//-------------------------------------------------------------------------
private bool ExtractImages(String PDFSourcePath, Audit sourceAudit, out Exception exception)
{
//List<System.Drawing.Image> ImgList = new List<System.Drawing.Image>();
exception = null;
iTextSharp.text.pdf.RandomAccessFileOrArray RAFObj = null;
iTextSharp.text.pdf.PdfReader PDFReaderObj = null;
iTextSharp.text.pdf.PdfObject PDFObj = null;
iTextSharp.text.pdf.PdfStream PDFStremObj = null;
int pageNumber = 0;
bool ok = false;
try
{
RAFObj = new iTextSharp.text.pdf.RandomAccessFileOrArray(PDFSourcePath);
PDFReaderObj = new iTextSharp.text.pdf.PdfReader(RAFObj, null);
for (int i = 0; i < PDFReaderObj.XrefSize; i++)
{
PDFObj = PDFReaderObj.GetPdfObject(i);
if ((PDFObj != null) && PDFObj.IsStream())
{
PDFStremObj = (iTextSharp.text.pdf.PdfStream)PDFObj;
iTextSharp.text.pdf.PdfObject subtype = PDFStremObj.Get(iTextSharp.text.pdf.PdfName.SUBTYPE);
if ((subtype != null) && subtype.ToString() == iTextSharp.text.pdf.PdfName.IMAGE.ToString())
{
try
{
iTextSharp.text.pdf.parser.PdfImageObject PdfImageObj =
new iTextSharp.text.pdf.parser.PdfImageObject((iTextSharp.text.pdf.PRStream)PDFStremObj);
System.Drawing.Image ImgPDF = PdfImageObj.GetDrawingImage();
pageNumber++;
if (pageNumber > PDFReaderObj.NumberOfPages)
return false;
}
catch (Exception)
{
}
}
}
}
pageNumber = 0;
PageCount = PDFReaderObj.NumberOfPages;
for (int i = 0; i < PDFReaderObj.XrefSize; i++)
{
PDFObj = PDFReaderObj.GetPdfObject(i);
if ((PDFObj != null) && PDFObj.IsStream())
{
PDFStremObj = (iTextSharp.text.pdf.PdfStream)PDFObj;
iTextSharp.text.pdf.PdfObject subtype = PDFStremObj.Get(iTextSharp.text.pdf.PdfName.SUBTYPE);
if ((subtype != null) && subtype.ToString() == iTextSharp.text.pdf.PdfName.IMAGE.ToString())
{
try
{
iTextSharp.text.pdf.parser.PdfImageObject PdfImageObj =
new iTextSharp.text.pdf.parser.PdfImageObject((iTextSharp.text.pdf.PRStream)PDFStremObj);
System.Drawing.Image ImgPDF = PdfImageObj.GetDrawingImage();
pageNumber++;
//using (Bitmap bmp = new Bitmap(ImgPDF))
//{
// SaveFrameFiles(PDFSourcePath, sourceAudit, bmp, pageNumber, PageCount);
try
{
SaveFrameFiles(PDFSourcePath, sourceAudit, (Bitmap)ImgPDF, pageNumber, PageCount);
}
catch (Exception)
{
ImgPDF.Dispose();
continue;
}
ImgPDF.Dispose();
ok = true;
//}
}
catch (Exception)
{ }
}
}
}
return ok;
}
catch (Exception)// ex
{
//log.LogMessage(PDFSourcePath + " Page" + pageNumber.ToString() + " ", ex);
return false;
}
finally
{
PDFReaderObj.Close();
PDFReaderObj.Dispose();
}
}
//-------------------------------------------------------------------------
public Audit CreateAuditAndMoveToTempFolder(string fileName, string directory, bool delete = true)
{
string sourceSHA1Hash = GetSHA1FromFile(fileName);
var tempFileName = directory + sourceSHA1Hash + Path.GetExtension(fileName);
var tempAuditFileName = directory + sourceSHA1Hash + ".audit";
