SIP SIMPLE Client SDK » Python-msrplib

h1. MSRP API

Message Session Relay Protocol (MSRP) is a protocol for transmitting a series of related Instant Messages in the context of a session. Message sessions are treated like any other media stream when set up via a rendezvous or session creation protocol such as the Session Initiation Protocol (SIP).  MSRP can be used for any application that requires a reliable stream of data, for example File Transfer or Desktop Sharing.

* MSRP sessions are defined in "RFC 4975":http://tools.ietf.org/html/rfc4975

* MSRP relay extension for NAT traversal is defined in "RFC 4976":http://tools.ietf.org/html/rfc4976

The MSRP protocol stack is implemented by "python-msrplib":http://download.ag-projects.com/SipClient package. 

@msrplib@ is based upon "twisted":http://twistedmatrix.com and "eventlet":http://download.ag-projects.com/SipClient/ and provides a set of classes for establishing and managing MSRP connections.

The library consists of the following modules:

*msrplib.transport*

 Defines @MSRPTransport@ class, which provides low level control over MSRP connections.

*msrplib.connect*

 Defines means to establish a connection, bind it, and provide an initialized @MSRPTransport@ instance.

*msrplib.session*

 Defines @MSRPSession@ class, which provides high level control over a MSRP connection.

*msrplib.protocol*

 Provides representation and parsing of MSRP entities - chunks and MSRP URIs.  

h2. URI

This class is implemented in the @msrplib.protocol@ module and is used to represent an MSRP URI.

h4. methods

*<notextile>__init__</notextile>*(_self_, *host*=@None@, *use_tls*=@False@, *user*=@None@, *port*=@None@, *session_id*=@None@, *parameters*=@None@, *credentials*=@None@)

>Constructs a new @URI@. All the arguments specified here are also attributes on the object. For more information on these attributes, see RFC4975.

>+_host_+:

>The hostname or IP address which forms the URI.

>+_use_tls_+:

>Whether this identifies an msrps or msrp URI.

>+_user_+:

>The user part of the URI.

>+_port_+:

>The port in the URI.

>+_session_id_+:

>The session identifier part of the URI.

>+_parameters_+:

>A @dict@ containing the parameters which will be appended to the URI.

>+_credentials_+:

>A @gnutls.interfaces.twisted.X509Credentials@ object which will be used if this identifies a TLS URI to authenticate to the other endpoint.

h2. MSRPRelaySettings

This class is implemented in the @msrplib.connect@ module and is used to specify the MSRP relay which will be used when connecting via a relay (using the @ConnectorRelay@ or @AcceptorRelay@ classes).

h4. methods

*<notextile>__init__</notextile>*(_self_, *domain*, *username*, *password*, *host*=@None@, *use_tls*=@False@, *port*=@None@, *credentials*=@None@)

>Constructs a new @URI@. All the arguments specified here are also attributes on the object. For more information on these attributes, see RFC4975.

>+_domain_+:

>The DNS domain in which to search for a MSRP relay using SRV queries.

>+_username_+:

>The username which will be used to authenticate to the relay.

>+_password_+:

>The password which will be used to authenticate to the relay.

>+_host_+:

>The hostname or IP address of the MSRP relay.

>+_use_tls_+:

>Whether this identifies an msrps or msrp URI.

>+_port_+:

>The port in the URI.

>+_credentials_+:

>A @gnutls.interfaces.twisted.X509Credentials@ object which will be used to authenticate to the relay if TLS is used.

h2. ConnectorDirect

This class is implemented in the @msrplib.connect@ module and is used for outbound MSRP connections without a relay.

h4. methods

*<notextile>__init__</notextile>*(_self_, *loger*=@None@, *use_sessmatch*=@False@)

 Constructs a new ConnectorDirect.

>+_logger_+:

 The logger which will be used for this MSRP connection. See the [[SipMSRPApi#Logging|Logging]] section for more information.

>+_use_sessmatch_+:

 Indicates if the connector should use the session matching mechanism defined by the "sessmatch draft version 10":http://tools.ietf.org/html/draft-ietf-simple-msrp-sessmatch-10

*prepare*(_self_, *local_uri*=@None@)

 This method returns a full local path - list of @protocol.URI@ instances, suitable to be put in SDP @'a:path'@ attribute.

