code review 4590047: mail: decode RFC 2047-encoded words, not phrases.

mail: decode RFC 2047-encoded words, not phrases.

[dsymonds, 2011-06-09 11:47:17]

Hello rsc (cc: golang-dev@googlegroups.com),

I'd like you to review this change to
https://go.googlecode.com/hg/

[r, 2011-06-09 12:05:10]

http://codereview.appspot.com/4590047/diff/5001/src/pkg/mail/message.go
File src/pkg/mail/message.go (right):

http://codereview.appspot.com/4590047/diff/5001/src/pkg/mail/message.go#newco...
src/pkg/mail/message.go:334: // RFC 2047 encoded-word starts with =?, ends with
?=, and has two other ?s.
hyphen is in the wrong place.
RFC 2047-encoded word ....

[dsymonds, 2011-06-09 12:38:45]

On Jun 9, 2011 10:05 PM, <r@golang.org> wrote:

> http://codereview.appspot.com/4590047/diff/5001/src/pkg/mail/message.go
> File src/pkg/mail/message.go (right):
>
>
http://codereview.appspot.com/4590047/diff/5001/src/pkg/mail/message.go#newco...
> src/pkg/mail/message.go:334: // RFC 2047 encoded-word starts with =?,
>
> ends with ?=, and has two other ?s.
> hyphen is in the wrong place.
> RFC 2047-encoded word ....

It's in the right place. It is a token in the RFC 2047 grammar called
"encoded-word".

[rsc, 2011-06-09 14:10:29]

LGTM

Drop the email address (put a fake one in)

[bradfitz, 2011-06-09 15:22:06]

We need to come up with a better charset conversion story.

At least the xml package provides a hook that you can use to plug in
http://code.google.com/p/go-charset/ but that's somewhat ad-hoc.  I agree
it'd be nicer to have a more universal solution.


On Thu, Jun 9, 2011 at 8:18 AM, Martin Capitanio <m@capitanio.org> wrote:

> Is there a plan for other character encodings? Would there
> be at some point a tables in the utf-8 package or just some hooks
> for external project(s)?
>
> Martin
>

[rsc, 2011-06-09 15:41:16]

On Thu, Jun 9, 2011 at 11:18, Martin Capitanio <m@capitanio.org> wrote:
> Is there a plan for other character encodings? Would there
> be at some point a tables in the utf-8 package or just some hooks
> for external project(s)?

My plan was to wait for all the other character sets to die.

Russ

[dsymonds, 2011-06-09 22:23:01]

On Fri, Jun 10, 2011 at 12:10 AM, Russ Cox <rsc@golang.org> wrote:

> Drop the email address (put a fake one in)

I'd prefer to leave it in. It's a verbatim copy from the RFC.

[rsc, 2011-06-09 22:25:01]

On Thu, Jun 9, 2011 at 18:22, David Symonds <dsymonds@golang.org> wrote:
> On Fri, Jun 10, 2011 at 12:10 AM, Russ Cox <rsc@golang.org> wrote:
>
>> Drop the email address (put a fake one in)
>
> I'd prefer to leave it in. It's a verbatim copy from the RFC.

oh, okay.  that's fine.

[dsymonds, 2011-06-09 22:46:25]

On Fri, Jun 10, 2011 at 1:51 AM, Martin Capitanio <m@capitanio.org> wrote:

> In the mean time perhaps a small most-of-the-crap2utf-8 decoder would
> be great. People in the rest of the world could at least decipher
> their incoming messages.

I think ISO-8859-1 and UTF-8 should cover the vast majority, which is
what we've got now. Are there other encodings you know of that are
widely used by email programs?


Dave.

[dsymonds, 2011-06-09 22:47:34]

*** Submitted as http://code.google.com/p/go/source/detail?r=b7cebe5f19c7 ***

mail: decode RFC 2047-encoded words, not phrases.

R=rsc, r, bradfitz
CC=golang-dev
http://codereview.appspot.com/4590047

[rsc, 2011-06-09 22:49:06]

> I think ISO-8859-1 and UTF-8 should cover the vast majority, which is
> what we've got now. Are there other encodings you know of that are
> widely used by email programs?

It depends who is sending you email.
I've seen a lot of the other 8859 code pages
from various places in Europe, depending on
what version of Windows is being used.
But I think this is fine for now.  As Brad says,
we can worry about character sets in general
at some other point.

Russ