On Sun, 3 Jun 2007 18:21:03 +0900 Charles Plessy <charles-debian-nospam@plessy.org> wrote: > > http://bibiserv.techfak.uni-bielefeld.de/spool/download/bibiserv_1180771945_17521/rnahybrid-2.1-src.tar.gz > > as you suspected, the link name changes everytime: > > http://bibiserv.techfak.uni-bielefeld.de/spool/download/bibiserv_1180862250_19693/rnahybrid-2.1-src.tar.gz But the wildcard was specifically designed to match - it allows for changes in the numbers, exactly as above. The only difference in the two URL's is: 1180771945_17521 becomes 1180862250_19693. The .*_.* matches that perfectly, everytime. It could be optimised to match only numbers but as there is no way of saying that the future script will only use numbers, it's best to leave it as a match-all wildcard. http://bibiserv.techfak.uni-bielefeld.de/spool/download/bibiserv_.*_.*/rnahybrid- (.*)-src.tar.gz > To access it I have to go through a page named > http://bibiserv.techfak.uni-bielefeld.de/cgi-bin/download.pl, and I am > affraid that one of its CGI parameters is rnahybrid-2.1-src.tar.gz. Not true - the directory tree still exists even if it is renamed on the next run. That URL can still be scanned. > The solution is therefore: rm debian/watch :) There is no reason for that. Have you actually tried that pattern in debian/watch with uscan? uscan scans directories, hence the name. It gets a directory list from the server, scans those against the pattern and sees if there is a match. The file is still accessible. -- Neil Williams ============= http://www.data-freedom.org/ http://www.nosoftwarepatents.com/ http://www.linux.codehelp.co.uk/
Attachment:
pgpjoUX5WvWer.pgp
Description: PGP signature