XML compression using GZip (Urgent)

Question

Hi ,
I am compressing XML on the fly using GZip and storing it in a Database. When required we uncompress using same utility and everything works fine until we encounter a large XML (more than 1200 Bytes). What happens then is XML is Truncated and final few tags are missing in the uncompressed XML. I am not sure if the Compression is truncation or the Uncompress ....
Please look into the code sinppets and lemme know if you find any bugs there....
CompressMethod............
String someXML = "....long xml....";
try {
int length = someXML .trim().length();
ByteArrayOutputStream baos =
new ByteArrayOutputStream(length);
ByteArrayInputStream bis = new ByteArrayInputStream(filename.trim().getBytes());
GZIPOutputStream gos = new GZIPOutputStream(baos);
byte[] buffer1 = new byte[9092];
int bytesRead;
while ((bytesRead = bis.read(buffer1)) != -1) {
gos.write(buffer1, 0, bytesRead);
bis.close();
gos.close();
return baos.toByteArray();
For Uncompressing ..................
ByteArrayInputStream bai = new ByteArrayInputStream(filename);
byte[] buffer = new byte[9092];
GZIPInputStream gos =
new GZIPInputStream(bai);
gos.read(buffer,0,buffer.length);
gos.close();
orig = new String(buffer,"iso-8859-1");
Any Help at the earliest is appriciated and REWARDED !!!!!!!

Answer 1

I'm not quite sure why you are doing some of the things your are. It might be your haste in getting a code "snippet" to show.
String someXML = "....long xml....";...
int length = someXML .trim().length();...
ByteArrayInputStream(filename.trim().getBytes());Is your xml really in a known String like "someXML" or is it being read from the file represented by "filename"? If it is in a string and you wish to find the length of the trimmed article, then trim it first and reference only the trimmed String from then on. If you take the length of a temporarily trimmed String and then use the non-trimmed String you may be cutting data off of the end.
Also, if your xml is in a String there is no need to have a read loop. Just write your byte array in one call.
The length of the trimmed string is the only thing that I see that could cause your loss of data.