C# pdf.PdfReader類代碼示例

        public static string ExtractTextFromPdf(string path)
        {
            ITextExtractionStrategy its = new iTextSharp.text.pdf.parser.LocationTextExtractionStrategy();

            using (PdfReader reader = new PdfReader(path))
            {
                StringBuilder text = new StringBuilder();

                for (int page = 1; page <= reader.NumberOfPages; page++)
                {
                    ITextExtractionStrategy strategy = new SimpleTextExtractionStrategy();

                    string currentText = PdfTextExtractor.GetTextFromPage(reader, page, strategy);

                    currentText = Encoding.UTF8.GetString(ASCIIEncoding.Convert(Encoding.Default, Encoding.UTF8, Encoding.Default.GetBytes(currentText)));
                    text.Append(currentText);
                }

                System.IO.StreamWriter file = new System.IO.StreamWriter("c:\\test.txt");
                file.WriteLine(text);

                file.Close();

                return text.ToString();
            }
        }

        // 1. save XFA form XML stream to a file to inspect form fields
        public void SaveDatasetsXml(string readerPath)
        {
            using (var reader = new PdfReader(readerPath))
            {
                var xfa = new XfaForm(reader);
                XmlDocument doc = xfa.DomDocument;
                //// remove namespace so XML is easier to read
                //if (!string.IsNullOrEmpty(doc.DocumentElement.NamespaceURI))
                //{
                //    doc.DocumentElement.SetAttribute("xmlns", "");
                //    var temp = new XmlDocument();
                //    temp.LoadXml(doc.OuterXml);
                //    doc = temp;
                //}

                var sb = new StringBuilder();
                var xSettings = new XmlWriterSettings() 
                {
                    Indent = true,
                    WriteEndDocumentOnClose = true
                };
                using (var writer = XmlWriter.Create(sb, xSettings))
                {
                    doc.Save(writer);
                }
                File.WriteAllText(GetXmlPath(readerPath), sb.ToString());
            }
        }

        public static byte[] MergePdfs(IEnumerable<byte[]> inputFiles)
        {
            MemoryStream outputStream = new MemoryStream();
            Document document = new Document();
            PdfWriter writer = PdfWriter.GetInstance(document, outputStream);
            document.Open();
            PdfContentByte content = writer.DirectContent;

            foreach (byte[] input in inputFiles)
            {
                PdfReader reader = new PdfReader(input);
                for (int i = 1; i <= reader.NumberOfPages; i++)
                {
                    document.SetPageSize(reader.GetPageSizeWithRotation(i));
                    document.NewPage();
                    PdfImportedPage page = writer.GetImportedPage(reader, i);
                    int rotation = reader.GetPageRotation(i);
                    if (rotation == 90 || rotation == 270)
                    {
                        content.AddTemplate(page, 0, -1f, 1f, 0, 0,
                            reader.GetPageSizeWithRotation(i).Height);
                    }
                    else
                    {
                        content.AddTemplate(page, 1f, 0, 0, 1f, 0, 0);
                    }
                }
            }

            document.Close();

            return outputStream.ToArray();
        }

        private void button2_Click(object sender, EventArgs e)
        {
            OpenFileDialog dlg = new OpenFileDialog();
            string filepath;
            dlg.Filter = "PDF Files(*.PDF)|*.PDF|All Files(*.*)|*.*";
            if(dlg.ShowDialog()==DialogResult.OK)
            {
                filepath = dlg.FileName.ToString();
            
            string strtext = string.Empty;
            try
            {
                PdfReader reader = new PdfReader(filepath);
                for(int page=1;page<=reader.NumberOfPages;page++)
                {
                    ITextExtractionStrategy its = new iTextSharp.text.pdf.parser.LocationTextExtractionStrategy();
                    String s = PdfTextExtractor.GetTextFromPage(reader, page, its);
                    s = Encoding.UTF8.GetString(ASCIIEncoding.Convert(Encoding.Default, Encoding.UTF8, Encoding.Default.GetBytes(s)));
                    strtext = strtext + s;
                    richTextBox1.Text = strtext;

