On 8/10/07, Stefano Bagnara <[EMAIL PROTECTED]> wrote:
> Robert Burrell Donkin ha scritto:
> > On 8/10/07, Stefano Bagnara <[EMAIL PROTECTED]> wrote:
> >> Robert Burrell Donkin (JIRA) ha scritto:
> >>>     [ 
> >>> https://issues.apache.org/jira/browse/JSIEVE-6?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12518975
> >>>  ]
> >>>
> >>> Robert Burrell Donkin commented on JSIEVE-6:
> >>> --------------------------------------------
> >>>
> >>> There is also the design issue of where the address list parsing should 
> >>> be done.
> >>>
> >>> Adding a getAddresses(headerName) method to MailAdapter may allow the 
> >>> adapter to perform the parsing. The adapter may be able to optimise this 
> >>> or may be able to perform more intelligently parsing. However, this is at 
> >>> the cost of a wider API. Implementing in Address would have the advantage 
> >>> that it would only need to be done correctly in one place.
> >> I'm slightly on the MailAdapter.getAddresses side.
> >>
> >> If I understand it correctly this way we don't have to add the parsing
> >> to the jSieve core and we could use mime4j on the james adapter side to
> >> support this feature. (maybe this also means that it would be better to
> >> move the james adapter to the james server source tree ?)
> >
> > this does seem best
> >
> > MimeMessage contains address parsing code for the common headers. IMHO
> > it would be acceptable for implementations to offer a service limited
> > to common address headers anyway. the implementations in jsieve have
> > access to a MimeMessage representation. the JAMES version could offer
> > a more comprehensive service using Mime4J.
> >
> > but what to do when the address cannot be parsed? i favour logging and
> > returning an empty list.
>
> I guess this problem belongs to the matcher/mailet using jsieve, right?
>
> In this case I think the best way is to simply throw an exception and
> then document the use of standard exception
> onMatchException/onMailetException handling in james server:
> http://wiki.apache.org/james/HandlingExceptions
>
> I think the exception handling is one of the most interesting features
> and yet less used features we have in the mailet processing engine.

from RFC3028:

> Implementations MUST restrict the address test to headers that
> contain addresses, but MUST include at least From, To, Cc, Bcc,
> Sender, Resent-From, Resent-To, and SHOULD include any other header
> that utilizes an "address-list" structured header body.

so, i now agree that throwing an exception is the right behaviour

- robert

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

Reply via email to