aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* setup.py: reorder data filesArthur Zamarin2022-11-011-1/+2
| | | | | | | | | Store all data files (files that should be installed into `/usr` outside python lib location) in one main "data" directory, with direct subdirs matching the once installed on system. This is preparation for moving to flit backend. Signed-off-by: Arthur Zamarin <arthurzam@gentoo.org>
* ebd: generate eapi bash lib cache when registering EAPI objectsTim Harder2019-07-231-2/+2
| | | | | | | | | | This lets us use cached versions for all generated libs while running from the git repo directly allowing things like cache regen or other highly bash-centric processes to run near install speed. Previously cache regen spawned from a pkgcore git repo would take lots of additional time due to the ebd having to dynamically generate and source the requisite bash EAPI libs for every package.
* gitignore: make C ext pattern more specificTim Harder2019-02-201-1/+1
| | | | To avoid matching the dolib.so and newlib.so helpers.
* git: ignore some more generated filesTim Harder2015-07-131-0/+2
|
* generate bash function lists for the dist tarballTim Harder2015-07-091-0/+3
| | | | | | | Also, remove global function list from git. The scripts to generate the lists will be run on demand when running tools within the git repo; otherwise, the generated files will be used if they exist in tarballs and installed instances.
* add doc/generated to gitignoreTim Harder2015-04-101-0/+1
|
* use generic glob for ignoring C extension shared objectsTim Harder2015-03-291-2/+4
|
* add .tox to gitignoreTim Harder2015-03-291-0/+1
|
* doc: first step towards having better man pagesTim Harder2015-01-251-2/+0
| | | | | | | | | | | | | | Now we only generate each script's synopsis and options from the related argparse code instead of generating the whole file. This allows us to easily add content to the man pages. Further down the road, we might move away from auto-generated man pages for scripts entirely and instead rely on generating option and synopsis dumps that can be compared manually against the current content of a man page. This is mostly because it would be nice to allow extra content to be added for various subcommands which in the current method is tricky without adding lots of extraneous content to the code.
* gitignore: only ignore certain root repo dirsTim Harder2015-01-101-3/+3
|
* ignore 2to3 cache filesTim Harder2014-12-141-0/+1
|
* doc: convert NEWS and README files to rstTim Harder2014-11-011-1/+0
|
* ignore vim swap filesTim Harder2013-11-261-2/+2
|
* Add basic news conversion, intended for integration into the man website.Brian Harring2012-03-311-0/+1
|
* plugin cache format v3Brian Harring2011-11-261-1/+1
| | | | | | | | 1) add enough information that unneccessary imports are avoided 2) allow plugins to specify python namespace path, rather than requiring object 3) refactor into easier to maintain code 4) move the cache back to 'plugincache'; header is enough to handle new versions 5) work around broken plugins, rather than breaking everything
* make sphinx doc's authorative; switch name to docBrian Harring2011-08-061-9/+6
|
* pull in initial argparse translator; subcommands are next neededBrian Harring2011-08-061-0/+1
|
* update to filter out sphinx/doc intermediatesBrian Harring2011-03-181-0/+9
|
* update gitignore, wipe bzrignoreBrian Harring2010-12-281-0/+2
|
* add gitignore/attributes for this projectBrian Harring2010-12-261-0/+6