On Sun, 2006-10-01 at 14:40 +0200, Pierre Schmitz wrote: > Am Sonntag, 1. Oktober 2006 14:17 schrieb Benoit Chesneau: > > I just put the source code of archlinuxfr.org website online. We use the > > django code of archlinux.org. To get packages I create a script to get > > them : pkgdb.rb. It use information from db. You could find it here : > > http://projet.archlinuxfr.org/browser/trunk > > > > or here : > > http://projet.archlinuxfr.org/svn/trunk/ > > Right, I know your script. But I fear it does not behave like the one used by > archlinux.org. e.g. maintainer-infos are missing and it fails on some scripts > using too much bash-scripting. ;-) And you cannot search for included files. > To achiebe this you would hae to parse the built packages. > > So I think reinventing the wheel would not be a good idea here. > > > > > > > > > But yes could be usefull to have a dump of official db. Maybe in xml. Or > > better a webservice that access to all packages. If anyone is > > interrested about it maybe I could start to work on it ? > > Yeah, a xml-export would be really great. But for now an official dump of the > package-db would be very helpfull. And this could be done within minutes. > Just create a cron-job which dumps the package-table and save it to a public > accessable place. > > _______________________________________________ > arch mailing list > [email protected] > http://www.archlinux.org/mailman/listinfo/arch
The new script work with bash scripting. Sorry I forgot to send the new one :$. But it works as you could see on http://archlinuxfr.org/packages ;) I think you're right. It would be better to work with a dump of database. So all informations are here like the list of file istalled. I thought about a webservice to support out of date package feature and such thing :) What do you think about it ? - b. _______________________________________________ arch mailing list [email protected] http://www.archlinux.org/mailman/listinfo/arch