Audit audit = new Audit(fileName, sourceSHA1Hash);
if (CanAccess(fileName))
{
File.Copy(fileName, tempFileName, true);
if (delete)
{
iOHelper.DeleteFile(fileName);
}
audit.Save(tempAuditFileName);
}
return audit;
}
//-------------------------------------------------------------------------
public Audit Clone()
{
var newAudit = new Audit()
{
fileName = this.fileName,
SHA1Hash = this.SHA1Hash,
type = this.type,
error = this.error,
dataFileName = this.dataFileName,
dataFileSHA1Hash = this.dataFileSHA1Hash,
sourceFilePath = this.sourceFilePath,
sourceFileName = this.sourceFileName,
sourceSHA1Hash = this.sourceSHA1Hash,
archiveFileName = this.archiveFileName,
sourcePage = this.sourcePage
};
return newAudit;
}
//-------------------------------------------------------------------------
public void MoveFileFromTemp(string dir, string incomingFile, Audit audit, string destFileName)
{
string tempFileName = dir + Path.GetFileName(incomingFile);
string tempAuditFileName = GetFileAuditName(tempFileName);
string destAuditFileName = GetFileAuditName(destFileName);
try
{
audit.Save(destAuditFileName);
File.Copy(tempFileName, destFileName, true);
iOHelper.DeleteFile(tempAuditFileName);
iOHelper.DeleteFile(tempFileName);
}
catch (Exception ex)
{
log.LogMessage(ex);
}
}
//-------------------------------------------------------------------------
public void MoveBadFile(string dir, string ManualNextProccessingFolder, string fileName, Audit audit, string exMessage)
{
if (!Directory.Exists(ManualNextProccessingFolder))
{
Directory.CreateDirectory(ManualNextProccessingFolder);
}
audit.MadeFailedAudit(exMessage);
var destFileName = ManualNextProccessingFolder + audit.sourceSHA1Hash + Path.GetExtension(fileName);
MoveFileFromTemp(dir, fileName, audit, destFileName);
}
//-------------------------------------------------------------------------
public string GetFileForRecognize(string fileName, string tempdirectory, bool showLog = true)
{
//bool useErrFolder = false;
if (string.IsNullOrEmpty(tempdirectory))
{
//useErrFolder = true;
tempdirectory = OcrAppConfig.TempEdFolder;
}
if (showLog) log.LogMessage("Get file for recognize: " + fileName);
if (!File.Exists(fileName))
{
if (showLog) log.LogMessage("File does not exist " + fileName);
return null;
}
string fileNameAudit = GetFileAuditName(fileName);
var sha1hash = GetSHA1FromFile(fileName);
if (File.Exists(fileNameAudit) && fileName.Contains(sha1hash) && fileName.Contains(tempdirectory))
{
return fileName;
}
if (fileName.Contains(sha1hash) && fileName.Contains(tempdirectory))
{
iOHelper.DeleteFile(fileNameAudit);
var audit = new Audit(fileName, GetSHA1FromFile(fileName));
audit.Save(GetFileAuditName(fileName));
return fileName;
}
string destfileName = tempdirectory + sha1hash + Path.GetExtension(fileName);
string destfileNameAudit = GetFileAuditName(destfileName);//The file exists.
