indexing filesystem- was (Re: [MLUG - DISCUSSION] Dumb Perl Question)

["Rick Buford" <EMAIL:PROTECTED>]

This kind of leads into something I've been wanting to ask this group. I've 
been asked several times if there is an indexed filesystem for linux similar 
to what VMS uses. Since I don't abla VMS, the best I can gather is that what 
they're looking for is a filesystem that maintains its own indexed metadata 
on every file, along with version info. I'm sure if I've munged up what we're 
looking for too much, Mr. Wolfe will correct me...

Rick 
When we were young, we were told that "everybody else is doing it" was a
really stupid reason to do something. Now it's the standard reason for
picking a particular software package. -- Barry Gehm



---------- Original Message -----------
From: Curtis Jewell <EMAIL:PROTECTED>
To: EMAIL:PROTECTED
Sent: Tue, 7 May 2002 15:48:09 -0500 (Central Daylight Time)
Subject: Re: [MLUG - DISCUSSION] Dumb Perl Question

> On Tue, 7 May 2002, Davis, Ryan Wiley (UMC-Student) wrote:
> 
> > Okay,
> >
> > So I am looking for a bit of perl to put into a CGI and I am kindof lost
> > on exactly how to do it. Let me describe.
> >
> > I want the perl to read the filenames in a certain sub directory of my
> > web account. There are a few files 10-30 in this directory. I would then
> > like the perl code to output (into HTMLof course) the listing of said
> > directory and a link to access all these files.
> >
> > I'm currently messing with different ways of doing this, but wanted to
> > know if any of you monkeys out there have already done something like
> > this and have a quick and easy solution.
> 
> I don't have "known good" Perl code, but you can use DBI with the
> DBD::ExampleP driver to make your file system act like a database. It's
> "rows" are file entries, it's "columns" are the different things
> that fstat returns about a file, and it's "tables" are directories.
> 
> Fun, huh? [little known answer, I see. How I know it is coming up.]
> 
> It's a trick I read about in a Perl book. The problem is that the 
> book discusses perl 5.002 (!!!!), so the example code I'm taking 
> this from is a REALLY OLD version of DBI, but ExampleP still comes 
> with the newest versions of DBI, so I'm extrapolating to recent 
> versions (something greater than DBI 1.00 and/or perl 5.004)...
> 
> I think this is the correct code, YMMV.
> 
>   use DBI 1.15;
> 
>   $dbh = DBI->connect("dbi:ExampleP", undef, undef, {RaiseError => 
> 0});  # I don't THINK username and password are required for ExampleP.
> 
>   $sth = $dbh->prepare('select * from ?');
>   # Selects all possible data from a directory.
>   # I don't know how complicated SQL passed to DBD::ExampleP can
>   # be without breaking it, so be careful.
> 
>   $files = $sth->selectall_hashref('/usr');
>   # returns an arrayref with it's values being hashrefs.
>   #   The keys of the hashref are column names.
>   #   The values are the data. (an individual file in this case)
>   # This call was introduced in DBI 1.15 (note "use" above)
>   # The usage is correct for 1.15 to 1.19 AFAIK -
>   #   1.20, I'm told, (looking at search.cpan.org) changes how
>   #   this routine is used. I've never used it, so I don't know
>   #   how to change it.
>   # If you're stuck with lower than 1.15 (current is 1.21), look for
>   #   some other way to figure out what the column names are.
>   #   I know two column names are name and atime.
> 
>   @files = @$files;
> 
>   foreach $file (@files) {
>     %file = %$file;
>     ... # do your thing!
> 
>   }
> 
> -- 
> Curtis Jewell
> http://www.curtisjewell.com/          EMAIL:PROTECTED
> http://curtis.livejournal.com/        EMAIL:PROTECTED
> http://new-york.utica.mission.net/    EMAIL:PROTECTED
> 
> --
> To unsubscribe, go to http://mlug.missouri.edu/members/edit.php
> 
> Archives are available at http://mlug.missouri.edu/list-archives/
------- End of Original Message -------

--
To unsubscribe, go to http://mlug.missouri.edu/members/edit.php

Archives are available at http://mlug.missouri.edu/list-archives/