On Thu, 25 Aug 2005 06:08, Steve Holdoway wrote: > On Wed, August 24, 2005 10:32 pm, Christopher Sawtell said: > > On Wed, 24 Aug 2005 13:37, Nick Rout wrote: > >> On Wed, 24 Aug 2005 08:40:31 +1200 > >> > >> Roy Britten wrote: > >> > On Tue, 2005-08-23 at 19:33 +0200, Robert Himmelmann wrote: > >> > > wget http://laby.toybox.de/download2.php?fileid=15 does the same. In > >> > > http there is some part of the header which specifies the filename > >> > >> of > >> > >> > > whatever is being transmitted. They use Apache/1.3.33 with PHP > >> > > (obviously). I would think that they forgot this field when they > >> > >> wrote > >> > >> > > the script. (You have to specify it) Firefox probably uses in cases > >> > > like this the URL it had before the redirect and wget the one after. > >> > > Classify it as a bug in their script and rename the file. > >> > > >> > man wget > >> > ... > >> > -O file > >> > --output-document=file > >> > The documents will not be written to the appropriate files, but > >> > >> all > >> > >> > will be concatenated together and written to file. If file > >> > >> already > >> > >> > exists, it will be overwritten. If the file is -, the documents > >> > will be written to standard output. Including this option > >> > >> automat- > >> > >> > ically sets the number of tries to 1. Note that when > >> > >> --output-docu- > >> > >> > ment is specified, --convert-links is ignored. > >> > > >> > If you're only downloading one file, it's effectively a rename. > >> > >> Yep I am familiar with the -O option. The program i am interfacing with > >> isn't. > >> > >> the script I am writing is an ebuild. It exports a URL to portageand > >> portage uses the url to download a file that is expected to behave > >> normally, ie be properly named. > > > > This is a known wget mis-feature. See:- > > http://bugs.gentoo.org/show_bug.cgi?id=103537 > > and > > http://wget.sunsite.dk/development.shtml > > First item on the list. > > I'm sure patches would be very welcome :-) > > If it's a bug, then how are you going to implement a standard file naming > scheme across http and ftp? if there is a line such as this:- Content-disposition: filename=laby_1.0.1.tar.gz in the http headers then use the filename mentioned. otherwise use the name of the file on server as is done currently.
I have not yet examined the wget code so cannot comment on the exact details. -- CS
