----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviewboard.asterisk.org/r/3628/#review12224 -----------------------------------------------------------
/trunk/res/res_pjsip_pubsub.c <https://reviewboard.asterisk.org/r/3628/#comment22300> I'm not against the 'real'/'virtual' nomenclature, but would 'parent'/'child' be better/clearer? /trunk/res/res_pjsip_pubsub.c <https://reviewboard.asterisk.org/r/3628/#comment22299> I'd go ahead and doxygen these as well. /trunk/res/res_pjsip_pubsub.c <https://reviewboard.asterisk.org/r/3628/#comment22301> Parent/child aside, I do enjoy the name of this union /trunk/res/res_pjsip_pubsub.c <https://reviewboard.asterisk.org/r/3628/#comment22297> There's a lot of things later on in the code that seem to really want that resource to exist. Is there any risk here that request_uri->user (and thus resource) is empty or NULL? If this is empty/NULL, what happens when things request the resource later? - Matt Jordan On June 18, 2014, 4:12 p.m., Mark Michelson wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviewboard.asterisk.org/r/3628/ > ----------------------------------------------------------- > > (Updated June 18, 2014, 4:12 p.m.) > > > Review request for Asterisk Developers. > > > Bugs: ASTERISK-23865 > https://issues.asterisk.org/jira/browse/ASTERISK-23865 > > > Repository: Asterisk > > > Description > ------- > > This is step one towards implementing resource lists in Asterisk. Once > resource list subscriptions are implemented, an ast_sip_subscription may not > directly correspond to an underly pjsip_evsub structure. Currently, > subscription handlers will gladly try to reach underneath the > ast_sip_subscription to get at the PJSIP-specific structures. This set of > changes seeks to abstract the pubsub API further so that subscription > handlers do not have any interaction with PJSIP. > > The changes outlined at > https://wiki.asterisk.org/wiki/display/AST/PJSIP+Subscription+Abstraction+Plan > provide a fairly accurate overview of what this review request entails. Some > differences are: > * I was not able to completely separate subscribers and notifiers into > separate structures. Instead, they are now separate sub-structures within the > ast_sip_subscription_handler struct. The reason for this is that PJSIP is not > flexible enough to register separate entities for clients and servers. > * The ast_sip_subscription_accept(), ast_sip_subscription_reject(), and > ast_sip_subscription_terminate() calls ended up not being needed. > * All functions on the page that have a body parameter as a const char * are > implemented to use a pjsip_msg_body instead. This is a case where the > underlying PJSIP structure does not need to be abstracted, and having a > parsed structure rather than a text blob will benefit everyone. > * One thing that is not on that page, but that you will find in this review, > is that we now store the SUBSCRIBE request that starts a subscription on the > subscription dialog. This allows subscription handlers to be able to retrieve > the values of certain headers from the SUBSCRIBE if they wish. This is > necessary for the exten_state handler currently since it wants the content of > the User-Agent header. > > > Diffs > ----- > > /trunk/res/res_pjsip_xpidf_body_generator.c 416652 > /trunk/res/res_pjsip_pubsub.exports.in 416652 > /trunk/res/res_pjsip_pubsub.c 416652 > /trunk/res/res_pjsip_pidf_body_generator.c 416652 > /trunk/res/res_pjsip_mwi.c 416652 > /trunk/res/res_pjsip_exten_state.c 416652 > /trunk/include/asterisk/res_pjsip_pubsub.h 416652 > > Diff: https://reviewboard.asterisk.org/r/3628/diff/ > > > Testing > ------- > > Since no new functionality has been added, I just ran the current gamut of > testsuite tests (tests/channels/pjsip/subscriptions). They all pass. > > > Thanks, > > Mark Michelson > >
-- _____________________________________________________________________ -- Bandwidth and Colocation Provided by http://www.api-digital.com -- asterisk-dev mailing list To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-dev
