-# Copyright (C) 2015-2020 all contributors <meta@public-inbox.org>
+# Copyright (C) 2015-2021 all contributors <meta@public-inbox.org>
# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
# based on notmuch, but with no concept of folders, files or flags
#
# Read-only search interface for use by the web and NNTP interfaces
package PublicInbox::Search;
use strict;
-
-# values for searching
+use v5.10.1;
+use parent qw(Exporter);
+our @EXPORT_OK = qw(retry_reopen int_val get_pct xap_terms);
+use List::Util qw(max);
+use POSIX qw(strftime);
+use Carp ();
+
+# values for searching, changing the numeric value breaks
+# compatibility with old indices (so don't change them it)
use constant {
- TS => 0, # Received: header in Unix time (IMAP INTERNALDATE)
- YYYYMMDD => 1, # Date: header for searching in the WWW UI
- DT => 2, # Date: YYYYMMDDHHMMSS
+ TS => 0, # Received: in Unix time (IMAP INTERNALDATE, JMAP receivedAt)
+ YYYYMMDD => 1, # redundant with DT below
+ DT => 2, # Date: YYYYMMDDHHMMSS (IMAP SENT*, JMAP sentAt)
+
+ # 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
- # REPLYCNT => 4, # IMAP ANSWERED
+ # REPLYCNT => ?, # IMAP ANSWERED
+
+ # SCHEMA_VERSION history
+ # 0 - initial
+ # 1 - subject_path is lower-cased
+ # 2 - subject_path is id_compress in the index, only
+ # 3 - message-ID is compressed if it includes '%' (hack!)
+ # 4 - change "Re: " normalization, avoid circular Reference ghosts
+ # 5 - subject_path drops trailing '.'
+ # 6 - preserve References: order in document data
+ # 7 - remove references and inreplyto terms
+ # 8 - remove redundant/unneeded document data
+ # 9 - disable Message-ID compression (SHA-1)
+ # 10 - optimize doc for NNTP overviews
+ # 11 - merge threads when vivifying ghosts
+ # 12 - change YYYYMMDD value column to numeric
+ # 13 - fix threading for empty References/In-Reply-To
+ # (commit 83425ef12e4b65cdcecd11ddcb38175d4a91d5a0)
+ # 14 - fix ghost root vivification
+ # 15 - see public-inbox-v2-format(5)
+ # further bumps likely unnecessary, we'll suggest in-place
+ # "--reindex" use for further fixes and tweaks:
+ #
+ # public-inbox v1.5.0 adds (still SCHEMA_VERSION=15):
+ # * "lid:" and "l:" for List-Id searches
+ #
+ # 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 $QP_FLAGS;
+our %X = map { $_ => 0 } qw(BoolWeight Database Enquire QueryParser Stem Query);
our $Xap; # 'Search::Xapian' or 'Xapian'
-my $ENQ_ASCENDING;
+our $NVRP; # '$Xap::'.('NumberValueRangeProcessor' or 'NumberRangeProcessor')
+
+# ENQ_DESCENDING and ENQ_ASCENDING weren't in SWIG Xapian.pm prior to 1.4.16,
+# let's hope the ABI is stable
+our $ENQ_DESCENDING = 0;
+our $ENQ_ASCENDING = 1;
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;
- $X{$_} = $Xap.'::'.$_ for (keys %X);
- # ENQ_ASCENDING doesn't seem exported by SWIG Xapian.pm,
- # so lets hope this part of the ABI is stable because it's
- # just an integer:
- $ENQ_ASCENDING = $x eq 'Xapian' ?
- 1 : Search::Xapian::ENQ_ASCENDING();
+ # `version_string' was added in Xapian 1.1
+ my $xver = eval('v'.eval($x.'::version_string()')) //
+ eval('v'.eval($x.'::xapian_version_string()'));
- # for Smsg:
- *PublicInbox::Smsg::sortable_unserialise =
- $Xap.'::sortable_unserialise';
+ # 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);
+
+ *sortable_serialise = $x.'::sortable_serialise';
+ *sortable_unserialise = $x.'::sortable_unserialise';
# n.b. FLAG_PURE_NOT is expensive not suitable for a public
# website as it could become a denial-of-service vector
# FLAG_PHRASE also seems to cause performance problems chert
# a prefix common in patch emails
our $LANG = 'english';
-use constant {
- # SCHEMA_VERSION history
- # 0 - initial
- # 1 - subject_path is lower-cased
- # 2 - subject_path is id_compress in the index, only
- # 3 - message-ID is compressed if it includes '%' (hack!)
