aboutsummaryrefslogtreecommitdiffstats
path: root/netlib/tcp.py
Commit message (Collapse)AuthorAgeFilesLines
* Made attribute optional (as it is in pyOpenSSL)kronick2014-07-291-1/+4
| | | See https://github.com/pyca/pyopenssl/commit/0d7e8a1af28ab22950b21afa3fd451cec7dd5fdc -- It looks like this constant isn't set on some platforms (including Raspberry Pi's libssl)
* add support for ctx.load_verify_locations, refs mitmproxy/mitmproxy#174Maximilian Hils2014-05-211-2/+4
|
* make cert and key mandatoryMaximilian Hils2014-05-151-2/+2
|
* extract cert creation to be accessible in handle_sni callbacksMaximilian Hils2014-05-151-4/+11
|
* fix #28Maximilian Hils2014-03-101-1/+3
|
* Support Ephemeral Diffie-HellmanAldo Cortesi2014-03-071-1/+6
|
* Handler convert_to_ssl now takes a key object, not a path.Aldo Cortesi2014-03-051-1/+1
|
* Use PyOpenSSL's underlying ffi interface to get current cipher for connections.Aldo Cortesi2014-03-021-3/+13
|
* Cleanups, tests, and no-cover directives for code sections we can't test.Aldo Cortesi2014-03-021-3/+5
|
* Cipher suite selection for client connections, improved error handlingAldo Cortesi2014-02-271-3/+16
|
* allow super() on TCPServer, add thread names for better debuggingMaximilian Hils2014-02-151-3/+6
|
* adjust netlib.wsgi to reflect changes in mitmproxys flow formatMaximilian Hils2014-02-051-1/+1
|
* store used sni in TCPClient, add equality check for tcp.AddressMaximilian Hils2014-02-041-1/+7
|
* move StateObject back into libmproxyMaximilian Hils2014-01-311-16/+5
|
* remove subclassing of tuple in tcp.Address, move StateObject into netlibMaximilian Hils2014-01-301-14/+31
|
* re-add server attribute to BaseHandlerMaximilian Hils2014-01-281-1/+3
|
* add tcp.Address to unify ipv4/ipv6 address handlingMaximilian Hils2014-01-281-14/+42
|
* fix tests, remove duplicate codeMaximilian Hils2014-01-091-54/+37
|
* change parameter names to reflect changesMaximilian Hils2014-01-091-16/+13
|
* merge smurfix/ipv6, add ipv6 support for TCPServer, add ipv6 testMaximilian Hils2013-12-131-7/+8
|\
| * tab fixMatthias Urlichs2013-12-121-1/+1
| |
| * support ipv6Matthias Urlichs2013-12-121-4/+4
|/
* Fix race condition in test suite.Aldo Cortesi2013-12-081-1/+0
|
* Domain checks for persistent cert store is now irrelevant.Aldo Cortesi2013-12-081-2/+3
| | | | | We no longer store these on disk, so we don't care about path components.
* Merge pull request #19 from rouli/ciphersuitesAldo Cortesi2013-12-071-1/+3
|\ | | | | adding cipher list selection option to BaseHandler
| * adding cipher list selection option to BaseHandlerIsrael Nir2013-08-211-1/+3
| |
* | Merge branch 'fix_invalid_tcp_close'Aldo Cortesi2013-12-081-0/+4
|\ \
| * | Fix client certificate request feature.Aldo Cortesi2013-12-081-2/+2
| | |
| * | compatibility fixes for windowsMaximilian Hils2013-08-191-1/+2
| | |
| * | fix windows bugsMaximilian Hils2013-08-171-0/+3
| |/
* / Handle interrupted system call errors.Aldo Cortesi2013-08-251-1/+7
|/
* Revamp dummy cert generation.Aldo Cortesi2013-08-121-1/+2
| | | | We no longer use on-disk storage - we just keep the certs in memory.
* Correct order of precedence for SSL errors.Aldo Cortesi2013-08-111-2/+2
|
* Better handling of cert errors.Aldo Cortesi2013-08-101-0/+3
|
* Clarify the interface for flush and close methods.Aldo Cortesi2013-07-301-5/+5
|
* Merge pull request #16 from mitmproxy/fix_socket_bufferAldo Cortesi2013-07-281-3/+12
|\ | | | | attempt to fix 'half-duplex' TCP close sequence
| * attempt to fix 'half-duplex' TCP close sequenceMaximilian Hils2013-06-171-3/+12
| |
* | Add socket binding support for TCPClientAndrey Plotnikov2013-07-071-1/+4
|/
* Add a request_client_cert argument to server SSL conversion.Aldo Cortesi2013-05-131-4/+16
| | | | | | | | | | By default, we now do not request the client cert. We're supposed to be able to do this with no negative effects - if the client has no cert to present, we're notified and proceed as usual. Unfortunately, Android seems to have a bug (tested on 4.2.2) - when an Android client is asked to present a certificate it does not have, it hangs up, which is frankly bogus. Some time down the track we may be able to make the proper behaviour the default again, but until then we're conservative.
* Fix exception hierarchy.Aldo Cortesi2013-05-051-2/+2
|
* 100% test coverage.Aldo Cortesi2013-03-031-1/+1
|
* Make sni_handler an argument to BaseHandler.convert_to_sslAldo Cortesi2013-02-251-20/+15
|
* Housekeeping and cleanup, some minor argument name changes.Aldo Cortesi2013-02-241-5/+5
|
* More accurate description of an HTTP read error, make pyflakes happy.Aldo Cortesi2013-02-241-2/+2
|
* 100% test coverageAldo Cortesi2013-01-271-14/+7
|
* Bump unit tests, fix two serious wee buglets discovered.Aldo Cortesi2013-01-261-2/+2
|
* Introduce the mock module to improve unit tests.Aldo Cortesi2013-01-261-0/+6
| | | | | There are a few socket corner-cases that are incredibly hard to reproduce in a unit test suite, so we use mock to trigger the exceptions instead.
* Extract TCP test utilities into netlib.testAldo Cortesi2013-01-251-5/+6
|
* Improve error signalling for client certificates.Aldo Cortesi2013-01-201-3/+6
|
* Beef up client certificate handling substantially.Aldo Cortesi2013-01-201-1/+9
|