>+_local_uri_+:

 This attribute will be used to construct the local path, but other than that it is not used anywhere else in case of the ConnectorDirect. If not provided, one

 will be automatically generated

*complete*(_self_, *full_remote_path*)

 This method establishes the connection and binds it (sends an empty chunk to verify each other's From-Path and To-Path). It returns @transport.MSRPTransport@ instance, ready to read and send chunks.

>+_full_remote_path_+:

 A list of @protocol.URI@ instances, obtained by parsing @'a:path'@ put in SDP by the remote party.

*cleanup*(_self_)

 This method cleans up after @prepare()@; it should be called if @complete()@ will not be called for whatever reason.

h2. AcceptorDirect

This class is implemented in the @msrplib.connect@ module and is used for inbound MSRP connections without using a relay.

h4. methods

*<notextile>__init__</notextile>*(_self_, *loger*=@None@, *use_sessmatch*=@False@)

>Constructs a new AcceptorDirect.

>+_logger_+:

 The logger which will be used for this MSRP connection. See the [[SipMSRPApi#Logging|Logging]] section for more information.

>+_use_sessmatch_+:

 Indicates if the connector should use the session matching mechanism defined by the "sessmatch draft version 10":http://tools.ietf.org/html/draft-ietf-simple-msrp-sessmatch-10

*prepare*(_self_, *local_uri*=@None@)

 This method starts listening on a socket identified by the parameters in the @local_uri@ argument. It returns a full local path - list of @protocol.URI@ instances, suitable to be put in SDP @'a:path'@ attribute.

>+_local_uri_+:

 This attribute will be used to construct the local path and to listen for incomming connections. If not provided, one

 will be automatically generated. Note that the object may be changed in place: for example, if the port specified is 0,

 a random one will be selected and the object will be updated accordingly.

*complete*(_self_, *full_remote_path*)

 This method waits for an incoming connection and a chunk sent by the other party. It returns @transport.MSRPTransport@ instance, ready to read and send chunks.

>+_full_remote_path_+:

 A list of @protocol.URI@ instances, obtained by parsing @'a:path'@ put in SDP by the remote party. This is checked agains the From-Path header in the binding chunk.

*cleanup*(_self_)

 This method cleans up after @prepare()@; it should be called if @complete()@ will not be called for whatever reason.

h2. RelayConnection

This class is implemented in the @msrplib.connect@ module and is used for inbound and outbound MSRP connections using a relay.

h4. methods

*<notextile>__init__</notextile>*(_self_, *relay*, *mode*, *loger*=@None@, *use_sessmatch*=@False@)

 Constructs a new RelayConnection.

>+_relay_+:

 An instance of @MSRPRelaySettings@ which identifies the relay which is to be used.

>+_mode_+:

 A string indicating if this connection should be _active_, an empty SEND should be sent when @complete@ is called, or                         _passive_, where it will wait for one.

>+_logger_+:

 The logger which will be used for this MSRP connection. See the [[SipMSRPApi#Logging|Logging]] section for more information.

>+_use_sessmatch_+:

 Indicates if the connector should use the session matching mechanism defined by the "sessmatch draft version 10":http://tools.ietf.org/html/draft-ietf-simple-msrp-sessmatch-10

*prepare*(_self_, *local_uri*=@None@)

 This method returns a full local path - list of @protocol.URI@ instances, suitable to be put in SDP @'a:path'@ attribute.

>+_local_uri_+:

 This attribute will be used to construct the local path, but other than that it is not used anywhere else in case of the ConnectorRelay. If not provided, one

 will be automatically generated

*complete*(_self_, *full_remote_path*)

 This method establishes the connection to the relay and binds it (sends an empty chunk or waits for one, depending on the mode, to verify each other's From-Path and To-Path). It returns @transport.MSRPTransport@ instance, ready to read and send chunks.

>+_full_remote_path_+:

 A list of @protocol.URI@ instances, obtained by parsing @'a:path'@ put in SDP by the remote party.