- # 4 - change "Re: " normalization, avoid circular Reference ghosts
- # 5 - subject_path drops trailing '.'
- # 6 - preserve References: order in document data
- # 7 - remove references and inreplyto terms
- # 8 - remove redundant/unneeded document data
- # 9 - disable Message-ID compression (SHA-1)
- # 10 - optimize doc for NNTP overviews
- # 11 - merge threads when vivifying ghosts
- # 12 - change YYYYMMDD value column to numeric
- # 13 - fix threading for empty References/In-Reply-To
- # (commit 83425ef12e4b65cdcecd11ddcb38175d4a91d5a0)
- # 14 - fix ghost root vivification
- # 15 - see public-inbox-v2-format(5)
- # further bumps likely unnecessary, we'll suggest in-place
- # "--reindex" use for further fixes and tweaks
- #
- # public-inbox v1.5.0 adds (still SCHEMA_VERSION=15):
- # * "lid:" and "l:" for List-Id searches
- SCHEMA_VERSION => 15,
-};
-
# note: the non-X term prefix allocations are shared with
# Xapian omega, see xapian-applications/omega/docs/termprefixes.rst
my %bool_pfx_external = (
our @HELP = (
's:' => 'match within Subject e.g. s:"a quick brown fox"',
'd:' => <<EOF,
-date range as YYYYMMDD e.g. d:19931002..20101002
-Open-ended ranges such as d:19931002.. and d:..20101002
-are also supported
-EOF
- 'dt:' => <<EOF,
-date-time range as YYYYMMDDhhmmss (e.g. dt:19931002011000..19931002011200)
+match date-time range, git "approxidate" formats supported
+Open-ended ranges such as `d:last.week..' and
+`d:..2.days.ago' are supported
EOF
'b:' => 'match within message body, including text attachments',
'nq:' => 'match non-quoted text within message body',
'dfpre:' => 'match pre-image git blob ID',
'dfpost:' => 'match post-image git blob ID',
'dfblob:' => 'match either pre or post-image git blob ID',
+ 'rt:' => <<EOF,
+match received time, like `d:' if sender's clock was correct
+EOF
);
chomp @HELP;
sub xdir ($;$) {
my ($self, $rdonly) = @_;
- if ($self->{ibx_ver} == 1) {
- "$self->{inboxdir}/public-inbox/xapian" . SCHEMA_VERSION;
- } else {
- my $dir = "$self->{inboxdir}/xap" . SCHEMA_VERSION;
- return $dir if $rdonly;
-
- my $shard = $self->{shard};
- defined $shard or die "shard not given";
- $dir .= "/$shard";
+ if ($rdonly || !defined($self->{shard})) {
+ $self->{xpfx};
+ } else { # v2 + extindex only:
+ "$self->{xpfx}/$self->{shard}";
}
}
-sub _xdb ($) {
+# returns all shards as separate Xapian::Database objects w/o combining
+sub xdb_shards_flat ($) {
my ($self) = @_;
- my $dir = xdir($self, 1);
- my ($xdb, $slow_phrase);
- my $qpf = \($self->{qp_flags} ||= $QP_FLAGS);
- if ($self->{ibx_ver} >= 2) {
- foreach my $shard (<$dir/*>) {
- -d $shard && $shard =~ m!/[0-9]+\z! or next;
- my $sub = $X{Database}->new($shard);
- if ($xdb) {
- $xdb->add_database($sub);
- } else {
- $xdb = $sub;
- }
- $slow_phrase ||= -f "$shard/iamchert";
- }
+ my $xpfx = $self->{xpfx};
+ my (@xdb, $slow_phrase);
+ load_xapian();
+ $self->{qp_flags} //= $QP_FLAGS;
+ if ($xpfx =~ m!/xapian[0-9]+\z!) {
+ @xdb = ($X{Database}->new($xpfx));
+ $self->{qp_flags} |= FLAG_PHRASE() if !-f "$xpfx/iamchert";
} else {
- $slow_phrase = -f "$dir/iamchert";
- $xdb = $X{Database}->new($dir);
+ opendir(my $dh, $xpfx) or return (); # not initialized yet
+ # We need numeric sorting so shard[0] is first for reading
+ # Xapian metadata, if needed
+ my $last = max(grep(/\A[0-9]+\z/, readdir($dh))) // return ();
+ for (0..$last) {
+ my $shard_dir = "$self->{xpfx}/$_";
+ push @xdb, $X{Database}->new($shard_dir);
+ $slow_phrase ||= -f "$shard_dir/iamchert";
+ }
+ $self->{qp_flags} |= FLAG_PHRASE() if !$slow_phrase;
}
- $$qpf |= FLAG_PHRASE() unless $slow_phrase;
- $xdb;
+ @xdb;
+}
+
+# v2 Xapian docids don't conflict, so they're identical to
+# NNTP article numbers and IMAP UIDs.
