Merge branch 'main' into improve-build-for-android-integrate-with-cpu…

…_features-library

.github/workflows/autotools-macos.yml

@@ -5,6 +5,7 @@ on:
   #push:
   schedule:
     - cron: 0 20 * * *
+  workflow_dispatch:
 jobs:
 
   brew:
@@ -42,7 +43,7 @@ jobs:
 
     - name: Configure Tesseract
       run: |
-           ./configure '--disable-shared' '--disable-openmp' '--disable-doc' '--with-pic' 'CXX=${{ matrix.config.cxx }}' 'CXXFLAGS=-g -O2' "PKG_CONFIG_PATH=$(brew --prefix)/opt/icu4c/lib/pkgconfig:$(brew --prefix)/opt/libarchive/lib/pkgconfig:$(brew --prefix)/opt/libffi/lib/pkgconfig"
+           ./configure '--disable-shared' '--disable-openmp' '--disable-doc' '--with-pic' 'CXX=${{ matrix.config.cxx }}' 'CXXFLAGS=-g -O2'
 
     - name: Make and Install Tesseract
       run: |
@@ -130,6 +131,9 @@ jobs:
     - name: Install Macports
       run: |
         curl -LO https://raw.githubusercontent.com/GiovanniBussi/macports-ci/master/macports-ci; source ./macports-ci install
+        # --remove-brew does not remove the Homebrew entries in bin,
+        # so remove them now.
+        rm -v $(brew --prefix)/bin/*
 
     - name: Install Dependencies
       run: |

.github/workflows/autotools-openmp.yml

@@ -0,0 +1,83 @@
+name: autotools-openmp
+# autotools on Ubuntu - run benchmark test. '--enable-openmp' no training tools
+on:
+  #push:
+  #schedule:
+  #  - cron: 0 20 * * *
+  workflow_dispatch:
+jobs:
+
+  linux:
+    runs-on: ${{ matrix.config.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        config:
+          - { name: 18.04-openmp, os: ubuntu-18.04 }
+          - { name: 20.04-openmp, os: ubuntu-20.04 }
+
+    steps:
+    - uses: actions/checkout@v2
+      with:
+        submodules: recursive
+
+    - name: Download fonts, tessdata and langdata required for tests
+      run: |
+           git clone https://github.com/egorpugin/tessdata tessdata_unittest
+           cp tessdata_unittest/fonts/* test/testing/
+           mv tessdata_unittest/* ../
+
+    - name: Install dependencies
+      run: |
+           sudo apt-get update
+           sudo apt-get install autoconf libleptonica-dev -y
+           sudo apt-get install libpango1.0-dev -y
+           sudo apt-get install cabextract libarchive-dev -y
+           sudo apt-get install libcurl4-openssl-dev libcurl4 curl -y
+
+    - name: Setup Tesseract
+      run: |
+           mkdir -p m4
+           ./autogen.sh
+
+    - name: Configure Tesseract
+      run: |
+           ./configure '--disable-shared' '--enable-openmp' '--disable-doc' 'CXX=g++' 'CXXFLAGS=-g -O2'
+           grep -i OpenMP config.log
+
+    - name: Make and Install Tesseract
+      run: |
+           make
+           sudo make install
+
+    - name: Setup for Tesseract benchmark using image from issue 263 fifteen times in a list file
+      run: |
+           wget -O i263_speed.jpg https://cloud.githubusercontent.com/assets/9968625/13674495/ac261db4-e6ab-11e5-9b4a-ad91d5b4ff87.jpg
+           printf 'i263_speed.jpg\n%.0s' {1..15} > benchmarks.list
+
+    - name: Run Tesseract using image from issue 263 with tessdata_fast
+      run: |
+           lscpu
+           free
+           g++ --version
+           tesseract -v
+           time tesseract benchmarks.list - --tessdata-dir ../tessdata_fast > /dev/null 2>&1
+           echo "tessdata_fast"
+
+    - name: Run Tesseract using image from issue 263 with tessdata_fast and OpenMP Thread Limit
+      run: |
+           for lmt in {1..3}; do
+                time OMP_THREAD_LIMIT=$lmt tesseract benchmarks.list - --tessdata-dir ../tessdata_fast > /dev/null 2>&1 && echo "OMP_THREAD_LIMIT=" $lmt "tessdata_fast"
+           done
+
+    - name: Run Tesseract using image from issue 263 with tessdata_best and OpenMP Thread Limit
+      run: |
+           for lmt in {1..3}; do
+                time OMP_THREAD_LIMIT=$lmt tesseract benchmarks.list - --tessdata-dir ../tessdata_best > /dev/null 2>&1 && echo "OMP_THREAD_LIMIT=" $lmt "tessdata_best"
+           done
+
+    - name: Run Tesseract using image from issue 263 with tessdata and OpenMP Thread Limit
+      run: |
+           for lmt in {1..3}; do
+                time OMP_THREAD_LIMIT=$lmt tesseract benchmarks.list - --tessdata-dir ../tessdata > /dev/null 2>&1 && echo "OMP_THREAD_LIMIT=" $lmt "tessdata"
+           done

