msgmap: add tmp_clone to create an anonymous copy

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

diff --git a/lib/PublicInbox/Msgmap.pm b/lib/PublicInbox/Msgmap.pm
index 3fb3805fe82abea0c23684890d1f33004e6815d5..78922d3693c9647e0b11c3dc7fc300ce1a985147 100644 (file)
--- a/lib/PublicInbox/Msgmap.pm
+++ b/lib/PublicInbox/Msgmap.pm
@@ -1,5 +1,5 @@
-# Copyright (C) 2015 all contributors <meta@public-inbox.org>
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2015-2018 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 
 # bidirectional Message-ID <-> Article Number mapping for the NNTP
 # and web interfaces.  This is required for implementing stable article
@@ -12,6 +12,7 @@ use strict;
 use warnings;
 use DBI;
 use DBD::SQLite;
+use File::Temp qw(tempfile);
 
 sub new {
        my ($class, $git_dir, $writable) = @_;
@@ -45,26 +46,35 @@ sub new_file {
        $self;
 }
 
+# used to keep track of used numeric mappings for v2 reindex
+sub tmp_clone {
+       my ($self) = @_;
+       my ($fh, $fn) = tempfile(EXLOCK => 0);
+       $self->{dbh}->sqlite_backup_to_file($fn);
+       my $tmp = ref($self)->new_file($fn, 1);
+       $tmp->{dbh}->do('PRAGMA synchronous = OFF');
+       $tmp->{tmp_name} = $fn; # SQLite won't work if unlinked, apparently
+       $fh = undef;
+       $tmp;
+}
+
 # n.b. invoked directly by scripts/xhdr-num2mid
 sub meta_accessor {
        my ($self, $key, $value) = @_;
-       use constant {
-               meta_select => 'SELECT val FROM meta WHERE key = ? LIMIT 1',
-               meta_update => 'UPDATE meta SET val = ? WHERE key = ? LIMIT 1',
-               meta_insert => 'INSERT INTO meta (key,val) VALUES (?,?)',
-       };
 
+       my $sql = 'SELECT val FROM meta WHERE key = ? LIMIT 1';
        my $dbh = $self->{dbh};
        my $prev;
-       defined $value or
-               return $dbh->selectrow_array(meta_select, undef, $key);
+       defined $value or return $dbh->selectrow_array($sql, undef, $key);
 
-       $prev = $dbh->selectrow_array(meta_select, undef, $key);
+       $prev = $dbh->selectrow_array($sql, undef, $key);
 
        if (defined $prev) {
-               $dbh->do(meta_update, undef, $value, $key);
+               $sql = 'UPDATE meta SET val = ? WHERE key = ? LIMIT 1';
+               $dbh->do($sql, undef, $value, $key);
        } else {
-               $dbh->do(meta_insert, undef, $key, $value);
+               $sql = 'INSERT INTO meta (key,val) VALUES (?,?)';
+               $dbh->do($sql, undef, $key, $value);
        }
        $prev;
 }
@@ -82,18 +92,18 @@ sub created_at {
 sub mid_insert {
        my ($self, $mid) = @_;
        my $dbh = $self->{dbh};
-       use constant MID_INSERT => 'INSERT INTO msgmap (mid) VALUES (?)';
-       my $sth = $self->{mid_insert} ||= $dbh->prepare(MID_INSERT);
+       my $sql = 'INSERT OR IGNORE INTO msgmap (mid) VALUES (?)';
+       my $sth = $self->{mid_insert} ||= $dbh->prepare($sql);
        $sth->bind_param(1, $mid);
-       $sth->execute;
+       return if $sth->execute == 0;
        $dbh->last_insert_id(undef, undef, 'msgmap', 'num');
 }
 
 sub mid_for {
        my ($self, $num) = @_;
        my $dbh = $self->{dbh};
-       use constant MID_FOR => 'SELECT mid FROM msgmap WHERE num = ? LIMIT 1';
-       my $sth = $self->{mid_for} ||= $dbh->prepare(MID_FOR);
+       my $sth = $self->{mid_for} ||=
+               $dbh->prepare('SELECT mid FROM msgmap WHERE num = ? LIMIT 1');
        $sth->bind_param(1, $num);
        $sth->execute;
        $sth->fetchrow_array;
@@ -102,8 +112,8 @@ sub mid_for {
 sub num_for {
        my ($self, $mid) = @_;
        my $dbh = $self->{dbh};
-       use constant NUM_FOR => 'SELECT num FROM msgmap WHERE mid = ? LIMIT 1';
-       my $sth = $self->{num_for} ||= $dbh->prepare(NUM_FOR);
+       my $sth = $self->{num_for} ||=
+               $dbh->prepare('SELECT num FROM msgmap WHERE mid = ? LIMIT 1');
        $sth->bind_param(1, $mid);
        $sth->execute;
        $sth->fetchrow_array;
@@ -112,8 +122,8 @@ sub num_for {
 sub minmax {
        my ($self) = @_;
        my $dbh = $self->{dbh};
-       use constant NUM_MINMAX => 'SELECT MIN(num),MAX(num) FROM msgmap';
-       my $sth = $self->{num_minmax} ||= $dbh->prepare(NUM_MINMAX);
+       my $sth = $self->{num_minmax} ||=
+               $dbh->prepare('SELECT MIN(num),MAX(num) FROM msgmap');
        $sth->execute;
         $sth->fetchrow_array;
 }
@@ -138,12 +148,19 @@ sub mid_prefixes {
 sub mid_delete {
        my ($self, $mid) = @_;
        my $dbh = $self->{dbh};
-       use constant MID_DELETE => 'DELETE FROM msgmap WHERE mid = ?';
-       my $sth = $dbh->prepare(MID_DELETE);
+       my $sth = $dbh->prepare('DELETE FROM msgmap WHERE mid = ?');
        $sth->bind_param(1, $mid);
        $sth->execute;
 }
 
+sub num_delete {
+       my ($self, $num) = @_;
+       my $dbh = $self->{dbh};
+       my $sth = $dbh->prepare('DELETE FROM msgmap WHERE num = ?');
+       $sth->bind_param(1, $num);
+       $sth->execute;
+}
+
 sub create_tables {
        my ($dbh) = @_;
        my $e;
@@ -175,14 +192,20 @@ sub id_batch {
 }
 
 # only used for mapping external serial numbers (e.g. articles from gmane)
-# see scripts/xhdr-num2mid for usage
+# see scripts/xhdr-num2mid or PublicInbox::Filter::RubyLang for usage
 sub mid_set {
        my ($self, $num, $mid) = @_;
        my $sth = $self->{mid_set} ||= do {
-               my $sql = 'INSERT INTO msgmap (num, mid) VALUES (?,?)';
-               $self->{dbh}->prepare($sql);
+               $self->{dbh}->prepare(
+                       'INSERT OR IGNORE INTO msgmap (num,mid) VALUES (?,?)');
        };
        $sth->execute($num, $mid);
 }
 
+sub DESTROY {
+       my ($self) = @_;
+       delete $self->{dbh};
+       unlink $self->{tmp_name} if defined $self->{tmp_name};
+}
+
 1;