在 Azure Cloud Service 中运行

rem   Script to clean up previous ABBYY FineReader Engine and Licensing Service packages
echo Cleaning up before launching service >> ".\CleanUpOnStart.log" 2>&1
PowerShell -ExecutionPolicy Unrestricted .\StartupScripts\CleanUpOnStart.ps1 >> ".\CleanUpOnStart.log" 2>&1
echo Cleaning up beafore launching service. >> ".\CleanUpOnStart.log" 2>&1
 
exit /B %errorlevel%

Write-Host "CleanUpOnStart.ps1 started...";
# 本地存储路径（参见 ServiceDefinition.csdef）
$local_storage_name = "LocalStorage1";
[System.Reflection.Assembly]::LoadWithPartialName("Microsoft.WindowsAzure.ServiceRuntime");
$local_storage_path = ([Microsoft.WindowsAzure.ServiceRuntime.RoleEnvironment]::GetLocalResource($local_storage_name)).RootPath.TrimEnd('\\');
# 如果许可服务进程正在运行，则将其停止
$processes = Get-Process;
foreach( $process in $processes ) {
    if( $process.Name -eq "LicesingService" ) {
        Stop-Process $process;
    }
}
 
# 清理并重新创建 fre_packages 文件夹
$destination_path = Join-Path -Path $local_storage_path -ChildPath 'fre_packages';
if( Test-Path -Path $destination_path ) {
    Remove-Item $destination_path -Recurse -Force;
}
New-Item -ItemType Directory -Path $destination_path;
# 清理并重新创建 Input 文件夹
$destination_path = Join-Path -Path $local_storage_path -ChildPath 'Input';
if( Test-Path -Path $destination_path ) {
    Remove-Item $destination_path -Recurse -Force;
}
New-Item -ItemType Directory -Path $destination_path;
# 清理并重新创建 Results 文件夹
$destination_path = Join-Path -Path $local_storage_path -ChildPath 'Results';
if( Test-Path -Path $destination_path ) {
    Remove-Item $destination_path -Recurse -Force;
}
New-Item -ItemType Directory -Path $destination_path;
Write-Host ("Resource folder was cleaned up.");

rem   用于准备相关模块的脚本，以便连接到存储帐户并下载 ABBYY FineReader Engine 和许可服务包
 
echo Preparing PoSh Modules >> ".\PreparePoShModules.log" 2>&1
PowerShell -ExecutionPolicy Unrestricted .\StartupScripts\PreparePoShModules.ps1 >> ".\PreparePoShModules.log" 2>&1
echo PoSh Modules wer prepared. >> ".\PreparePoShModules.log" 2>&1
 
exit /B %errorlevel%

# 安装模块
Write-Host "PreparePoShModules.ps1 started...";
# 安装用于下载 Azure 模块的 NuGet 提供程序
Write-Host "Installing NuGet package provider...";
Install-PackageProvider -Name NuGet -MinimumVersion 2.8.5.201 -Force;
 
# 安装用于操作 Blob 存储的 Azure Storage 模块
Write-Host "Installing Az.Storage module...";
Install-Module -Name Az.Storage -Scope CurrentUser -Repository PSGallery -Force -AllowClobber;
Write-Host ("PoSh modules were prepared.");

rem   用于准备 ABBYY FineReader Engine 包的脚本
echo Preparing Library >> ".\PrepareLibrary.log" 2>&1
PowerShell -ExecutionPolicy Unrestricted .\StartupScripts\PrepareLibrary.ps1 >> ".\PrepareLibrary.log" 2>&1
echo Library was prepared. >> ".\PrepareLibrary.log" 2>&1
exit /B %errorlevel%

