Problem while Calling a CGI pgm From Java with code sample

Question

Hey guys,
I am calling a CGI program from java servlet, while calling cgi program I am passing encoded(Base64) content via post process,
My problem is the encoded data is not posted as expected, the encoded data is corrupting. But when I send encoded data in a text file, cgi program is perfectly decoding and working fine.
Here I am doing Base64 encoding as per requirement, I cannot avoid this encoding.
My doubt is about OutputStreamWriter constructor argument , In OutputStream Constructor I am passing one argument is OutputStream object and another argument is encoding type. I tried with ASCII, US-ASCII & UTF-8 .
My code is as follows, please help me to resolve this issue.
URL url = new URL("CGI server path");
URLConnection urlConnection = url.openConnection();
urlConnection.setDoOutput(true);
OutputStream os = urlConnection.getOutputStream();
BufferedOutputStream buffer = new BufferedOutputStream(os);
OutputStreamWriter  writer = new
                                               OutputStreamWriter(buffer, "US-ASCII");
writer.write(encodedPDF-Content);
writer.write("\r\n");
writer.flush();
writer.close();
here encodedPDF-Content is String and it's size is 9565 bytes

Answer 1

Whenever you read something in java into string (with Reader implementation) it expects source to contain text in encoding you specified. It then decodes it and makes 16 bit unicode string from it. Whenever you store string using Writers it does reverse operation with specified encoding (may be different than this which you used to read source) and stores text as a sequence of bytes made of 16 bit unicode string. So, passing text back and forth between programs with the help of files or I/O you can make mistake at both reading and writing encoding. Check for it.
Now, when C programm or other application reads file it may take another assumptions about encoding or may even completly ignore it and read source as a binary file. Then, if you have a source text as a file and have to pass it to other application never do it using Reader/Writer. User raw InputStream/OutputStream instead what will preserve all information unchanged.
here encodedPDF-Content is String and it's size is 9565 byteHow id you get this info? String.length() gets you how many chars is in it, it will be half the number of bytes. If you see your input file beeing 9565 bytes long, see my above statements.