summaryrefslogtreecommitdiffstats
path: root/pym
Commit message (Collapse)AuthorAgeFilesLines
* _process_slot_conflicts: do one slot at a timeZac Medico2012-06-161-36/+37
| | | | | This allows us to avoid redundant processing of slots that have already been processed by a previous call.
* depgraph: split out _handle_slot_conflict methodZac Medico2012-06-161-120/+129
|
* depgraph: simplify _select_pkg_highest_availableZac Medico2012-06-161-8/+0
| | | | | | This hunk of code is no longer needed since the content of _highest_pkg_cache is discarded each time a package is added to the graph.
* depgraph: _autounmask in _highest_pkg_cache keyZac Medico2012-06-161-1/+1
|
* Unmerge symlink listed as dir for bug #421165.Zac Medico2012-06-151-4/+5
|
* Add USE + license autounmask for bug #420847.Zac Medico2012-06-152-8/+13
|
* test_autounmask: add case for bug #420847Zac Medico2012-06-131-0/+11
|
* Fix test failure with Python 2.6.Arfrever Frehtes Taifersar Arahesis2012-06-131-1/+1
|
* repoman: handle trailing newlines betterMike Frysinger2012-06-122-3/+17
| | | | | | | | Automatically strip trailing newlines from the ChangeLog, and be better about not adding them in the first place (still not perfect, but getting there). Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* repoman: update cat/pkg info in header all the timeMike Frysinger2012-06-122-25/+133
| | | | | | | | | | | There are edge cases where repoman's changelog code is not as good as the existing echangelog. Mostly related to out of date headers. Have the code check the header in more cases not just for missing lines, but also outdated values all the time. While we're at it, write some tests! Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* test_match_from_list: fix _pkg_str usageZac Medico2012-06-121-3/+3
|
* dbapi: construct _pkg_str during match on-demandZac Medico2012-06-112-15/+49
|
* config: pass tolerant param to getconfig moreZac Medico2012-06-111-5/+8
|
* Move post_src_install metadata code to one func.Zac Medico2012-06-113-43/+39
|
* QueryCommand: add get_db() classmethodZac Medico2012-06-111-3/+7
|
* Package: move SLOT validation to _pkg_strZac Medico2012-06-102-15/+18
|
* Add a _get_slot_re() function.Zac Medico2012-06-103-4/+9
|
* Package: use _eapi_attrsZac Medico2012-06-102-4/+8
|
* _get_pv_re: use _eapi_attrsZac Medico2012-06-101-13/+20
|
* dep: remove old "DEPEND SYNTAX" commentsZac Medico2012-06-101-14/+0
|
* use_reduce: use _eapi_attrsZac Medico2012-06-102-3/+5
|
* _get_eapi_attrs: move to eapi moduleZac Medico2012-06-102-32/+31
|
* dep: move regex constants to top of fileZac Medico2012-06-101-16/+17
|
* _get_atom_wildcard_re: use _eapi_attrsZac Medico2012-06-101-16/+22
|
* _get_atom_re: share cache between more EAPIsZac Medico2012-06-101-4/+6
|
* _get_useflag_re: use _eapi_attrsZac Medico2012-06-101-20/+28
|
* _get_usedep_re: use _eapi_attrsZac Medico2012-06-101-33/+43
|
* Atom: use _eapi_attrs in constructorZac Medico2012-06-101-10/+15
|
* _get_atom_re: handle many combinationsZac Medico2012-06-101-19/+43
| | | | | | A namedtuple of _eapi_attrs is used to hash atom regular expressions, making it easy to handle many different combinations, as will be necessary for the addition of new features such as abi-slot deps.
* cache/sqlite.py: dynamically add columns to tableZac Medico2012-06-101-2/+40
|
* InheritEclass: avoid false positive in func regexZac Medico2012-06-081-5/+6
|
* depgraph: pass use_reduce result to dep_checkZac Medico2012-06-083-42/+38
| | | | This allows us to avoid extraneous use_reduce/paren_enclose calls.
* InheritEclass: avoid false positive in func regexZac Medico2012-06-071-2/+5
|
* parse_layout_conf: fix cache-formats pms fallbackZac Medico2012-06-051-3/+3
| | | | | This fixes an issue where egencache would not generate metadata/cache unless the directory already existed.
* Handle more import exceptions involving xml.Zac Medico2012-06-043-5/+24
| | | | See http://bugs.python.org/issue14988
* InheritEclass: exempt auto/libtool for multilibZac Medico2012-06-031-1/+1
|
* InheritEclass: exempt autotools for libtoolZac Medico2012-06-031-1/+2
|
* Package: calc visiblity on demandZac Medico2012-06-031-15/+34
| | | | | | This delays visiblity calculations until after FakeVartree applies its metadata tweaks, so commit 650c960be3170bade8bb63ddedcd4796c75ec374 will work as intended.
* FakeVartree: use KEYWORDS from unbuilt ebuildZac Medico2012-06-031-1/+1
| | | | | | This allows us to avoid triggering the code from bug #350285 (see commit b7de3d1fb10e40b4ef03462f8a145968689ac587) in cases when the KEYWORDS from the unbuilt ebuild would unmask the package.
* Delete workaround for http://bugs.python.org/issue14007.Arfrever Frehtes Taifersar Arahesis2012-06-031-10/+0
|
* InheritEclass: remove unused inherited_api codeZac Medico2012-06-011-14/+0
|
* InheritEclass: exempt eclasses that export src_*Zac Medico2012-06-011-3/+35
| | | | | We really need to annotate exceptions like these directly in the eclasses, but hardcode them for now.
* InheritEclass: kde4-base exemptionsZac Medico2012-06-011-2/+2
|
* InheritEclass: add stable / conservative modeZac Medico2012-06-012-2/+29
|
* config: whitelist PORTAGE_COMPRESS, don't save itZac Medico2012-06-011-1/+2
|
* config: lazy VirtualsManager instantiationZac Medico2012-06-011-3/+10
|
* config: lazy MaskManager instantiationZac Medico2012-06-011-6/+14
|
* config: lazy KeywordsManager instantiationZac Medico2012-06-011-5/+18
| | | | | | This allows a config instance to be instantiated by portageq, while accessing /etc/portage/package.keywords only on demand, solving bug #418475 by avoiding PermissionDenied exceptions when possible.
* portageexit: skip SANDBOX_ON checkZac Medico2012-06-011-2/+1
| | | | | This check is redundant, since the portdbapi constructor already has special SANDBOX_ON / SANDBOX_WRITE handling.
* grablines: raise PermissionDenied if appropriateZac Medico2012-06-011-2/+13
|
generated by cgit v1.2.3-1-g7c22 (git 2.25.1) at 2024-07-04 16:11:03 +0000