Re: Extracting Arabic text from a PDF

[Hesham G.]

Thanks Andreas ...

I first agreed with you that there's no need to redefine the Encoding like 
this :
String yourString = new String("pasted text from word","UTF-8");
As a String already contains Unicode, so Bulgarian or Arabic text surely 
appears okay inside that String. Am I right ?
But anyway I tried what you asked me and it produced question-marks "???? 
????" in the pdf :
                                           try {
                        byte[] arabic_bytes = "بسم الله".getBytes();  // Arabic 
Text.
                        String arabic_text = new String( arabic_bytes, "UTF-8" 
);
                        PDDocument pdfFile = new PDDocument();
			String filePath = "C:\\new_arabic.pdf";			PDTrueTypeFont font = PDTrueTypeFont.loadTTF( pdfFile, new 
File("fonts/times.ttf") );  // Windows Arabic font
                        //font.setEncoding( new WinAnsiEncoding() );

                        PDPage page = new PDPage();
			PDPageContentStream contentStream = new PDPageContentStream( pdfFile, 
page );

                        contentStream.beginText();
                        contentStream.setFont( font, 10 );
                        contentStream.moveTextPositionByAmount( 50, 720 );
                        contentStream.drawString( arabic_text );
                        contentStream.moveTextPositionByAmount( 0, -12 );
                        contentStream.endText();
                        contentStream.close();

                        pdfFile.addPage( page );
                        pdfFile.save( filePath );
                        pdfFile.close();
                } catch (IOException e) {
                        e.printStackTrace();
                } catch (COSVisitorException e) {
                        e.printStackTrace();
                }

I think the main problem is when writing that String inside the PDF which 
needs an Arabic font(Which I already did) + A suitable Encoding to use when 
writing the text in the PDF, as we do when we write the String inside a text 
file. Am I right ?
OutputStreamWriter fout = new OutputStreamWriter( new FileOutputStream( 
textFile ), "UTF-8" );

I checked the class WinAnsiEncoding ... I don't understand a lot about 
encoding, but is there a way to define the Bulgarian/Arabic characters 
inside that class using the method addCharacterEncoding(int code, COSName 
name); ?


Thanks ,
Hesham