summaryrefslogtreecommitdiffstats
path: root/pym
Commit message (Collapse)AuthorAgeFilesLines
...
* Fix --rebuilt-binaries code to return existing_node where appropriate.Zac Medico2010-05-271-2/+2
|
* Tweak --rebuilt-binaries code to solve bug #306659 once again (logicZac Medico2010-05-271-2/+4
| | | | was reverted by the --rebuilt-binaries-timestamp patch).
* Add --rebuilt-binaries-timestamp optionSebastian Luther2010-05-272-4/+42
| | | | | | | This option makes emerge ignore binaries that would have been used for --rebuilt-binaries, if they are older than the given timestamp. Binaries are only reinstalled if they have a newer BUILD_TIME than the installed package (not only unequal) with this option.
* Bug #310947 - When expanding categories for atoms inside select_files,Zac Medico2010-05-271-2/+19
| | | | | make sure packages are available besides the installed one when preferring the category of an installed package.
* Make --deselect work for sets (bug 268798)Sebastian Luther2010-05-273-14/+20
|
* emerge --help: Add --ignore-default-opts (closes bug 246491)Sebastian Luther2010-05-271-0/+8
|
* Add _emerge.Package.iuse.get_missing_iuse to return missing IUSE flags. ↵Sebastian Luther2010-05-252-2/+15
| | | | Restore original behavior of is_valid_flag.
* _emerge.depgraph._show_unsatisfied_dep(): Do an ealier missing IUSE check ↵Sebastian Luther2010-05-251-10/+7
| | | | and simplify some related code
* _emerge.Package.iuse.is_valid_flags should return a list of flags missing ↵Sebastian Luther2010-05-251-2/+3
| | | | from IUSE
* Simplify code by passing a list of flags to Package.iuse.is_valid_flagSebastian Luther2010-05-251-5/+1
|
* Allow _emerge.Package.iuse.is_valid_flag() to take a list of flagsSebastian Luther2010-05-251-4/+9
|
* _emerge.depgraph._show_unsatisfied_dep(): properly check if something ↵Sebastian Luther2010-05-251-1/+1
| | | | belongs into missing_use
* Make parent_use parameter for portage.dep.Atom.violated_conditionals() ↵Sebastian Luther2010-05-252-17/+18
| | | | optional (for evaluated atoms)
* Fix traceback in _show_unsatisfied_dep, reported on forums:Zac Medico2010-05-251-1/+2
| | | | | | | | http://forums.gentoo.org/viewtopic-t-829445.html?sid=58b3b5cec88b8e2ffebc9991c90a473e File "/usr/lib/portage/pym/_emerge/depgraph.py", line 2273, in _show_unsatisfied_dep if not (violated_atom.use.enabled or violated_atom.use.disabled): AttributeError: 'NoneType' object has no attribute 'enabled'
* _emerge.depgraph._show_slot_collision_notice(): Improve atom list prunningSebastian Luther2010-05-221-6/+12
|
* _emerge.depgraph._show_unsatisfied_dep(): Suggest changing USE on the parent ↵Sebastian Luther2010-05-211-1/+29
| | | | if possible
* Add portage.dep.Atom.violated_conditionals()Sebastian Luther2010-05-211-0/+39
|
* _emerge.Package.__str__(): handle operation==uninstallSebastian Luther2010-05-211-0/+6
|
* Fix depgraph.schedulerGraph() to break DepPriority.satisfied attributesZac Medico2010-05-201-0/+10
| | | | which reference installed Package instances.
* If a dependency is installed but masked, try to pull in an unmasked versionZac Medico2010-05-201-4/+23
| | | | even when --deep is not specified. Thanks to Daniel Robbins for reporting.
* Use os.environ.get() instead of os.getenv(), to hopefully avoid issuesZac Medico2010-05-201-4/+5
| | | | with python-3.2 reported by Arfrever.
* Make the non --update package selection logic avoid installed packages ifZac Medico2010-05-201-1/+5
| | | | they are masked. Thanks to Daniel Robbins for reporting.
* Make the non --update package selection logic avoid installed packages ifZac Medico2010-05-201-1/+1
| | | | they are masked. Thanks to Daniel Robbins for reporting.
