summaryrefslogtreecommitdiffhomepage
path: root/tests/test_transport.py
AgeCommit message (Collapse)Author
2017-06-05test transport security options can be set to defaultsPierce Lopez
ensures all defaults key/cipher/digest etc types are supported
2017-05-31Additional house style formatting tweaks, mostly re: removal of line ↵Jeff Forcier
continuations
2016-06-09Issue #537 reproduction test and fixKrzysztof Rusek
2016-04-22Merge branch '1.15' into 1.16Jeff Forcier
2016-04-22Hack in a sleep() to avoid race conditions during timeout test.Jeff Forcier
(HOPEFULLY) closes #612
2015-11-03Merge branch 'master' into 502-intJeff Forcier
2015-10-30Merge branch 'master' into 596-intJeff Forcier
2015-03-20Commit 838e02ab42 changed the type of the exec command string on python3Anselm Kruis
from unicode to bytes. This commit adapts the test suite accordingly.
2015-03-20According to RFC 4254 sec 6.5 the "command" string of an "exec" channelAnselm Kruis
request is a byte-string. Previously paramiko assumed "command" to be UTF-8 encoded. Invalid UTF-8 sequences caused an UnicodeDecodeError. This commit changes a test case to uses a non UTF-8 string and fixes the bug.
2015-02-24Expose handshake timeout in the transport API.Olle Lundberg
This is a reimplementation of #62.
2014-12-17Suggest a MIN_WINDOW_SIZE and MIN_PACKET_SIZEJeff Quast
Not fully confident with this change, though I will describe my findings fully in the pull request. The OpenSSH client requests a maximum packet size of 16384, but this MIN_PACKET_SIZE value of 32768 causes its request to be "clamped" up to 32768, later causing an error to stderr on the OpenSSH client. Suggest then, to delineate MIN_WINDOW_SIZE from MIN_PACKET_SIZE, as they are applied. I don't think there is any minimum value of MIN_PACKET_SIZE, however we can suggest a value of 4096 for now.
2014-12-17Use modern api to check if event is set.Olle Lundberg
Since we are a python2.6+ code base now, we want to be as forward compatible as possible.
2014-11-25Merge upstream branch 'master' into add_sha2_supportMatthias Witte
Conflicts: paramiko/transport.py tests/test_transport.py
2014-09-10Merge branch 'master' into 216-intJeff Forcier
Conflicts: paramiko/transport.py paramiko/util.py tests/test_client.py
2014-09-08Merge branch 'master' into 372-intJeff Forcier
Conflicts: paramiko/channel.py tests/test_util.py
2014-09-07Convert Channel into a context managerMichael Williamson
2014-08-15Remove all occurences of ParamikoTest.Olle Lundberg
Sorry paramiko, it's time to put on the big boy pants. You no longer support old as hell versions of python.
2014-08-15Strip whitespace.Olle Lundberg
2014-08-14Add sanitation methods for window and packet size.Olle Lundberg
2014-08-13Change window and packet size to match opensshs'.Olle Lundberg
Update tests to match the new numbers.
2014-07-16Include sha2 changes in testsMatthias Witte
- let _compute_key default default to sha1 if local_mac is not set instead of setting local_mac explicitly in the unit test - add tests for KexGexSHA256
2014-07-15Fix transport testMatthias Witte
2014-03-07Fix import * and a bunch of PEP8 formattingScott Maxwell
2013-11-19Remove byte conversions and unhexlify calls that we only needed for Py2.5 ↵Scott Maxwell
support and use the `b` byte string marker instead
2013-11-19Change all exceptions to modern format (not Py2.5 compatible)Scott Maxwell
2013-11-02Fix some deprecation and resource warningsScott Maxwell
2013-11-01Fixes for Python 2.5 and Python 3.2 supportScott Maxwell
2013-11-01Eliminate all uses of b'' syntax to allow for Python 2.5 supportScott Maxwell
2013-10-31More type conversionScott Maxwell
2013-10-31More type fixupsScott Maxwell
2013-10-30Convert and detect types properly, use helper constants, use StringIO and rangeScott Maxwell
2013-10-30Fix message sendingScott Maxwell
Create constants for byte messages, implement asbytes so many methods can take Message and key objects directly and split get_string into get_text and get_binary. Also, change int handling to use mpint with a flag whenever the int is greater than 32 bits.
2013-10-30Fix dict iters, sorts, exceptions, bytes renames and tuple argsScott Maxwell
2013-10-30Use test_path to avoid relative path issuesScott Maxwell
2013-10-30Fix importsScott Maxwell
2013-09-27Fixed a typo in the license header of most filesJeff Forcier
Conflicts: paramiko/proxy.py
2012-10-15Refactoring: it's a thing. Re #85Jeff Forcier
2012-10-15Only override unittest assertions when absolutely necessary.Jeff Forcier
Fixes #85
2009-07-19fix my email address to be the current one.Robey Pointer
2009-02-17bug 300536: allow a really long banner before the SSH handshake.Robey Pointer
2008-06-03[project @ robey@lag.net-20080604053906-vz5toqvlp5miqy1x]Robey Pointer
merge deadlog bugfix from dwayne litzenberger.
2008-03-23[project @ robey@lag.net-20080324065111-qv8icii767fqex9j]Robey Pointer
fix some unit tests for windows
2008-03-22[project @ robey@lag.net-20080323020709-ub8ya3xkh1ebywfa]Robey Pointer
add exit_status_ready for justin cook.
2008-02-17[project @ robey@lag.net-20080218051229-xtylb1poe246k2ci]Robey Pointer
merge patch from Dwayne Litzenberger to fix unit tests on python 2.3.
2008-01-23[project @ robey@lag.net-20080124014121-95bt9p4n4k8lw72i]Robey Pointer
remove some now-unneeded code.
2008-01-23[project @ robey@lag.net-20080124013849-jno9xkgwvvqrvuov]Robey Pointer
split auth tests into their own file, and clean up the remaining transport tests a bit (use existing refactoring).
2007-12-30[project @ robey@lag.net-20071231052950-8h599bnez3sgbf2e]Robey Pointer
patch from david guerizec for direct-tcpip forwarding support, and a unit test added by yours truly.
2007-12-30[project @ robey@lag.net-20071230220403-9c9735x0tfr88dd8]Robey Pointer
oops, fix typo.
2007-11-20[project @ robey@lag.net-20071120052527-hloi0b30yngbay0x]Robey Pointer
add send_ready() and a unit test.
2007-10-28[project @ robey@lag.net-20071029030344-9adfzb9ulfodtepu]Robey Pointer
bug 157205: select() doesn't notify incoming stderr data, because stderr's pipe isn't hooked up to the fileno() BufferedPipe. to fix, i added an "or" pipe-event that can be triggered by either stdout or stderr, and hooked them both up to fileno(). added a unit test for the bug and one for the "or" pipe.