+# https://trac.xapian.org/wiki/FAQ/MultiDatabaseDocumentID
+sub mdocid {
+ my ($nshard, $mitem) = @_;
+ my $docid = $mitem->get_docid;
+ int(($docid - 1) / $nshard) + 1;
+}
+
+sub mset_to_artnums {
+ my ($self, $mset) = @_;
+ my $nshard = $self->{nshard};
+ [ map { mdocid($nshard, $_) } $mset->items ];
}
sub xdb ($) {
my ($self) = @_;
- $self->{xdb} ||= do {
- load_xapian();
- _xdb($self);
+ $self->{xdb} // do {
+ my @xdb = $self->xdb_shards_flat or return;
+ $self->{nshard} = scalar(@xdb);
+ my $xdb = shift @xdb;
+ $xdb->add_database($_) for @xdb;
+ $self->{xdb} = $xdb;
};
}
sub new {
my ($class, $ibx) = @_;
ref $ibx or die "BUG: expected PublicInbox::Inbox object: $ibx";
- my $self = bless {
- inboxdir => $ibx->{inboxdir},
- altid => $ibx->{altid},
- ibx_ver => $ibx->version,
- }, $class;
- my $dir = xdir($self, 1);
- $self->{over_ro} = PublicInbox::Over->new("$dir/over.sqlite3");
+ my $xap = $ibx->version > 1 ? 'xap' : 'public-inbox/xapian';
+ my $xpfx = "$ibx->{inboxdir}/$xap".SCHEMA_VERSION;
+ my $self = bless { xpfx => $xpfx }, $class;
+ $self->{altid} = $ibx->{altid} if defined($ibx->{altid});
$self;
}
$self; # make chaining easier
}
+# Convert git "approxidate" ranges to something usable with our
+# Xapian indices. At the moment, Xapian only offers a C++-only API
+# and neither the SWIG nor XS bindings allow us to use custom code
+# to parse dates (and libgit2 doesn't expose git__date_parse, either,
+# so we're running git-rev-parse(1)).
+# This replaces things we need to send to $git->date_parse with
+# "\0".$strftime_format.['+'|$idx]."\0" placeholders
+sub date_parse_prepare {
+ my ($to_parse, $pfx, $range) = @_;
+ # are we inside a parenthesized statement?