*cleanup*(_self_)

 This method cleans up after @prepare()@; it should be called if @complete()@ will not be called for whatever reason.

h2. MSRPTransport

This class is implemented in the @msrplib.transport@ module and provies low level access to the MSRP connection. This class should not be constructed directly, but rather its intances should be obtained by using the various connector/acceptor classes documented above.

h4. methods

*make_chunk*(_self_, *transaction_id*=@None@, *method*=@'SEND'@, *code*=@None@, *comment*=@None@, *data*=@''@, *contflag*=@None@, *start*=@1@, *end*=@None@, *length*=@None@, *message_id*=@None@)

 Creates a new chunk (@protocol.MSRPData@ instance), which is either an MSRP request or a response. Proper @From-Path@, @To-Path@, @Byte-Range@ and @Message-ID@ headers are added based on MSRPTransport's state and the parameters provided. Use _data_ for payload, and _start_/_end_/_length_ to generate the @Byte-Range@ header.

>+_transaction_id_+:

 The transaction id which will be put in the chunk. If it is not provided, one will be randomly generated.

>+_method_+:

 The method of the new MSRP request, or @None@ if a response is required.

>+_code_+:

 The code of the new MSRP response, or @None@ if a request is required.

>+_code_+:

 The comment of the new MSRP response, or @None@ if a request is required or a comment on the response is not.

>+_data_+:

 The payload of the new chunk, or an empty string if no payload is required.

>+_contflag_+:

 MSRP chunk's continuation flag (@'$'@, @'+'@ or @'#'@). Default is @'$'@ for a full message, unless a partial @SEND@ chunk required, in which case it should be set to @'+'@. If @None@ is provided, either @'$'@ or @'+'@ will be used depending on whether this chunk seems to carry the last part of the message data.

>+_start_+:

 The first byte's index within the whole message payload this chunk will carry as its own payload.

>+_end_+:

 The last byte's index within the whole message payload this chunk will carry as its own payload. Note that this is an inclusive index. If this is not provided, it's computed based on the number of bytes in data and @start@.

>+_length_+:

 The total number of bytes of the whole message payload. If this is not provided, it is computed as if this chunk carries that last part of the message payload.

>+_message_id_+:

 The ID of the message this chunk is part of. If it is not provided, one will be randomly generated.

*write_chunk*(_self_, *chunk*, *wait*=@True@)

 Writes the chunk provided to the underlying socket.

>+_chunk_+:

 The chunk which is to be written, an instance of @MSRPData@.

>+_wait_+:

 If @True@, waits for the operation to complete.

*write_response*(_self_, *chunk*, *code*, *comment*, *wait*=@True@):

 Creates a response which is suitable as a reply to the specified chunk and writes it to the underlying socket.

>+_chunk_+:

 The chunk for which to create a response.

>+_code_+:

 The status code of the response which is to be created.

>+_comment_+:

 The comment of the response which is to be created.

>+_wait_+:

 If @True@, waits for the operation to complete.

*read_chunk*(_self_, *size*=@None@)

 Waits for a new chunk and returns it.

 If there was an error, closes the connection and raises @ChunkParseError@.

 In case of unintelligible input, loses the connection and returns @None@.

 When the connection is closed, raises the reason of the closure (e.g. @ConnectionDone@).

 If the data already read exceeds _size_, stops reading the data and returns

 a "virtual" chunk, i.e. the one that does not actually correspond the the real

 MSRP chunk. Such chunks have Byte-Range header changed to match the number of

 bytes read and continuation that is @'+'@; they also possess @segment@ attribute,

 an integer, starting with 1 and increasing with every new segment of the chunk.

 Note, that _size_ only hints when to interrupt the segment but does not affect

 how the data is read from socket. You may have segments bigger than _size_ and it's

 legal to set _size_ to zero (which would mean return a chunk as long as you get

 some data, regardless how small).

>+_size_+:

 The hint towards how much to read from the socket. If the data already read is larger, then all the data will be returned, even if it exceeds @size@ bytes.

