private async Task ConvertFiles(string id, List <RepositoryFile> files, PaketDIP paket, string rootFolder, string tempFolder, JobContext context)
{
// Skip empty collections
if (files.Count == 0)
{
return;
}
// Create the list with conversion files.
// This list will contain the splitted file names for processing
// This list does not contain files that didn't have the flag exported or should be skipped
var conversionFiles = pdfManipulator.ConvertToConversionFiles(files.ToList(), tempFolder, true);
var sw = new Stopwatch();
sw.Start();
var parallelism = Settings.Default.DocumentTransformParallelism;
Log.Information("Starting parallel document transform for-each-loop with parallelism of {parallelism} for {Count} files of archiveRecordId or orderId {id}",
parallelism, files.Count, id);
var supportedFileTypesForRendering = await renderEngine.GetSupportedFileTypes();
await conversionFiles.ParallelForEachAsync(async conversionFile =>
{
var file = new FileInfo(conversionFile.FullName);
Log.Information("Start conversion for file: {file} for archive record or order id {id}", file, id);
conversionFile.ConvertedFile = await ConvertFile(file, supportedFileTypesForRendering, context);
}, parallelism, true);
// Now stich back files that were possibly splitted
pdfManipulator.MergeSplittedFiles(conversionFiles);
// Update the metadata.xml for all the converted files
// As speed is not an issue, we're not doing it in parallel
foreach (var conversionFile in conversionFiles)
{
var file = new FileInfo(conversionFile.FullName);
if (string.IsNullOrEmpty(conversionFile.ParentId))
{
MetadataXmlUpdater.UpdateFile(file, new FileInfo(conversionFile.ConvertedFile), paket, rootFolder);
}
// Delete the original file, if the convertedFile exists and is not the same as the original file.
// In case of PDF the name of the original and converted file could be the same. --> PDF to PDF with OCR
if (file.Exists && conversionFile.ConvertedFile != file.FullName)
{
file.Delete();
}
}
sw.Stop();
Log.Information("Finished parallel document transform for-each-loop with parallelism of {parallelism} for {Count} files of archiveRecordId or orderId {id} in {TotalSeconds}",
parallelism, files.Count, id, sw.Elapsed.TotalSeconds);
}
private void DeletePremisFiles(List <DateiDIP> premisFiles)
{
foreach (var file in premisFiles)
{
var path = GetPath(file.Id, paketToConvert.Inhaltsverzeichnis.Ordner, rootFolder);
var fileName = Path.Combine(path, file.Name);
if (File.Exists(fileName))
{
MetadataXmlUpdater.RemoveFile(new FileInfo(fileName), paketToConvert, rootFolder);
File.Delete(fileName);
}
}
}
private void CreatePdfFile(List <FileInfo> imageFileList, string pdfFileName)
{
var pdfDocument = new Document();
var parents = new DateiParents();
foreach (var imageFile in imageFileList)
{
AddPage(pdfDocument, imageFile.FullName);
pdfDocument.FreeMemory();
parents = MetadataXmlUpdater.RemoveFile(imageFile, paketToConvert, rootFolder);
imageFile.Delete();
}
pdfDocument.Save(pdfFileName);
MetadataXmlUpdater.AddFile(new FileInfo(pdfFileName), parents);
}
private async Task <string> ConvertFile(FileInfo file, PaketDIP paket, string tempFolder)
{
if (!file.Exists)
{
throw new FileNotFoundException($"Unable to find file {file.FullName}", file.FullName);
}
var supportedFileTypesForRendering = await renderEngine.GetSupportedFileTypes();
if (!supportedFileTypesForRendering.Contains(file.Extension.Replace(".", "").ToLowerInvariant()))
{
return(file.FullName);
}
var targetExtension = GetTargetExtension(file);
var convertedFile = await renderEngine.ConvertFile(file.FullName, targetExtension);
MetadataXmlUpdater.UpdateFile(file, new FileInfo(convertedFile), paket, tempFolder);
return(convertedFile);
}