lei: retry_reopen on read-only Xapian access

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

diff --git a/lib/PublicInbox/Search.pm b/lib/PublicInbox/Search.pm
index 8e4cce33f5f6eb267d480da59e3f6b59a9d01fe5..59a5a3b006755967699f1731e19039d61574d81d 100644 (file)
--- a/lib/PublicInbox/Search.pm
+++ b/lib/PublicInbox/Search.pm
@@ -9,13 +9,14 @@ 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)
+       TS => 0, # Received: in Unix time (IMAP INTERNALDATE, JMAP receivedAt)
        YYYYMMDD => 1, # Date: header for searching in the WWW UI
-       DT => 2, # Date: YYYYMMDDHHMMSS
+       DT => 2, # Date: YYYYMMDDHHMMSS (IMAP SENT*, JMAP sentAt)
 
        # added for public-inbox 1.6.0+
        BYTES => 3, # IMAP RFC822.SIZE
@@ -275,7 +276,9 @@ sub date_parse_prepare {
 
        # 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-$3";
@@ -283,14 +286,27 @@ sub date_parse_prepare {
                                # to parse anyways for "d+" below
                        } else {
                                push @$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";
+                               }
                        }
-                       $r[0] = "\0%Y%m%d$#$to_parse\0";
-                       $r[1] = "\0%Y%m%d+\0";
+                       $r[0] = "$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;
-                               $x = "\0%Y%m%d$#$to_parse\0";
+                               if ($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') {
@@ -337,7 +353,7 @@ sub query_argv_to_string {
        my (undef, $git, $argv) = @_;
        my $to_parse;
        my $tmp = join(' ', map {;
-               if (s!\b(d|rt|dt):([[:print:]]+)\z!date_parse_prepare(
+               if (s!\b(d|rt|dt):(\S+)\z!date_parse_prepare(
                                                $to_parse //= [], $1, $2)!sge) {
                        $_;
                } elsif (/\s/) {
@@ -357,7 +373,7 @@ sub query_approxidate {
        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])?}{
+       $_[2] =~ s{([^$DQ]*)([$DQ][^$DQ]*[$DQ])?}{
                ($terms, $phrase) = ($1, $2);
                $terms =~ s!\b(d|rt|dt):(\S+)!
                        date_parse_prepare($to_parse //= [], $1, $2)!sge;
@@ -370,7 +386,7 @@ sub query_approxidate {
 sub mset {
        my ($self, $query_string, $opts) = @_;
        $opts ||= {};
-       my $qp = $self->{qp} //= qparse_new($self);
+       my $qp = $self->{qp} //= $self->qparse_new;
        my $query = $qp->parse_query($query_string, $self->{qp_flags});
        _do_enquire($self, $query, $opts);
 }
@@ -390,16 +406,16 @@ sub retry_reopen {
                # Exception: The revision being read has been discarded -
                # you should call Xapian::Database::reopen()
                if (ref($@) =~ /\bDatabaseModifiedError\b/) {
-                       warn "reopen try #$i on $@\n";
+                       warn "# reopen try #$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 {
@@ -463,7 +479,7 @@ sub mset_to_smsg {
 sub stemmer { $X{Stem}->new($LANG) }
 
 # read-only
-sub qparse_new ($) {
+sub qparse_new {
        my ($self) = @_;
 
        my $xdb = xdb($self);
@@ -481,7 +497,7 @@ sub qparse_new ($) {
        $cb->($qp, $NVRP->new(DT, 'dt:'));
 
        # for IMAP, undocumented for WWW and may be split off go away
-       $cb->($qp, $NVRP->new(BYTES, 'bytes:'));
+       $cb->($qp, $NVRP->new(BYTES, 'z:'));
        $cb->($qp, $NVRP->new(TS, 'rt:'));
        $cb->($qp, $NVRP->new(UID, 'uid:'));
 
@@ -516,7 +532,7 @@ EOF
 
 sub help {
        my ($self) = @_;
-       $self->{qp} //= qparse_new($self); # parse altids
+       $self->{qp} //= $self->qparse_new; # parse altids
        my @ret = @HELP;
        if (my $user_pfx = $self->{-user_pfx}) {
                push @ret, @$user_pfx;