if (File.Exists(fileNameAudit))// && fileName.Contains(sha1hash)
{
//try
//{
// File.Move(fileName, destfileName);
// File.Move(fileNameAudit, destfileNameAudit);
//}
//catch (Exception ex)
//{
// log.LogMessage(ex);
// return null;
//}
try
{
if (!File.Exists(destfileName))
{
File.Move(fileName, destfileName);
}
else
{
File.Delete(fileName);
}
}
catch (Exception ex)
{
log.LogMessage(ex);
}
try
{
if (!File.Exists(destfileNameAudit))
{
File.Move(fileNameAudit, destfileNameAudit);
}
else
{
File.Delete(fileNameAudit);
}
}
catch (Exception ex)
{
log.LogMessage(ex);
}
return destfileName;
}
else
{
log.LogMessage("File " + fileNameAudit + " not exists");
FileInfo fi = new FileInfo(fileName);
DateTime begEx = DateTime.Parse(fi.CreationTime.ToString());
TimeSpan ts = DateTime.Now - begEx;
if (ts.TotalSeconds < 2)
return "";
}
CreateAuditAndMoveToTempFolder(fileName, tempdirectory);
return destfileName;
}
//-------------------------------------------------------------------------
private void FileInTempFolder(string fileName)
{
Exception exception;
var auditFileName = utils.GetFileAuditName(fileName);
Audit sourceAudit = new Audit(auditFileName, out exception);
if (sourceAudit == null || !File.Exists(auditFileName))
{
log.LogMessage("___" + "FileInTempFolder (sourceAudit == null) : true");
if (exception != null)
log.LogMessage(exception);
File.Copy(fileName, defaults.InputFolder + Path.GetFileName(fileName), true);//если файл без файла-аудита -> перемещение в папку InputFolder
iOHelper.DeleteFile(fileName);
return;
}
Result = new ProcessingIncomingFileResult()
{
SourceFileName = fileName,
FileName = fileName,
Audit = sourceAudit
};
}
public Recognize(string fileName, Defaults defaults)
{
this.defaults = defaults;
if (String.IsNullOrEmpty(fileName) || !File.Exists(fileName))
{
exception = new Exception("File name is incorrect: " + fileName);
NotifyUpdated(NotifyKey.Exception, exception);
return;
}
FileName = fileName;
FileNameAudit = fileName.Replace(new FileInfo(fileName).Extension, ".audit");
#region Get Bitmap
PDFLibNet.PDFWrapper pdfDoc = null;
try
{
Image = Utils.GetBitmapFromFile(FileName, ref pdfDoc);
Image = Utils.NormalizeBitmap(Image);
Image = RecognitionTools.GetMonohromeNoIndexBitmap(Image);
}
catch (Exception ex)
{
exception = ex;
NotifyUpdated(NotifyKey.Exception, exception);
return;
}
finally
{
if (pdfDoc != null)
{
pdfDoc.Dispose();
pdfDoc = null;
}
}
#endregion
#region Get audit file
FileAudit = Utils.GetAuditFromFile(FileNameAudit, out exception);
if (FileAudit == null)
{
NotifyUpdated(NotifyKey.Exception, exception);
return;
}
#endregion
regionsExt = RecognitionTools.GetRegionsExt(out sheetIdentifiers, Defaults.ManualConfigsFolder);
}
//-------------------------------------------------------------------------
private void SaveFrameFiles(string fileName, Audit sourceAudit, Bitmap entryBitmap, int pageNumber, int pageCount)
{
if (workGetFramesFromPic)
return;
workGetFramesFromPic = true;
try
{
var tempdir = OcrAppConfig.TempFolder + "SaveFrame\\";
iOHelper.CreateDirectory(tempdir);
var temp = tempdir + sourceAudit.sourceSHA1Hash + "_" + pageNumber + ".tiff";
//iOHelper.DeleteFile(temp);
