One other quick thought ~ when i transfer files between linux and windows in ASCII mode it usually doesn't do any character set translation, just converts CR to CRLF. If this is your only problem then the old dos2unix utility might undo the CRLF conversions. You could try running it through that utility and then see if the new copy will decompress. -Jeremy On 4/24/07, Peter McLarty <p.mclarty@xxxxxxxxxx> wrote:
Have you tried running the file command ie file abc.Z to see what unix thinks of it if it says its text I guess you have a nice lot of rubbish of which you may or may not be able to salvage anything, if not then you should know what sort of compression utility you require to unpack it Cheers Peter ------------------------------ *From:* Dennis Williams [mailto:oracledba.williams@xxxxxxxxx] *Sent:* Wednesday, 25 April 2007 04:59 AM *To:* WLJohnson@xxxxxxxxxxxxxxxxxxx *Cc:* oracle-l *Subject:* Re: uncompress: corrupt input Bill, My understanding is that a compressed file uses all 8 bits of a byte. My understanding is the ASCII setting of FTP causes only 7 bits of each byte to be transferred. Therefore, that is my best guess as to why you have a corrupt compressed file. I think you've done all the reasonable actions to resolve this aside from considering whether maybe this file was misnamed and it really isn't compressed, just renamed with a "Z" suffix. One other thing you could consider, depending on how important this file is to you. An Oracle export file is basically a text file. You could consider reading the file byte-by-byte to understand what you have. This may involve understanding just how the common compression tools (starting with compress) actually work at the byte level. You could create a similar export, then compress it the way you think this file was compressed and start comparing byte for byte. I think export files have some generic headers that would help you with your comparison. There are other Internet groups that probably have people who could explain some of the mechanics of "compress". I would imagine that is pretty well documented somewhere. Depending on what you learn (whether the file is truly corrupted or mostly salvageable) you might write a program to uncompress what you can salvage or whatever. This might take a few days, depending on what you find. Not sure what your schedule is. Dennis Williams
-- Jeremy Schneider Chicago, IL http://www.ardentperf.com/category/technical