.github/workflows/cmake-win64.yml

@@ -5,138 +5,154 @@ on:
   #push:
   schedule:
     - cron: 0 23 * * *
+  workflow_dispatch:
+
+env:
+  ILOC: d:/a/local
 
 jobs:
   build:
-    name: ${{ matrix.config.name }}
-    runs-on: ${{ matrix.config.os }}
-    strategy:
-      fail-fast: false
-      matrix:
-        config:
-        - {
-            name: "Windows Latest MSVC - cmake",
-            os: windows-latest,
-            cc: "cl",
-            cxx: "cl",
-            environment_script: "C:/Program Files (x86)/Microsoft Visual Studio/2019/Enterprise/VC/Auxiliary/Build/vcvars64.bat"
-          }
-
+    name: cmake-win64
+    runs-on: windows-latest
     steps:
-      - name: Checkout code
-        uses: actions/checkout@v2
+      - uses: ilammy/setup-nasm@v1
+      - name: "Checkout ${{ github.ref }} ( ${{ github.sha }} )"
+        uses: actions/checkout@v3
         with:
           submodules: recursive
-
-      - name: Build and Install leptonica dependencies
-        shell: cmd
+      - name: Get the version
+        id: get_version
+        run: echo ::set-output name=VERSION::${GITHUB_REF/refs\/tags\//}
+      - name: Setup Instalation Location
         run: |
-             mkdir d:/a/local
-             set PKG_CONFIG_PATH=d:/a/local/lib/pkgconfig
-             echo "PKG_CONFIG_PATH=d:/a/local/lib/pkgconfig" >> $GITHUB_ENV
-
-      - name: Build and Install zlib
+             mkdir ${{env.ILOC}}
+      - name: Build and Install zlib-ng
         shell: cmd
         run: |
-             curl -sSL -o zlib1211.zip https://zlib.net/zlib1211.zip
-             unzip.exe zlib1211.zip
-             cd zlib-1.2.11
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             git clone --depth 1 https://github.com/zlib-ng/zlib-ng.git
+             cd zlib-ng
+             cmake -Bbuild -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DBUILD_SHARED_LIBS=OFF -DZLIB_COMPAT=ON -DZLIB_ENABLE_TESTS=OFF -DINSTALL_UTILS=OFF
+             cmake -Bbuild -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DZLIB_COMPAT=ON -DZLIB_ENABLE_TESTS=OFF -DINSTALL_UTILS=OFF
              cmake --build build --config Release --target install
+             cd ..
 
       - name: Build and Install libpng
         shell: cmd
         run: |
              curl -sSL -o lpng1637.zip https://download.sourceforge.net/libpng/lpng1637.zip
-             unzip.exe lpng1637.zip
+             unzip.exe  -qq lpng1637.zip
              cd lpng1637
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
-             cmake --build build --config Release --target install
-
-      - name: Build and Install webp
-        shell: cmd
-        run: |
-             git clone --depth 1 https://github.com/webmproject/libwebp.git && cd libwebp
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DPNG_TESTS=OFF -DPNG_SHARED=OFF
              cmake --build build --config Release --target install
+             cd ..
 
       - name: Build and Install giflib
         shell: cmd
         run: |
              curl -sSL -o giflib-master.zip https://codeload.github.com/xbmc/giflib/zip/master
-             unzip giflib-master.zip
+             unzip -qq giflib-master.zip
              cd giflib-master
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             cmake -Bbuild -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}}
              cmake --build build --config Release --target install
+             cd ..
 
       - name: Build and Install libjpeg
         shell: cmd
         run: |
              git clone --depth 1 https://github.com/libjpeg-turbo/libjpeg-turbo.git
              cd libjpeg-turbo
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DWITH_TURBOJPEG=OFF -DENABLE_SHARED=OFF
              cmake --build build --config Release --target install
+             cd ..
 