                }
                reader.Close();
            }
            catch(Exception ex)
            {
                MessageBox.Show(ex.Message);
            }
        
           }
            
        }

        static void Main(string[] args)
        {
            List<SchoolTestResult> result = new List<SchoolTestResult>();

            string[] urls = ConfigurationManager.AppSettings["urls"].Split(',');
            Console.WriteLine("Reading " + urls.Length + " files");

            foreach (string url in urls)
            {
                Uri uri = new Uri(url);
                PdfReader reader = new PdfReader(uri);

                try
                {
                    String text = PdfTextExtractor.GetTextFromPage(reader, 1);
                    result.AddRange(new AssessmentReader().parse(text));
                }
                catch (Exception e)
                {
                    Console.WriteLine("Error Reading " + url);
                    Console.WriteLine(e);
                }
            }

            string jsonOutput = JsonConvert.SerializeObject(result, Formatting.Indented);
            string csvOutput = new FileHelperEngine<SchoolTestResult>().WriteString( result );

            Console.WriteLine(jsonOutput);
            Console.WriteLine(csvOutput);

            //Console.WriteLine(text);
            Console.ReadLine();
        }

 // ---------------------------------------------------------------------------
 public void Write(Stream stream)
 {
     // Use old example to create PDF
     MovieTemplates mt = new MovieTemplates();
     byte[] pdf = Utility.PdfBytes(mt);
     using (ZipFile zip = new ZipFile())
     {
         using (MemoryStream ms = new MemoryStream())
         {
             // step 1
             using (Document document = new Document())
             {
                 // step 2
                 PdfWriter writer = PdfWriter.GetInstance(document, ms);
                 // step 3
                 document.Open();
                 // step 4
                 PdfPTable table = new PdfPTable(2);
                 PdfReader reader = new PdfReader(pdf);
                 int n = reader.NumberOfPages;
                 PdfImportedPage page;
                 for (int i = 1; i <= n; i++)
                 {
                     page = writer.GetImportedPage(reader, i);
                     table.AddCell(Image.GetInstance(page));
                 }
                 document.Add(table);
             }
             zip.AddEntry(RESULT, ms.ToArray());
         }
         zip.AddEntry(Utility.ResultFileName(mt.ToString() + ".pdf"), pdf);
         zip.Save(stream);
     }
 }

 /**
  * Creates new PdfStamperImp.
  * @param reader reads the PDF
  * @param os the output destination
  * @param pdfVersion the new pdf version or '\0' to keep the same version as the original document
  * @param append
  * @param conformanceLevel PDF/A conformance level of a new PDF document
  * @throws DocumentException on error
  * @throws IOException
  */
 internal PdfAStamperImp(PdfReader reader, Stream os, char pdfVersion, bool append, PdfAConformanceLevel conformanceLevel)
     : base(reader, os, pdfVersion, append)
 {
     ((IPdfAConformance)pdfIsoConformance).SetConformanceLevel(conformanceLevel);
     PdfAWriter.SetPdfVersion(this, conformanceLevel);
     ReadPdfAInfo();
 }

        private static String GetPdfContent(string path)
        {
            var text = new StringBuilder();

            using (var pdfReader = new PdfReader(path))
            {
                for (var page = 1; page <= pdfReader.NumberOfPages; page++)
                {
                    ITextExtractionStrategy strategy = new SimpleTextExtractionStrategy(); //  LocationTextExtractionStrategy();

                    var currentText = PdfTextExtractor.GetTextFromPage(
                        pdfReader,
                        page,
                        strategy);

                    currentText =
                        Encoding.UTF8.GetString(Encoding.Convert(
                            Encoding.Default,
                            Encoding.UTF8,
                            Encoding.Default.GetBytes(currentText)));

                    text.Append(currentText);
                }
            }
            return text.ToString();
        }

