->cur;
 	switch (yych) {
-	case ':':
-		++in->cur;
-		goto yy32;
-	default:	goto yy9;
+		case '+':
+		case '-':
+			++in->cur;
+			goto yy32;
+		case 'Z':
+			++in->cur;
+			goto yy33;
+		default: goto yy7;
 	}
 yy32:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-		++in->cur;
-		goto yy33;
-	default:	goto yy9;
+		case '0':
+		case '1':
+			++in->cur;
+			goto yy34;
+		case '2':
+			++in->cur;
+			goto yy35;
+		default: goto yy7;
 	}
 yy33:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		in->yyt2 = in->cur;
-		goto yy34;
-	default:	goto yy9;
+		case '\n':
+			++in->cur;
+			goto yy36;
+		default: goto yy7;
 	}
 yy34:
 	yych = *in->cur;
 	switch (yych) {
-	case '+':
-	case '-':
-		++in->cur;
-		goto yy35;
-	case 'Z':
-		++in->cur;
-		goto yy36;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy37;
+		default: goto yy7;
 	}
 yy35:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-		++in->cur;
-		goto yy37;
-	case '2':
-		++in->cur;
-		goto yy38;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+			++in->cur;
+			goto yy37;
+		default: goto yy7;
 	}
 yy36:
-	yych = *in->cur;
-	switch (yych) {
-	case '\n':
-		++in->cur;
-		goto yy39;
-	default:	goto yy9;
-	}
-yy37:
-	yych = *in->cur;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy41;
-	default:	goto yy9;
-	}
-yy38:
-	yych = *in->cur;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-		++in->cur;
-		goto yy41;
-	default:	goto yy9;
-	}
-yy39:
 	y1 = in->yyt1;
 	z1 = in->yyt2;
 	y2 = in->yyt2 - 15;
@@ -641,44 +665,44 @@ yy39:
         outc(out, '\n');
         goto loop;
     }
-#line 645 "gen/re2c/05__datetime-tdfa0.c"
-yy41:
+#line 669 "gen/re2c/05__datetime-tdfa0.c"
+yy37:
 	yych = *in->cur;
 	switch (yych) {
-	case ':':
-		++in->cur;
-		goto yy42;
-	default:	goto yy9;
+		case ':':
+			++in->cur;
+			goto yy38;
+		default: goto yy7;
 	}
-yy42:
+yy38:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-		++in->cur;
-		goto yy43;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+			++in->cur;
+			goto yy39;
+		default: goto yy7;
 	}
-yy43:
+yy39:
 	yych = *in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		++in->cur;
-		goto yy36;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			++in->cur;
+			goto yy33;
+		default: goto yy7;
 	}
 }
 #line 32 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"

benchmarks/submatch_dfa_aot/pregen/re2c/05__datetime-tdfa1.c

@@ -1,5 +1,6 @@
 /* Generated by re2c */
 #line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 #include <assert.h>
 #include <stdlib.h>
 #include "common.h"
@@ -24,10 +25,16 @@ typedef struct {
     char *cur;
     char *mar;
     char *tok;
-    char *yyt1;
+    
+#line 30 "gen/re2c/05__datetime-tdfa1.c"
+char *yyt1;
 char *yyt2;
+#line 34 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
     
+#line 36 "gen/re2c/05__datetime-tdfa1.c"
+#line 35 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
     taglistpool_t tlp;
     int eof;
 } input_t;
@@ -36,6 +43,9 @@ static inline void taglistpool_clear(taglistpool_t *tlp, input_t *in)
 {
     tlp->next = tlp->head;
     
+#line 47 "gen/re2c/05__datetime-tdfa1.c"
+#line 43 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
+
 }
 
 static inline void taglistpool_init(taglistpool_t *tlp)
@@ -73,10 +83,13 @@ static inline void taglist(taglist_t **ptl, const char *b, const char *t, taglis
     *ptl = tl;
 }
 
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 #line 4 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
+#line 91 "gen/re2c/05__datetime-tdfa1.c"
 #define YYMAXFILL 26
+#line 6 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
 static inline int fill(input_t *in, size_t need)
@@ -92,8 +105,11 @@ static inline int fill(input_t *in, size_t need)
     in->cur -= free;
     in->mar -= free;
     in->tok -= free;
-    if (in->yyt1) in->yyt1 -= free;
+    
+#line 110 "gen/re2c/05__datetime-tdfa1.c"
+if (in->yyt1) in->yyt1 -= free;
 if (in->yyt2) in->yyt2 -= free;
+#line 21 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
 
     in->lim += fread(in->lim, 1, free, stdin);
@@ -114,13 +130,20 @@ static inline void init_input(input_t *in)
     in->cur = in->lim;
     in->mar = in->lim;
     in->tok = in->lim;
-    in->yyt1 = 0;
+    
+#line 135 "gen/re2c/05__datetime-tdfa1.c"
+in->yyt1 = 0;
 in->yyt2 = 0;
+#line 41 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 
     
