Subject:
|
Re: perl file copy question
|
Newsgroups:
|
lugnet.off-topic.geek
|
Date:
|
Wed, 5 Oct 2005 22:04:22 GMT
|
Viewed:
|
2450 times
|
| |
|
|
In lugnet.off-topic.geek, Ross Crawford wrote:
> In lugnet.off-topic.geek, Larry Pieniazek wrote:
> > I have a need from within a perl module to copy a large (and variable, it's
> > wildcard driven) number of files from one directory to another. TMTOWTDI of
> > course. What's the best way under the following constraints?
> >
> > - the OS is linux
> > - The wildcard varies from run to run so it's going to need to be a variable
> > (that seems like no big deal at all just mentioning it)
> > - I want to avoid system() or backtick, as I do not want to spawn a subprocess,
> > too expensive in the context I am in
> > - I prefer not to schlep around with digging up filenames and separating them
> > from the directories from what glob returned
> > - I prefer to use only standard perl and the default module set, not stuff I'll
> > have to install, getting stuff installed is a major pain (it took months to get
> > DBI and DBD onto the production systems)
> >
> >
> > I thought of using glob($wildcardExpression) to get an array of filehandles and
> > then using File::Copy but where I broke down there is that copy wants two
> > filehandles or file names and the target names aren't individually known,
> > necssarily, unless I dug into the filehandle to find the name, then separated
> > the directory off or something.
>
> Try File::NCopy (although I dont think it's a standard module) -
> http://search.cpan.org/~gabor/File-NCopy0.32/lib/File/NCopy.pm
>
> Note however that you can easily obtain directory and filename from the path
> string - you don't need the file handle (I'm pretty sure File::Basename is
> standard):
>
> use File::Basename;
>
> $fullname = "/usr/local/src/perl-5.6.1.tar.gz";
>
> $file = basename($fullname);
> # file="perl-5.6.1.tar.gz"
>
> $dir = dirname($fullname);
> # dir="/usr/local/src"
>
> > This is something that seems easy in .sh or .ksh. I bet it's easy here too I'm
> > just not finding it, although I have been reading docs and searching for it for
> > a while.
>
> It is easy - if you're able to call the OS. The OS is, after all, designed as a
> program for handling files. But without that as an option, it becomes harder.
>
> HTH
>
> ROSCO
that does help! (meanwhile I got a version working that uses system(), just as a
stopgap...)
However I really don't want to use system() (or any process spawning things) and
since I want to wildcard aren't I going to have to use glob() to get an array of
filehandles? However,
http://perldoc.perl.org/functions/glob.html
says glob returns a list of filename expansions... do they mean names rather
than handles? If so, then ya, all set.
|
|
Message has 2 Replies:
 | | Re: perl file copy question
|
| (...) Does this code help? It shows the outputs. Decided to have a fiddle with what Ross posted to see what everything did. Tim use File::Glob ':glob'; use File::Basename; $ddn="destination directory"; @t=(<g:/docs/in_prog...bes/*.*>); foreach $f (...) (19 years ago, 5-Oct-05, to lugnet.off-topic.geek)
|
Message is in Reply To:
| | Re: perl file copy question
|
| (...) Try File::NCopy (although I dont think it's a standard module) - (URL) however that you can easily obtain directory and filename from the path string - you don't need the file handle (I'm pretty sure File::Basename is standard): use (...) (19 years ago, 5-Oct-05, to lugnet.off-topic.geek)
|
5 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