* Use Atom.unevaluated_atom in depgraph._show_slot_collision_notice()Sebastian Luther2010-05-201-1/+1
|
* Add portage.dep.Atom.unevaluated_atom to not lose the original use dep in ↵Sebastian Luther2010-05-201-2/+10
| | | | Atom.evaluate_conditionals()
* Add _emerge.Package.__str__() for nicer output formattingSebastian Luther2010-05-201-0/+28
|
* Support specifying exclude list as multiple arguments.Michał Górny2010-05-192-3/+3
| | | | | | | Instead of relying on user to provide the whole exclude list as a space-separated list in a single argument (which is uncomfortable), allow him/her to specify '--exclude' multiple times, joining the provided atom lists together.
* Only count "merge" nodes when determining whether or not to triggerZac Medico2010-05-161-1/+1
| | | | parallel-fetch.
* Fix --resume to support --exclude.Zac Medico2010-05-151-0/+4
|
* Don't use finally with atomic_ofstream since that prevents it fromZac Medico2010-05-131-12/+6
| | | | aborting when appropriate.
* Add a break_hardlinks parameter for tbz2 recompose methods, and enable itZac Medico2010-05-131-3/+26
| | | | | | | | | | by default. Since bindbapi.aux_update() calls this method, this changes the default behavior for all modifications to binary package metadata. This makes it safe to use hardlinks to create cheap snapshots of the repository, which is useful for solving race conditions on binhosts as described here: http://code.google.com/p/chromium-os/issues/detail?id=3225.
* Define __all__.Zac Medico2010-05-131-1/+6
|
* Use bytes literals in instead of _unicode_encode() where appropriateZac Medico2010-05-131-13/+13
| | | | (works with python-2.6).
* Add missing newline from previous commit.Zac Medico2010-05-061-1/+1
|
* Bug #317775 - In depgraph.display() use writemsg_stdout() to avoidZac Medico2010-05-061-7/+4
| | | | UnicodeEncodeError.
* Eliminate redundant os.sep in vardbapi.getpath().Zac Medico2010-05-061-1/+1
|
* Eliminate redundant os.sep in vardbapi._bump_mtime. Thanks toZac Medico2010-05-061-1/+3
| | | | grobian for reporting.
* Make myopts["--exclude"] a plain string in order to avoid issues with mtimedbZac Medico2010-05-052-3/+5
| | | | unpickling.
* Make --exclude use whitespace to separate atoms since atoms are typicallyZac Medico2010-05-042-3/+3
| | | | separated by whitespace.
* Make vardbapi.counter_tick_core() create the counter file's parent directoryZac Medico2010-05-031-0/+3
| | | | | if necessary, in order to avoid FileNotFound error. Thanks to Fabio Erculiani <lxnay@g.o> for reporting.
* Use keep_going local var.Michał Górny2010-05-031-1/+1
| | | | Use keep_going local var instead of checking self.myopts once again.
* Allow -C to remove packages referenced by sets in world_setsSebastian Luther2010-04-291-10/+3
|
* Support per-package bashrc files.Michał Górny2010-04-293-3/+6
| | | | | | | Support loading per-package bashrc files from the /etc/portage/env directory. The invidual files should reside in a directory matching ${CATEGORY}, with names matching ${PN}, ${PN}:${SLOT}, ${P} and/or ${PF} (which are applied in that order).
* Bail out immediately if --exclude is given a bad atom.Zac Medico2010-04-291-2/+2
|
* Tweak packages_with_invalid_use_config handling:Zac Medico2010-04-291-4/+8
| | | | | | * Ignore IUSE since those can't be changed by the user. * Only add unbuilt ebuilds to the list since USE can't be changed for built ebuilds.
* Add --exclude optionSebastian Luther2010-04-293-0/+66
|
* Fix typo from previous commit.Zac Medico2010-04-281-1/+1
|
* Inside bintree.inject(), discard cached metadata to ensure thatZac Medico2010-04-281-0/+4
| | | | _pkgindex_entry doesn't return stale metadata.
* Remove old portage module namespace compatibility that's been deprecatedZac Medico2010-04-2825-82/+0
| | | | since portage-2.1.6.
* Use shlex_split for splitting PORTDIR_OVERLAY.Zac Medico2010-04-272-17/+18
|