*check_incoming_SEND_chunk*(_self_, *chunk*)

 Checks the 'To-Path' and 'From-Path' of the incoming SEND chunk.

 Returns None is the paths are valid for this connection.

 If an error is detected an MSRPError is created and returned.

h2. MSRPData

This class is implemented in the @msrplib.protocol@ module and represents an MSRP chunk, either a request or a response.

h4. attributes

 The following attributes are read-only.

*content_type*

>The MIME type of the payload carried by this chunk, which is stored in the Content-Type header.

*message_id*

>The ID of the message this chunk is part of, which is stored in the Message-ID header.

*byte_range*

>A 3-tuple containing the start, end and length values (in this order) from the Byte-Range header.

*status*

>The value of the Status header.

*failure_report*

>The value of the Failure-Report header, whether it exists or it is implied; one of @'yes'@, @'no'@, @'partial'@.

*success_report*

>The value of the Success-Report header, whether it exists or it is implied; one of @'yes'@ or @'no'@.

*size*

>The size of the payload of this chunk, in bytes.

h4. methods

*<notextile>__init__</notextile>*(_self_, *transaction_id*, *method*=@None@, *code*=@None@, *comment*=@None@, *headers*=@None@, *data*=@''@, *contflag*=@'$'@)

>Initializes a new MSRPData instance. All the arguments are also available as attributes.

>+_transaction_id_+:

>The transaction identified of this chunk.

>+_method_+:

>The method of this chunk if it a request and @None@ if it is a response.

>+_code_+:

>The status code of this chunk if it is a response and @None@ if it is a request.

>+_comment_+:

>The comment of this chunk if it is a response and @None@ if it is a request.

>+_headers_+:

>A dict containing the headers which are to be added to this chunk, or @None@ if no headers are to be added.

>+_data_+:

>The payload of this chunk, or an empty string if no payload is to be added.

>+_contflag_+:

>The MSRP continuation flag of this chunk, one of @'$'@, @'#'@ or @'+'@.

*copy*(_self_)

>Returns a new MSRPData instance with exactly the same attributes as this object.

*add_header*(_self_, *header*)

>Add a MSRP header to this chunk.

>+_header_+:

>The header object which is to be added to this chunk.

*verify_headers*(_self_)

>Verifies that the chunk contains a To-Path and a From-Path header and that all the headers are valid.

*encode_start*(_self_)

>Returns a string containing the MSRP header of this chunk.

*encode_end*(_self_, *continuation*)

>Returns a string containing the MSRP end line of this chunk: 7 commas followed by the transaction identifier and the continuation flag specified.

>+_continuation_+:

>The continuation flag which is to be used in the end line.

*encode*(_self_):

>Returns a string containing the whole of this MSRP chunk.

h2. OutgoingFile

This class is implemented in the @msrplib.session@ module and represents a file which is to be sent via MSRP.

h4. attributes

*headers*

>A dict which maps header names to header objects. These headers will be sent in the MSRP message used to send the file.

h4. methods

*<notextile>__init__</notextile>*(_self_, *fileobj*, *size*, *content_type*=@None@, *position*=@0@, *message_id*=@None@)

>Initializes a new OutgoingFile using the specified file object. All arguments are available as attributes, except for @content_type@ which will be part of the @headers@ attribute.

>+_fileobj_+:

>A file-like object which will be used for reading the data to be sent.

>+_size_+:

>The size of the whole file.

>+_content_type_+:

>The MIME type associated with the file's data. If provided, it will be added as a Content-Type header.

>+_position_+:

>The position within the file from which data will be sent. The file object must already be seeked to this position.

>+_message_id_+:

>The message ID which will be used for sending this file.

h2. MSRPSession

This class is implemented in the @msrplib.session@ module and provides a high level API for MSRP sessions.

h4. methods

*<notextile>__init__</notextile>*(_self_, *msrptransport*, *accept_types*=@['*']@, *on_incoming_cb*=@None@)

 Initializes MSRPSession instance over the specified transport. The incoming chunks are reported through the _on_incoming_cb_ callback.

>+_msrptransport_+:

 An instance of @MSRPTransport@ over which this session will operate.

