summaryrefslogtreecommitdiffstats
path: root/pym
Commit message (Collapse)AuthorAgeFilesLines
...
* portdbapi: don't convert _pkg_str to strZac Medico2012-05-131-1/+2
|
* dblink: use _pkg_str + vercmp, not pkgcmpZac Medico2012-05-131-5/+9
|
* dep_check: use _pkg_str + vercmp, not pkgcmpZac Medico2012-05-131-15/+21
|
* _pkg_str: validate repo (move code from Package)Zac Medico2012-05-134-16/+15
|
* unmerge: use cpv_sort_key instead of pkgcmpZac Medico2012-05-131-10/+12
|
* _pkg_str: make version -r0 consistent with inputZac Medico2012-05-131-1/+4
|
* cpv_sort_key: use _pkg_strZac Medico2012-05-131-8/+18
|
* best_match_to_list: use _pkg_strZac Medico2012-05-131-7/+5
|
* match_from_list: avoid multiple meanings of mycpvZac Medico2012-05-131-2/+2
|
* cpvequal: use _pkg_strZac Medico2012-05-132-10/+22
|
* _pms_eapi_re: require whitespace before commentZac Medico2012-05-131-1/+1
| | | | | See discussion on gentoo-pms mailing list here: http://archives.gentoo.org/gentoo-pms/msg_1f76a55d04b4f2a3721cddc611434aad.xml
* match_from_list: preserve input in resultZac Medico2012-05-131-4/+6
| | | | | This fixes an unintended change in behavior from commit f31320b67c9f593a2a8592e1a4e547f5f641943a.
* Use _pkg_str.version more.Zac Medico2012-05-131-4/+14
|
* Don't cache vercmp results.Zac Medico2012-05-131-14/+1
| | | | | It's probably not very helpful to cache vercmp results, since identical version comparisons probably don't recur very often.
* _pkg_str: add version attribute for comparisonsZac Medico2012-05-135-23/+34
| | | | | This attribute can be passed directly into vercmp, avoiding the need to generate this string many times.
* depgraph: minimize match_from_list operationsZac Medico2012-05-131-58/+64
| | | | | | In _iter_match_pkgs, call match_from_list on one cpv at a time, in order to avoid unnecessary match_from_list comparisons on versions that are never yielded from this method.
* SimpleResolverTestCase: handle PYTHONHASHSEEDZac Medico2012-05-121-1/+2
|
* _pkg_str: add slot and repo attributesZac Medico2012-05-129-40/+64
|
* match_from_list: handle Package.UNKNOWN_REPOZac Medico2012-05-122-4/+8
|
* _pkg_str: add cpv attribute for match_from_listZac Medico2012-05-121-0/+2
|
* Check type of xml.etree.ElementTree.TreeBuilder before applyingArfrever Frehtes Taifersar Arahesis2012-05-131-4/+10
| | | | workaround for http://bugs.python.org/issue14007.
* test_digraph: fix get_cycles for PYTHONHASHSEEDZac Medico2012-05-122-7/+15
|
* test_digraph: fix bfs for PYTHONHASHSEED=randomZac Medico2012-05-121-4/+39
|
* test_stackLists: fix for PYTHONHASHSEED="random"Zac Medico2012-05-121-2/+2
|
* Add workaround for http://bugs.python.org/issue14007.Arfrever Frehtes Taifersar Arahesis2012-05-121-0/+4
|
* Add a _pkg_str class to cache catpkgsplit resultsZac Medico2012-05-126-13/+68
| | | | | This will compensate for the removal of the catpkgsplit cache in commit 68888b0450b1967cb70673a5f06b04c167ef879c.
* catpkgsplit: don't cache resultsZac Medico2012-05-121-7/+1
| | | | | | | | The cache has been subtly broken since commit eb2056631021a04b62c228206e44376f5c7a81ba when the eapi parameter was added. If necessary, we can add a cpv class (or something like that) in order to avoid redundant catpkgsplit calls in places like match_from_list.