+#line 141 "gen/re2c/05__datetime-tdfa1.c"
+#line 42 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
+
     taglistpool_init(&in->tlp);
     in->eof = 0;
 }
+#line 81 "../../../benchmarks/submatch_dfa_aot/src/re2c/common.re"
 
 
 static inline void free_input(input_t *in)
@@ -154,6 +177,7 @@ int main(int argc, char **argv)
     POST;
     return 0;
 }
+#line 1 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
 
 
 static int lex(input_t *in, Output *out)
@@ -164,390 +188,390 @@ static int lex(input_t *in, Output *out)
 loop:
     in->tok = in->cur;
 
-#line 168 "gen/re2c/05__datetime-tdfa1.c"
+#line 192 "gen/re2c/05__datetime-tdfa1.c"
 {
 	char yych;
 	if ((in->lim - in->cur) < 26) if (fill(in, 26) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case 0x00:	goto yy2;
-	case '0':
-		in->yyt1 = in->cur;
-		goto yy6;
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':
-		in->yyt1 = in->cur;
-		goto yy7;
-	default:	goto yy4;
+		case 0x00: goto yy1;
+		case '0':
+			in->yyt1 = in->cur;
+			goto yy4;
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9':
+			in->yyt1 = in->cur;
+			goto yy5;
+		default: goto yy2;
 	}
-yy2:
+yy1:
 	++in->cur;
 #line 3 "../../../benchmarks/submatch_dfa_aot/src/re2c/include/fill.re"
 	{ return 0; }
-#line 195 "gen/re2c/05__datetime-tdfa1.c"
-yy4:
+#line 219 "gen/re2c/05__datetime-tdfa1.c"
+yy2:
 	++in->cur;
-yy5:
+yy3:
 #line 20 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"
 	{ return 1; }
-#line 201 "gen/re2c/05__datetime-tdfa1.c"
-yy6:
+#line 225 "gen/re2c/05__datetime-tdfa1.c"
+yy4:
 	yych = *(in->mar = ++in->cur);
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy8;
-	default:	goto yy5;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy6;
+		default: goto yy3;
+	}
+yy5:
+	yych = *(in->mar = ++in->cur);
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy8;
+		default: goto yy3;
+	}
+yy6:
+	yych = *++in->cur;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy9;
+		default: goto yy7;
 	}
 yy7:
-	yych = *(in->mar = ++in->cur);
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy10;
-	default:	goto yy5;
-	}
+	in->cur = in->mar;
+	goto yy3;
 yy8:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy11;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy10;
+		default: goto yy7;
 	}
 yy9:
-	in->cur = in->mar;
-	goto yy5;
+	yych = *++in->cur;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy11;
+		default: goto yy7;
+	}
 yy10:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy12;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy12;
+		default: goto yy7;
 	}
 yy11:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy13;
-	default:	goto yy9;
+		case '-': goto yy13;
+		default: goto yy7;
 	}
 yy12:
-	yych = *++in->cur;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy14;
-	default:	goto yy9;
-	}
-yy13:
-	yych = *++in->cur;
-	switch (yych) {
-	case '-':	goto yy16;
-	default:	goto yy9;
-	}
-yy14:
 	++in->cur;
 	if ((in->lim - in->cur) < 22) if (fill(in, 22) != 0) return 1;
 	yych = *in->cur;
 	switch (yych) {
-	case '-':	goto yy16;
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy14;
-	default:	goto yy9;
+		case '-': goto yy13;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy12;
+		default: goto yy7;
+	}
+yy13:
+	yych = *++in->cur;
+	switch (yych) {
+		case '0': goto yy14;
+		case '1': goto yy15;
+		default: goto yy7;
+	}
+yy14:
+	yych = *++in->cur;
+	switch (yych) {
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy16;
+		default: goto yy7;
+	}
+yy15:
+	yych = *++in->cur;
+	switch (yych) {
+		case '0':
+		case '1':
+		case '2': goto yy16;
+		default: goto yy7;
 	}
 yy16:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':	goto yy17;
-	case '1':	goto yy18;
-	default:	goto yy9;
+		case '-': goto yy17;
+		default: goto yy7;
 	}
 yy17:
 	yych = *++in->cur;
 	switch (yych) {
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy19;
-	default:	goto yy9;
+		case '0': goto yy18;
+		case '1':
+		case '2': goto yy19;
+		case '3': goto yy20;
+		default: goto yy7;
 	}
 yy18:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':	goto yy19;
-	default:	goto yy9;
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy21;
+		default: goto yy7;
 	}
 yy19:
 	yych = *++in->cur;