-      - name: Build and Install openjpeg
+      - name: Build and Install webp
         shell: cmd
         run: |
-             git clone --depth 1 https://github.com/uclouvain/openjpeg.git
-             cd openjpeg
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             git clone --depth 1 https://github.com/webmproject/libwebp.git
+             cd libwebp
+             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DWEBP_BUILD_ANIM_UTILS=OFF -DWEBP_BUILD_CWEBP=OFF -DWEBP_BUILD_DWEBP=OFF -DWEBP_BUILD_GIF2WEBP=OFF -DWEBP_BUILD_IMG2WEBP=OFF -DWEBP_BUILD_VWEBP=OFF -DWEBP_BUILD_WEBPMUX=OFF -DWEBP_BUILD_WEBPINFO=OFF -DWEBP_BUILD_EXTRAS=OFF
              cmake --build build --config Release --target install
              cd ..
 
       - name: Build and Install jbigkit
         shell: cmd
         run: |
-             git clone --depth 1 https://github.com/zdenop/jbigkit
-             cd jbigkit-2.1
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             git clone --depth 1 https://github.com/zdenop/jbigkit.git
+             cd jbigkit
+             cmake -Bbuild -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DBUILD_PROGRAMS=OFF -DBUILD_TOOLS=OFF -DCMAKE_WARN_DEPRECATED=OFF
              cmake --build build --config Release --target install
              cd ..
 
       - name: Build and Install zstd
         shell: cmd
         run: |
              git clone --depth 1 https://github.com/facebook/zstd.git
-             cd zstd
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             cd zstd/build/cmake
+             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DZSTD_BUILD_PROGRAMS=OFF -DBUILD_TESTING=OFF
              cmake --build build --config Release --target install
              cd ..
 
       - name: Build and Install libtiff
         shell: cmd
         run: |
-             git clone --depth 1 https://gitlab.com/libtiff/libtiff
+             git clone --depth 1 https://gitlab.com/libtiff/libtiff.git
              cd libtiff
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -Dtiff-tools=OFF -Dtiff-tests=OFF -Dtiff-contrib=OFF -Dtiff-docs=OFF
+             cmake --build build --config Release --target install
+             cd ..
+
+      - name: Build and Install openjpeg
+        shell: cmd
+        run: |
+             git clone --depth 1 https://github.com/uclouvain/openjpeg.git
+             cd openjpeg
+             cmake -Bbuild -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DBUILD_CODEC=OFF -DBUILD_TESTING=OFF -DBUILD_DOC=OFF -DCMAKE_WARN_DEPRECATED=OFF
              cmake --build build --config Release --target install
              cd ..
 
       - name: Build and Install leptonica
         shell: cmd
         run: |
              echo "Building leptonica..."
-             git clone --depth 1 https://github.com/DanBloomberg/leptonica.git && cd leptonica
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DBUILD_PROG=OFF -DSW_BUILD=OFF -DBUILD_SHARED_LIBS=ON -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             git clone --depth 1 https://github.com/DanBloomberg/leptonica.git
+             cd leptonica
+             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DSW_BUILD=OFF -DBUILD_PROG=OFF -DBUILD_SHARED_LIBS=ON
+             cmake --build build --config Release --target install
+
+      - name: Build and Install libarchive
+        shell: cmd
+        run: |
+             git clone --depth 1 https://github.com/libarchive/libarchive.git
+             cd libarchive
+             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DENABLE_TEST=OFF
              cmake --build build --config Release --target install
 
+      - name: Remove not needed tools Before building tesseract
+        shell: cmd
+        run: >
+             rm -Rf ${{env.ILOC}}/bin/*.exe
+
       - name: Build and Install tesseract
         shell: cmd
         run: |
-             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DBUILD_TRAINING_TOOLS=OFF -DSW_BUILD=OFF -DBUILD_SHARED_LIBS=ON -DOPENMP_BUILD=OFF -DCMAKE_PREFIX_PATH=d:/a/local -DCMAKE_INSTALL_PREFIX=d:/a/local
+             REM cmake -E env CXXFLAGS="/Qpar /fp:fast"
+             cmake -Bbuild -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH=${{env.ILOC}} -DCMAKE_INSTALL_PREFIX=${{env.ILOC}} -DSW_BUILD=OFF -DBUILD_SHARED_LIBS=ON -DENABLE_LTO=ON -DBUILD_TRAINING_TOOLS=OFF -DFAST_FLOAT=ON -DGRAPHICS_DISABLED=ON -DOPENMP_BUILD=ON
              cmake --build build --config Release --target install
 