+ my $end = $range =~ s/([\)\s]*)\z// ? $1 : '';
+ my @r = split(/\.\./, $range, 2);
+
+ # expand "d:20101002" => "d:20101002..20101003" and like
+ # n.b. git doesn't do YYYYMMDD w/o '-', it needs YYYY-MM-DD
+ # We upgrade "d:" to "dt:" to iff using approxidate
+ if ($pfx eq 'd') {
+ my $fmt = "\0%Y%m%d";
+ if (!defined($r[1])) {
+ if ($r[0] =~ /\A([0-9]{4})([0-9]{2})([0-9]{2})\z/) {
+ push @$to_parse, "$1-$2-$3";
+ # we could've handled as-is, but we need
+ # to parse anyways for "d+" below
+ } else {
+ push @$to_parse, $r[0];
+ if ($r[0] !~ /\A[0-9]{4}-[0-9]{2}-[0-9]{2}\z/) {
+ $pfx = 'dt';
+ $fmt = "\0%Y%m%d%H%M%S";
+ }
+ }
+ $r[0] = "$fmt+$#$to_parse\0";
+ $r[1] = "$fmt+\0";
+ } else {
+ for my $x (@r) {
+ next if $x eq '' || $x =~ /\A[0-9]{8}\z/;
+ push @$to_parse, $x;
+ if ($x !~ /\A[0-9]{4}-[0-9]{2}-[0-9]{2}\z/) {
+ $pfx = 'dt';
+ }
+ $x = "$fmt$#$to_parse\0";
+ }
+ if ($pfx eq 'dt') {
+ for (@r) {
+ s/\0%Y%m%d/\0%Y%m%d%H%M%S/;
+ s/\A([0-9]{8})\z/${1}000000/;
+ }
+ }
+ }
+ } elsif ($pfx eq 'dt') {
+ if (!defined($r[1])) { # git needs gaps and not /\d{14}/
+ if ($r[0] =~ /\A([0-9]{4})([0-9]{2})([0-9]{2})
+ ([0-9]{2})([0-9]{2})([0-9]{2})\z/x) {
+ push @$to_parse, "$1-$2-$3 $4:$5:$6";
+ } else {
+ push @$to_parse, $r[0];
+ }
+ $r[0] = "\0%Y%m%d%H%M%S$#$to_parse\0";
+ $r[1] = "\0%Y%m%d%H%M%S+\0";
+ } else {
+ for my $x (@r) {
+ next if $x eq '' || $x =~ /\A[0-9]{14}\z/;
+ push @$to_parse, $x;
+ $x = "\0%Y%m%d%H%M%S$#$to_parse\0";
+ }
+ }
+ } else { # "rt", let git interpret "YYYY", deal with Y10K later :P
+ for my $x (@r) {
+ next if $x eq '' || $x =~ /\A[0-9]{5,}\z/;
+ push @$to_parse, $x;
+ $x = "\0%s$#$to_parse\0";
+ }
+ $r[1] //= "\0%s+\0"; # add 1 day
+ }
+ "$pfx:".join('..', @r).$end;
+}
+
+sub date_parse_finalize {
+ my ($git, $to_parse) = @_;
+ # git-rev-parse can handle any number of args up to system
+ # limits (around (4096*32) bytes on Linux).
+ my @r = $git->date_parse(@$to_parse);
+ # n.b. git respects TZ, times stored in SQLite/Xapian are always UTC,
+ # and gmtime doesn't seem to do the right thing when TZ!=UTC
+ my ($i, $t);
+ $_[2] =~ s/\0(%[%YmdHMSs]+)([0-9\+]+)\0/
+ $t = $2 eq '+' ? ($r[$i]+86400) : $r[$i=$2+0];
+ $1 eq '%s' ? $t : strftime($1, gmtime($t))/sge;
+}
+
+# n.b. argv never has NUL, though we'll need to filter it out
+# if this $argv isn't from a command execution
+sub query_argv_to_string {
+ my (undef, $git, $argv) = @_;
+ my $to_parse;
+ my $tmp = join(' ', map {;
+ if (s!\b(d|rt|dt):(\S+)\z!date_parse_prepare(
+ $to_parse //= [], $1, $2)!sge) {
+ $_;
+ } elsif (/\s/) {
+ s/(.*?)\b(\w+:)// ? qq{$1$2"$_"} : qq{"$_"};
+ } else {
+ $_
+ }
+ } @$argv);
+ date_parse_finalize($git, $to_parse, $tmp) if $to_parse;
+ $tmp
+}
+
+# this is for the WWW "q=" query parameter and "lei q --stdin"
+# it can't do d:"5 days ago", but it will do d:5.days.ago
+sub query_approxidate {
+ my (undef, $git) = @_; # $_[2] = $query_string (modified in-place)
+ my $DQ = qq<"\x{201c}\x{201d}>; # Xapian can use curly quotes
+ $_[2] =~ tr/\x00/ /; # Xapian doesn't do NUL, we use it as a placeholder
+ my ($terms, $phrase, $to_parse);
+ $_[2] =~ s{([^$DQ]*)([$DQ][^$DQ]*[$DQ])?}{
+ ($terms, $phrase) = ($1, $2);
+ $terms =~ s!\b(d|rt|dt):(\S+)!