string[] ss = Directory.GetFiles(tempdir);
for (int j = 0; j < ss.Length; j++)
{
try
{
iOHelper.DeleteFile(ss[j]);
}
catch (Exception)
{
}
}
//Bitmap bmp = (Bitmap)entryBitmap.Clone();// new Bitmap(entryBitmap.Width, entryBitmap.Height, PixelFormat.Format24bppRgb);
//bmp = (Bitmap)bmp.GetThumbnailImage(bmp.Width, bmp.Height, null, IntPtr.Zero);
//bmp.SetResolution(entryBitmap.VerticalResolution, entryBitmap.HorizontalResolution);
//bmp.Save("SaveFrameFiles.bmp", ImageFormat.Bmp);
entryBitmap.Save(temp, ImageFormat.Tiff);
//bmp.Save(temp, ImageFormat.Tiff);
//bmp.Dispose();
var sha1Hash = utils.GetSHA1FromFile(temp);
var tempFileName = OcrAppConfig.TempFramesFolder + sha1Hash + ".tiff";
var tempFileNameAudit = OcrAppConfig.TempFramesFolder + sha1Hash + ".audit";
if (File.Exists(tempFileName) && File.Exists(tempFileNameAudit))
{
log.LogMessage("Frame exists page " + pageNumber.ToString() + " of " + pageCount.ToString() + " from " + sourceAudit.sourceFileName);
iOHelper.DeleteFile(temp);
//iOHelper.DeleteDirectory(tempdir, false);
workGetFramesFromPic = false;
return;
}
var audit = sourceAudit.GetFrameAudit(tempFileName, sha1Hash, pageNumber);
audit.Save(tempFileNameAudit);
log.LogMessage("Get frame page " + pageNumber.ToString() + " of " + pageCount.ToString() + " from " + sourceAudit.sourceFileName);
//File.Move(temp, tempFileName);
File.Copy(temp, tempFileName, true);
//System.Windows.Forms.Application.DoEvents();
//iOHelper.DeleteFile(temp);
//iOHelper.DeleteDirectory(tempdir, false);
//entryBitmap.Dispose();//недопустимый параметр
workGetFramesFromPic = false;
}
catch (Exception)
{
workGetFramesFromPic = false;
}
}
//-------------------------------------------------------------------------
private void GetFramesFromPic(string fileName, Audit sourceAudit, out Exception exception)
{
exception = null;
if (workGetFramesFromPic)
return;
workGetFramesFromPic = true;
Bitmap entryBitmap;
try
{
entryBitmap = GetPicFile(fileName);
}
catch (Exception ex)
{
exception = ex;
workGetFramesFromPic = false;
return;
}
int i = 0;
try
{
Guid[] guids = entryBitmap.FrameDimensionsList;
FrameDimension fd = new FrameDimension(guids[0]);
PageCount = entryBitmap.GetFrameCount(fd);
for (i = 0; i < PageCount; i++)
{
entryBitmap.SelectActiveFrame(fd, i);
//SaveFrameFiles(fileName, sourceAudit, entryBitmap, i + 1, PageCount);
var tempdir = OcrAppConfig.TempFolder + "SaveFrame\\";
iOHelper.CreateDirectory(tempdir);
string[] ss = Directory.GetFiles(tempdir);
for (int j = 0; j < ss.Length; j++)
{
try
{
iOHelper.DeleteFile(ss[j]);
}
catch (Exception)
{
}
}
var temp = tempdir + sourceAudit.sourceSHA1Hash + "_" + i + 1 + ".tiff";
//iOHelper.DeleteFile(temp);
//Bitmap bmp = (Bitmap)entryBitmap.Clone();// new Bitmap(entryBitmap.Width, entryBitmap.Height, PixelFormat.Format24bppRgb);
//bmp = (Bitmap)bmp.GetThumbnailImage(bmp.Width, bmp.Height, null, IntPtr.Zero);
//bmp.SetResolution(entryBitmap.VerticalResolution, entryBitmap.HorizontalResolution);
//bmp.Save("SaveFrameFiles.bmp", ImageFormat.Bmp);
entryBitmap.Save(temp, ImageFormat.Tiff);
//bmp.Save(temp, ImageFormat.Tiff);
//bmp.Dispose();
var sha1Hash = utils.GetSHA1FromFile(temp);
//проверить на ""!!!
