Mailinglist Archive: zypp-devel (225 mails)
| < Previous | Next > |
Re: [zypp-devel] Re: [SoC-student] libzypp HTTP download failover
- From: Duncan Mac-Vicar Prett <dmacvicar@xxxxxxx>
- Date: Mon, 28 Apr 2008 11:12:50 +0200
- Message-id: <48159512.6040208@xxxxxxx>
Gerard Farràs i Ballabriga wrote:
http://en.opensuse.org/Libzypp/Programmers_Guide
then I would start looking how the different layers interact:
in the lowest level you have plain curl. This is wrapped by MediaManager (in media directory), which has different backends, curl is just one of them. Then there is MediaSetAccess on top which adds the ability to manage multiple media sets (CDs) and set checkers when the wrong media is inserted. Then there is Fetcher which allows to queue jobs, and can look if the requested files are already present in some local cache. Then there are the repo/Downloaders which are Fetcher themselves, and they have the specific logic on how to download one type of repository. Then you have RepoMediaAccess that knows more specific information about the repository itself, and sets the correct media verifier with the repository metadata media information.
The idea is that the upper layers provide most of the functionality using the lower ones. But the design always allow you to use low level pieces directly (most of the time)
Duncan
--
To unsubscribe, e-mail: zypp-devel+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: zypp-devel+help@xxxxxxxxxxxx
2008/4/28, Duncan Mac-Vicar Prett <dmacvicar@xxxxxxx>:http://en.opensuse.org/Libzypp/Design (needs some update but most is still valid)
May be compiling/installing libzypp from svn trunk in a prefix, then
compiling zypper against this libzypp.
This step is "already" done ;-) . What next ?
Thanks,
Gerard
http://en.opensuse.org/Libzypp/Programmers_Guide
then I would start looking how the different layers interact:
in the lowest level you have plain curl. This is wrapped by MediaManager (in media directory), which has different backends, curl is just one of them. Then there is MediaSetAccess on top which adds the ability to manage multiple media sets (CDs) and set checkers when the wrong media is inserted. Then there is Fetcher which allows to queue jobs, and can look if the requested files are already present in some local cache. Then there are the repo/Downloaders which are Fetcher themselves, and they have the specific logic on how to download one type of repository. Then you have RepoMediaAccess that knows more specific information about the repository itself, and sets the correct media verifier with the repository metadata media information.
The idea is that the upper layers provide most of the functionality using the lower ones. But the design always allow you to use low level pieces directly (most of the time)
Duncan
--
To unsubscribe, e-mail: zypp-devel+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: zypp-devel+help@xxxxxxxxxxxx
| < Previous | Next > |