| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
|
|
|
| |
For some utterly moronic reason, something somewhere in the
stack (e.g. tox or pip) keeps insisting on installing an ancient pip
version 7 which fails in other depressing ways, so force it to not be
idiotic.
*rage*
|
|
|
|
| |
Probably excessive but let's try it for now.
|
|\ |
|
| |
| |
| |
| |
| | |
Pull requests should be tested automatically so add a .travis.yml file
to setup TravisCI testing.
|
| |
| |
| |
| | |
Python 3.6 is available so add it to tox.ini so it can be tested.
|
|/
|
|
|
|
|
|
|
|
| |
Recreating the sdist tarball takes some time when calling tox. So skip
the sdist creation (which will be implicitly done when
"usedevelop=True" is set) and do the "setup.py develop" step instead
which is recommended in the tox documentation[1].
[1]
http://tox.readthedocs.io/en/latest/example/general.html#avoiding-expensive-sdist
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
When a new dependency is discovered, its target may have already been
placed on the TODO list via a different source (and maybe even analysed
for dependencies), e.g. if find_dependencies() was previously called on
it, or if it was previously found via recursion. In this case we still
need to notify listeners of the dependency, even though we won't add it
to the TODO list.
|
|
|
|
|
|
|
|
| |
Avoid any errors if find_dependencies() is called repeatedly.
Previously this could fail when recursion was enabled, because
a commit could be processed via recursion, and then again via
a subsequent call to the TODO list, at which point the sanity
check that each blame line hadn't been seen before would fail.
|
| |
|
|
|
|
| |
Avoid calling hex() repeatedly
|
| |
|
| |
|
| |
|
|
|
|
| |
This helper can be reused elsewhere, e.g. in git-explode.
|
|
|
|
|
| |
It has nothing to do with DependencyDetector, and this will let us
reuse it from git-explode too.
|
|
|
|
|
| |
For some reason, Chrome sometimes passes the URL with the #
URL-encoded as %23.
|
| |
|
|
|
|
|
|
| |
pygit2.revparse_single(annotated_tag) returns a pygit2.Tag object, and
this extra layer of indirection has to be resolved to obtain the
pygit2.Commit.
|
|
|
|
|
| |
webcola requires d3 which is not there before bundle.js is loaded.
But why wasn't this ever a problem before?
|
| |
|
| |
|
|
|
|
| |
This is cleaner and will allow it to be reused elsewhere.
|
| |
|
|
|
|
| |
Apparently filename is more standard for Python modules.
|
|
|
|
|
|
| |
Bug was discovered by PEP8 :-)
Sem-Ver: bugfix
|
| |
|
|
|
|
| |
Sem-Ver: api-break
|
|\
| |
| | |
Fix noty depdendency to 2.4.1, version 3 is incompatible.
|
|/ |
|
|\
| |
| | |
Apply dwm1945's patch
|
|/
|
| |
Apply dwm1945's patch from the issue about KeyError.
|
|
|
|
|
|
|
|
|
|
| |
- import latest git master
https://github.com/tgdwyer/WebCola/issues/203#issuecomment-271311782
- use hacky workaround for d3 import
https://github.com/tgdwyer/WebCola/issues/145#issuecomment-271316856
- update for new makeEdgeBetween API
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If we analyse multiple commits for dependencies via the CLI, e.g.
git deps master~3..master
or
git deps master~4 master
then we typically want to know not just what the dependencies are,
but which of the input commits caused each dependency. So use the
same two-column output format which we already used with --recurse.
This also happens to be the same format understood by tsort(1).
|
|
|
|
|
|
|
| |
Now revision ranges like A..B or A...B are also accepted, and then git
deps will analyse the dependencies for all revisions in that range.
In the CLI invocation, if a dependency is encountered more than once
from separate commits, it will only be output once.
|
|\
| |
| | |
Add .snapcraft.yaml which will produce a strict snap for CLI use
|
|/ |
|
|\
| |
| | |
add debian install guide
|
|/ |
|
|\
| |
| | |
Explicitly require python2
|
|/ |
|
|\
| |
| | |
Account for Diff API changes in pygit2 version 0.22.1.
|
|/
|
|
| |
Signed-off-by: Michael Stefaniuc <mstefani@winehq.org>
|
| |
|