--- trunk/bin/BackupPC_updatedb 2005/06/23 12:22:21 8 +++ trunk/bin/BackupPC_updatedb 2005/08/28 18:02:58 92 @@ -1,319 +1,599 @@ -#!/usr/local/bin/perl -$| = 1; +#!/usr/local/bin/perl -w use strict; -use DBI; use lib "__INSTALLDIR__/lib"; + +use DBI; use BackupPC::Lib; use BackupPC::View; use Data::Dumper; use Getopt::Std; +use Time::HiRes qw/time/; +use File::Pid; +use POSIX qw/strftime/; + use constant BPC_FTYPE_DIR => 5; +my $debug = 0; +$|=1; + +my $start_t = time(); + +my $pidfile = new File::Pid; + +if (my $pid = $pidfile->running ) { + die "$0 already running: $pid\n"; +} elsif ($pidfile->pid ne $$) { + $pidfile->remove; + $pidfile = new File::Pid; +} +$pidfile->write; +print STDERR "$0 using pid ",$pidfile->pid," file ",$pidfile->file,"\n"; + +my $t_fmt = '%Y-%m-%d %H:%M:%S'; my $hosts; my $bpc = BackupPC::Lib->new || die; my %Conf = $bpc->Conf(); my $TopDir = $bpc->TopDir(); -my @beenThere = (); +my $beenThere = {}; + +my $dsn = $Conf{SearchDSN} || die "Need SearchDSN in config.pl\n"; +my $user = $Conf{SearchUser} || ''; +my $index_path = $Conf{HyperEstraierIndex}; +$index_path = $TopDir . '/' . $index_path; +$index_path =~ s#//#/#g; -my $dbh = DBI->connect("dbi:SQLite:dbname=$TopDir/$Conf{SearchDB}", "", "", { RaiseError => 1, AutoCommit => 0 }); + +my $dbh = DBI->connect($dsn, $user, "", { RaiseError => 1, AutoCommit => 0 }); my %opt; -if ( !getopts("cdu", \%opt ) ) { +if ( !getopts("cdm:v:i", \%opt ) ) { print STDERR <prepare(qq{ + SELECT + files.id AS fid, + hosts.name AS hname, + shares.name AS sname, + -- shares.share AS sharename, + files.backupnum AS backupnum, + -- files.name AS filename, + files.path AS filepath, + files.date AS date, + files.type AS type, + files.size AS size, + files.shareid AS shareid, + backups.date AS backup_date + FROM files + INNER JOIN shares ON files.shareID=shares.ID + INNER JOIN hosts ON hosts.ID = shares.hostID + INNER JOIN backups ON backups.num = files.backupNum and backups.hostID = hosts.ID AND backups.shareID = shares.ID + $where + }); + + $sth->execute(@_); + my $results = $sth->rows; + + if ($results == 0) { + print " - no files, skipping\n"; + return; + } + + my $dot = int($results / 15) || 1; + + print " $results ($dot/#)"; -if ($opt{c}) - { - print "creating database..."; + sub fmt_date { + my $t = shift || return; + my $iso = BackupPC::Lib::timeStamp($t); + $iso =~ s/\s/T/; + return $iso; + } + + my $max = int($results / $dot); + + print ", opening index $index_path..."; + use HyperEstraier; + my $db = HyperEstraier::Database->new(); + +# unless ($hest_db) { +# print " open reader"; +# $hest_db = HyperEstraier::Database->new(); +# +# } + + + $db->open($index_path, $HyperEstraier::Database::DBWRITER | $HyperEstraier::Database::DBCREAT); + + my $added = 0; + + while (my $row = $sth->fetchrow_hashref()) { + + my $fid = $row->{'fid'} || die "no fid?"; + my $uri = 'file:///' . $fid; + + my $id = $db->uri_to_id($uri); + next unless ($id == -1); + + # create a document object + my $doc = HyperEstraier::Document->new; + + # add attributes to the document object + $doc->add_attr('@uri', $uri); + + foreach my $c (@{ $sth->{NAME} }) { + $doc->add_attr($c, $row->{$c}) if ($row->{$c}); + } + + #$doc->add_attr('@cdate', fmt_date($row->{'date'})); + + # add the body text to the document object + my $path = $row->{'filepath'}; + $doc->add_text($path); + $path =~ s/(.)