c#抽取pdf文档标题（2）

  public class IETitle
     {
         public static List<WordInfo> WordsInfo = new List<WordInfo>();

         private static string pdfcontent;
         public static HandleResult GetTitle(string path, string realtitle)
         {
             WordsInfo.Clear();

             string content = string.Empty;
             try
             {
                 content = ITextSharpLib.ExtractTextFromPdf(path);
             }
             catch
             {
                 try
                 {
                     content = PDFBoxLib.Pdf2txt(path);
                 }
                 catch (Exception ex)
                 {

                 }
             }

             pdfcontent = content;

             PDFBoxLib.HandleContent(path);

             //处理字符

             Word w = new Word();
             w.MakeWord(WordsInfo);

             Line line = new Line();
             line.MakeLine(w);

             //处理行
             Block block = new Block();
             block.MakeBlock(line);

             //获取全部的文本
             string text = string.Empty;

             try
             {
                 text = ITextSharpLib.ExtractTextFromPdf(path, );
             }
             catch (Exception ex)
             {
                 text = content;
             }

             HandleResult title = new HandleResult() { Title = "" };

             try
             {
                 var sentences = text.Split('\n');

                 InfoExtract ie = new InfoExtract(sentences, text);

                 title = ie.ExtractTitle(block, realtitle);

             }
             catch (Exception ex)
             {
                 Logger.Debug(ex.Message);
             }

             return title;
         }
     }

上面就是获取标题的整体逻辑代码。29行，是调用pdfboxLib，读取pdf第一页内容：

  public static string HandleContent(string fileName, int pageIndex = )
         {
             try
             {
                 PDDocument document = null;
                 try
                 {
                     document = PDDocument.load(fileName);
                     List allPages = document.getDocumentCatalog().getAllPages();

                     int size = pageIndex ==  ? allPages.size() : ;

                     for (int i = ; i < size; i++)
                     {
                         var page = (PDPage)allPages.get(i);

                         var contents = page.getContents();

                         PrintTextLocatins2 printer = new PrintTextLocatins2();

                         if (contents != null)
                         {
                             printer.processStream(page, page.findResources(), page.getContents().getStream());
                         }
                     }
                 }
                 catch (Exception ex)
                 {
                 }
                 finally
                 {
                     if (document != null)
                     {
                         document.close();
                     }
                 }
             }
             catch (Exception ex)
             {

             }
             return "";
         }

第23行 printer.processStream方法，会触发自定义类PrintTextLocation2类中的字符处理方法 processTextPosition：

  public class PrintTextLocatins2 : PDFTextStripper
     {
         private static int BOLD_F_NUM = ;
         private static String[] BOLD_FLAGS = { "Bold", "CAJ FNT04" };
         private static int ITALIC_F_NUM = ;
         private static String[] ITALIC_FLAGS = { "Italic", "CAJ FNT03" };

         private static bool IsBold(String font)
         {
             int i;
             for (i = ; i < BOLD_F_NUM; i++)
                 if (font.Contains(BOLD_FLAGS[i]))
                     return true;
             return false;
         }

         private static bool IsItalic(String font)
         {
             int i;
             for (i = ; i < ITALIC_F_NUM; i++)
                 if (font.Contains(ITALIC_FLAGS[i]))
                     return true;
             return false;
         }

         public PrintTextLocatins2()
         {
             base.setSortByPosition(false);
         }
         protected override void processTextPosition(TextPosition text)
         {

             WordInfo info = new WordInfo()
             {
                 X = text.getX(),
                 Y = text.getY(),
                 XDirAdj = text.getXDirAdj(),
                 YDirAdj = text.getYDirAdj(),
                 FontSize = text.getFontSize(),
                 Xscale = text.getXScale(),
                 Yscale = text.getYScale(),
                 Height = text.getHeight(),
                 Space = text.getWidthOfSpace(),
                 Width = text.getWidth(),

                 Subfont = text.getFont().getSubType(),
                 Basefont = text.getFont().getBaseFont(),
                 IsBold = IsBold(text.getFont().getBaseFont()),
                 IsItalic = IsItalic(text.getFont().getBaseFont()),

                 XSize = (int)(text.getFontSize() * text.getXScale()),

                 YSize = (int)(text.getFontSize() * text.getYScale()),

                 Word = text.getCharacter()
             };

             if (info.Space.ToString() == "非数字")
             {
                 info.Space = ;
             }

             IETitle.WordsInfo.Add(info);
         }
     }

这样我们就利用pdfbox收集了pdf文档的字符信息。