Write-Host "PrepareLibrary.ps1 started...";
# 包含 ABBYY FineReader Engine 和许可服务包的容器
$container_name = 'fre-lib';
# 存储帐户连接字符串 
$connection_string = "<connection_string_to_frestorage>";
Write-Host "Configuration";
Write-Host ("Container name: " + $container_name);
Write-Host ("Connection string: " + $storage_account);
# 连接到存储对象
Write-Host ("Connecting to storage account...");
$storage_account = New-AzStorageContext -ConnectionString $connection_string;
# 获取容器中的 blob（本质上就是文件）
Write-Host ("Getting blobs from container...");
$fre_blobs = Get-AzStorageBlob -Container $container_name -Context $storage_account;
# 本地存储路径（参见 ServiceDefinition.csdef）
$local_storage_name = "LocalStorage1";
[System.Reflection.Assembly]::LoadWithPartialName("Microsoft.WindowsAzure.ServiceRuntime");
$local_storage_path = ([Microsoft.WindowsAzure.ServiceRuntime.RoleEnvironment]::GetLocalResource($local_storage_name)).RootPath.TrimEnd('\\');
$destination_path = Join-Path -Path $local_storage_path -ChildPath 'fre_packages';
# 下载唯一需要的 blob
Write-Host ("Downloading FRE package from container...");
foreach( $blob in $fre_blobs ) {
    Write-Host $blob.Name;
    if( $blob.Name -eq "LibraryPackage.zip" ) {
        Write-Host ("Downloading blob: " + $blob.Name + "to " + $destination_path);
        Get-AzStorageBlobContent -Container $container_name -Blob $blob.Name `
            -Destination (Join-Path -Path $destination_path -ChildPath $blob.Name) -Context $storage_account;
    } else {
        Write-Host ("Downloading blob " + $blob.Name + " was skipped.");
    }
}
# 解压已下载的包（旧版库将被强制覆盖）
Write-Host ("Unzipping library package...");
Expand-Archive -Path (Join-Path -Path $destination_path -ChildPath "LibraryPackage.zip") `
    -DestinationPath (Join-Path -Path $destination_path -ChildPath "LibraryPackage") -Force;
Write-Host ("Library package was prepared.");

rem   用于准备许可服务包并启动 LicensingService.exe 的脚本
echo Preparing LS >> ".\PrepareLS.log" 2>&1
PowerShell -ExecutionPolicy Unrestricted .\StartupScripts\PrepareLS.ps1 >> ".\PrepareLS.log" 2>&1
echo LS was prepared. >> ".\PrepareLS.log" 2>&1
exit /B %errorlevel%

Write-Host "PrepareLS.ps1 started...";
# 包含 ABBYY FineReader Engine 和许可服务包的容器
$container_name = 'fre-lib';
# 存储帐户连接字符串
$connection_string = "<connection_string_to_frestorage>";
Write-Host "Configuration";
Write-Host ("Container name: " + $container_name);
Write-Host ("Connection string: " + $storage_account);
# 连接到存储对象
Write-Host ("Connecting to storage account...");
$storage_account = New-AzStorageContext -ConnectionString $connection_string;
# 获取 blob（实际上就是文件）
Write-Host ("Getting blobs from container...");
$fre_blobs = Get-AzStorageBlob -Container $container_name -Context $storage_account;
# 本地存储路径（参见 ServiceDefinition.csdef）
$local_storage_name = "LocalStorage1";
[System.Reflection.Assembly]::LoadWithPartialName("Microsoft.WindowsAzure.ServiceRuntime");
$local_storage_path = ([Microsoft.WindowsAzure.ServiceRuntime.RoleEnvironment]::GetLocalResource($local_storage_name)).RootPath.TrimEnd('\\');
$destination_path = Join-Path -Path $local_storage_path -ChildPath 'fre_packages'; 
# 下载许可服务包
Write-Host ("Downloading Licensing Service package from container...");
foreach( $blob in $fre_blobs ) {
    Write-Host $blob.Name;
    if( $blob.Name -eq "LSPackage.zip" ) {
        Write-Host ("Downloading blob: " + $blob.Name + "to " + $destination_path);
        Get-AzStorageBlobContent -Container $container_name -Blob $blob.Name `
            -Destination (Join-Path -Path $destination_path -ChildPath $blob.Name) -Context $storage_account;
    } else {
        Write-Host ("Downloading blob " + $blob.Name + " was skipped.");
    }
}
# 解压已下载的包
Write-Host ("Unzipping library package...");
Expand-Archive -Path (Join-Path -Path $destination_path -ChildPath "LSPackage.zip") `
    -DestinationPath (Join-Path -Path $destination_path -ChildPath "LSPackage") -Force;
# 为许可证创建文件夹
$program_data_path = [System.Environment]::ExpandEnvironmentVariables("%programdata%");
$path = Join-Path -Path $program_data_path -ChildPath "ABBYY\SDK\12\Licenses";
New-Item -ItemType Directory -Path $path -Force;
# 以 standalone 模式启动许可服务
Write-Host ("Starting Licensing Service...");
 
