summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* cleanup (if we sync_fetch something, we really want it to succeed)Pascal Rigaux2006-11-151-10/+6
|
* use copy_and_ownPascal Rigaux2006-11-151-1/+1
|
* create copy_and_own, and use itPascal Rigaux2006-11-152-10/+19
|
* - create _update_medium__get_descriptions_local() out of ↵Pascal Rigaux2006-11-151-34/+50
| | | | | | | _update_medium_first_pass__local() - create _update_medium__get_descriptions_remote() out of _update_medium_first_pass__remote()
* create _update_medium_first_pass__remote() out of _update_medium_first_pass()Pascal Rigaux2006-11-151-141/+152
|
* use statedir_* funcs, but make them safe firstPascal Rigaux2006-11-151-18/+16
|
* removing $medium->{name}.cache was introduced in commit 33830, but i can'tPascal Rigaux2006-11-151-2/+1
| | | | | find out the reason why... dropping
* create statedir_names and use itPascal Rigaux2006-11-151-2/+6
|
* create statedir_descriptions() and use itPascal Rigaux2006-11-151-12/+11
|
* move comment where it should bePascal Rigaux2006-11-151-2/+2
|
* create write_urpmi_cfg() out of write_config()Pascal Rigaux2006-11-151-3/+8
|
* reorderPascal Rigaux2006-11-151-2/+4
|
* simplifyPascal Rigaux2006-11-141-4/+2
|
* reorderPascal Rigaux2006-11-141-3/+4
|
* log the fact that hdlist/synthesis is up-to-datePascal Rigaux2006-11-141-0/+2
|
* create $urpm->{md5sum_modified} which is a finer flag than $urpm->{modified},Pascal Rigaux2006-11-141-11/+14
| | | | | write_config is no more called because md5sum changed :)
* - don't pass $basename to get_md5sum()Pascal Rigaux2006-11-141-5/+6
| | | | | - cleanup
* create write_MD5SUM() out of write_config()Pascal Rigaux2006-11-141-7/+13
|
* don't allow calling ->add_media or ->add_distrib_media with unitialised ↵Pascal Rigaux2006-11-141-2/+2
| | | | $urpm->{media}
* create may_reconfig_urpmi() out of _update_medium_first_pass()Pascal Rigaux2006-11-141-14/+20
|
* create _read_cachedir_pubkey() out of _update_medium_first_pass()Pascal Rigaux2006-11-141-16/+26
|
* get rid of one $basenamePascal Rigaux2006-11-141-2/+1
|
* trying to get rid of $basenamePascal Rigaux2006-11-141-7/+10
|
* use statedir_synthesis()Pascal Rigaux2006-11-141-1/+2
|
* create statedir_hdlist_or_synthesis() and use itPascal Rigaux2006-11-141-8/+8
|
* create statedir_list() and cachedir_list()Pascal Rigaux2006-11-141-12/+20
|
* create cachedir_hdlist() and use itPascal Rigaux2006-11-141-16/+20
|
* (packing => 1) is used by every callers of _parse_hdlistPascal Rigaux2006-11-141-12/+9
|
* create hdlist_or_synthesis_for_virtual_medium() and use itPascal Rigaux2006-11-141-11/+15
|
* factorize codePascal Rigaux2006-11-141-6/+3
|
* create statedir_hdlist() and statedir_synthesis() and use themPascal Rigaux2006-11-141-33/+42
|
* rewrite a littlePascal Rigaux2006-11-141-9/+16
|
* simplifyPascal Rigaux2006-11-141-10/+12
|
* simplify code using select_media_by_name()Pascal Rigaux2006-11-141-14/+5
|
* use ->select_media_by_name & ->remove_media instead of ->select_media & ↵Pascal Rigaux2006-11-131-5/+4
| | | | ->remove_selected_media
* create remove_media out of remove_selected_media, and cleanupPascal Rigaux2006-11-131-8/+10
|
* create select_media_by_name which return the list of matching media.Pascal Rigaux2006-11-131-30/+31
| | | | | | the goal is to stop using $medium->{modified} to mean selected (eg: ->remove_selected_media used in urpmi.removemedia)
* cleanupPascal Rigaux2006-11-131-4/+7
|
* fix replacing "next" with "return" when _update_medium_first_pass() was ↵Pascal Rigaux2006-11-131-1/+1
| | | | introduced
* fix replacing "next"s with "return" after moving code in previous commitPascal Rigaux2006-11-131-1/+1
|
* create _update_medium_first_pass__local() out of _update_medium_first_pass()Pascal Rigaux2006-11-131-180/+196
|
* localize $dir to the block we're sure it has a meaning (and it's now no more ↵Pascal Rigaux2006-11-131-2/+2
| | | | used outside)
* this can't hurt: {url} ends with $dir, and it may even help in case if ↵Pascal Rigaux2006-11-131-1/+1
| | | | non-local media
* factorize handling of reconfig.urpmi, do it early so we don't need to redo ↵Pascal Rigaux2006-11-131-28/+24
| | | | the function
* fix error when introducing _update_medium_first_pass: there was a "redo ↵Pascal Rigaux2006-11-131-2/+2
| | | | MEDIA", so it's not a function call, it must be return after redoing
* cleanupPascal Rigaux2006-11-131-2/+3
|
* more clearPascal Rigaux2006-11-131-4/+4
|
* add a commentPascal Rigaux2006-11-131-0/+1
|
* simplifyPascal Rigaux2006-11-131-17/+10
|
* factorize code into ↵Pascal Rigaux2006-11-131-27/+23
| | | | _read_existing_synthesis_and_hdlist_if_same_time_and_msize() and _read_existing_synthesis_and_hdlist()
generated by cgit v1.2.1 (git 2.21.0) at 2024-11-14 01:38:28 +0000