summaryrefslogtreecommitdiff
path: root/pysnmp/carrier/asyncore
Commit message (Collapse)AuthorAgeFilesLines
* PEP-8 long lines and dunders (#245)Ilya Etingof2019-02-265-49/+127
| | | | This patch massively reformats the whole codebase mainly wrapping long lines and eliminating dundered private attributes.
* PEP8 optimize imports (#242)Ilya Etingof2019-02-126-20/+26
| | | | To make them PEP8-compliant
* Fix misnamed constantsIlya Etingof2019-02-101-7/+7
| | | | | This is a follow up to 588b9b902d191d8010cb6b247fcb07887d59542c fixing a couple of improperly named constants.
* Uppercase global constants (#238)Ilya Etingof2019-02-106-53/+55
| | | | | | | | | This is a massive patch essentially upper-casing global/class attributes that mean to be constants. Some previously exposed constants have been preserved for compatibility reasons (notably, in `hlapi`), though the rest might break user code relying on pysnmp 4.
* Drop Python < 2.6 except statement compatibility trickIlya Etingof2019-02-092-23/+23
|
* Fix IPV6_TRANSPARENT socket optionIlya Etingof2019-02-071-1/+1
|
* Do not set IP_PKTINFO socket option on IPv6 socketIlya Etingof2019-02-071-2/+1
| | | | | Trying to understand why sendmsg() fails on a transparent IPv6 socket
* Improve IPV6 debuggingIlya Etingof2019-02-071-1/+1
|
* Add `IPV6_V6ONLY` IPV6 socket optionIlya Etingof2019-02-071-0/+1
| | | | SO says this is required for proper UDP socket multi-homing.
* Emit low-level sendmsg/recvmsg debuggingIlya Etingof2019-02-071-0/+7
| | | | | When running in transparent proxy mode, log syscall parameters to aid troubleshooting
* Extend copyright notice to year 2019Ilya Etingof2018-12-306-5/+66
|
* Fix hlapi/v1arch request timeoutIlya Etingof2018-12-291-2/+1
| | | | The retrying mechanism never worked, apparently.
* Fix typosIlya Etingof2018-10-171-1/+1
|
* many backward-compatibility aids droppedIlya Etingof2018-02-251-61/+0
|
* copyright notice extendedv4.4.4Ilya Etingof2018-01-036-6/+6
|
* migrated references from SourceForge (#110)Ilya Etingof2017-11-176-6/+6
|
* minor pep8 improvementsIlya Etingof2017-05-272-4/+4
|
* email changed, copyright extended to the year 2017Ilya Etingof2017-01-126-6/+6
|
* pep8 reformattedIlya Etingof2016-04-026-21/+41
|
* pep8 fixesIlya Etingof2016-03-301-0/+2
|
* copyright updatedelie2015-12-296-6/+6
|
* copyright notice added to source codeelie2015-11-206-6/+36
|
* some more random linter warnings fixedelie2015-10-171-1/+1
|
* linted for bad-whitespace and some other issueselie2015-10-171-2/+2
|
* linted the bad-continuation issue as well as some otherselie2015-10-171-3/+2
|
* multiple-statements-in-one-line linted outelie2015-10-173-16/+27
|
* linted for trailing whitespaceselie2015-10-174-24/+28
|
* public API moved to AbstractTransportelie2015-10-161-11/+0
|
* a couple of misnamed symbols fixedelie2015-10-161-2/+2
|
* - original asynsock transport and AsyncsockDispatcher renamed intoelie2015-07-056-9/+9
| | | | | | asyncore and AsyncoreDispatcher respectively to provide better hint to fellow devs on the underlying transport being used - backward compatibility preserved through dummy asynsock symbols
* cosmeticselie2015-03-081-1/+0
|
* base addressType definedelie2015-03-081-0/+1
|
* - The asyncore-based transport subsystem extended to support POSIXelie2015-01-204-21/+92
| | | | | | | | | | | | sendmsg()/recvmsg() based socket communication what could be used, among other things, in the context of a transparent SNMP proxy application. Technically, the following features were brought into pysnmp with this update: * Sending SNMP packets from a non-local IP address * Receiving IP packets for non-local IP addresses * Responding to SNMP requests from exactly the same IP address the query was sent to. This proves to be useful when listening on both primary and secondary IP interfaces.
* pylint'edelie2014-06-172-2/+2
|
* fix to asyncore-based transport not to use asyncore's cheap inheritanceelie2014-06-141-0/+2
| | | | from socket object what caused warnings.
* new public DgramSocketTransport.getLocalAddress() returns local endpointelie2014-01-081-3/+3
| | | | address underlying BSD socket is currently bound to
* * broadcast socket option can now be enabled with the .enableBroadcast()elie2013-12-121-2/+11
| | | | | | | call for any datagram-based transport (namely, UDP and UDP6) * AbstractTransportDispatcher's jobStarted() and jobFinished() methods now accept optional 'count' parameter which is a way for an app to indicate how many responses are expected or have been processed in bulk
* fixes to normalizeAddress()elie2013-09-252-2/+2
|
* IPv6 address normalization moved into UDP6 transportelie2013-09-242-10/+15
|
* forcibly zero ZoneID, FlowID and ScopeID components to make itelie2013-09-241-0/+10
| | | | comparable with user-specified IPv6 endpoint
* sockets now configured with SO_REUSEADDR option to fix possibleelie2013-09-021-0/+1
| | | | Windows error 10048.
* fix to asyncore-based transport to work only with AsynsockDispatcher'selie2013-07-031-6/+21
| | | | socket map and not to touch default asyncore's one
* move common methods to AbstractTransportelie2013-06-301-7/+1
|
* exception re-raising improved at asynsock/twisted dispatchers so thatelie2013-06-261-1/+2
| | | | original traceback is preserved and reported
* disptacher vs transport compatibiilty verification moved intoelie2013-06-211-1/+2
| | | | base transport class
* raise KeyboardInterrupt for backward compatibilityelie2013-06-081-0/+2
|
* unused variables, imports; wrong indentation; undefined prototype attribueselie2013-06-042-2/+1
| | | | all the things that bothers linter
* missing imports, non-existing variables as suggested by linterelie2013-06-031-4/+1
|
* config.addSocketTransport() helper renamed into config.addTransport()elie2013-06-011-0/+2
| | | | | | and improved by automatically instantiating compatible TransportDispatcher making it dispatcher-agnostic. As an additional bonus, application may not call registerTransportDispatcher() as it would be called by addTransport().
* Asyncore-based dispatcher attempts to use poll() whenever availableelie2013-05-261-2/+8
| | | | | on the platform. It would help handling a really large number (>1024) of file descriptors.
View Lorry Controller Status