[ 
http://issues.apache.org/jira/browse/AXIS2C-269?page=comments#action_12433594 ] 
            
James Clark commented on AXIS2C-269:
------------------------------------

The output stream needs flush and close methods as well.

> Try to clean up axis2_stream
> ----------------------------
>
>                 Key: AXIS2C-269
>                 URL: http://issues.apache.org/jira/browse/AXIS2C-269
>             Project: Axis2-C
>          Issue Type: Improvement
>          Components: util
>    Affects Versions: Current (Nightly)
>            Reporter: James Clark
>            Priority: Minor
>
> For doing encryption and signing right, it will be important to have a clean, 
> minimal stream interface.  axis2_stream is a bit crufty at the moment.
> Specifically, the get_type method is bogus: if you have an OO type (which 
> stream should be), then you shouldn't have a method returning a type-code for 
> each implementation.  It only seems to be used in 
> axis2_http_transport_utils_on_data_request. I don't understand what that's 
> doing, but I'm fairly sure get_type is not the best way to it.
> I would also suggest having separate input and output streams (unless there 
> are really cases where you need to have a random access stream that can be 
> both read and written).
> Having both get_char and read in an interface is a bad idea.  Have just read. 
>  You can implement get_char as a function on top of read.
> unget is  a bad idea on your fundamental stream type.  If you really need it, 
> create a lookahead_stream derived from stream that provides this additional 
> operation and forwards to an underlying regular stream.  Providing a peek 
> method is better than unget (it's clear what the semantics of multiple peeks 
> are).

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: 
http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to