On Mon, 4 Mar 2002, Erik Price wrote: > On Monday, March 4, 2002, at 05:54 PM, Anthony de Almeida Lopes wrote: > > > Does anyone know of a way to download all of the pacakges from the web. > > I want to use a resumable downloader (IE or SpeedDownload) to get them > > since I have a slow connection?
You can get part of the way there with something like this: % find /sw/fink/dists/stable/main/finkinfo | \ grep 'info$' | xargs egrep '^Source: (ht|f)tp://' | head The problem there is that [a] the argument list could quickly get too long for 'grep' to digest efficiently (or at all), and [b] the results will have placeholders, such as (first sample for me): /sw/fink/dists/stable/main/finkinfo/base/anacron-2.3-3.info:Source: http://prdownloads.sourceforge.net/anacron/%n-%v.tar.gz So you need to go in and swap out %n for the package name, and %v for the version info. If anyone were ambitious, this could be folded into a Perl/LWP thing that ran a lot faster, did that substitution automatically, and saved everything somewhere appropriate. Once you have that list, it could either have curl or wget grab the files or, since you're working in Perl anyway, it could just grab them inline. None of this should be particularly difficult to implement, but most of these features partly or wholly overlap with what Fink does already, so it might make more sense to implement it not as a free standing program, but as an extension to the existing Fink command: Usage: fink [options] command [package...] fink install pkg1 [pkg2 ...] fink dl pkg1 [pkg2 ...] fink dl-all Where these new 'dl' and 'dl-all' options allow you to download the source tarballs for the named packages, or for all packages. (Being able to get fink to download .debs would be nice too...) > Fink uses curl, which supports resuming downloads AFAIK. If you don't > like curl, you can point Fink at the package 'wget', and use that > instead by temporarily renaming /usr/bin/curl, as Fink will use wget if > it is available and curl is not. Is there a way to get Fink to prefer 'wget' *without* messing around with system commands? I didn't see anything relevant in fink.conf, and I don't want to delete / move a command that system functionality might depend on. Where in the Fink source does it reference curl, anyway? Is it safe to manually change that in the source, or are the dependencies more subtle than I'm expecting? -- Chris Devers [EMAIL PROTECTED] Apache / mod_perl / http://homepage.mac.com/chdevers/resume/ _______________________________________________ Fink-beginners mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/fink-beginners
