- SàT can now handle components
- plugin have now a "modes" key in PLUGIN_INFO where they declare if they can be used with clients and or components. They default to be client only.
- components are really similar to clients, but with some changes in behaviour:
* component has "entry point", which is a special plugin with a componentStart method, which is called just after component is connected
* trigger end with a different suffixes (e.g. profileConnected vs profileConnectedComponent), so a plugin which manage both clients and components can have different workflow
* for clients, only triggers of plugins handling client mode are launched
* for components, only triggers of plugins needed in dependencies are launched. They all must handle component mode.
* component have a sendHistory attribute (False by default) which can be set to True to allow saving sent messages into history
* for convenience, "client" is still used in method even if it can now be a component
* a new "component" boolean attribute tells if we have a component or a client
* components have to add themselve Message protocol
* roster and presence protocols are not added for components
* component default port is 5347 (which is Prosody's default port)
- asyncCreateProfile has been renamed for profileCreate, both to follow new naming convention and to prepare the transition to fully asynchronous bridge
- createProfile has a new "component" attribute. When used to create a component, it must be set to a component entry point
- jp: added --component argument to profile/create
- disconnect bridge method is now asynchronous, this way frontends can know when disconnection is finished
- new PI_* constants for PLUGIN_INFO values (not used everywhere yet)
- client/component connection workflow has been moved to their classes instead of being a host methods
- host.messageSend is now client.sendMessage, and former client.sendMessage is now client.sendMessageData.
- identities are now handled in client.identities list, so it can be updated dynamically by plugins (in the future, frontends should be able to update them too through bridge)
- profileConnecting* profileConnected* profileDisconnected* and getHandler now all use client instead of profile
/!\ not everything is working yet, and specially groupblogs are broken /!\
- renamed bridge api to use prefixed methods (e.g. psSubscribeToMany instead of subscribeToMany in PubSub)
- (xep-0060): try to find a default PubSub service, and put it in client.pubsub_service
- (xep-0060): extra dictionary can be used in bridge method for RSM and other options
- (xep-0060): XEP_0060.addManagedNode and XEP_0060.removeManagedNode allow to easily catch notifications for a specific node
- (xep-0060): retractItem manage "notify" attribute
- (xep-0060): new signal psEvent will be used to transmit notifications to frontends
- (xep-0060, constants): added a bunch of useful constants
- (xep-0163): removed personalEvent in favor of psEvent
- (xep-0163): addPEPEvent now filter non PEP events for in_callback
- (xep-0277): use of new XEP-0060 plugin's addManagedNode
- (xep-0277): fixed author handling for incoming blogs: author is the human readable name, author_jid it jid, and author_jid_verified is set to True is the jid is checked
- (xep-0277): reworked data2entry with Twisted instead of feed, item_id can now be specified, <content/> is changed to <title/> if there is only content
- (xep-0277): comments are now managed here (core removed from groupblog)
- (xep-0277): (comments) node is created if needed, default pubsub service is used if available, else PEP
- (xep-0277): retract is managed
- renamed variables nammed after reserved word/module
- removed class variables which are not needed
- removed external parenthesis from asserts
- in mam.MAMPrefs, removed default value for default argument, as None can't be used
- (groupblog) extended docstring for DeferredItems and DeferredItemsFromMany
- removed external parenthesis in assertions
- added blank line after first line of docstrings
- replaced "__" prefixes by "_"
- renamed variabled named after the reserverd word "max" to "max_"
- hashing algorithm of XEP-0115 has been including in core
- our own hash is still calculated by XEP-0115 and can be regenerated with XEP_0115.recalculateHash
- old discovery methods have been removed. Now the following methods are used:
- hasFeature: tell if a feature is available for an entity
- getDiscoInfos: self explaining
- getDiscoItems: self explaining
- findServiceEntities: return all available items of an entity which given (category, type)
- findFeaturesSet: search for a set of features in entity + entity's items
all these methods are asynchronous, and manage cache automatically
- XEP-0115 manage in a better way hashes, and now use a trigger for presence instead of monkey patch
- new FeatureNotFound exception, when we want to do something which is not available
- refactored client initialisation sequence, removed client.initialized Deferred
- added constant APP_URL
- test_plugin_xep_0033.py has been temporarly deactivated, the time to adapt it
- lot of cleaning
- add or update some methods docstrings
- add GroupBlog.__fillCommentsElement to factorize the code responsible to build the comments link
- do not mix up attributes names between the item own service/node and the comments service/node
Implementation should follow the following formal specification:
"title" and "content" data can be passed in raw, xhtml or rich format.
When we receive from a frontend a new/updated microblog item:
- keys "title" or "content" have to be escaped (disable HTML tags)
- keys "title_rich" or "content_rich" have to be converted from the current syntax to XHTML
- keys "title_xhtml" or "content_xhtml" have to be cleaned from unwanted XHTML content
Rules to deal with concurrent keys:
- existence of both "*_xhtml" and "*_rich" keys must raise an exception
- existence of both raw and ("*_xhtml" or "*_rich") is OK
As the storage always need raw data, if it is not given by the user it can be
extracted from the "*_rich" or "*_xhtml" data (remove the XHTML tags).
When a frontend wants to edit a blog post that contains XHTML title or content,
the conversion is made from XHTML to the current user-defined syntax.
- plugin text_syntaxes: added "text" syntax (using lxml)
- _() is no more installed in __builtin__
- instead, there is a new sat.core.i18n module
- added D_() method for deferred translation
- languageSwitch method allow to dynamically change translation language
- import gettext is tested against ImportError, and dummy methods are used when not available (mainly useful for Libervia)
