Re: [Nmh-workers] mhshow display bug

[Paul Fox]

ken wrote:
 > part       text/plain                 898
 > >OK, I can reproduce it now.  "%%l" obviously works in the common case
 > >but I don't if it should be changed all the time, or not used when
 > >there's a ct->c_termproc, i.e.,
 > >
 > >    if (concatsw  &&  ! ct->c_termproc)
 > >
 > >Or if parse_display_string() show look for the empty buffer when
 > >appending c_termproc?  (Definitely my last choice.)
 > 
 > Okay, I think I understand the problem now.
 > 
 > With -concat, it's assumed that things aren't going to do their own
 > pagination (like I said earlier, the interface here is sort-of changed)
 > So defaulting back to "%lmoreproc" seems wrong if -concat is set but
 > c_termproc is also set.  But how about:
 > 
 > if (concatsw)
 >    if (ct->c_termproc)
 >      buffer = "%lcat"
 >    else
 >      buffer = "%l"
 > else
 >    ...

sure, that works.  more completely:
+       if (concatsw) {
+           if (ct->c_termproc)
+               snprintf(buffer, sizeof(buffer), "%%lcat %%F");
+           else
+               snprintf(buffer, sizeof(buffer), "%%l");
+       } else {

it does subvert the mhshow-charset-<charset> feature somewhat for the
-concat case, in that if someone might still want to invoke an xterm,
or openoffice, to display some rare character set that their normal
pager won't handle readily, they can still do it, but they won't get a
pager by default as they would in the past.  i guess that's okay.  at
a minimum, the xterm example in the man page should change to say: 
        xterm -fn '....' -e "%s | less"

it's probably easier to add a pager where it's needed than to remove
it where it's not, so yes, i've convinced myself you're right.  :-)

paul
----------------------
 paul fox, address@hidden (arlington, ma, where it's 48.7 degrees)