Hilko Bengen | 3 Jul 12:09 2008
Picon

Re: character set conversion

Gunnar Ritter <gunnarr <at> acm.org> writes:

> so I've implemented that and not escaping here too. The only
> difference is that I don't do it in interactive mode, as we agreed
> previously.

I still get the old behavior.

$ echo äöü | LANG=C ./mailx hbengen
Invalid or incomplete multibyte or wide character
"/home/hbengen/dead.letter" 1/7
. . . message not sent.

It looks as if you introduced a typo in your changes to sendout.c:

>  <at>  <at>  -1001,10 +1040,21  <at>  <at>  try:	if ((nmtf = infix(hp, mtf, dosign)) == NULL) {
>  			hp->h_charset = NULL;
>  			goto hloop;
>  		}
> -		if (ncs && *ncs && (errno == EILSEQ || errno == EINVAL)) {
> -			rewind(mtf);
> -			wantcharset = ncs;
> -			goto loop;
> +		if (errno == EILSEQ || errno == EINVAL) {
> +			if (ncs && *ncs) {
> +				rewind(mtf);
> +				wantcharset = ncs;
> +				goto loop;
> +			}
> +			if (wantcharset && value("interactive") != NULL) {
(Continue reading)

Gunnar Ritter | 4 Jul 08:17 2008
Picon

Re: character set conversion

Hilko Bengen <bengen@...> wrote:

> Gunnar Ritter <gunnarr@...> writes:
>
> > so I've implemented that and not escaping here too. The only
> > difference is that I don't do it in interactive mode, as we agreed
> > previously.
>
> I still get the old behavior.
>
> $ echo äöü | LANG=C ./mailx hbengen
> Invalid or incomplete multibyte or wide character
> "/home/hbengen/dead.letter" 1/7
> . . . message not sent.
>
> It looks as if you introduced a typo in your changes to sendout.c:

Right. Thank you very much for testing and finding this.
I've fixed it in the CVS repository now.

By the way, for testing, you can send messages directly
to an mbox file on local disk with "mailx ./file", so
you don't need to clutter up your mailbox with these.
Also, LANG is the weakest of the locale variables and
gets overridden by everything else. So to enforce a
locale, LC_ALL=C is better since it is the strongest.

Gunnar

-------------------------------------------------------------------------
(Continue reading)

Hilko Bengen | 4 Jul 08:54 2008
Picon

Re: character set conversion

Gunnar Ritter <gunnarr@...> writes:

>> It looks as if you introduced a typo in your changes to sendout.c:
>
> Right. Thank you very much for testing and finding this.
> I've fixed it in the CVS repository now.

Fine. I'm going to upload an updated package to the Debian repository,
then.

> By the way, for testing, you can send messages directly to an mbox
> file on local disk with "mailx ./file", so you don't need to clutter
> up your mailbox with these. 

Ah, great. I wasn't aware of that.

-Hilko

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
Gunnar Ritter | 29 Jul 00:44 2008
Picon

Heirloom mailx 12.4 released

Hi,

Heirloom mailx 12.4 has been released and is available
from <http://heirloom.sourceforge.net/mailx.html> as
usual. Changes are:

[12.4] released 7/29/08
* With the "-E" command line option or if the "skipemptybody" variable is
  set, outgoing messages that contain no text in their first or only part
  are not sent but silently discarded (contributed by Eygene Ryabinkin).
* Support for SMTP AUTH PLAIN has been added (contributed by Joe Feise).
  Set "smtp-auth=plain" to use it.
* When the text of an outgoing message contains illegal byte sequences
  in non-interactive mode, do not refuse sending it but send it with
  the "application/octet-stream" content type instead (thanks for the
  report and further help to Hilko Bengen).
* When an attachment that would have a "text/something" content type
  contains illegal byte sequences, it is now reliably sent out with the
  "application/octet-stream" content type instead (bugreport by
  Hilko Bengen).
* The values of the "sendcharsets" variable are now separately traversed
  for each part of a multipart message, so that if e.g. the main body can
  be represented in ISO-8859-1 but an attachment text needs UTF-8, these
  two encodings will be chosen, respectively.
* Fixed a bug that caused messages to be truncated with IMAP servers that
  use LF as line ending in message data, such as Google Mail (reported by 
  Matthew L. Shobe).
* Do not run filename expansion for IMAP or POP3 mailboxes names, making
  it possible to select mailboxes that contain both brackets and spaces
  in their names (reported by Matthew L. Shobe).
(Continue reading)


Gmane