>+_accept_types_+:

 A list of MIME types which are acceptable over this session. If data is received with a Content-Type header which doesn't match this list, a negative response is sent back and the application does not get the received data. The special strings @'*'@ and @'<type>/*'@ can be used to match multiple MIME types.

>+_on_incoming_cb_+:

 A function which receives two arguments, both optional with default values of @None@: _chunk_ and _error_. This will be called when a new chunk is received.

*send_chunk*(_self_, *chunk*, *response_cb*=@None@)

 Sends the specified chunk and reports the result via the _response_cb_ callback.

 When _response_cb_ argument is present, it will be used to report

 the transaction response to the caller. When a response is received or generated

 locally, _response_cb_ is called with one argument. The function must do something

 quickly and must not block, because otherwise it would block the reader greenlet.

 If no response was received after @RESPONSE_TIMEOUT@ seconds,

* 408 response is generated if Failure-Report was @'yes'@ or absent

* 200 response is generated if Failure-Report was @'partial'@ or @'no'@

 Note that it's rather wasteful to provide _response_cb_ argument other than @None@

 for chunks with Failure-Report='no' since it will always fire 30 seconds later

 with 200 result (unless the other party is broken and ignores Failure-Report header)

 If sending is not possible, @MSRPSessionError@ is raised.

>+_chunk_+:

 The chunk which is to be sent. It must be an instance of @MSRPData@ which represents a request.

>+_response_cb_+:

 A function receiving a single argument which will be the response received for the sent chunk.

*deliver_chunk*(_self_, *chunk*, *event*=@None@)

 Sends the specified chunk and waits for the transaction response (if Failure-Report header is not @'no'@).

 Returns the transaction response if it's a success or raise @MSRPTransactionError@ if it's not.

 If chunk's Failure-Report is @'no'@, returns @None@ immediately.

>+_chunk_+:

 The chunk which is to be sent. It must be an intance of @MSRPData@ which represents a request.

>+_event_+:

 The @eventlet.coros.event@ object which will be used to wait for a response. It's send method will be called when a response is received. If it is not provided, one will be constructed automatically and used.

*send_file*(_self_, *outgoing_file*)

 Adds the specified file to the queue of files to be sent. The method returns immediately.

>+_outgoing_file_+:

 An instance of @OutgoingFile@ which represents the file to be sent.

*shutdown*(_self_, *sync*=@True@)

 Sends the messages already in queue then closes the connection.

h2. MSRPServer

This class is implemented in the @msrplib.connect@ module.

MSRPServer solves the problem with AcceptorDirect: concurrent using of 2

or more AcceptorDirect instances on the same non-zero port is not possible.

If you initialize() those instances, one after another, one will listen on

the socket and another will get BindError.

MSRPServer avoids the problem by sharing the listening socket between multiple connections.

It has a slightly different interface from AcceptorDirect, so it cannot be considered a drop-in

replacement.

It manages the listening sockets and binds incoming requests.

h4. methods

*<notextile>__init__</notextile>*(_self_, *logger*)

 Constructs a new MSRPServer which will be using the specifed logger for its connections.

>+_logger_+:

 The default logger which will be used for this object's MSRP connections. See the [[SipMSRPApi#Logging|Logging]] section for more information.

*prepare*(_self_, _local_uri_=@None@, _logger_=@None@)

 Starts a listening port specified by _local_uri_ if there isn't one on that port/interface already.

 Adds _local_uri_ to the list of expected URIs, so that incoming connections featuring this URI won't be rejected.

 If _logger_ is provided, it uses it for this connection instead of the one specified in @__init___@.

>+_local_uri_+:

 The URI which will be added to the list of expected URIs. Connections from this URI will be accepted.

>+_logger_+:

 The logger which will be used for connections from the specified URI. See the [[SipMSRPApi#Logging|Logging]] section for more information.

*complete*(_self_, _full_remote_path_)

 Waits until one of the incoming connections binds using the provided _full_remote_path_.

 Returns the connected and bound @MSRPTransport@ instance.

