summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Add test proving basic behavior of #18222.8Jeff Forcier2022-06-031-0/+18
|
* Changelog re #1822, closes #1822Jeff Forcier2022-06-031-0/+7
|
* split comment to linesyoav2022-06-031-1/+2
|
* Close socket on failureyoav2022-06-031-0/+3
| | | | | | | As mentioned in socket docs: "Sockets are automatically closed when they are garbage-collected, but it is recommended to close() them explicitly, or to use a with statement around them." Resolve #1126
* Skip tests requiring invoke if it's not installedMichał Górny2022-06-031-1/+6
| | | | | | Since invoke is an optional dependency and only one group of tests require it, skip them gracefully rather than failing if it's not present.
* Fix broken unittesty skipTest in locale decoratorJeff Forcier2022-06-031-3/+2
|
* sftp tests: Replace unittest partsStanislav Levin2022-06-032-4/+4
| | | | | | | | | | | | | | | The original PR https://github.com/paramiko/paramiko/pull/992 introduced several tests for sftp functionality. These tests made use of unittest's stuff like `assertTrue` and `assertEqual` because at that moment the tests were grouped under the `unittest.TestCase`-based class (`SFTPTest`). Before PR merge `unittest.TestCase` was refactored out from sftp tests (667bd74b139ed86f9b261d3abf5b6042ba80920b) but PR was not updated. The sftp tests are marked with `slow` and that's why they are not failed in CI (slow tests are excluded by default). Fixes: https://github.com/paramiko/paramiko/issues/1941 Signed-off-by: Stanislav Levin <slev@altlinux.org>
* Streamline default test name on CircleJeff Forcier2022-06-031-4/+4
|
* Clarify SFTPClient.open() docs re: actual behavior of BufferedFileJeff Forcier2022-03-111-2/+2
| | | | Closes #2000
* blackenJeff Forcier2022-02-251-1/+6
|
* Comment and changelog re #1985Jeff Forcier2022-02-252-0/+7
|
* Add six to `install_requires`Sondre Lillebø Gundersen2022-02-251-1/+1
|
* No idea why the lack of this blew up on circle but not locallyi-hate-my-lifeJeff Forcier2021-11-281-1/+1
| | | | Probably the locales differ? sigh
* Cut 2.8.12.8.1Jeff Forcier2021-11-282-1/+2
|
* Fix changelog referenceJeff Forcier2021-11-281-3/+3
|
* Changelog format fix re #992Jeff Forcier2021-11-281-1/+1
|
* Credit re #992Jeff Forcier2021-11-281-1/+1
|
* Blacken code for #992Jeff Forcier2021-11-284-15/+22
|
* Cleanup re #992Jeff Forcier2021-11-282-46/+7
| | | | | | - remove apparently vestigial bits - remove Python 2.6 related bits - use pytest.skip instead of custom skip
* Add changelog entry for #992Martin Packman2021-11-281-0/+4
|
* Fix failure in listdir when server uses a localeMartin Packman2021-11-283-12/+51
| | | | | | | | | | | | | | | Fixes #985 SFTPAttributes uses the locale-aware %b directive for the abbreviated month name with time.strftime, but was not decoding the result on Python 2.7. Add a strftime alias in py3compat that will always return unicode, and resolve the mixing of bytes and text in SFTPAttributes methods. Add a test at the listdir level, and a more specific test for the SFTPAttributes asbytes method.
* Add new requireNonAsciiLocale test decoratorMartin Packman2021-11-281-0/+48
|
* Expose Python 2.6 compatible test skip decoratorMartin Packman2021-11-281-1/+39
|
* Changelog re #1024Jeff Forcier2021-11-281-0/+3
|
* Catch TypeError and UnsupportedAlgorithm when using Cryptography to load ↵Jeff Forcier2021-11-282-4/+9
| | | | | | | private key material Prior to this change, these exceptions bubble up as-is instead of becoming SSHException instances like most other key-loading errors
* Changelog closes #1257, re #1266Jeff Forcier2021-11-281-0/+7
|
* Tests proving #1257Jeff Forcier2021-11-281-1/+31
|
* Patch missed RSAKey for some reason. Fixed.Jeff Forcier2021-11-281-4/+3
|
* Roll back ABC using kit from contributed patchJeff Forcier2021-11-281-6/+3
| | | | | | - blew up good for me for reasons I don't entirely grok - also feels like too big of a change to do in a patch release - is not a critical piece of the bugfix - NotImplementedError is fine for now
* Blacken patch.Jeff Forcier2021-11-281-4/+4
|
* Check all data on keys when comparing for equalityJosh Snyder2021-11-285-14/+29
| | | | | This eliminates the possibility that a hash collision results in two PKey objects that == each other.
* Enhance tests of existing/general key cmp/hash behaviorJeff Forcier2021-11-281-0/+22
| | | | Not sure why these were never explicitly tested
* Pre-patch test proving security flaw re: 32bit key hash comparisonsJeff Forcier2021-11-284-2/+49
|
* basic alpine i386 dockerfileJeff Forcier2021-11-281-0/+10
|
* Fix silly but apparently nonfatal dev-req self-install typoJeff Forcier2021-11-281-1/+1
|
* Changelog re security flaw email, re #908Jeff Forcier2021-11-281-0/+24
|
* Blacken for hostkeys fixJeff Forcier2021-11-281-4/+4
|
* test_hostkeys: test SubDict.__delitem__()Lans Zhang2021-11-281-0/+17
| | | | | | | SubDict.__delitem__() would trigger the KeyError exception if a key type to be deleted is not present. Signed-off-by: Lans Zhang <jia.zhang@windriver.com>
* hostkeys: fix raising KeyError in SubDict.__delitem__()Lans Zhang2021-11-281-0/+1
| | | | | | | If the specified key type is found out, the exception KeyError should not happen. Signed-off-by: Lans Zhang <jia.zhang@windriver.com>
* Merge branch '2.7' into 2.8Jeff Forcier2021-10-151-4/+6
|\
| * Merge branch '2.6' into 2.72.7Jeff Forcier2021-10-151-4/+6
| |\
| | * Merge branch '2.5' into 2.62.6Jeff Forcier2021-10-151-4/+6
| | |\
| | | * Enhance language around SSHClient.close()2.5Jeff Forcier2021-10-151-4/+6
| | | |
* | | | blackenJeff Forcier2021-10-091-2/+4
| | | |
* | | | Cut 2.8.02.8.0Jeff Forcier2021-10-092-1/+2
| | | |
* | | | Nix travis from tasksJeff Forcier2021-10-091-3/+0
| | | |
* | | | Classify bug as major in case I don't cut a bugfix for 2.7Jeff Forcier2021-10-091-2/+2
| | | |
* | | | For some reason :rfc: blows up where :rfc-reference: does notJeff Forcier2021-10-091-2/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Only when running as part of my fullstack release task, specifically when I load up sphinx+releases to parse the changelog. Makes no sense at all but no time to really dig more...
* | | | Massage upstream 'release.all' task too re: custom publishJeff Forcier2021-10-091-5/+10
| | | |
* | | | Less confusing name of overriding task pleaseJeff Forcier2021-10-081-2/+2
| | | |
View Lorry Controller Status