var tempFileName = OcrAppConfig.TempFramesFolder + sha1Hash + ".tiff";
var tempFileNameAudit = OcrAppConfig.TempFramesFolder + sha1Hash + ".audit";
log.LogMessage("Get frame page " + (i + 1).ToString() + " of " + pageCount.ToString() + " from " + sourceAudit.sourceFileName);
if (File.Exists(tempFileName) && File.Exists(tempFileNameAudit))
{
log.LogMessage("Frame exists page " + (i + 1).ToString() + " of " + pageCount.ToString() + " from " + sourceAudit.sourceFileName);
try
{
iOHelper.DeleteFile(temp);
}
catch (Exception)
{
}
//iOHelper.DeleteDirectory(tempdir, false);
//entryBitmap.Dispose();
//return;
continue;
}
var audit = sourceAudit.GetFrameAudit(tempFileName, sha1Hash, i + 1);
audit.Save(tempFileNameAudit);
//File.Move(temp, tempFileName);
File.Copy(temp, tempFileName, true);
//System.Windows.Forms.Application.DoEvents();
//try
//{
// iOHelper.DeleteFile(temp);
//}
//catch (Exception ex)
//{
// log.LogMessage(fileName + " Page" + (i + 1).ToString() + " (for) ", ex);
// continue;
//}
//iOHelper.DeleteDirectory(tempdir, false);
//entryBitmap.Dispose();//недопустимый параметр
}
entryBitmap.Dispose();
workGetFramesFromPic = false;
}
catch (Exception ex)
{
workGetFramesFromPic = false;
entryBitmap.Dispose();
log.LogMessage(fileName + " Page" + (i + 1).ToString() + " ", ex);
}
}
//-------------------------------------------------------------------------
private void GetFramesFromPDF(string fileName, Audit sourceAudit, out Exception exception)
{
PdfReader doc = null;
PDFLibNet.PDFWrapper pdfDoc = null;
int pageNumber = 0;
exception = null;
//ImageExtractor
try
{
if (ExtractImages(fileName, sourceAudit, out exception))
return;
////for (int i = 0; i < list.Count; i++)
////{
//// var item = list[i];
//// using (Bitmap bmp = new Bitmap(item))
//// {
//// SaveFrameFiles(fileName, sourceAudit, bmp, i + 1, list.Count);
//// }
////}
////list.Clear();
//doc = new PdfReader(fileName);
//for (pageNumber = 1; pageNumber <= doc.NumberOfPages; pageNumber++)
//{
// //using (Bitmap bmp = GetBitmapFromPDFPage(doc, pageNumber))
// //{
// // SaveFrameFiles(fileName, sourceAudit, bmp, pageNumber, doc.NumberOfPages);
// //}
// PdfDictionary dict = doc.GetPageN(pageNumber);
// PdfDictionary res = (PdfDictionary)(PdfReader.GetPdfObject(dict.Get(PdfName.RESOURCES)));
// PdfDictionary xobj = (PdfDictionary)(PdfReader.GetPdfObject(res.Get(PdfName.XOBJECT)));
// if (xobj != null)
// {
// foreach (PdfName name in xobj.Keys)
// {
// PdfObject obj = xobj.Get(name);
// if (obj.IsIndirect())
// {
// PdfDictionary tg = (PdfDictionary)(PdfReader.GetPdfObject(obj));
// PdfName subtype = (PdfName)(PdfReader.GetPdfObject(tg.Get(PdfName.SUBTYPE)));
// if (PdfName.IMAGE.Equals(subtype))
// {
// int xrefIdx = ((PRIndirectReference)obj).Number;
// PdfObject pdfObj = doc.GetPdfObject(xrefIdx);
// PdfStream str = (PdfStream)(pdfObj);
// iTextSharp.text.pdf.parser.PdfImageObject pdfImage =
// new iTextSharp.text.pdf.parser.PdfImageObject((PRStream)str);
// System.Drawing.Image ImgPDF = pdfImage.GetDrawingImage();
// using (Bitmap bmp = new Bitmap(ImgPDF))
// {
