Rocco Rutte [Mon, 15 Jun 2009 14:28:03 +0000 (16:28 +0200)]
pager: intermediate fix for wrapping long header lines with $smart_wrap set
The header folding algorithm outputs a word without spaces and longer
than $wrap as-is. The pager however tries to break it. With $smart_wrap
unset, it simply breaks at $wrap regardless if there's a space or
not. With $smart_wrap set it tried to find a space and break if the next
word is too long. This logic doesn't work for folded header lines which
always start with space. Thus, the output would always contain the
folding whitespace on a line by itself detaching the header value from
the header name.
As an intermediate fix we don't try to be smart if the line begins with
space or tab.
Rocco Rutte [Fri, 12 Jun 2009 23:08:01 +0000 (01:08 +0200)]
Rewrite header folding
We now distinct between sending and display case. For display,
we always use tabs for folding for readability; for sending we
now correctly fold using whitespace found in the header.
Rocco Rutte [Wed, 3 Jun 2009 20:35:32 +0000 (22:35 +0200)]
Fix several f=f bugs
It fixes problems with space-handling in general (trailing spaces for
display, un-space-stuffing), fixes quote prefix for replies and lowers
FLOWED_MAX so we don't run into line length trouble too early.
Rocco Rutte [Mon, 1 Jun 2009 08:36:50 +0000 (10:36 +0200)]
Treat address groups as no recipients
When sending with Sendmail or SMTP we exclude address groups
anyway, so treat these addresses as not present when checking
for valid recipients before sending in the compose menu.
Rocco Rutte [Sun, 31 May 2009 12:46:40 +0000 (14:46 +0200)]
SMTP: Don't send "undisclosed-recipients" in RCPT TO
This happened for Bcc only messages. "undisclosed-recipients" is added
in To: as guard against MTAs that leak BCCs if no To:/Cc: header is
given. See Debian #529090.
Rocco Rutte [Sun, 31 May 2009 10:54:19 +0000 (12:54 +0200)]
Turn trailing \r\n to \n for qp-encoded messages.
RFC2045 (sect. 6.7, (1) general 8bit representation)
states that neither CR nor LF of the trailing CRLF may
be qp-encoded. So we ignore trailing qp-encoded CRs.
Rocco Rutte [Sun, 31 May 2009 09:32:06 +0000 (11:32 +0200)]
f=f: Strip trailing spaces for interoperability. Closes #3001.
If we make a fixed reply to a flowed message we remove trailing
spaces and can now safely allow $indent_string to be used which
is what users expect.
Second, if we make a flowed reply to a fixed message we also
strip trailing spaces since from format=fixed we assume all
lines are fixed (i.e. we don't want to errorneously make fixed
lines flowed).
Rocco Rutte [Sun, 31 May 2009 08:40:23 +0000 (10:40 +0200)]
Add $honor_disposition to not display attachments inline. Closes #3170.
Previously all MIME parts were displayed inline regardless of the
Content-Disposition header (if present). With $honor_disposition
set, only inline parts are displayed inline, attachments can only
be viewed from the attachments menu.
Rocco Rutte [Fri, 29 May 2009 19:24:29 +0000 (21:24 +0200)]
Add flags to mutt_read_line() for EOL-stripping and continuation support
We use these to stop supporting EOL-escaping with \ which was
wrong in most cases (e.g. $history_file), support is kept for
reading config and mailcap files.
Rocco Rutte [Fri, 29 May 2009 14:21:15 +0000 (16:21 +0200)]
Fix pager search prompt for reverse search
When doing a reverse search for the first time after switching to the
pager, SearchBack is always 0 and thus gave the wrong prompt, change to
using op directly. While I'm at it, unify search prompts.
Rocco Rutte [Thu, 28 May 2009 14:37:44 +0000 (16:37 +0200)]
SMTP: don't interactively authenticate without a password in batch/mailx mode
Otherwise, if we don't have a password and curses is not running, the
SASL authentication callback will crash mutt. We now abort if the
password wasn't given in $smtp_url.
Brendan Cully [Thu, 28 May 2009 06:14:48 +0000 (23:14 -0700)]
gnutls: do not ignore certificate problems.
Specifically, do not allow revoked certificates or those signed by non-CAs
unless they have been explicitly stored in the trusted certificate file.
Thanks to Miroslav Lichvar.
Brendan Cully [Tue, 26 May 2009 00:31:27 +0000 (17:31 -0700)]
Fix a serious oversight validating TLS certificates.
If any certificate in a chain presented by a server was accepted, the
connection was allowed without verifying that the presented
certificate was actually signed by the certificate in the chain.
Rocco Rutte [Wed, 13 May 2009 10:21:43 +0000 (12:21 +0200)]
Actually reopen mbox/mmdf files in case we find modifications. Closes #2725.
Since the file is changed while we have it open in ctx->fp, we likely
get wrong information when parsing the mailbox. Now we explicitely close
and (re)open it.
Fix maildir times in mailboxes browser. Closes #626.
Instead of using the (pointless) mtime of the top-level maildir folder,
use the latest mtime of the "new" and "cur" subdirectories. Maildir
folders in the mailboxes list can now be properly sorted by date. This
does not affect the directory browser. Also see #2421.