get rid of unnecessary bytes::length usage

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

diff --git a/lib/PublicInbox/OverIdx.pm b/lib/PublicInbox/OverIdx.pm
index 4a39bf5346285fc9dd3c464545cd48e3b6db02c6..e0893337e7843461740ad607ecff7aa5e7dd9b6f 100644 (file)
--- a/lib/PublicInbox/OverIdx.pm
+++ b/lib/PublicInbox/OverIdx.pm
@@ -1,4 +1,4 @@
-# Copyright (C) 2018-2020 all contributors <meta@public-inbox.org>
+# Copyright (C) 2018-2021 all contributors <meta@public-inbox.org>
 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 
 # for XOVER, OVER in NNTP, and feeds/homepage/threads in PSGI
@@ -243,26 +243,6 @@ sub link_refs {
        $tid;
 }
 
-sub parse_references ($$$) {
-       my ($smsg, $hdr, $mids) = @_;
-       my $refs = references($hdr);
-       push(@$refs, @$mids) if scalar(@$mids) > 1;
-       return $refs if scalar(@$refs) == 0;
-
-       # prevent circular references here:
-       my %seen = ( $smsg->{mid} => 1 );
-       my @keep;
-       foreach my $ref (@$refs) {
-               if (length($ref) > PublicInbox::MID::MAX_MID_SIZE) {
-                       warn "References: <$ref> too long, ignoring\n";
-                       next;
-               }
-               push(@keep, $ref) unless $seen{$ref}++;
-       }
-       $smsg->{references} = '<'.join('> <', @keep).'>' if @keep;
-       \@keep;
-}
-
 # normalize subjects so they are suitable as pathnames for URLs
 # XXX: consider for removal
 sub subject_path ($) {
@@ -283,7 +263,11 @@ sub add_overview {
        my ($self, $eml, $smsg) = @_;
        $smsg->{lines} = $eml->body_raw =~ tr!\n!\n!;
        my $mids = mids_for_index($eml);
-       my $refs = parse_references($smsg, $eml, $mids);
+       my $refs = $smsg->parse_references($eml, $mids);
+       $mids->[0] //= do {
+               $smsg->{mid} //= '';
+               $eml->{-lei_fake_mid};
+       };
        my $subj = $smsg->{subject};
        my $xpath;
        if ($subj ne '') {
@@ -473,10 +457,14 @@ sub dbh_close {
 
 sub create {
        my ($self) = @_;
-       unless (-r $self->{filename}) {
+       my $fn = $self->{filename} // do {
+               croak('BUG: no {filename}') unless $self->{dbh};
+               return;
+       };
+       unless (-r $fn) {
                require File::Path;
-               require File::Basename;
-               File::Path::mkpath(File::Basename::dirname($self->{filename}));
+               my ($dir) = ($fn =~ m!(.*?/)[^/]+\z!);
+               File::Path::mkpath($dir);
        }
        # create the DB:
        PublicInbox::Over::dbh($self);
@@ -531,7 +519,7 @@ sub eidx_prep ($) {
        my ($self) = @_;
        $self->{-eidx_prep} //= do {
                my $dbh = $self->dbh;
-               $dbh->do(<<"");
+               $dbh->do(<<'');
 INSERT OR IGNORE INTO counter (key) VALUES ('eidx_docid')
 
                $dbh->do(<<'');
@@ -568,11 +556,9 @@ CREATE TABLE IF NOT EXISTS eidx_meta (
                # Currently used for "-extindex --reindex" for Xapian
                # data, but may be used in more places down the line.
                $dbh->do(<<'');
-CREATE TABLE IF NOT EXISTS eidxq (
-       docid INTEGER PRIMARY KEY NOT NULL
-)
+CREATE TABLE IF NOT EXISTS eidxq (docid INTEGER PRIMARY KEY NOT NULL)
 
-               $dbh;
+               1;
        };
 }
 
@@ -670,6 +656,26 @@ UPDATE over SET ddd = ? WHERE num = ?
        $sth->execute;
 }
 
+sub merge_xref3 { # used for "-extindex --dedupe"
+       my ($self, $keep_docid, $drop_docid, $oidhex) = @_;
+       my $oidbin = pack('H*', $oidhex);
+       my $sth = $self->{dbh}->prepare_cached(<<'');
+UPDATE OR IGNORE xref3 SET docid = ? WHERE docid = ? AND oidbin = ?
+
+       $sth->bind_param(1, $keep_docid);
+       $sth->bind_param(2, $drop_docid);
+       $sth->bind_param(3, $oidbin, SQL_BLOB);
+       $sth->execute;
+
+       # drop anything that conflicted
+       $sth = $self->{dbh}->prepare_cached(<<'');
+DELETE FROM xref3 WHERE docid = ? AND oidbin = ?
+
+       $sth->bind_param(1, $drop_docid);
+       $sth->bind_param(2, $oidbin, SQL_BLOB);
+       $sth->execute;
+}
+
 sub eidxq_add {
        my ($self, $docid) = @_;
        $self->dbh->prepare_cached(<<'')->execute($docid);
@@ -684,4 +690,22 @@ DELETE FROM eidxq WHERE docid = ?
 
 }
 
+# returns true if we're vivifying a message for lei/store that was
+# previously external-metadata only
+sub vivify_xvmd {
+       my ($self, $smsg) = @_;
+       my @docids = $self->blob_exists($smsg->{blob});
+       my @vivify_xvmd;
+       for my $id (@docids) {
+               if (my $cur = $self->get_art($id)) {
+                       # already indexed if bytes > 0
+                       return if $cur->{bytes} > 0;
+                       push @vivify_xvmd, $id;
+               } else {
+                       warn "W: $smsg->{blob} #$id gone (bug?)\n";
+               }
+       }
+       $smsg->{-vivify_xvmd} = \@vivify_xvmd;
+}
+
 1;