 	switch (yych) {
-	case '-':	goto yy20;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy21;
+		default: goto yy7;
 	}
 yy20:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':	goto yy21;
-	case '1':
-	case '2':	goto yy22;
-	case '3':	goto yy23;
-	default:	goto yy9;
+		case '0':
+		case '1': goto yy21;
+		default: goto yy7;
 	}
 yy21:
 	yych = *++in->cur;
 	switch (yych) {
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy24;
-	default:	goto yy9;
+		case 'T': goto yy22;
+		default: goto yy7;
 	}
 yy22:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy24;
-	default:	goto yy9;
+		case '0':
+		case '1': goto yy23;
+		case '2': goto yy24;
+		default: goto yy7;
 	}
 yy23:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':	goto yy24;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy25;
+		default: goto yy7;
 	}
 yy24:
 	yych = *++in->cur;
 	switch (yych) {
-	case 'T':	goto yy25;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3': goto yy25;
+		default: goto yy7;
 	}
 yy25:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':	goto yy26;
-	case '2':	goto yy27;
-	default:	goto yy9;
+		case ':': goto yy26;
+		default: goto yy7;
 	}
 yy26:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy28;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5': goto yy27;
+		default: goto yy7;
 	}
 yy27:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':	goto yy28;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy28;
+		default: goto yy7;
 	}
 yy28:
 	yych = *++in->cur;
 	switch (yych) {
-	case ':':	goto yy29;
-	default:	goto yy9;
+		case ':': goto yy29;
+		default: goto yy7;
 	}
 yy29:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':	goto yy30;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5': goto yy30;
+		default: goto yy7;
 	}
 yy30:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy31;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy31;
+		default: goto yy7;
 	}
 yy31:
 	yych = *++in->cur;
 	switch (yych) {
-	case ':':	goto yy32;
-	default:	goto yy9;
+		case '+':
+		case '-':
+			in->yyt2 = in->cur;
+			goto yy32;
+		case 'Z':
+			in->yyt2 = in->cur;
+			goto yy33;
+		default: goto yy7;
 	}
 yy32:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':	goto yy33;
-	default:	goto yy9;
+		case '0':
+		case '1': goto yy34;
+		case '2': goto yy35;
+		default: goto yy7;
 	}
 yy33:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy34;
-	default:	goto yy9;
+		case '\n': goto yy36;
+		default: goto yy7;
 	}
 yy34:
 	yych = *++in->cur;
 	switch (yych) {
-	case '+':
-	case '-':
-		in->yyt2 = in->cur;
-		goto yy35;
-	case 'Z':
-		in->yyt2 = in->cur;
-		goto yy36;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy37;
+		default: goto yy7;
 	}
 yy35:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':	goto yy37;
-	case '2':	goto yy38;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3': goto yy37;
+		default: goto yy7;
 	}
 yy36:
-	yych = *++in->cur;
-	switch (yych) {
-	case '\n':	goto yy39;
-	default:	goto yy9;
-	}
-yy37:
-	yych = *++in->cur;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy41;
-	default:	goto yy9;
-	}
-yy38:
-	yych = *++in->cur;
-	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':	goto yy41;
-	default:	goto yy9;
-	}
-yy39:
 	++in->cur;
 	y1 = in->yyt1;
 	z1 = in->yyt2;
@@ -575,38 +599,38 @@ yy39:
         outc(out, '\n');
         goto loop;
     }
-#line 579 "gen/re2c/05__datetime-tdfa1.c"
-yy41:
+#line 603 "gen/re2c/05__datetime-tdfa1.c"
+yy37:
 	yych = *++in->cur;
 	switch (yych) {
-	case ':':	goto yy42;
-	default:	goto yy9;
+		case ':': goto yy38;
+		default: goto yy7;
 	}
-yy42:
+yy38:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':	goto yy43;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5': goto yy39;
+		default: goto yy7;
 	}
-yy43:
+yy39:
 	yych = *++in->cur;
 	switch (yych) {
-	case '0':
-	case '1':
-	case '2':
-	case '3':
-	case '4':
-	case '5':
-	case '6':
-	case '7':
-	case '8':
-	case '9':	goto yy36;
-	default:	goto yy9;
+		case '0':
+		case '1':
+		case '2':
+		case '3':
+		case '4':
+		case '5':
+		case '6':
+		case '7':
+		case '8':
+		case '9': goto yy33;
+		default: goto yy7;
 	}
 }
 #line 32 "../../../benchmarks/submatch_dfa_aot/src/re2c/05__datetime.re"