james-mime4j-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Oleg Kalnichevski <ol...@apache.org>
Subject Re: Address list parser expects unfolded field body. Does that make sense?
Date Fri, 27 Nov 2009 21:21:23 GMT
Markus Wiederkehr wrote:
> On Fri, Nov 27, 2009 at 2:37 PM, Oleg Kalnichevski <olegk@apache.org> wrote:
>> Oleg Kalnichevski wrote:
>>> Markus et al
>>>
>>> The address list parser currently chokes on folded field values that are
>>> otherwise perfectly valid. That seems somewhat illogical to me. It really
>>> took me a while to figure out what was wrong with the address list until I
>>> stumbled upon a commend about the parser expecting unfolded fields. The very
>>> cryptic exception message did not really help either.
>>>
>>> What is the reason for this restriction? It is because folded values are
>>> difficult to parse with jjtree? Should not we unfold field values
>>> automatically prior to feeding them to the parser?
> 
> As for the reason, I don't know, that was before my time..
> 

And also before mine


>>> Oleg
>>>
>> DelegatingFieldParser#parse method does not automatically unfold the field
>> body, which actually seems like a bug to me. What is the expected behavior
>> of this method?
> 
>>>From what I see the unfolding happens in
> AbstractField.parse(ByteSequence, String) at line 155. The call
> hierarchy leads to AbstractField.parse(ByteSequence) and
> MessageBuilder.field(Field)..
> 
> I guess you should use AbstractField.parse(ByteSequence).
> 

That still leaves us with DefaultFieldParser and DelegatingFieldParser 
that produce incorrect results, at least in my opinion.

Are there any objections to changing the behavior of these classes?

Oleg

> Markus
> 


Mime
View raw message