feat: add ARM64 support

.github/workflows/ci.yml

@@ -30,3 +30,15 @@ jobs:
       run: npm run test:integration:py38
     - name: Run integration test against Runtime NodeJS 16
       run: npm run test:integration:node16
+  arm64:
+    name: AARCH64
+    runs-on: ubuntu-latest
+    defaults:
+      run:
+        working-directory: './'
+    steps:
+    - uses: actions/checkout@v2
+    - uses: pguyot/arm-runner-action@v2
+      with:
+        commands: |
+          docker build --platform linux/arm64 -t tesseract-lambda-layer -f Dockerfile.al2 .

AARCH64.md

@@ -0,0 +1,33 @@
+qemu-user-static [1]
+
+## How-To
+
+**Setup machone to be able to build-multi architecture images**
+
+Download qemu binaries and register binfmt_misc entries:
+
+```shell
+# check host architecture (only x86_64 supported atm)
+uname -m
+> x86_64
+# enable the execution of different multi-architecture containers by QEMU
+docker run --rm --privileged multiarch/qemu-user-static --reset -p yes
+```
+
+**Build images for amd64 and aarch64**
+
+```shell
+docker buildx build --platform linux/amd64,linux/arm64 -t tesseract-lambda-layer -f Dockerfile.al2 .
+```
+
+```
+export CONTAINER=$(docker run -d tesseract-lambda-layer-aarch64 false)
+docker cp $CONTAINER:/opt/build-dist ready-to-use/amazonlinux-2-aarch64
+docker rm $CONTAINER
+unset CONTAINER
+```
+
+
+---
+
+[1]: https://github.com/multiarch/qemu-user-static

Dockerfile.al2

@@ -1,5 +1,5 @@
 ## Builds binaries for Amazonlinux 2
-FROM lambci/lambda-base-2:build
+FROM public.ecr.aws/sam/build-provided.al2:1.62
 
 ARG LEPTONICA_VERSION=1.82.0
 ARG TESSERACT_VERSION=5.2.0
@@ -17,7 +17,7 @@ ARG TESSERACT_DATA_VERSION=4.1.0
 RUN yum makecache fast; yum clean all && yum -y update && yum -y upgrade; yum clean all && \
     yum install -y yum-plugin-ovl; yum clean all && yum -y groupinstall "Development Tools"; yum clean all
 
-RUN yum -y install gcc gcc-c++ make autoconf aclocal automake libtool \
+RUN yum -y install clang gcc-c++ make autoconf aclocal automake libtool \
     libjpeg-devel libpng-devel libtiff-devel zlib-devel \
     libzip-devel freetype-devel lcms2-devel libwebp-devel \
     libicu-devel tcl-devel tk-devel pango-devel cairo-devel; yum clean all
@@ -26,7 +26,7 @@ WORKDIR ${TMP_BUILD}/leptonica-build
 RUN curl -L https://github.com/DanBloomberg/leptonica/releases/download/${LEPTONICA_VERSION}/leptonica-${LEPTONICA_VERSION}.tar.gz | tar xz && cd ${TMP_BUILD}/leptonica-build/leptonica-${LEPTONICA_VERSION} && \
     ./configure --prefix=${LEPTONICA} && make && make install && cp -r ./src/.libs /opt/liblept
 
-RUN echo "/opt/leptonica/lib" > /etc/ld.so.conf.d/leptonica.conf && ldconfig
+RUN echo "/opt/leptonica/lib" > /etc/ld.so.conf.d/leptonica.conf && /usr/sbin/ldconfig
 
 WORKDIR ${TMP_BUILD}/autoconf-build
 RUN curl https://ftp.gnu.org/gnu/autoconf-archive/autoconf-archive-${AUTOCONF_ARCHIVE_VERSION}.tar.xz | tar xJ && \
@@ -35,7 +35,7 @@ RUN curl https://ftp.gnu.org/gnu/autoconf-archive/autoconf-archive-${AUTOCONF_AR
 WORKDIR ${TMP_BUILD}/tesseract-build
 RUN curl -L https://github.com/tesseract-ocr/tesseract/archive/${TESSERACT_VERSION}.tar.gz | tar xz && \
     cd tesseract-${TESSERACT_VERSION} && ./autogen.sh  && PKG_CONFIG_PATH=/opt/leptonica/lib/pkgconfig LIBLEPT_HEADERSDIR=/opt/leptonica/include \
-    ./configure --prefix=${TESSERACT} --with-extra-includes=/opt/leptonica/include --with-extra-libraries=/opt/leptonica/lib 'CXXFLAGS=-mavx2' && make && make install
+    ./configure --prefix=${TESSERACT} --with-extra-includes=/opt/leptonica/include --with-extra-libraries=/opt/leptonica/lib && make && make install
 