$licensing_service_app = Join-Path -Path $destination_path -ChildPath "LSPackage\LicensingService.exe";
Start-Process -FilePath $licesing_service_app -ArgumentList "/standalone";
Write-Host ("LS package was prepared and LS was started.");

namespace WorkerRole1.Engine
{
    public class FileWriter : FREngine.IFileWriter, IDisposable
    {
        public FileWriter(string _resultBlobName, string _fileExtension)
        {
            resultBlobName = _resultBlobName;
            fileExtension = _fileExtension;
        }
        public void Open(string fileName, ref int bufferSize)
        {
            stream = new MemoryStream();
        }
        public void Write(byte[] data)
        {
            stream.Write(data, 0, data.Length);
        }
        public void Close() 
        {
            // 在 <OutputContainerName> 中创建指向新 blob 的连接
            // 处理结果将存储到其中
            BlobClient resultBlobClient = new BlobClient(Config.ConnectionString, Config.OutputContainerName,
                resultBlobName + fileExtension);
            // 覆盖现有文件
            resultBlobClient.DeleteIfExists();
            // 将位置设为 0，以便从头开始写入文件
            stream.Position = 0;
            resultBlobClient.Upload(stream);
 
            stream.Close();
        }
        public void Dispose()
        {
            // 释放时关闭内存流，以便在数据写入后仍可访问它
            stream.Close();
        }
        private string resultBlobName;
        private string fileExtension;
        private MemoryStream stream;
    }
}

namespace WorkerRole1
{
    public class WorkerRole : RoleEntryPoint
    {
        private readonly CancellationTokenSource cancellationTokenSource = new CancellationTokenSource();
        private readonly ManualResetEvent runCompleteEvent = new ManualResetEvent(false);
        Processor processor = new Processor();
 
        public override void Run()
        {
            Trace.TraceInformation("WorkerRole1 is running");
 
            try
            {
                this.RunAsync(this.cancellationTokenSource.Token).Wait();
            }
            finally
            {
                this.runCompleteEvent.Set();
            }
        }
        public override bool OnStart()
        {
            // 设置最大并发连接数
            ServicePointManager.DefaultConnectionLimit = 12;
 
            // 将 TLS 设置为 1.2，以便能够连接到存储帐户
            System.Net.ServicePointManager.SecurityProtocol = System.Net.SecurityProtocolType.Tls12;
 
            // 有关如何处理配置更改的信息
            // 请参阅 MSDN 主题：https://go.microsoft.com/fwlink/?LinkId=166357
 
            bool result = base.OnStart();
            Trace.TraceInformation("WorkerRole1 has been started");
 
            processor.LoadEngine();
            return result;
        }
 
        public override void OnStop()
        {
            processor.UnloadEngine();
            Trace.TraceInformation("WorkerRole1 is stopping");
            this.cancellationTokenSource.Cancel();
            this.runCompleteEvent.WaitOne();
            base.OnStop();
            Trace.TraceInformation("WorkerRole1 has stopped");
        }
 
        private async Task RunAsync(CancellationToken cancellationToken)
        {
            // TODO: 使用你自己的逻辑替换以下内容
            while (!cancellationToken.IsCancellationRequested)
            {
                Trace.TraceInformation("Working");
 
                try
                {
                    // 创建容器并返回容器客户端对象
                    BlobContainerClient inputContainerClient = new BlobContainerClient(Config.ConnectionString, Config.InputContainerName);
                    foreach (BlobItem blobItem in inputContainerClient.GetBlobs())
                    {
                        Trace.TraceInformation("\t" + blobItem.Name);
                        BlobClient blobClient = new BlobClient(Config.ConnectionString, Config.InputContainerName, blobItem.Name);
 