+ date_parse_prepare($to_parse //= [], $1, $2)!sge;
+ $terms.($phrase // '');
+ }sge;
+ date_parse_finalize($git, $to_parse, $_[2]) if $to_parse;
+}
+
# read-only
-sub query {
+sub mset {
my ($self, $query_string, $opts) = @_;
$opts ||= {};
- if ($query_string eq '' && !$opts->{mset}) {
- $self->{over_ro}->recent($opts);
- } else {
- my $qp = qp($self);
- my $qp_flags = $self->{qp_flags};
- my $query = $qp->parse_query($query_string, $qp_flags);
- $opts->{relevance} = 1 unless exists $opts->{relevance};
- _do_enquire($self, $query, $opts);
- }
+ my $qp = $self->{qp} //= $self->qparse_new;
+ my $query = $qp->parse_query($query_string, $self->{qp_flags});
+ _do_enquire($self, $query, $opts);
}
sub retry_reopen {
- my ($self, $cb, $arg) = @_;
+ my ($self, $cb, @arg) = @_;
for my $i (1..10) {
if (wantarray) {
- my @ret;
- eval { @ret = $cb->($arg) };
+ my @ret = eval { $cb->($self, @arg) };
return @ret unless $@;
} else {
- my $ret;
- eval { $ret = $cb->($arg) };
+ my $ret = eval { $cb->($self, @arg) };
return $ret unless $@;
}
# Exception: The revision being read has been discarded -
# you should call Xapian::Database::reopen()
if (ref($@) =~ /\bDatabaseModifiedError\b/) {
- warn "reopen try #$i on $@\n";
reopen($self);
} else {
# let caller decide how to spew, because ExtMsg queries
# get wonky and trigger:
# "something terrible happened at .../Xapian/Enquire.pm"
- die;
+ Carp::croak($@);
}
}
- die "Too many Xapian database modifications in progress\n";
+ Carp::croak("Too many Xapian database modifications in progress\n");
}
sub _do_enquire {
my ($self, $query, $opts) = @_;
- retry_reopen($self, \&_enquire_once, [ $self, $query, $opts ]);
+ retry_reopen($self, \&_enquire_once, $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 ($self, $query, $opts) = @_;
my $xdb = xdb($self);
+ if (defined(my $eidx_key = $opts->{eidx_key})) {
+ $query = $X{Query}->new(OP_FILTER(), $query, 'O'.$eidx_key);
+ }
+ if (defined(my $uid_range = $opts->{uid_range})) {
+ my $range = $X{Query}->new(OP_VALUE_RANGE(), UID,
+ sortable_serialise($uid_range->[0]),
+ sortable_serialise($uid_range->[1]));
+ $query = $X{Query}->new(OP_FILTER(), $query, $range);
+ }
my $enquire = $X{Enquire}->new($xdb);
$enquire->set_query($query);
$opts ||= {};
- my $desc = !$opts->{asc};
- if (($opts->{mset} || 0) == 2) {
- $enquire->set_docid_order($ENQ_ASCENDING);
+ my $rel = $opts->{relevance} // 0;
+ if ($rel == -2) { # ORDER BY docid/UID (highest first)
$enquire->set_weighting_scheme($X{BoolWeight}->new);
- } elsif ($opts->{relevance}) {
- $enquire->set_sort_by_relevance_then_value(TS, $desc);
- } else {
- $enquire->set_sort_by_value_then_relevance(TS, $desc);
+ $enquire->set_docid_order($ENQ_DESCENDING);
+ } elsif ($rel == -1) { # ORDER BY docid/UID (lowest first)
+ $enquire->set_weighting_scheme($X{BoolWeight}->new);
+ $enquire->set_docid_order($ENQ_ASCENDING);