// SaveFrameFiles(fileName, sourceAudit, bmp, pageNumber, doc.NumberOfPages);
// }
// }
// else if (PdfName.FORM.Equals(subtype) || PdfName.GROUP.Equals(subtype))
// {
// //images.AddRange(GetImagesFromPdfDict(tg, doc));
// }
// }
// }
// }
//}
try
{
pdfDoc = GetPDFDoc(fileName);
}
catch (Exception ex)
{
exception = ex;
return;
}
int i = 0;
if (pdfDoc != null)
{
PageCount = pdfDoc.PageCount;
for (i = 0; i < PageCount; i++)
{
var bmp = GetBitmapFromPDFPage(pdfDoc, i + 1);
SaveFrameFiles(fileName, sourceAudit, bmp, i + 1, PageCount);
bmp.Dispose();
}
}
}
catch (Exception ex)
{
log.LogMessage(fileName + " Page" + pageNumber.ToString() + " ", ex);
}
finally
{
if (doc != null)
{
doc.Close();
doc.Dispose();
}
if (pdfDoc != null)
{
pdfDoc.Dispose();
}
}
}
public void GetFrames(string fileName, Audit sourceAudit)
{
exception = null;
if (IsPDFFile(fileName))
{
GetFramesFromPDF(fileName, sourceAudit, out exception);
}
else
{
GetFramesFromPic(fileName, sourceAudit, out exception);
}
}
public Recognize(string fileName, OcrAppConfig defaults, CancellationToken token, bool auto = false
, bool normalizeBitmap = true)
{
SetCancellationToken(token);
this.Auto = auto;
this.defaults = defaults;
recTools.frameFileName = fileName;
//defaults.frameFileName = fileName;
if (String.IsNullOrEmpty(fileName) || !File.Exists(fileName))
{
Exception = new Exception("File name is incorrect: " + fileName);
log.LogMessage(Exception);
NotifyUpdated(ExceptionEvent, Exception, null);
return;
}
FileName = fileName;
AuditFileName = utils.GetFileAuditName(fileName);
#region Get Bitmap
FramesAndBitmap fab = new FramesAndBitmap();
Bitmap = fab.GetBitmapFromFile(FileName);
if (fab.Exception != null)
{
Exception = fab.Exception;
log.LogMessage(Exception);
NotifyUpdated(ExceptionEvent, Exception, null);
return;
}
//bool bppIndexed1 = false;
//if (Bitmap.PixelFormat == System.Drawing.Imaging.PixelFormat.Format1bppIndexed)
// bppIndexed1 = true;
if (normalizeBitmap)
Bitmap = recTools.NormalizeBitmap(Bitmap, out exception);
if (exception != null)
{
log.LogMessage(Exception);
NotifyUpdated(ExceptionEvent, Exception, null);
return;
}
//if (exception != null)
//{
// if (auto)
// {
// fab.Exception = exception;
// Audit = new Audit(AuditFileName, out exception);
// exception = fab.Exception;
// Audit.error = exception.Message;
// if (defaults.MoveToNextProccessingFolderOnSheetIdentifierError)
// {
// utils.MoveBadFile(OcrAppConfig.TempFramesFolder, defaults.ManualNextProccessingFolder, fileName, Audit, fab.Exception.Message);//сохраняем файл в папку для дальнейшей обработки и удаляем из ТЕМПА
// File.Delete(fileName);
// }
// else
// {
// }
// return;
// }
//}
//else if (exception != null)
//{
//Bitmap = recTools.GetMonohromeNoIndexBitmap(Bitmap, false, true);
//Bitmap.SetResolution(96, 96);
#endregion
#region Get audit file
Audit = new Audit(AuditFileName, out exception);
if (Audit == null || !File.Exists(AuditFileName))
{
log.LogMessage(Exception);
NotifyUpdated(ExceptionEvent, Exception, null);
return;
}
#endregion
regionsList = recTools.GetAllRegions(ConfigsFolder);
regionsListFLEX = recTools.GetAllRegionsFLEX(ConfigsFolder);
}