// ---------------------------------------------------------------------------
    public void Write(Stream stream) {
      // use one of the previous examples to create a PDF
      MovieTemplates mt = new MovieTemplates();
      // Create a reader
      byte[] pdf = Utility.PdfBytes(mt);
      PdfReader reader = new PdfReader(pdf);
      // loop over all the pages in the original PDF
      int n = reader.NumberOfPages;      
      using (ZipFile zip = new ZipFile()) {
        for (int i = 0; i < n; ) {
          string dest = string.Format(RESULT, ++i);
          using (MemoryStream ms = new MemoryStream()) {
// We'll create as many new PDFs as there are pages
            // step 1
            using (Document document = new Document()) {
              // step 2
              using (PdfCopy copy = new PdfCopy(document, ms)) {
                // step 3
                document.Open();
                // step 4
                copy.AddPage(copy.GetImportedPage(reader, i));
              }
            }
            zip.AddEntry(dest, ms.ToArray());
          }
        }
        zip.AddEntry(Utility.ResultFileName(mt.ToString() + ".pdf"), pdf);
        zip.Save(stream);       
      }
    }

        private String GetPdfText(String year, String servantID)
        {
            HttpWebRequest request = (HttpWebRequest)WebRequest.Create("https://www1.fgmaiss.com.br/contabil/relflh/flhrel037pdf.php?prc=MjkuMTM4LjM3Ny8wMDAxLTkz&etd=MDE=&ser=" + servantID + "&ano=" + year + "&fv1=300&fv2=000&fv3=000&fv4=000&fv5=000&fv00=500&fv01=186&fv02=008&fv03=012&fv04=071&fv05=007&fv06=073&fv07=189&fv08=406&fv09=039&fv10=041&fv11=000&fv12=000&fv13=000&fv14=000&fv15=000");
            request.Host = "www1.fgmaiss.com.br";
            request.CookieContainer = new CookieContainer();
            request.CookieContainer.Add(new Cookie("8jpo2jrlp3005q3lj3qsbf5hq7PESQGRIDfindGrid", "") { Domain = request.Host });
            request.CookieContainer.Add(new Cookie("PHPSESSID", Cookie) { Domain = request.Host });
            ServicePointManager.SecurityProtocol = SecurityProtocolType.Ssl3;
            ServicePointManager.ServerCertificateValidationCallback = delegate { return true; };

            ServicePointManager.Expect100Continue = true;

            //salva conteúdo da resposta (pdf) http em um arquivo
            var responseStream = request.GetResponse().GetResponseStream();
            String pdfFilePath = "c:/parsal/" + servantID +  year.ToString() + ".pdf";
            var pdfOut = File.Create(pdfFilePath);
            responseStream.CopyTo(pdfOut);
            pdfOut.Close();

            //resultado
            StringBuilder pdfText = new StringBuilder();
            PdfReader pdfReader = new PdfReader(pdfFilePath);
            for (int i = 1; i <= pdfReader.NumberOfPages; i++)
                pdfText.Append(PdfTextExtractor.GetTextFromPage(pdfReader, i));

            return pdfText.ToString();
        }

 public bool WatermarkPDF_SN(string SourcePdfPath, string OutputPdfPath, string WatermarkPath, int positionX, int positionY, int WatermarkHeight, int WatermarkWidth, out string msg)
 {
     try
     {
         PdfReader reader = new PdfReader(SourcePdfPath);
         PdfStamper stamp = new PdfStamper(reader, new FileStream(OutputPdfPath, FileMode.Create));
         int n = reader.NumberOfPages;
         int i = 0;
         PdfContentByte under;
         iTextSharp.text.Image im = iTextSharp.text.Image.GetInstance(WatermarkPath);
         im.SetAbsolutePosition(positionX, positionY);
         im.ScaleAbsolute(WatermarkWidth, WatermarkHeight);
         while (i < n)
         {
             i++;
             under = stamp.GetUnderContent(i);
             under.AddImage(im, true);
         }
         stamp.Close();
         reader.Close();
     }
     catch (Exception ex)
     {
         msg = ex.Message;
         return false;
     }
     msg = "��ˮӡ�ɹ���";
     return true;
 }

        public byte[] FillIRS941Form(ReportPayrollCompanyTotal payrollData)
        {
            string iRS941FormTemplate = HttpContext.Current.Server.MapPath("~/FormTemplates/IRS2015Form941.pdf");
            byte[] formResults;

            using (PdfReader pdfReader = new PdfReader(iRS941FormTemplate))
            {
                using (MemoryStream ms = new MemoryStream())
                {
                    using (PdfStamper pdfStamper = new PdfStamper(pdfReader, ms))
                    {
                        AcroFields pdfFormFields = pdfStamper.AcroFields;
                        var s = pdfFormFields.GetSignatureNames();
                        var e = pdfFormFields.GetBlankSignatureNames();