/$1 /g; + $doc->add_hidden_text($path); + + print STDERR $doc->dump_draft,"\n" if ($debug > 1); + + # register the document object to the database + $db->put_doc($doc, $HyperEstraier::Database::PDCLEAN); + + $added++; + if ($added % $dot == 0) { + print "$max "; + $max--; + } + + } + + print "sync $added new files"; + $db->sync(); + print ", close"; + $db->close(); + + my $dur = (time() - $t) || 1; + printf(" [%.2f/s new %.2f/s dur: %s]\n", + ( $results / $dur ), + ( $added / $dur ), + fmt_time($dur) + ); +} + +#---- /subs ---- + + +## update index ## +if ($opt{i} || ($index_path && ! -e $index_path)) { + # update all + print "force update of HyperEstraier index "; + print "importing existing data" unless (-e $index_path); + print "by -i flag" if ($opt{i}); + print "\n"; + hest_update(); +} + +## create tables ## +if ($opt{c}) { + sub do_index { + my $index = shift || return; + my ($table,$col,$unique) = split(/_/, $index); + $unique ||= ''; + $index =~ s/,/_/g; + $dbh->do(qq{ create $unique index $index on $table($col) }); + } + + print "creating tables...\n"; - $dbh->do( - q{ - create table hosts - ( ID INTEGER PRIMARY KEY, - name VARCHAR(30) NOT NULL, - IP VARCHAR(20) NOT NULL + $dbh->do(qq{ + create table hosts ( + ID SERIAL PRIMARY KEY, + name VARCHAR(30) NOT NULL, + IP VARCHAR(15) ); - } - ); + }); - $dbh->do( - q{ - create table shares - ( ID INTEGER PRIMARY KEY, - hostID INTEGER NOT NULL, - name VARCHAR(30) NOT NULL, - share VARCHAR(200) NOT NULL, - localpath VARCHAR(200) + $dbh->do(qq{ + create table shares ( + ID SERIAL PRIMARY KEY, + hostID INTEGER NOT NULL references hosts(id), + name VARCHAR(30) NOT NULL, + share VARCHAR(200) NOT NULL, + localpath VARCHAR(200) ); - } - ); + }); - $dbh->do( - q{ - create table backups - ( hostID INTEGER NOT NULL, - num INTEGER NOT NULL, - date DATE, - type CHAR(1), - PRIMARY KEY(hostID, num) + $dbh->do(qq{ + create table backups ( + hostID INTEGER NOT NULL references hosts(id), + num INTEGER NOT NULL, + date integer NOT NULL, + type CHAR(4) not null, + shareID integer not null references shares(id), + size integer not null, + PRIMARY KEY(hostID, num, shareID) ); - } - ); + }); + + #do_index('backups_hostid,num_unique'); - $dbh->do( - q{ - create table dvds - ( ID INTEGER PRIMARY KEY, - num INTEGER NOT NULL, - name VARCHAR(255) NOT NULL, - mjesto VARCHAR(255) + $dbh->do(qq{ + create table dvds ( + ID SERIAL PRIMARY KEY, + num INTEGER NOT NULL, + name VARCHAR(255) NOT NULL, + mjesto VARCHAR(255) ); - } - ); + }); - $dbh->do( - q{ - create table files - ( ID INTEGER NOT NULL PRIMARY KEY, - shareID INTEGER NOT NULL, - backupNum INTEGER NOT NULL, - name VARCHAR(255) NOT NULL, - path VARCHAR(255) NOT NULL, - fullpath VARCHAR(255) NOT