+ } elsif ($rel == 0) {
+ $enquire->set_sort_by_value_then_relevance(TS, !$opts->{asc});
+ } else { # rel > 0
+ $enquire->set_sort_by_relevance_then_value(TS, !$opts->{asc});
}
- 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)
+
+ # `mairix -t / --threads' or JMAP collapseThreads
+ if ($opts->{threads} && has_threadid($self)) {
+ $enquire->set_collapse_key(THREADID);
+ }
+ $enquire->get_mset($opts->{offset} || 0, $opts->{limit} || 50);
+}
+
+sub mset_to_smsg {
+ my ($self, $ibx, $mset) = @_;
+ my $nshard = $self->{nshard};
+ my $i = 0;
+ my %order = map { mdocid($nshard, $_) => ++$i } $mset->items;
+ my @msgs = sort {
+ $order{$a->{num}} <=> $order{$b->{num}}
+ } @{$ibx->over->get_all(keys %order)};
+ wantarray ? ($mset->get_matches_estimated, \@msgs) : \@msgs;
}
# read-write
sub stemmer { $X{Stem}->new($LANG) }
# read-only
-sub qp {
+sub qparse_new {
my ($self) = @_;
- my $qp = $self->{query_parser};
- return $qp if $qp;
my $xdb = xdb($self);
- # new parser
- $qp = $X{QueryParser}->new;
+ my $qp = $X{QueryParser}->new;
$qp->set_default_op(OP_AND());
$qp->set_database($xdb);
- $qp->set_stemmer($self->stemmer);
+ $qp->set_stemmer(stemmer($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
+ $cb->($qp, $NVRP->new(BYTES, 'z:'));
+ $cb->($qp, $NVRP->new(TS, 'rt:'));
+ $cb->($qp, $NVRP->new(UID, 'uid:'));
while (my ($name, $prefix) = each %bool_pfx_external) {
$qp->add_boolean_prefix($name, $_) foreach split(/ /, $prefix);
while (my ($name, $prefix) = each %prob_prefix) {
$qp->add_prefix($name, $_) foreach split(/ /, $prefix);
}
-
- $self->{query_parser} = $qp;
+ $qp;
}
sub help {
my ($self) = @_;
- $self->qp; # parse altids
+ $self->{qp} //= $self->qparse_new; # parse altids
my @ret = @HELP;
if (my $user_pfx = $self->{-user_pfx}) {
push @ret, @$user_pfx;
\@ret;
}
+sub int_val ($$) {
+ my ($doc, $col) = @_;
+ my $val = $doc->get_value($col) or return; # undefined is '' in Xapian
+ sortable_unserialise($val) + 0; # PV => IV conversion
+}
+
+sub get_pct ($) { # mset item
+ # Capped at "99%" since "100%" takes an extra column in the
+ # thread skeleton view. <xapian/mset.h> says the value isn't
+ # very meaningful, anyways.
+ my $n = $_[0]->get_percent;
+ $n > 99 ? 99 : $n;
+}
+
+sub xap_terms ($$;@) {
+ my ($pfx, $xdb_or_doc, @docid) = @_; # @docid may be empty ()
+ my %ret;
+ my $end = $xdb_or_doc->termlist_end(@docid);
+ my $cur = $xdb_or_doc->termlist_begin(@docid);
+ for (; $cur != $end; $cur++) {
+ $cur->skip_to($pfx);
+ last if $cur == $end;
+ my $tn = $cur->get_termname;
+ $ret{substr($tn, length($pfx))} = undef if !index($tn, $pfx);
+ }
+ wantarray ? sort(keys(%ret)) : \%ret;
+}
+
+# get combined docid from over.num:
+# (not generic Xapian, only works with our sharding scheme)
+sub num2docid ($$) {
+ my ($self, $num) = @_;
+ my $nshard = $self->{nshard};
+ ($num - 1) * $nshard + $num % $nshard + 1;
+}
+
1;