diff options
Diffstat (limited to 'lib/PublicInbox')
-rw-r--r-- | lib/PublicInbox/MID.pm | 27 | ||||
-rw-r--r-- | lib/PublicInbox/OverIdx.pm | 4 | ||||
-rw-r--r-- | lib/PublicInbox/SearchIdx.pm | 4 |
3 files changed, 25 insertions, 10 deletions
diff --git a/lib/PublicInbox/MID.pm b/lib/PublicInbox/MID.pm index 14089f91..d7a42c38 100644 --- a/lib/PublicInbox/MID.pm +++ b/lib/PublicInbox/MID.pm @@ -7,7 +7,7 @@ use strict; use warnings; use base qw/Exporter/; our @EXPORT_OK = qw/mid_clean id_compress mid2path mid_mime mid_escape MID_ESC - mids references/; + mids references mids_for_index/; use URI::Escape qw(uri_escape_utf8); use Digest::SHA qw/sha1_hex/; require PublicInbox::Address; @@ -54,11 +54,10 @@ sub mid2path { # Only for v1 code paths: sub mid_mime ($) { mids($_[0]->header_obj)->[0] } -sub mids ($) { - my ($hdr) = @_; +# only intended for Message-ID and X-Alt-Message-ID +sub extract_mids { my @mids; - my @v = $hdr->header_raw('Message-Id'); - foreach my $v (@v) { + for my $v (@_) { my @cur = ($v =~ /<([^>]+)>/sg); if (@cur) { push(@mids, @cur); @@ -66,7 +65,23 @@ sub mids ($) { push(@mids, $v); } } - uniq_mids(\@mids); + \@mids; +} + +sub mids ($) { + my ($hdr) = @_; + my @mids = $hdr->header_raw('Message-Id'); + uniq_mids(extract_mids(@mids)); +} + +# we allow searching on X-Alt-Message-ID since PublicInbox::NNTP uses them +# to placate some clients, and we want to ensure NNTP-only clients can +# import and index without relying on HTTP endpoints +sub mids_for_index ($) { + my ($hdr) = @_; + my @mids = $hdr->header_raw('Message-Id'); + my @alts = $hdr->header_raw('X-Alt-Message-ID'); + uniq_mids(extract_mids(@mids, @alts)); } # last References should be IRT, but some mail clients do things diff --git a/lib/PublicInbox/OverIdx.pm b/lib/PublicInbox/OverIdx.pm index 01ca6f11..189bd21d 100644 --- a/lib/PublicInbox/OverIdx.pm +++ b/lib/PublicInbox/OverIdx.pm @@ -13,7 +13,7 @@ use warnings; use base qw(PublicInbox::Over); use IO::Handle; use DBI qw(:sql_types); # SQL_BLOB -use PublicInbox::MID qw/id_compress mids references/; +use PublicInbox::MID qw/id_compress mids_for_index references/; use PublicInbox::SearchMsg qw(subject_normalized); use Compress::Zlib qw(compress); use PublicInbox::Search; @@ -256,7 +256,7 @@ sub add_overview { lines => $lines, blob => $oid, }, 'PublicInbox::SearchMsg'; - my $mids = mids($mime->header_obj); + my $mids = mids_for_index($mime->header_obj); my $refs = parse_references($smsg, $mid0, $mids); my $subj = $smsg->subject; my $xpath; diff --git a/lib/PublicInbox/SearchIdx.pm b/lib/PublicInbox/SearchIdx.pm index aed3875a..b2d71a1f 100644 --- a/lib/PublicInbox/SearchIdx.pm +++ b/lib/PublicInbox/SearchIdx.pm @@ -12,7 +12,7 @@ use warnings; use base qw(PublicInbox::Search PublicInbox::Lock); use PublicInbox::MIME; use PublicInbox::InboxWritable; -use PublicInbox::MID qw/mid_clean id_compress mid_mime mids/; +use PublicInbox::MID qw/mid_clean id_compress mid_mime mids_for_index/; use PublicInbox::MsgIter; use Carp qw(croak); use POSIX qw(strftime); @@ -344,7 +344,7 @@ sub add_xapian ($$$$$) { sub add_message { # mime = Email::MIME object my ($self, $mime, $bytes, $num, $oid, $mid0) = @_; - my $mids = mids($mime->header_obj); + my $mids = mids_for_index($mime->header_obj); $mid0 = $mids->[0] unless defined $mid0; # v1 compatibility unless (defined $num) { # v1 $self->_msgmap_init; |