Zip creation fails for one-letter filenames

Feb 17, 2009 at 4:12 AM
Our product has a large set of files, some zipped data (one data file per zip file, same name for the inner file as the zip), some various image types.  The files have names with length varying from 0 to about 25 characters, and we'd like to keep them the same across all types for consistency.  We're using DotNetZip for the zipped ones, which has worked well, except that DotNetZip throws an exception for the one-digit filenames.  We changed the files being zipped to be named "prefix_<character>" (the zip file is still just named "<character>") as a workaround, but this creates an extra check for the client using the data on every file loaded (and there are *lots* of them).  We can't really do them manually (with, say, WinZip) because it's a huge automated data generation and processing pipeline, and we don't want to change the names of all of the other files.  Is this a known issue?  Do you know if it will be fixed soon?
Feb 17, 2009 at 5:48 AM
Hello Leilani

no, this problem has never been reported.
I need some additional information from you. 
What version of DotNetZip are you using?
Can you show me the stacktrace of the exception that is thrown?

Just now I tried to zip up a file called "a" into a zipfile called "" and it worked.
I also tried a numerical digit:  putting a file called "2" into a zipfile called "".  It also worked.

I'm sure there is something I haven't understood about what you are doing.  If you can provide some additional information I will look into it.
Feb 17, 2009 at 9:04 PM
Actually, it turns out this was with an older version (the one that was available when we started generating those files, I'm guessing), and the one-letter names work now.  Sorry for the runaround; I assumed the group generating the data was using the latest version :)
Feb 17, 2009 at 10:49 PM
No problem.  Glad it's working for you.