 WORKDIR /opt
 RUN mkdir -p ${DIST}/lib && mkdir -p ${DIST}/bin && \

continous-integration/index-aarch64.ts

@@ -0,0 +1,75 @@
+import * as lambda from 'aws-cdk-lib/aws-lambda';
+import * as nodelambda from 'aws-cdk-lib/aws-lambda-nodejs';
+import { Architecture, CfnLayerVersion, Code, Runtime } from 'aws-cdk-lib/aws-lambda';
+import * as path from 'path';
+import { App, DockerImage, Duration, Stack } from 'aws-cdk-lib';
+
+
+const app = new App();
+const stack = new Stack(app, 'tesseract-lambda-ci');
+const pathToLayerSource = path.resolve(__dirname, '..');
+/**
+ * Test setup and artifacts for AL 2
+ */
+const al2Layer = new lambda.LayerVersion(stack, 'al2-layer', {
+    code: Code.fromAsset(pathToLayerSource, {
+    bundling: {
+        image: DockerImage.fromBuild(pathToLayerSource, { file: 'Dockerfile.al2', platform: 'linux/arm64' }),
+        command: ['/bin/bash', '-c', 'cp -r /opt/build-dist/. /asset-output/'],
+    },
+    }),
+    description: 'AL2 Tesseract Layer',
+});
+stack.renameLogicalId(stack.getLogicalId(al2Layer.node.defaultChild as CfnLayerVersion), 'al2layer')
+
+new lambda.Function(stack, 'python3.8', {
+    code: lambda.Code.fromAsset(path.resolve(__dirname, 'lambda-handlers/py38'),
+    {
+        bundling: {
+            image: DockerImage.fromRegistry('public.ecr.aws/sam/build-python3.8:latest'),
+            command: ['/bin/bash', '-c', [
+                'pip install -r requirements.txt -t /asset-output/',
+                'cp faust.png /asset-output',
+                'cp handler.py /asset-output',
+            ].join(' && ')],
+        }
+    }),
+    runtime: Runtime.PYTHON_3_8,
+    architecture: Architecture.ARM_64,
+    layers: [al2Layer],
+    functionName: `py38`,
+    memorySize: 512,
+    timeout: Duration.seconds(30),
+    handler: 'handler.main',
+});
+
+
+new nodelambda.NodejsFunction(stack, 'node16', {
+    bundling: {
+        nodeModules: ['tesseractocr'],
+        commandHooks: {
+            beforeInstall() {
+                return [];
+              },
+            beforeBundling(inputDir: string, outputDir: string): string[] {
+              return [
+                `cp ${inputDir}/faust.png ${outputDir}`,
+              ];
+            },
+            afterBundling(inputDir: string, outputDir: string): string[] {
+              return [];
+            },
+        },
+    },
+    depsLockFilePath: path.resolve(__dirname, 'lambda-handlers/node16/package-lock.json'),
+
+    runtime: Runtime.NODEJS_16_X,
+    entry: path.resolve(__dirname, 'lambda-handlers/node16/index.js'),
+    architecture: Architecture.ARM_64,
+    layers: [al2Layer],
+    functionName: `node16`,
+    memorySize: 512,
+    timeout: Duration.seconds(30),
+    handler: 'handler',
+});
+

continous-integration/package.json

@@ -17,7 +17,8 @@
     "url": "https://github.com/bweigel/aws-lambda-tesseract-layer.git"
   },
   "scripts": {
-    "synth": "npx cdk synth --app \"npx ts-node index-all.ts\"",
+    "synth": "npx cdk synth --app \"npx ts-node index.ts\"",
+    "synth:aarch64": "npx cdk synth --app \"npx ts-node index-aarch64.ts\"",
     "test:integration:py38": "sam local invoke -t cdk.out/tesseract-lambda-ci.template.json py38 --no-event > py38-test-output.txt && cat py38-test-output.txt | grep -Eiv \"(fail|error|exception)\"",
     "test:integration:node16": "sam local invoke -t cdk.out/tesseract-lambda-ci.template.json node16 --no-event > node16-test-output.txt && cat node16-test-output.txt | grep -Eiv \"(fail|error|exception)\"",
     "bundle:binary": "cp -r cdk.out/$(cat cdk.out/tesseract-lambda-ci.template.json | jq -r '.Resources.al2layer.Metadata.\"aws:asset:path\"')/. ../ready-to-use/amazonlinux-2"

