Quick search on the domain of doom came up with:

http://msdn.microsoft.com/library/default.asp?URL=/library/books/dnjet/c1_body_16.htm

Disclaimer: I didn't browse too deep (This site dislikes Netscape a bit!) so I don't know if they actually provide enough detail for implementation; however, this seems to be the best bet.

That'll bring you into their manual on the "Jet" (aka "Access") file format; scroll down in the left frame to find the appropriate chapters. I don't know of a module to do this already, but a DBD::MDB object would probably be the way to go (possible derived from DBD::CSV?)...?

You might also want to look into the formats used by e.g. MS Word; they (Microsoft) tend to use a chunk-structured file format (called OLE like everything else?) similar (...) to PNG or TIFF, so Access likely is quite similar; work toward parsing Word files might have a lot of useful info. for you. One starting-point might be WV

Another option might be some kind of MDB2CSV|XML filter :-)

Good luck!

Amendment: Today's post has an actual ref to an MDB2CSV converter, which should be 'importable' to DBD::CSV. Voila! Thanks to jeroenes for the info

If they are static files, your best bet is whip up a Win32 Perl script to extract the databases from the gzips, load them up using DBI::ODBC, and them write them out again in the format or database of your choice. Is there a reason you need to actually perform the extraction on Solaris?

   MeowChow                                               
                print $/='"',(`$^X\144oc $^X\146aq1`)[-2]

See that little box next to the word search at the top of the page? Try typing 'access' in there and have a look at the results.

(Methinks I assumed to much...)

Update:It appears I was a little hasty. Without opening the file, you might just need OLE::Storage to read the raw OLE storage streams, but it would be complicated. It would certainly make you question whether it was worth it.

Have a think about what runrig has suggested.

*~-}hotyopa{-~*

We've determined through the CB that he wants to access the data from gzipped files on Unix. DBI won't help, you'd need something like Parse::Access which doesn't AFAIK exist yet.

Any reason why you can't get the files to a PC with Access on it, then use DBI/DBD::ODBC or Win32::ODBC? Or export the data to a delimited or other format? Maybe even write a script to unzip the files and Net::FTP them to a Windoze box with Access on it? Automate any necessary commands on the Windoze box with Net::Telnet or maybe Expect (I've never used Net::Telnet, but it looks like if you use it you don't need Expect...)?

As I recall you don't actually need access installed to use *.mdb files. I _think_ I recall someone saying they had used DBD::ODBC to read those files under a unix system. However you may (or may not, try it) need an ODBC driver, but there is supposedly (according to the docs), a ODBC driver included with the DBD::ODBC source. Otherwise there are links to other cross-platform ODBC drivers in those docs.

Notice how he says he wants to be able to do this *without running Access*. He wants to parse the database file directly.

Philosophy can be made out of anything. Or less -- Jerry A. Fodor