 If no such binding was made within @MSRPBindSessionTimeout.seconds@, @MSRPBindSessionTimeout@ is raised.

>+_full_remote_path_+:

 A list of @protocol.URI@ instances, obtained by parsing @'a:path'@ put in SDP by the remote party.

*cleanup*(_self_, _local_uri_)

 Removes _local_uri_ from the list of expected URIs.

>+_local_uri_+:

 The URI which is to be removed from the list of expected URIs.

h2. Headers

The MSRP headers are represented using objects from the @msrplib.protocol module@. All MSRP header object provide three properties:

 _name_:

 The name of the header, as it appears in MSRP chunks.

>+_encode_+:

 The MSRP encoded header value, as it appears in MSRP chunks.

>+_decode_+:

 The high-level object representing the header value.

All headers can be constructed by passing either the encoded or decoded objects to @__init__@. The following headers are provided:

h3. ToPathHeader

The name of the header is @'To-Path'@ and the decoded value is a deque of @URI@ objects.

h3. FromPathHeader

The name of the header is @'From-Path'@ and the decoded value is a deque of @URI@ objects.

h3. MessageIDHeader

The name of the header is @'Message-ID'@ and the decoded value is the string containing the message ID.

h3. SuccessReportHeader

The name of the header is @'Success-Report'@ and the decoded value is one of @'yes'@ or @'no'@.

h3. FailureReportHeader

The name of the header is @'Failure-Report'@ and the decoded value is one of @'yes'@, @'partial'@ or @'no'@.

h3. ByteRangeHeader

The name of the header is @'Byte-Range'@ and the decoded value is a 3-tuple containing the start, end and length values.

h4. attributes

*fro*

>The start value of the header: the index within the whole message payload where a chunk's payload starts.

*end*

>The end value of the header: the index within the whole message payload where a chunk's payload ends. Note that this index is inclusive.

*length*

>The total size of the message payload.

h3. StatusHeader

The name of the header is @'Status'@ and the decoded value is a 2-tuple containing the status code and comment.

h4. attributes

*code*

>The code component of the header.

*comment*

>The comment component of the header.

h3. ExpiresHeader

The name of the header is @'Expires'@ and the decoded value is an integer.

h3. MinExpiresHeader

The name of the header is @'Min-Expires'@ and the decoded value is an integer.

h3. MaxExpiresHeader

The name of the header is @'Max-Expires'@ and the decoded value is an integer.

h3. UsePathHeader

The name of the header is @'Use-Path'@ and the decoded value is a deque of @URI@ objects.

h3. WWWAuthenticateHeader

The name of the header is @'WWW-Authenticate'@ and the decoded value is a dictionary of the parameters within the header.

h3. AuthorizationHeader

The name of the header is @'Authorization'@ and the decoded value is a dictionary of the parameters within the header.

h3. AuthenticationInfoHeader

The name of the header is @'Authentication-Info'@ and the decoded value is a dictionary of the parameters within the header.

h3. ContentTypeHeader

The name of the header is @'Content-Type'@ and the decoded value is the string representing the MIME type.

h3. ContentIDHeader

The name of the header is @'Content-ID'@ and the decoded value is the string with the value of the header.

h3. ContentDescriptionHeader

The name of the header is @'Content-Description'@ and the decoded value is the string with the value of the header.

h3. ContentDispositionHeader

The name of the header is @'Content-Disposition'@ and the decoded value is a list with two elements: the disposition and a dict with the parameters.

h2. Logging

Logging is done throughout the library using objects defined by the application, called loggers. If logging is not desired, the @application.python.Null@ object of @python-application@ can be used. These loggers must define the following methods:

h4. methods

*received_new_chunk*(*data*, *transport*, *chunk*)

>This method is called when the start of a new chunk is received.

>+_data_+:

>The data which came along with the start of the chunk.

>+_transport_+:

>The @MSRPTransport@ instance on which the chunk was received.

>+_chunk_+:

>The actual chunk object.

*received_chunk_data*(*data*, *transport*, *transaction_id*)

>This method is called when data is received as part of a chunk previously announced via *received_new_chunk*.

>+_data_+:

>The data received as part of the chunk.

