The main hurdle seems to be the design of the javax.mail headers
structure but I don't think it is insurmountable...
All the setters / getters need to call one common setHeader() /
getHeader() method which does the encoding / unencoding and folding /
unfolding. We then need setRawHeader() / getRawHeader() methods to set /
get raw data for higher level "structured header fields" as per RFC 822.
It would be up to the application to provide specific lexical parsing of
these headers. I've personally not ever come across them, but the RFC
gives "address" as an example.
I'll have a go and let you know how things pan out.
On 24/06/2010 15:27, James Carman wrote:
Well, if we can make things better (more correct), then we should try.
Do you already have code that does this stuff? If so, create a JIRA
with a patch and we can look into including it. Contributions are
*always* welcome, and I am not trying to say that in a snotty way. It
really does sound like you're onto something we can do to add value
here (if we're not already). If you need help creating the patch,
just let me know.
On Thu, Jun 24, 2010 at 8:30 AM, Clive Haworth
<[email protected]> wrote:
Thanks for the pointer.
I grabbed the source to MimeMessage.java and had a look. It seems that the
"standard" headers are US-ASCII encoded automatically when you use the
specific setter. For example setRecipient(), setReplyTo() etc.. However if
you use the generic setHeader() you have to do this yourself. The other
thing is that none of the methods do the automatic folding for long lines.
Too complicated by half, if you ask me. This level of detail should be
handled by javax.mail and applications should not have to worry about it.
Worse is the javax.mail documentation: use the source Luke, as they say.
On 23/06/2010 14:22, James Carman wrote:
Commons Email is a "wrapper" around the JavaMail API. I would imagine
that the JavaMail provider(s) would be responsible for this.
On Wed, Jun 23, 2010 at 9:19 AM, Clive Haworth
<[email protected]> wrote:
I need some clarification on what is and isn't done to mail headers by
commons email, both setting and getting.
RFC 822 forces all email headers to use characters in the US-ASCII range
1
through 127. Headers not natively in US-ASCII need to be encoded using
RFC
2047.
Additionally RFC 822 forces header lines longer than 998 characters to be
folded.
Does commons email handle all this automatically in ALL cases, some cases
(and if so which) or not at all?
Regards,
Clive Haworth
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
