Receiving multiple message types - Protocol Buffers

The old Google Groups will be going away soon, but your browser is incompatible with the new version.
Message from discussion Receiving multiple message types
Date: Mon, 19 Mar 2012 12:46:16 -0700 (PDT)
From: Jim Morris <wolfma...@gmail.com>
To: protobuf@googlegroups.com
Message-ID: <16384733.3636.1332186376922.JavaMail.geo-discussion-forums@pbcr5>
In-Reply-To: <da806534-cc56-49bb-8950-1450fc11f006@do4g2000vbb.googlegroups.com>
References: <da806534-cc56-49bb-8950-1450fc11f006@do4g2000vbb.googlegroups.com>
Subject: Re: Receiving multiple message types
MIME-Version: 1.0
Content-Type: multipart/mixed; 
	boundary="----=_Part_3634_29960742.1332186376921"

------=_Part_3634_29960742.1332186376921
Content-Type: multipart/alternative; 
	boundary="----=_Part_3635_32395466.1332186376921"

------=_Part_3635_32395466.1332186376921
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit

There are several ways to do it, on eis called unions and the other is to 
use extensions. My blog entry here covers the latter and has links to the 
union method.

http://blog.wolfman.com/articles/2011/11/23/how-to-implement-polymorphic-protocol-buffers-in-java

You are on the right track though you just put all the messages as optional 
in the one message with the header, so you only need to decode once.

On Sunday, March 18, 2012 3:41:34 PM UTC-7, Jens Rantil wrote:
>
> Hi, 
>
> I'm a newcomer to protobuf and am trying to wrap my head around some 
> things. Let's say I have a message receiver of some kind. It can 
> receive different types requests. All messages are encoded in protobuf 
> and my receiver has all *.proto message definitions necessary be able 
> to decode all incoming requests. 
>
> My question is, does protobuf come with any bells and whistles that 
> let me decode a random incoming message, given all possible message 
> definitions? If not, how do you generally cope with this situation? 
> Because I feel it's a common one. 
>
> I guess one possible solution is to create another message definition: 
> message GenericMessage { 
>   enum MessageType { 
>     SET_NAME = 0; 
>     CREATE_CUSTOMER = 1; 
>   } 
>   required MessageType type = 0; 
>   required bytes protobuf_data = 1; 
> } 
>
> However, this would require me to do two deserializations. Is this the 
> prefered way to go? I'm curious to hear you input. Please pinpoint me 
> if this has been answered previously in forum or in the documentation. 
>
> Regards, 
> Jens


------=_Part_3635_32395466.1332186376921
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 7bit

There are several ways to do it, on eis called unions and the other is to use extensions. My blog entry here covers the latter and has links to the union method.<br><br>http://blog.wolfman.com/articles/2011/11/23/how-to-implement-polymorphic-protocol-buffers-in-java<br><br>You are on the right track though you just put all the messages as optional in the one message with the header, so you only need to decode once.<br><br>On Sunday, March 18, 2012 3:41:34 PM UTC-7, Jens Rantil wrote:<blockquote class="gmail_quote" style="margin: 0;margin-left: 0.8ex;border-left: 1px #ccc solid;padding-left: 1ex;">Hi,
<br>
<br>I'm a newcomer to protobuf and am trying to wrap my head around some
<br>things. Let's say I have a message receiver of some kind. It can
<br>receive different types requests. All messages are encoded in protobuf
<br>and my receiver has all *.proto message definitions necessary be able
<br>to decode all incoming requests.
<br>
<br>My question is, does protobuf come with any bells and whistles that
<br>let me decode a random incoming message, given all possible message
<br>definitions? If not, how do you generally cope with this situation?
<br>Because I feel it's a common one.
<br>
<br>I guess one possible solution is to create another message definition:
<br>message GenericMessage {
<br>&nbsp; enum MessageType {
<br>&nbsp; &nbsp; SET_NAME = 0;
<br>&nbsp; &nbsp; CREATE_CUSTOMER = 1;
<br>&nbsp; }
<br>&nbsp; required MessageType type = 0;
<br>&nbsp; required bytes protobuf_data = 1;
<br>}
<br>
<br>However, this would require me to do two deserializations. Is this the
<br>prefered way to go? I'm curious to hear you input. Please pinpoint me
<br>if this has been answered previously in forum or in the documentation.
<br>
<br>Regards,
<br>Jens</blockquote>
------=_Part_3635_32395466.1332186376921--

------=_Part_3634_29960742.1332186376921--
Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
Account Options
