PDF on z/OS - Arabic in PDF not displayed on Windows
PDF on z/OS - Arabic in PDF not displayed on Windows
I have download the jZOS toolkit from Dovetail and executed the supplied sample successfully and produced the PDF on z/OS. I then edited the sample and changed some names to Arabic, just to see if arabic would be displayed in the PDF. The z/os codepage is 420 and I tried file.encoding=ISO-8859-6 (for windows 1256) and also ISO-8859-1. When the PDF is displayed under Windows, the Arabic is not displayed, being replaced by latin characters. Is this a problem with the Windows fonts or is there another setting I need in JZOS?
Re: PDF on z/OS - Arabic in PDF not displayed on Windows
I don't understand... the JZOS toolkit is an IBM product. Where did you download it? What specific sample program are you referring to?
Re: PDF on z/OS - Arabic in PDF not displayed on Windows
I m using the iTextPdfGenerator through jZOS. I downloaded this from the DOvetail website in the Case Study section:
http://www.dovetail.com/downloads/coz/index.html
Generation of multi-page PDF documents with Co:Z, Release 0.1.0 - May 15, 2012 Executables, Java source and JCL, coz-itext-0.1.0-xmit.zip
http://www.dovetail.com/downloads/coz/index.html
Generation of multi-page PDF documents with Co:Z, Release 0.1.0 - May 15, 2012 Executables, Java source and JCL, coz-itext-0.1.0-xmit.zip
Re: PDF on z/OS - Arabic in PDF not displayed on Windows
There could be two places where the encoding needs to be fixed:
1) Reading the input CSV data from a file or dataset into Java (Unicode) strings.
The sample iText program reads input CSV data using the sample FileFactory class.
If you are reading from a dataset (or DD), the sample code uses the default z/OS encoding, which is usually IBM-1047:
So, the first thing would be to check the input data to see if the characters that you want are properly represented in that codepage.
If you want a different codepage, you would need to modify the sample program to allow you to read data from a dataset in a different codepage.
2) The PDF template field's encoding and font.
I'm not sure how to change this. Best to rule out (1) as the problem first.
1) Reading the input CSV data from a file or dataset into Java (Unicode) strings.
The sample iText program reads input CSV data using the sample FileFactory class.
If you are reading from a dataset (or DD), the sample code uses the default z/OS encoding, which is usually IBM-1047:
Code: Select all
private static BufferedReader getZFileBufferedReader(String filename) throws IOException {
ZFile zFile = new ZFile(ZFile.getSlashSlashQuotedDSN(filename, true), "rt");
return new BufferedReader(new InputStreamReader(zFile.getInputStream(), ZUtil.getDefaultPlatformEncoding()));
}
If you want a different codepage, you would need to modify the sample program to allow you to read data from a dataset in a different codepage.
2) The PDF template field's encoding and font.
I'm not sure how to change this. Best to rule out (1) as the problem first.