NULL, - date TIMESTAMP NOT NULL, - type INTEGER NOT NULL, - size INTEGER NOT NULL, - dvdid INTEGER + $dbh->do(qq{ + create table files ( + ID SERIAL PRIMARY KEY, + shareID INTEGER NOT NULL references shares(id), + backupNum INTEGER NOT NULL, + name VARCHAR(255) NOT NULL, + path VARCHAR(255) NOT NULL, + date integer NOT NULL, + type INTEGER NOT NULL, + size INTEGER NOT NULL, + dvdid INTEGER references dvds(id) ); - } - ); - print "done\n"; - } - -if ($opt{d}) - { - print("deleting db first...\n"); - - $dbh->do( - q{ DELETE FROM hosts; } - ); - $dbh->do( - q{ DELETE FROM shares; } - ); - $dbh->do( - q{ DELETE FROM files;} - ); - $dbh->do( - q{ DELETE FROM dvds;} - ); - $dbh->do( - q{ DELETE FROM backups; } - ); - } + }); + + print "creating indexes:"; + + foreach my $index (qw( + hosts_name + backups_hostID + backups_num + shares_hostID + shares_name + files_shareID + files_path + files_name + files_date + files_size + )) { + print " $index"; + do_index($index); + } + print "...\n"; + + $dbh->commit; -#################################INSERT VALUES############################# +} + +## delete data before inseting ## +if ($opt{d}) { + print "deleting "; + foreach my $table (qw(files dvds backups shares hosts)) { + print "$table "; + $dbh->do(qq{ DELETE FROM $table }); + } + print " done...\n"; + + $dbh->commit; +} + +## insert new values ## # get hosts $hosts = $bpc->HostInfoRead(); -print Dumper($hosts); my $hostID; my $shareID; -foreach my $host_key (keys %{$hosts}) -{ - my $hostname = $hosts->{$host_key}->{'host'} || die "can't find host for $host_key"; - my $backups; - my $sql; - - $sql = q{ SELECT ID FROM hosts WHERE name=? }; - my $st = $dbh->prepare($sql); - $st->bind_param(1,$hosts->{$host_key}->{'host'}); - $st->execute(); - if (my $tmp = $st->fetchrow_hashref()) - { - $hostID = $tmp->{'ID'}; - } - else - { - $sql = q{ INSERT INTO hosts ( ID, name, IP) VALUES (NULL,"}. - $hosts->{$host_key}->{'host'}."\", \"". - $hosts->{$host_key}->{'ip'}."\");"; - - $dbh->do($sql); - $hostID = $dbh->func('last_insert_rowid'); - - } - $st->finish(); - print("processing host ".$hosts->{$host_key}->{'host'}.":\n"); + +my $sth; + +$sth->{insert_hosts} = $dbh->prepare(qq{ +INSERT INTO hosts (name, IP) VALUES (?,?) +}); + +$sth->{hosts_by_name} = $dbh->prepare(qq{ +SELECT ID FROM hosts WHERE name=? +}); + +$sth->{backups_count} = $dbh->prepare(qq{ +SELECT COUNT(*) +FROM backups +WHERE hostID=? AND num=? AND shareid=? +}); + +$sth->{insert_backups} = $dbh->prepare(qq{ +INSERT INTO backups (hostID, num, date, type, shareid, size) +VALUES (?,?,?,?,?,?) +}); + +$sth->{insert_files} = $dbh->prepare(qq{ +INSERT INTO files + (shareID, backupNum, name, path, date, type, size) + VALUES (?,?,?,?,?,?,?) +}); + +foreach my $host_key (keys %{$hosts}) { + + my $hostname = $hosts->{$host_key}->{'host'} || die "can't find host for $host_key"; + + $sth->{hosts_by_name}->execute($hosts->{$host_key}->{'host'}); + + unless (($hostID) = $sth->{hosts_by_name}->fetchrow_array()) { + $sth->{insert_hosts}->execute( + $hosts->{$host_key}->{'host'}, + $hosts->{$host_key}->{'ip'} + ); + + $hostID = $dbh->last_insert_id(undef,undef,'hosts',undef); + } + + print "host ".