X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FMsgmap.pm;h=dea957316609ffb63332098d0e994945447ca4bd;hb=35ff6bb106909b1c1232666a9792156dfa398ea8;hp=2f64d90c71f741cb100c3beacd124e2698260d71;hpb=7ee29d11b25b09b8919b9a125732e4c2c5784d28;p=public-inbox.git diff --git a/lib/PublicInbox/Msgmap.pm b/lib/PublicInbox/Msgmap.pm index 2f64d90c..dea95731 100644 --- a/lib/PublicInbox/Msgmap.pm +++ b/lib/PublicInbox/Msgmap.pm @@ -1,12 +1,18 @@ -# Copyright (C) 2015 all contributors -# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt) -# bidirectional Message-ID <-> Article Number mapping +# Copyright (C) 2015-2018 all contributors +# License: AGPL-3.0+ + +# bidirectional Message-ID <-> Article Number mapping for the NNTP +# and web interfaces. This is required for implementing stable article +# numbers for NNTP and allows prefix lookups for partial Message-IDs +# in case URLs get truncated from copy-n-paste errors by users. +# +# This is maintained by ::SearchIdx package PublicInbox::Msgmap; use strict; use warnings; -use fields qw(dbh mid_insert mid_for num_for num_minmax); use DBI; use DBD::SQLite; +use File::Temp qw(tempfile); sub new { my ($class, $git_dir, $writable) = @_; @@ -15,53 +21,70 @@ sub new { my $err = $!; -d $d or die "$d not created: $err"; } - my $f = "$d/msgmap.sqlite3"; + new_file($class, "$d/msgmap.sqlite3", $writable); +} + +sub dbh_new { + my ($f, $writable) = @_; my $dbh = DBI->connect("dbi:SQLite:dbname=$f",'','', { AutoCommit => 1, RaiseError => 1, PrintError => 0, + ReadOnly => !$writable, sqlite_use_immediate_transaction => 1, }); $dbh->do('PRAGMA case_sensitive_like = ON'); - my $self = fields::new($class); - $self->{dbh} = $dbh; + $dbh; +} + +sub new_file { + my ($class, $f, $writable) = @_; + return if !$writable && !-r $f; + + my $dbh = dbh_new($f, $writable); + my $self = bless { dbh => $dbh }, $class; if ($writable) { create_tables($dbh); + $dbh->begin_work; $self->created_at(time) unless $self->created_at; + $dbh->commit; } $self; } +# used to keep track of used numeric mappings for v2 reindex +sub tmp_clone { + my ($self) = @_; + my ($fh, $fn) = tempfile('msgmap-XXXXXXXX', EXLOCK => 0, TMPDIR => 1); + $self->{dbh}->sqlite_backup_to_file($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 + $tmp->{pid} = $$; + close $fh or die "failed to close $fn: $!"; + $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); - - $dbh->begin_work; - eval { - $prev = $dbh->selectrow_array(meta_select, undef, $key); - - if (defined $prev) { - $dbh->do(meta_update, undef, $value, $key); - } else { - $dbh->do(meta_insert, undef, $key, $value); - } - $dbh->commit; - }; - my $err = $@; - return $prev unless $err; + defined $value or return $dbh->selectrow_array($sql, undef, $key); + + $prev = $dbh->selectrow_array($sql, undef, $key); - $dbh->rollback; - die $err; + if (defined $prev) { + $sql = 'UPDATE meta SET val = ? WHERE key = ? LIMIT 1'; + $dbh->do($sql, undef, $value, $key); + } else { + $sql = 'INSERT INTO meta (key,val) VALUES (?,?)'; + $dbh->do($sql, undef, $key, $value); + } + $prev; } sub last_commit { @@ -77,18 +100,18 @@ 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'); } -use constant MID_FOR => 'SELECT mid FROM msgmap WHERE num = ? LIMIT 1'; sub mid_for { my ($self, $num) = @_; my $dbh = $self->{dbh}; - 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; @@ -97,8 +120,8 @@ 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; @@ -107,8 +130,8 @@ 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; } @@ -133,12 +156,19 @@ 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; @@ -155,6 +185,7 @@ sub create_tables { 'val VARCHAR(255) NOT NULL)'); } +# used by NNTP.pm sub id_batch { my ($self, $num, $cb) = @_; my $dbh = $self->{dbh}; @@ -168,4 +199,42 @@ sub id_batch { $nr; } +# only used for mapping external serial numbers (e.g. articles from gmane) +# see scripts/xhdr-num2mid or PublicInbox::Filter::RubyLang for usage +sub mid_set { + my ($self, $num, $mid) = @_; + my $sth = $self->{mid_set} ||= do { + $self->{dbh}->prepare( + 'INSERT OR IGNORE INTO msgmap (num,mid) VALUES (?,?)'); + }; + $sth->execute($num, $mid); +} + +sub DESTROY { + my ($self) = @_; + delete $self->{dbh}; + my $f = delete $self->{tmp_name}; + if (defined $f && $self->{pid} == $$) { + unlink $f or warn "failed to unlink $f: $!\n"; + } +} + +sub atfork_parent { + my ($self) = @_; + my $f = $self->{tmp_name} or die "not a temporary clone\n"; + delete $self->{dbh} and die "tmp_clone dbh not prepared for parent"; + my $dbh = $self->{dbh} = dbh_new($f, 1); + $dbh->do('PRAGMA synchronous = OFF'); +} + +sub atfork_prepare { + my ($self) = @_; + my $f = $self->{tmp_name} or die "not a temporary clone\n"; + $self->{pid} == $$ or + die "BUG: atfork_prepare not called from $self->{pid}\n"; + $self->{dbh} or die "temporary clone not open\n"; + # must clobber prepared statements + %$self = (tmp_name => $f, pid => $$); +} + 1;