search: allow testing with current xapian.git and 1.5.x

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

diff --git a/lib/PublicInbox/Search.pm b/lib/PublicInbox/Search.pm
index e6200bfb771d5aa8ff8814c4551de63d74138d80..b739faf1a50c7eae1ae781078a0621fa71e648c4 100644 (file)
--- a/lib/PublicInbox/Search.pm
+++ b/lib/PublicInbox/Search.pm
@@ -18,9 +18,9 @@ use constant {
        # added for public-inbox 1.6.0+
        BYTES => 3, # IMAP RFC822.SIZE
        UID => 4, # IMAP UID == NNTP article number == Xapian docid
+       THREADID => 5, # RFC 8474, RFC 8621
 
        # TODO
-       # THREADID => ?
        # REPLYCNT => ?, # IMAP ANSWERED
 
        # SCHEMA_VERSION history
@@ -47,26 +47,40 @@ use constant {
        #      public-inbox v1.5.0 adds (still SCHEMA_VERSION=15):
        #      * "lid:" and "l:" for List-Id searches
        #
-       #      v1.6.0 adds BYTES and UID values
+       #      v1.6.0 adds BYTES, UID and THREADID values
        SCHEMA_VERSION => 15,
 };
 
 use PublicInbox::Smsg;
 use PublicInbox::Over;
 my $QP_FLAGS;
-our %X = map { $_ => 0 } qw(BoolWeight Database Enquire
-                       NumberValueRangeProcessor QueryParser Stem);
+our %X = map { $_ => 0 } qw(BoolWeight Database Enquire QueryParser Stem);
 our $Xap; # 'Search::Xapian' or 'Xapian'
+my $NVRP; # '$Xap::'.('NumberValueRangeProcessor' or 'NumberRangeProcessor')
 my $ENQ_ASCENDING;
 
 sub load_xapian () {
        return 1 if defined $Xap;
-       for my $x (qw(Search::Xapian Xapian)) {
+       # n.b. PI_XAPIAN is intended for development use only.  We still
+       # favor Search::Xapian since that's what's available in current
+       # Debian stable (10.x) and derived distros.
+       for my $x (($ENV{PI_XAPIAN} // 'Search::Xapian'), 'Xapian') {
                eval "require $x";
                next if $@;
 
                $x->import(qw(:standard));
                $Xap = $x;
+
+               # `version_string' was added in Xapian 1.1
+               my $xver = eval('v'.eval($x.'::version_string()')) //
+                               eval('v'.eval($x.'::xapian_version_string()'));
+
+               # NumberRangeProcessor was added in Xapian 1.3.6,
+               # NumberValueRangeProcessor was removed for 1.5.0+,
+               # favor the older /Value/ variant since that's what our
+               # (currently) preferred Search::Xapian supports
+               $NVRP = $x.'::'.($x eq 'Xapian' && $xver ge v1.5 ?
+                       'NumberRangeProcessor' : 'NumberValueRangeProcessor');
                $X{$_} = $Xap.'::'.$_ for (keys %X);
 
                # ENQ_ASCENDING doesn't seem exported by SWIG Xapian.pm,
@@ -219,6 +233,12 @@ sub mdocid {
        int(($docid - 1) / $nshard) + 1;
 }
 
+sub mset_to_artnums {
+       my ($self, $mset) = @_;
+       my $nshard = $self->{nshard} // 1;
+       [ map { mdocid($nshard, $_) } $mset->items ];
+}
+
 sub xdb ($) {
        my ($self) = @_;
        $self->{xdb} ||= do {
@@ -305,6 +325,12 @@ sub _do_enquire {
        retry_reopen($self, \&_enquire_once, [ $self, $query, $opts ]);
 }
 
+# returns true if all docs have the THREADID value
+sub has_threadid ($) {
+       my ($self) = @_;
+       (xdb($self)->get_metadata('has_threadid') // '') eq '1';
+}
+
 sub _enquire_once { # retry_reopen callback
        my ($self, $query, $opts) = @{$_[0]};
        my $xdb = xdb($self);
@@ -320,14 +346,23 @@ sub _enquire_once { # retry_reopen callback
        } else {
                $enquire->set_sort_by_value_then_relevance(TS, $desc);
        }
+
+       # `mairix -t / --threads' or JMAP collapseThreads
+       if ($opts->{thread} && has_threadid($self)) {
+               $enquire->set_collapse_key(THREADID);
+       }
+
        my $offset = $opts->{offset} || 0;
        my $limit = $opts->{limit} || 50;
        my $mset = $enquire->get_mset($offset, $limit);
        return $mset if $opts->{mset};
-       my @msgs = map { PublicInbox::Smsg::from_mitem($_) } $mset->items;
-       return \@msgs unless wantarray;
-
-       ($mset->get_matches_estimated, \@msgs)
+       my $nshard = $self->{nshard} // 1;
+       my $i = 0;
+       my %order = map { mdocid($nshard, $_) => ++$i } $mset->items;
+       my @msgs = sort {
+               $order{$a->{num}} <=> $order{$b->{num}}
+       } @{$self->{over_ro}->get_all(keys %order)};
+       wantarray ? ($mset->get_matches_estimated, \@msgs) : \@msgs;
 }
 
 # read-write
@@ -343,15 +378,18 @@ sub qparse_new ($) {
        $qp->set_database($xdb);
        $qp->set_stemmer(stemmer($self));
        $qp->set_stemming_strategy(STEM_SOME());
-       $qp->set_max_wildcard_expansion(100);
-       my $nvrp = $X{NumberValueRangeProcessor};
-       $qp->add_valuerangeprocessor($nvrp->new(YYYYMMDD, 'd:'));
-       $qp->add_valuerangeprocessor($nvrp->new(DT, 'dt:'));
+       my $cb = $qp->can('set_max_wildcard_expansion') //
+               $qp->can('set_max_expansion'); # Xapian 1.5.0+
+       $cb->($qp, 100);
+       $cb = $qp->can('add_valuerangeprocessor') //
+               $qp->can('add_rangeprocessor'); # Xapian 1.5.0+
+       $cb->($qp, $NVRP->new(YYYYMMDD, 'd:'));
+       $cb->($qp, $NVRP->new(DT, 'dt:'));
 
        # for IMAP, undocumented for WWW and may be split off go away
-       $qp->add_valuerangeprocessor($nvrp->new(BYTES, 'bytes:'));
-       $qp->add_valuerangeprocessor($nvrp->new(TS, 'ts:'));
-       $qp->add_valuerangeprocessor($nvrp->new(UID, 'uid:'));
+       $cb->($qp, $NVRP->new(BYTES, 'bytes:'));
+       $cb->($qp, $NVRP->new(TS, 'ts:'));
+       $cb->($qp, $NVRP->new(UID, 'uid:'));
 
        while (my ($name, $prefix) = each %bool_pfx_external) {
                $qp->add_boolean_prefix($name, $_) foreach split(/ /, $prefix);