       - name: Display Tesseract Version and Test Command Line Usage
         shell: cmd
         run: |
-          git clone --depth 1 https://github.com/tesseract-ocr/tessconfigs
-          mkdir d:/a/local/share
-          move tessconfigs d:/a/local/share
-          curl -L https://github.com/tesseract-ocr/tessdata/raw/main/eng.traineddata --output d:/a/local/share/tessconfigs/eng.traineddata
-          curl -L https://github.com/tesseract-ocr/tessdata/raw/main/osd.traineddata --output d:/a/local/share/tessconfigs/osd.traineddata
-          set TESSDATA_PREFIX=d:/a/local/share/tessconfigs
-          set PATH=d:/a/local/bin;%PATH%
+          curl -L https://github.com/tesseract-ocr/tessdata/raw/main/eng.traineddata --output ${{env.ILOC}}/share/tessdata/eng.traineddata
+          curl -L https://github.com/tesseract-ocr/tessdata/raw/main/osd.traineddata --output ${{env.ILOC}}/share/tessdata/osd.traineddata
+          set TESSDATA_PREFIX=${{env.ILOC}}/share/tessdata
+          set PATH=${{env.ILOC}}/bin;%PATH%
           tesseract -v
           tesseract --list-langs
           tesseract test/testing/phototest.tif -
+          
+      - name: Upload Build Results
+        uses: actions/upload-artifact@v2
+        with:
+          name: tesseract-${{ steps.get_version.outputs.VERSION }}-VS2019_win64
+          path: ${{env.ILOC}}
+          retention-days: 5

.github/workflows/cmake.yml

@@ -50,23 +50,26 @@ jobs:
     steps:
       - name: Install compilers on Linux
         run: |
+             sudo apt-get update
              sudo apt-get install ${{ matrix.config.cxx }} -y
         if: runner.os == 'Linux'
 
-# sudo apt-get install libarchive-dev libcurl4-openssl-dev libcurl4 curl -y
       - name: Install dependencies on Linux
         run: |
            sudo apt-get install autoconf libleptonica-dev -y
+           sudo apt-get install libarchive-dev libcurl4-openssl-dev -y
            sudo apt-get install libpango1.0-dev -y
            sudo apt-get install cabextract -y
            sudo apt-get install ninja-build -y
+           cmake --version
         if: runner.os == 'Linux'
 
       - name: Install dependencies on macOS
         run: |
            brew install autoconf automake
            brew install leptonica
-           brew install cairo pango icu4c
+           brew install libarchive
+           brew install pango
            brew install cabextract
            brew install ninja
            ninja --version
@@ -147,15 +150,15 @@ jobs:
         run: |
              export "PKG_CONFIG_PATH=$GITHUB_WORKSPACE/build/inst/lib/pkgconfig/:$PKG_CONFIG_PATH"
              cd test
-             ${{ matrix.config.cxx }} -o basicapitest testing/basicapitest.cpp "-I$GITHUB_WORKSPACE/build/inst/include" "-L$GITHUB_WORKSPACE/build/inst/lib" `pkg-config --cflags --libs tesseract lept ` -pthread -std=c++11
+             ${{ matrix.config.cxx }} -o basicapitest testing/basicapitest.cpp "-I$GITHUB_WORKSPACE/build/inst/include" "-L$GITHUB_WORKSPACE/build/inst/lib" $(pkg-config --cflags --libs tesseract lept libarchive libcurl) -pthread -std=c++11
              ./basicapitest
         if: runner.os == 'Linux'
 
       - name: Build and run basicapitest (macOS)
         run: |
-             export "PKG_CONFIG_PATH=$GITHUB_WORKSPACE/build/inst/lib/pkgconfig/:$PKG_CONFIG_PATH"
+             export "PKG_CONFIG_PATH=$GITHUB_WORKSPACE/build/inst/lib/pkgconfig/:$(brew --prefix)/opt/libarchive/lib/pkgconfig:$(brew --prefix)/Library/Homebrew/os/mac/pkgconfig/11:$PKG_CONFIG_PATH"
              cd test
-             ${{ matrix.config.cxx }} -o basicapitest testing/basicapitest.cpp "-I$GITHUB_WORKSPACE/build/inst/include" "-L$GITHUB_WORKSPACE/build/inst/lib" `pkg-config --cflags --libs tesseract lept ` -lcurl -pthread -std=c++11
+             ${{ matrix.config.cxx }} -o basicapitest testing/basicapitest.cpp "-I$GITHUB_WORKSPACE/build/inst/include" "-L$GITHUB_WORKSPACE/build/inst/lib" $(pkg-config --cflags --libs tesseract lept libarchive libcurl) -pthread -std=c++11
              ./basicapitest
         if: runner.os == 'macOS'