$hosts->{$host_key}->{'host'}.": "; - # get backups for a host - my @backups = $bpc->BackupInfoRead($hostname); - foreach my $backup (@backups) - { - my $backupNum = $backup->{'num'}; - my @backupShares = (); + # get backups for a host + my @backups = $bpc->BackupInfoRead($hostname); + my $incs = scalar @backups; + print "$incs increments\n"; + + my $inc_nr = 0; + $beenThere = {}; + + foreach my $backup (@backups) { + + $inc_nr++; + last if ($opt{m} && $inc_nr > $opt{m}); + + my $backupNum = $backup->{'num'}; + my @backupShares = (); + + printf("%-10s %2d/%-2d #%-2d %s %5s/%5s files (date: %s dur: %s)\n", + $hosts->{$host_key}->{'host'}, + $inc_nr, $incs, $backupNum, + $backup->{type} || '?', + $backup->{nFilesNew} || '?', $backup->{nFiles} || '?', + strftime($t_fmt,localtime($backup->{startTime})), + fmt_time($backup->{endTime} - $backup->{startTime}) + ); + + my $files = BackupPC::View->new($bpc, $hostname, \@backups, 1); + foreach my $share ($files->shareList($backupNum)) { + + my $t = time(); + + $shareID = getShareID($share, $hostID, $hostname); + + $sth->{backups_count}->execute($hostID, $backupNum, $shareID); + my ($count) = $sth->{backups_count}->fetchrow_array(); + # skip if allready in database! + next if ($count > 0); + + # dump some log + print curr_time," ", $share; + + my ($f, $nf, $d, $nd, $size) = recurseDir($bpc, $hostname, $files, $backupNum, $share, "", $shareID); + + $sth->{insert_backups}->execute( + $hostID, + $backupNum, + $backup->{'endTime'}, + $backup->{'type'}, + $shareID, + $size, + ); + + print " commit"; + $dbh->commit(); + + my $dur = (time() - $t) || 1; + printf(" %d/%d files %d/%d dirs %0.2f MB [%.2f/s dur: %s]\n", + $nf, $f, $nd, $d, + ($size / 1024 / 1024), + ( ($f+$d) / $dur ), + fmt_time($dur) + ); + + hest_update($hostID, $shareID, $backupNum); + } - - if ($opt{u}) - { - my $sql = q{ - SELECT COUNT(*) AS broj - FROM backups - WHERE hostID=? AND - num=? - }; - - my $st = $dbh->prepare($sql); - $st->bind_param(1,$hostID); - $st->bind_param(2,$backupNum); - $st->execute(); - my $tmp = $st->fetchrow_hashref(); - $st->finish(); - if ($tmp->{'broj'} > 0) - { - next; - } } - - print "\tprocessing backup no. $backupNum..."; - my $sql = - q{ - INSERT INTO backups (hostID, num, date, type) - VALUES - (}.$hostID.",". $backupNum.q{, }.$backup->{'endTime'}.",\"". $backup->{'type'}.q{"); - }; - $dbh->do($sql); - - my $files = BackupPC::View->new($bpc, $hostname, \@backups); - @backupShares = $files->shareList($backupNum); - foreach my $share (@backupShares) - { - my @flattenedFiles = (); - clearBeenThereCache(); - print "\n\t\tprocessing share ".$share."..."; - $shareID = getShareID($share, $hostID, $hostname); - - @flattenedFiles = recurseDir($bpc, $hostname, \@backups, $backupNum, $share, ""); - print "done\n"; - print "\t\tinserting data into db..."; - foreach my $file (@flattenedFiles) - { - $dbh->do("INSERT INTO files(ID, shareID, backupNum, name, path, fullpath, date, type, size) VALUES " - ."