example/cdk/index.ts

@@ -1,41 +1,57 @@
-import * as lambda from '@aws-cdk/aws-lambda';
-import { Code, Runtime } from '@aws-cdk/aws-lambda';
-import { RestApi, LambdaIntegration } from '@aws-cdk/aws-apigateway';
+import * as lambda from 'aws-cdk-lib/aws-lambda';
+import { Architecture, Code, Runtime } from 'aws-cdk-lib/aws-lambda';
+import { RestApi, LambdaIntegration } from 'aws-cdk-lib/aws-apigateway';
 import * as path from 'path';
-import { App, BundlingDockerImage, Duration, Stack } from '@aws-cdk/core';
+import { App, DockerImage, Duration, Stack } from 'aws-cdk-lib';
 
 
 const app = new App();
-const stack = new Stack(app, 'tesseract-ocr-example-cdk-py38');
+const stack = new Stack(app, 'tesseract-ocr-example-cdk-py38', {tags: {'owner': 'bgenz'}});
 
 /**
  * Artifacts for AL 2
  */
-const al2Layer = new lambda.LayerVersion(stack, 'al2-layer', {
+const amdLayer = new lambda.LayerVersion(stack, 'amd-layer', {
     code: Code.fromAsset(path.resolve(__dirname, '../../ready-to-use/amazonlinux-2')),
-    description: 'AL2 Tesseract Layer',
+    description: 'AL2 Tesseract Layer - AMD64',
+});
+const aarchLayer = new lambda.LayerVersion(stack, 'aarch-layer', {
+    code: Code.fromAsset(path.resolve(__dirname, '../../ready-to-use/amazonlinux-2-aarch64')),
+    description: 'AL2 Tesseract Layer - AARCH64',
+});
+
+const ocrFnAmd = new lambda.Function(stack, 'python3.8-amd', {
+    code: lambda.Code.fromDockerBuild(path.resolve(__dirname, 'lambda-handlers'),
+    {
+        platform: 'linux/amd64',
+        file: 'Dockerfile',
+    }),
+    runtime: Runtime.PYTHON_3_8,
+    architecture: Architecture.X86_64,
+    layers: [amdLayer],
+    memorySize: 1024,
+    timeout: Duration.seconds(10),
+    handler: 'handler.main',
 });
 
-const ocrFn = new lambda.Function(stack, 'python3.8', {
-    code: lambda.Code.fromAsset(path.resolve(__dirname, 'lambda-handlers'),
+const ocrFnAarch = new lambda.Function(stack, 'python3.8-aarch', {
+    code: lambda.Code.fromDockerBuild(path.resolve(__dirname, 'lambda-handlers'),
     {
-        bundling: {
-            image: BundlingDockerImage.fromRegistry('lambci/lambda:build-python3.8'),
-            command: ['/bin/bash', '-c', [
-                'pip install -r requirements.txt -t /asset-output/',
-                'cp handler.py /asset-output',
-            ].join(' && ')],
-        }
+        platform: 'linux/arm64',
+        file: 'Dockerfile.arm',
     }),
     runtime: Runtime.PYTHON_3_8,
-    layers: [al2Layer],
+    architecture: Architecture.ARM_64,
+    layers: [aarchLayer],
     memorySize: 1024,
     timeout: Duration.seconds(10),
     handler: 'handler.main',
 });
 
 const api = new RestApi(stack, 'ocr-api');
-const ocr = api.root.addResource('ocr');
-ocr.addMethod('POST', new LambdaIntegration(ocrFn, {proxy: true}));
+const amd = api.root.addResource('amd');
+const arm = api.root.addResource('arm');
+amd.addMethod('POST', new LambdaIntegration(ocrFnAmd, {proxy: true}));
+arm.addMethod('POST', new LambdaIntegration(ocrFnAarch, {proxy: true}));
 
 

example/cdk/lambda-handlers/Dockerfile

@@ -0,0 +1,7 @@
+FROM public.ecr.aws/sam/build-python3.8:1.62.0
+
+WORKDIR /asset
+COPY requirements.txt /asset/requirements.txt
+COPY handler.py /asset/handler.py
+
+RUN pip install -r requirements.txt -t /asset/ --only-binary=:all:  --implementation cp

example/cdk/lambda-handlers/Dockerfile.arm

@@ -0,0 +1,7 @@
+FROM public.ecr.aws/sam/build-python3.8:1.62.0-arm64
+
+WORKDIR /asset
+COPY requirements.txt /asset/requirements.txt
+COPY handler.py /asset/handler.py
+
+RUN pip install -r requirements.txt -t /asset/ --only-binary=:all:  --implementation cp --platform manylinux2014_aarch64