imap+nntp: share COMPRESS implementation

[public-inbox.git] / lib / PublicInbox / MsgIter.pm

diff --git a/lib/PublicInbox/MsgIter.pm b/lib/PublicInbox/MsgIter.pm
index bb1dfeadc291857f993b68dd297da46de3a9728c..dd28417b70c043b092f231147d82580ba94e31cf 100644 (file)
--- a/lib/PublicInbox/MsgIter.pm
+++ b/lib/PublicInbox/MsgIter.pm
@@ -1,4 +1,4 @@
-# Copyright (C) 2016-2020 all contributors <meta@public-inbox.org>
+# Copyright (C) 2016-2021 all contributors <meta@public-inbox.org>
 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 
 # read-only utilities for Email::MIME
@@ -84,18 +84,30 @@ sub msg_part_text ($$) {
                # If forcing charset=UTF-8 failed,
                # caller will warn further down...
                $s = $part->body if $@;
+       } elsif ($err && $ct =~ m!\bapplication/octet-stream\b!i) {
+               # Some unconfigured/poorly-configured MUAs will set
+               # application/octet-stream even for all text attachments.
+               # Try to see if it's printable text that we can index
+               # and display:
+               $s = $part->body;
+               utf8::decode($s);
+               undef($s =~ /[^\p{XPosixPrint}\s]/s ? $s : $err);
        }
        ($s, $err);
 }
 
 # returns an array of quoted or unquoted sections
 sub split_quotes {
+       # some editors don't put trailing newlines at the end,
+       # make sure split_quotes can work:
+       $_[0] .= "\n" if substr($_[0], -1) ne "\n";
+
        # Quiet "Complex regular subexpression recursion limit" warning
        # in case an inconsiderate sender quotes 32K of text at once.
        # The warning from Perl is harmless for us since our callers can
        # tolerate less-than-ideal matches which work within Perl limits.
        no warnings 'regexp';
-       split(/((?:^>[^\n]*\n)+)/sm, shift);
+       split(/((?:^>[^\n]*\n)+)/sm, $_[0]);
 }
 
 1;