                        Trace.TraceInformation("\t Downloading blob to memory: " + blobItem.Name);
                        MemoryStream memoryStream = new MemoryStream();
                        blobClient.DownloadTo(memoryStream);
                        Trace.TraceInformation("\t Processing in FRE: " + blobItem.Name);
                        string resultBlobName = processor.ProcessImage(memoryStream, blobItem.Name);
                        Trace.TraceInformation("\t Result blob name in output container: " + resultBlobName);
                        Trace.TraceInformation("\t Deleting from input container: " + blobItem.Name);
                        blobClient.Delete();
                    }
                }
                catch (Exception error)
                {
                    Trace.TraceInformation("error: " + error.Message);
                }
 
                await Task.Delay(1000);
            }
        }
    }
}

namespace WorkerRole1
{
    class Processor : IDisposable
    {
        EngineLoader engineLoader = null;
        private void displayMessage(string text)
        {
            File.AppendAllText(".\\FRE.log", text + "\n");
            Trace.TraceInformation("\t" + text );
        }        
        private void loadProfile()
        {
            engineLoader.Engine.LoadPredefinedProfile("DocumentConversion_Accuracy");
        }        
        private void setupFREngine()
        {
            displayMessage("Loading predefined profile...");
            loadProfile();
        }
        public void LoadEngine()
        {
            try {
                if (engineLoader == null)
                {
                    // 与 Hello 示例中的 EngineLoader 相同
                    engineLoader = new EngineLoader();
                }
                setupFREngine();
            }
            catch (Exception error)
            {
                displayMessage("error: " + error.Message);
            }
        }
        public void UnloadEngine()
        {
            try
            {
                if (engineLoader != null)
                {
                    engineLoader.Dispose();
                    engineLoader = null;
                }
            }
            catch (Exception error)
            {
                displayMessage("error: " + error.Message);
            }
        }
 
        public string ProcessImage( MemoryStream inputMemoryStream, string inputBlobName )
        {
            FRDocument document = engineLoader.Engine.CreateFRDocument();
            string resultBlobName = "";
            try
            {
               document.PageFlushingPolicy = FREngine.PageFlushingPolicyEnum.PFP_KeepInMemory;
 
                // 将图像文件添加到文档中
                displayMessage("Loading image...");
                IntPtr handle = Marshal.AllocHGlobal(inputMemoryStream.GetBuffer().Length);
                Marshal.Copy(inputMemoryStream.GetBuffer(), 0, handle, inputMemoryStream.GetBuffer().Length);
                document.AddImageFileFromMemory(handle.ToInt64(), null, null);
 
                // 识别文档
                displayMessage("Recognizing...");
                document.Process(null);
 
                // 保存结果
                displayMessage("Saving results...");
                // 使用“balanced”方案将结果保存为 pdf
                // FREngine.PDFExportParams pdfParams = engineLoader.Engine.CreatePDFExportParams();
                // pdfParams.Scenario = FREngine.PDFExportScenarioEnum.PES_Balanced;
 
                WorkerRole1.Engine.FileWriter fileWriter = new WorkerRole1.Engine.FileWriter(inputBlobName, ".pdf");
                resultBlobName = inputBlobName + ".pdf";
                document.ExportToMemory(fileWriter, FREngine.FileExportFormatEnum.FEF_PDF, null);
 
            }
            catch (Exception error)
            {
                displayMessage("error: " + error.Message);
            }
            finally
            {
                // 关闭文档
                document.Close();
            }
 
            return resultBlobName;
        }
 
        public void Dispose()
        {
            UnloadEngine();
        }
    }
}

class Config
{
    // 在 ServiceDefinition.csdef 中定义的本地存储
    private static string LocalStorageName = "LocalStorage1";
    // Blob 容器的连接字符串
    public static readonly string ConnectionString = "<connection_string_to_frestorage>";
    // 存储中输入和输出容器的名称
    public static readonly string InputContainerName = "fre-input";
    public static readonly string OutputContainerName = "fre-output";
 
    // 返回 ABBYY FineReader Engine 的 Customer Project ID
    public static String GetCustomerProjectId()
    {
        return "<Your_Customer_Project_ID>";
    }
 
    // 返回许可证令牌名称
    public static String GetLicenseTokenName()
    {
        return "<Token_number>.ABBYY.ActivationToken";
    }
 
    // 返回许可证密码
    public static String GetLicensePassword()
    {
        return "<Your_Online_License_token_password>";
    }
    // 返回引擎路径
    public static String GetEngineFolder()
    {
        string engineSubfolder = "fre_packages\\LibraryPackage\\Bin64";
        string engineDllFolder = Path.Combine(RoleEnvironment.GetLocalResource(LocalStorageName).RootPath, engineSubfolder);
        return engineDllFolder;
    }
}