( NULL, $shareID, " - .$backupNum.", \"" - .$file->{'fileName'}."\", \"" - .$file->{'relPath'}. "\", \"" - .$file->{'fullPath'}."\", " - .$file->{'mtime'}.", " - .$file->{'type'}.", " - .$file->{'size'}.")" - ); - } - print "done\n"; - } - } - print "done.\n"; } +undef $sth; $dbh->commit(); $dbh->disconnect(); +print "total duration: ",fmt_time(time() - $start_t),"\n"; -sub haveBeenThere -{ - my ($where) = @_; - - foreach my $st (@beenThere) - { - if ($where eq $st) - { - return 1; - } - } - - push(@beenThere, $where); - return 0; -} - -sub clearBeenThereCache -{ - @beenThere = (); -} - -sub getShareID() - { - my ($share, $hostID, $hostname) = @_; - my $tmp; - - my $st = $dbh -> prepare(" SELECT shares.ID AS ID FROM shares WHERE hostID=? AND name=?"); - $st -> execute($hostID,$share); - my $tmp = $st->fetchrow_hashref(); - $st->finish(); - if ($tmp) - { - return $tmp->{'ID'}; - } - my $sql = - q{ - INSERT INTO shares(ID,hostID,name,share,localpath) - VALUES - (NULL,}. - "$hostID,\"$share\",\"//$hostname$share\",NULL);"; - $dbh->do($sql); - return $dbh->func('last_insert_rowid'); - } +$pidfile->remove; + +sub getShareID() { + + my ($share, $hostID, $hostname) = @_; + + $sth->{share_id} ||= $dbh->prepare(qq{ + SELECT ID FROM shares WHERE hostID=? AND name=? + }); + + $sth->{share_id}->execute($hostID,$share); + + my ($id) = $sth->{share_id}->fetchrow_array(); + + return $id if (defined($id)); + + $sth->{insert_share} ||= $dbh->prepare(qq{ + INSERT INTO shares + (hostID,name,share,localpath) + VALUES (?,?,?,?) + }); + + my $drop_down = $hostname . '/' . $share; + $drop_down =~ s#//+#/#g; + + $sth->{insert_share}->execute($hostID,$share, $drop_down ,undef); + return $dbh->last_insert_id(undef,undef,'shares',undef); +} + +sub found_in_db { + + my @data = @_; + shift @data; + + my ($key, $shareID,undef,$name,$path,$date,undef,$size) = @_; + + return $beenThere->{$key} if (defined($beenThere->{$key})); + + $sth->{file_in_db} ||= $dbh->prepare(qq{ + SELECT 1 FROM files + WHERE shareID = ? and + path = ? and + date = ? and + size = ? + LIMIT 1 + }); + + my @param = ($shareID,$path,$date,$size); + $sth->{file_in_db}->execute(@param); + my $rows = $sth->{file_in_db}->rows; + print STDERR "## found_in_db($shareID,$path,$date,$size) ",( $rows ? '+' : '-' ), join(" ",@param), "\n" if ($debug >= 3); + + $beenThere->{$key}++; + + $sth->{'insert_files'}->execute(@data) unless ($rows); + return $rows; +} #################################################### # recursing through filesystem structure and # # and returning flattened files list # #################################################### -sub recurseDir -{ - my ($bpc, $hostname, $backups, $backupNo, $share, $dir) = @_; - my @ret = (); - my $files = BackupPC::View->new($bpc, $hostname, $backups); - my $filesInBackup = $files->dirAttrib($backupNo, $share, $dir); - my $file_key = ""; +sub recurseDir($$$$$$$$) { - - # first, add all the entries in current directory - foreach $file_key (keys %{$filesInBackup}) - { - push(@ret, { - 'fileName' => $file_key, - 'relPath' => $filesInBackup->{$file_key}->{'relPath'}, - 'fullPath' => $filesInBackup->{$file_key}->{'fullPath'}, - 'sharePath'=> $filesInBackup->{$file_key}->{'sharePathM'}, - 'size' => $filesInBackup->{$file_key}->{'size'}, - 'mtime' => $filesInBackup->{$file_key}->{'mtime'}, - 'type' => $filesInBackup->{$file_key}->{'type'} - }); - } - - # then, recurse thru subfolders - foreach my $fold (@ret) - { - if ($fold->{'type'} == BPC_FTYPE_DIR && - haveBeenThere($fold->{'relPath'}) != 1 - ) - { - - push(@ret, - recurseDir($bpc, $hostname, $backups, $backupNo, $share, $fold->{'relPath'}) - ); - } - } - return @ret; + my ($bpc, $hostname, $files, $backupNum, $share, $dir, $shareID) = @_; + + print STDERR "\nrecurse($hostname,$backupNum,$share,$dir,$shareID)\n" if ($debug >= 1); + + my ($nr_files, $new_files, $nr_dirs, $new_dirs, $size) = (0,0,0,0,0); + + { # scope + my @stack; + + print STDERR "# dirAttrib($backupNum, $share, $dir)\n" if ($debug >= 2); + my $filesInBackup = $files->dirAttrib($backupNum, $share, $dir); + + # first, add all the entries in current directory + foreach my $path_key (keys %{$filesInBackup}) { + print STDERR "# file ",Dumper($filesInBackup->{$path_key}),"\n" if ($debug >= 3); + my @data = ( + $shareID, + $backupNum, + $path_key, + $filesInBackup->{$path_key}->{'relPath'}, + $filesInBackup->{$path_key}->{'mtime'}, + $filesInBackup->{$path_key}->{'type'}, + $filesInBackup->{$path_key}->{'size'} + ); + + my $key = join(" ", ( + $shareID, + $dir, + $path_key, + $filesInBackup->{$path_key}->{'mtime'}, + $filesInBackup->{$path_key}->{'size'} + )); + + my $found; + if (! defined($beenThere->{$key}) && ! ($found = found_in_db($key, @data)) ) { + print STDERR "# key: $key [", $beenThere->{$key},"]" if ($debug >= 2); + + if ($filesInBackup->{$path_key}->{'type'} == BPC_FTYPE_DIR) { + $new_dirs++ unless ($found); + print STDERR " dir\n" if ($debug >= 2); + } else { + $new_files++ unless ($found); + print STDERR " file\n" if ($debug >= 2); + } + $size += $filesInBackup->{$path_key}->{'size'} || 0; + } + + if ($filesInBackup->{$path_key}->{'type'} == BPC_FTYPE_DIR) { + $nr_dirs++; + + my $full_path = $dir . '/' . $path_key; + push @stack, $full_path; + print STDERR "### store to stack: $full_path\n" if ($debug >= 3); + +# my ($f,$nf,$d,$nd) = recurseDir($bpc, $hostname, $backups, $backupNum, $share, $path_key, $shareID) unless ($beenThere->{$key}); +# +# $nr_files += $f; +# $new_files += $nf; +# $nr_dirs += $d; +# $new_dirs += $nd; + + } else { + $nr_files++; + } + } + + print STDERR "## STACK ",join(", ", @stack),"\n" if ($debug >= 2); + + while ( my $dir = shift @stack ) { + my ($f,$nf,$d,$nd, $s) = recurseDir($bpc, $hostname, $files, $backupNum, $share, $dir, $shareID); + print STDERR "# $dir f: $f nf: $nf d: $d nd: $nd\n" if ($debug >= 1); + $nr_files += $f; + $new_files += $nf; + $nr_dirs += $d; + $new_dirs += $nd; + $size += $s; + } + } + + return ($nr_files, $new_files, $nr_dirs, $new_dirs, $size); }