>+_transport_+:

>The @MSRPTransport@ instance on which the chunk was received.

>+_transaction_id_+:

>The transaction ID which identifies the chunk for which data was received.

*received_chunk_end*(*data*, *transport*, *transaction_id*)

>This method is called when the last data of a chunk is received. The chunk was previously announced via *received_new_chunk*.

>+_data_+:

>The last data received as part of the chunk.

>+_transport_+:

>The @MSRPTransport@ instance on which the chunk was received.

>+_transaction_id_+:

>The transaction ID which identifies the chunk which was ended.

*sent_new_chunk*(*data*, *transport*, *chunk*)

>This method is called when the start of a new chunk is sent.

>+_data_+:

>The data which was sent along with the start of the chunk.

>+_transport_+:

>The @MSRPTransport@ instance on which the chunk was sent.

>+_chunk_+:

>The actual chunk object.

*sent_chunk_data*(*data*, *transport*, *transaction_id*)

>This method is called when data is sent as part of a chunk previously announced via *sent_new_chunk*.

>+_data_+:

>The data sent as part of the chunk.

>+_transport_+:

>The @MSRPTransport@ instance on which the chunk was sent.

>+_transaction_id_+:

>The transaction ID which identifies the chunk for which data was sent.

*sent_chunk_end*(*data*, *transport*, *transaction_id*)

>This method is called when the last data of a chunk is sent. The chunk was previously announced via *sent_new_chunk*.

>+_data_+:

>The last data sent as part of the chunk.

>+_transport_+:

>The @MSRPTransport@ instance on which the chunk was sent.

>+_transaction_id_+:

>The transaction ID which identifies the chunk which was ended.

*debug*(*message*)

>This method is called when a debug level message is sent by the library.

*info*(*message*)

>This method is called when a info level message is sent by the library.

*warn*(*message*)

>This method is called when a warning level message is sent by the library.

*error*(*message*)

>This method is called when a error level message is sent by the library.

*fatal*(*message*)

>This method is called when a fatal level message is sent by the library.

h2. Examples

h3. Creating an outbound connection

When creating an outbound connection, a relay is not usually used because NAT traversal is not a problem for the endpoing creating the connection. If one is nevertheless desired, a @ConnectorRelay@ can be used instead:

<pre>

from msrplib.connect import ConnectorDirect

from msrplib.session import MSRPSession

connector = ConnectorDirect()

full_local_path = connector.prepare()

try:

    # put full_local_path in SDP 'a:path' attribute of offer

    # get full_remote_path from remote's 'a:path: attribute of reply

    msrp_transport = connector.complete(full_remote_path)

except:

    connector.cleanup()

    raise

def handle_incoming(chunk=None, error=None):

    if error is not None:

       print 'Error: %s' % error

       session.shutdown()

    elif chunk is not None:

       print 'Got data: %s' % chunk.data

session = MSRPSession(msrp_transport, on_incoming_cb=handle_incoming)

session.send_chunk(msrp_transport.make_chunk(data='Hello world!'))

</pre>

h3. Waiting for an inbound connection

When creating an inbound connection, a relay must be used for NAT traversal. However, if one is not desired, an @AceptorDirect@ can be used instead:

<pre>

from msrplib.connect import RelayConnection, MSRPRelaySettings

from msrplib.session import MSRPSession

relay = MSRPRelaySettings(domain='example.org', username='user', password='pass')

connector = RelayConnection(relay, 'passive')

full_local_path = connector.prepare()

try:

    # get full_remote_path from remote's 'a:path: attribute of offer

    # put full_local_path in SDP 'a:path' attribute of reply

    msrp_transport = connector.complete(full_remote_path)

except:

    connector.cleanup()

    raise

def handle_incoming(chunk=None, error=None):

    if error is not None:

       print 'Error: %s' % error

       session.shutdown()

    elif chunk is not None:

       print 'Got data: %s' % chunk.data

session = MSRPSession(msrp_transport, on_incoming_cb=handle_incoming)

session.send_chunk(msrp_transport.make_chunk(data='Hello world!'))

</pre>