summaryrefslogtreecommitdiffhomepage
AgeCommit message (Collapse)Author
2023-01-18Streamline copyright stringJeff Forcier
2023-01-18Intersphinx to py36 not py27Jeff Forcier
2023-01-18Handpick remaining-ly relevant changes from #2110Jeff Forcier
Also make 'em f-strings, which AFAIK should be comparable Closes #2110
2023-01-16Changelog re #2165Jeff Forcier
2023-01-16Merge branch '2165-int' into 3.0Jeff Forcier
2023-01-16Stop ignoring warnings in the test suite for nowJeff Forcier
Most of what we put this in for is gone as we now use newer versions of most dependencies. The few that remain are acceptable levels of noise or only on certain versions (eg py36).
2023-01-16Break up a big int literalJeff Forcier
We could do this on a LOT of hex and similar literals, but that's a lot of work for little gain, let's just do that organically over time instead.
2023-01-16Nuke retry_on_signal, pointless on modern Py3Jeff Forcier
Additionally, other bits of code that retry EINTR can similarly just go away.
2023-01-16Stop stripping ProxyCommand none, make it NoneJeff Forcier
Also apparently the old, old test for this had the wrong issue number in it :(
2023-01-16flake8Jeff Forcier
2023-01-16Nuke unused private methodJeff Forcier
2023-01-16Punt a bunch of backwards incompatible TODOs to 4.0Jeff Forcier
2023-01-16Nuke now extraneous chmodJeff Forcier
2023-01-16Switch an os.open call to use kwargs now that we're py3 onlyJeff Forcier
Also shuffle comments around to be more local to subjects.
2023-01-16blackenJeff Forcier
2023-01-16well that's just super()!Jeff Forcier
2023-01-16No more (object)ionsJeff Forcier
2023-01-16s/%/fstrings/gJeff Forcier
Except in one spot where it was too complicated to bother for now XD
2023-01-16Twiddle a test to not shadow hex builtin (and nuke %)Jeff Forcier
2023-01-16Invoke 2 is now outJeff Forcier
2023-01-13flake8/blackJeff Forcier
2023-01-12No more PKey.__cmp__Jeff Forcier
2023-01-12Can't actually use a set here as it will destroy orderingJeff Forcier
2023-01-12Swap around __bytes__/__str__ for some classesJeff Forcier
Plus related updates to tests, docstrings, changelog
2023-01-12Start fixing up blatantly wrong-under-py3 docstring typesJeff Forcier
2023-01-12Upcoming alabaster release with modern sphinx stylesheet fixJeff Forcier
2023-01-12Modern sphinx autodoc settingsJeff Forcier
2023-01-12We live in the future nowJeff Forcier
2023-01-12Update a bunch of no/low impact python2/3 references/commentsJeff Forcier
2023-01-12No need for passthru dunder-init in Py3-only dict subclasses yeyJeff Forcier
2023-01-12Nix setup.cfgJeff Forcier
2023-01-12Last spot of six removalJeff Forcier
2023-01-11flake8Jeff Forcier
2023-01-11blackenJeff Forcier
2023-01-11Missed a spot re: b,u helpers now in utilJeff Forcier
2023-01-11No idea when self.fail worked or didn't. old code is old?Jeff Forcier
2023-01-11Import shuffleJeff Forcier
2023-01-11Put back the old faux-long(), kinda, in a few spots onlyJeff Forcier
2023-01-10Set up icecream in test harnessJeff Forcier
2023-01-10Nix old todoJeff Forcier
2023-01-10Improve performance by reducing expensive bytes conversionAlex Gaynor
In two core parts of the codebase, complex type-switch based code is used to convert a value to bytes. However, in all cases (except for one, fixed in this PR), the caller is specifying a Message instance. We can make this code much simpler by directly calling the correct method on Message. In #2110 this is measured to massively speed up large SFTP transfers.
2023-01-09Oh right that's why I had specific pins hereJeff Forcier
2023-01-09Some semi-missed removals of 'long'Jeff Forcier
- a couple spots where we wrote longs as int64 and ints as int. I _think_ the right thing to do here is write all ints as int64 given how Python handles these under 3. - docs bits
2023-01-09s/mock/unittest.mock/gJeff Forcier
2023-01-09py3compat.strftime -> time.strftimeJeff Forcier
2023-01-09Nuke outdated todoJeff Forcier
2023-01-09nuke py3compat.inputJeff Forcier
2023-01-09Remove sys.version_info checksJeff Forcier
2023-01-09Update encode/decodebytesJeff Forcier
2023-01-09py3compat.bytes -> builtin or literal dependingJeff Forcier
generated by cgit v1.2.3 (git 2.39.1) at 2025-02-23 11:33:53 +0000