                        pdfFormFields.SetField("f1_10_0_", "Suitland Road Baptist Church");

                        pdfStamper.FormFlattening = false;
                        pdfStamper.Close();
                    }

                    formResults = ms.GetBuffer();
                }
            }

            return formResults;
        }

        /// <summary>
        /// 根據文件名進行分類
        /// </summary>
        /// <param name="fileName"></param>
        /// <returns>0:正常，1：加密,2:異常</returns>
        public static int ClassifyByWjm(string fileDir, string fileName)
        {
            if (fileName == null || "".Equals(fileName)) return 2; //文件名為空，不存在全文

            //文件原來所在目錄
            string fullFileName = fileDir;

            fullFileName += ConstructFilePath(fileName, 1);
            fullFileName += fileName;

            if (!File.Exists(fullFileName)) return 2; //有文件名，但是文件不存在

            if (File.ReadAllBytes(fullFileName).Length == 0) return 2; //文件大小為0

            try
            {
                PdfReader reader = new PdfReader(fullFileName);
                if (reader.IsEncrypted()) return 1; //文件加密
            }
            catch
            {
                return 2;//PDF文件損壞
            }

            //正常
            return 0;
        }

        /*
         * Explodes a pdf
         * */
        public static List<string> explodePdf(String inFile, String extractor, String repair)
        {

            FileInfo f = new FileInfo(inFile);
            inFile = f.FullName;
            List<string> theParts = new List<string>();

            try
            {
                CMDUtil cmd = new CMDUtil();
                UtilManager.repairPDF(inFile, repair);
                PdfReader reader = new PdfReader(inFile);
                int n = reader.NumberOfPages;
                String str;
                cmd.explode(inFile, extractor, "1-" + n);
                for (int i = 0; i < n; i++)
                {
                    str = f.FullName.Replace(".pdf", "-x" + (i + 1) + ".pdf");
                    UtilManager.repairPDF(str, repair);
                    theParts.Add(str);
                }
            }
            catch (Exception e)
            {
                System.Diagnostics.Debug.WriteLine(e.StackTrace);
            }
            System.Diagnostics.Debug.WriteLine("Exploded: " + theParts);
            return theParts;
        }

 public void Sign(String src, String dest,
                  ICollection<X509Certificate> chain, X509Certificate2 pk,
                  String digestAlgorithm, CryptoStandard subfilter,
                  String reason, String location,
                  ICollection<ICrlClient> crlList,
                  IOcspClient ocspClient,
                  ITSAClient tsaClient,
                  int estimatedSize) {
     // Creating the reader and the stamper
     PdfReader reader = null;
     PdfStamper stamper = null;
     FileStream os = null;
     try {
         reader = new PdfReader(src);
         os = new FileStream(dest, FileMode.Create);
         stamper = PdfStamper.CreateSignature(reader, os, '\0');
         // Creating the appearance
         PdfSignatureAppearance appearance = stamper.SignatureAppearance;
         appearance.Reason = reason;
         appearance.Location = location;
         appearance.SetVisibleSignature(new Rectangle(36, 748, 144, 780), 1, "sig");
         // Creating the signature
         IExternalSignature pks = new X509Certificate2Signature(pk, digestAlgorithm);
         MakeSignature.SignDetached(appearance, pks, chain, crlList, ocspClient, tsaClient, estimatedSize,
                                    subfilter);
     }
     finally {
         if (reader != null)
             reader.Close();
         if (stamper != null)
             stamper.Close();
         if (os != null)
             os.Close();
     }
 }