* Add portage.eapi.eapi_allows_dots_in_use_flags().Arfrever Frehtes Taifersar Arahesis2012-05-122-11/+14
|
* Allow dots in package names in EAPI="4-python".Arfrever Frehtes Taifersar Arahesis2012-05-124-49/+99
|
* Fix handling of empty PORTDIR{,_OVERLAY} setting.Zac Medico2012-05-111-2/+2
|
* Cleanup profile-formats code.v2.2.0_alpha104Zac Medico2012-05-112-3/+9
|
* RepoConfig: sort __slots__Zac Medico2012-05-111-3/+3
|
* Avoid redundant layout.conf parsing.Zac Medico2012-05-113-7/+17
| | | | | | By passing the RepoConfigLoader instance into LocationsManager, we can re-use previously parsed layout.conf data. The RepoConfigLoader instance will also be useful for bug #414961.
* config: load repositories before profileZac Medico2012-05-111-5/+19
| | | | | This is useful for bug #414961, since we want to have access to the repositories when we're loading the profile.
* config: propagate SYNC to self.repositoriesZac Medico2012-05-112-9/+6
| | | | | | This resolves a circular dependency on the SYNC variable, which is useful for bug #414961, since we want to be able to initialize self.repositories before the SYNC variable is known.
* config: use shlex to split known_reposZac Medico2012-05-111-3/+3
|
* PollScheduler: use local EventLoop (thread safe)Zac Medico2012-05-106-27/+51
| | | | | | | For API consumers, this makes the doebuild() function compatible with threads, avoiding a ValueError raised by the signal module, as reported at http://bugs.sabayon.org/show_bug.cgi?id=3305. Classes derived from PollScheduler still use the signal module when possible.
* Don't write or trust cache for unsupported EAPIs.Zac Medico2012-05-106-80/+44
| | | | | | Since we're supposed to be able to efficiently obtain the EAPI from _parse_eapi_ebuild_head, we don't need to write or trust cache entries for unsupported EAPIs.
* Make REPOSITORY variable again available in global scope of ebuilds,Arfrever Frehtes Taifersar Arahesis2012-05-101-6/+3
| | | | | which use EAPI="4-python". This fixes regression introduced in commit 1c26d24dfc869a6f5ef426f7c02e1d25abcda34e.
* porttree.py: remove unused importsv2.2.0_alpha103Zac Medico2012-05-091-2/+0
|
* Refer to bug #402167 where appropriate.Zac Medico2012-05-091-1/+1
|
* doebuild_environment: clean up EAPI related codeZac Medico2012-05-091-31/+29
|
* EbuildMetadataPhase: handle empty EAPI from bashZac Medico2012-05-091-1/+2
|
* Split out eapi_invalid function.Zac Medico2012-05-092-39/+55
|
* Show unsatisfied REQUIRED_USE for bug #415125.Zac Medico2012-05-091-4/+9
|
* doebuild_environment: remove unused eapi parsingZac Medico2012-05-091-18/+2
|
* Parse EAPI with pattern from PMS section 7.3.1.Zac Medico2012-05-097-89/+123
| | | | | | | | This implements the specification that was approved in Gentoo's council meeting on May 8, 2012 (see bug #402167). The parse-eapi-ebuild-head FEATURES setting is now enabled by default, and causes non-conformant ebuilds to be treated as invalid. This behavior will soon become enabled unconditionally.
* COLLISION_IGNORE: support fnmatch patternsZac Medico2012-05-072-16/+9
| | | | | | This allows it to be combined with the COLLISION_IGNORE_UNOWNED variable from commit 33545ea18e8816addb3c54bb26a0cc788b8512e6, so only one variable is needed.
* COLLISION_IGNORE: handle prefixZac Medico2012-05-051-3/+4
|
* Support COLLISION_IGNORE_UNOWNED="*.pyc *.pyo"Krzysztof Pawlik2012-05-042-2/+11
| | | | This will fix bug #410691.