revZIP library...

Paul Dupuis paul at researchware.com
Thu Oct 20 11:31:51 EDT 2022


In addition to the revZIP library, whether building for 32 bit Windows 
standalone or 64-bit Standalones, still (as of LC 9.6.8) has a 2GB limit 
on archives it can open and a 2GB limit on what it can save, there 
appears to be NO mechanism to get the compressed size of an item (file) 
in the archive. Neither the revZipAddItemWithFile nor the 
revZipAddItemWithData, nor any other API in the dictionary lets you get 
the compressed size of something after compressing it.

This is especially annoying with the 2GB limit on saving an archive and 
that error only occurs when you call revZipCloseArchive at the end. That 
means if a user of your application tries to archive a list of file that 
are 50GB or some even larger number the code happily compresses 
everyone, going well over 2GB until it tries to close and save the 
archive at the end and ONLY after all that wasted time do you get an error.

I can obviously get the uncompressed size of each file (using detailed 
files) and keep a running total and IF that total hist 2GB then stop a 
process that MIGHT fail in the end. However, perhaps their uncompressed 
files are 2.5GB but when compressed would be 1.8gb and work.

Folks have already suggested abandoning the revZIP library and using 
shell to invoke native command line tools (although no call backs for a 
progress bar is possible with CLI tools, which, is a BIG disadvantage to 
that approach for our users). However, I really wish the revZIP library 
just worked right.

On a long shot, does anyone know of a clever trick using the revZIP 
library, to get the compressed size of an item after you compress it?

I suppose I could open the archive, compress 1 item, close the archive, 
check the size of the archive via detailed files, and then open the 
archive  add the nest compressed file, close, measure the size and so 
on, but I feel the overhead would slow down and already slow process 
even further.

My best option may be to just add up the uncompressed sizes of the files 
involved.

Any clever ideas welcome!




More information about the use-livecode mailing list