--- trunk2/lib/WebPAC.pm 2004/06/20 15:49:09 373 +++ trunk2/lib/WebPAC.pm 2004/09/21 19:08:11 459 @@ -9,6 +9,7 @@ use XML::Simple; use Template; use Log::Log4perl qw(get_logger :levels); +use Time::HiRes qw(time); use Data::Dumper; @@ -29,16 +30,19 @@ =head2 new -This will create new instance of WebPAC using configuration specified by C. +Create new instance of WebPAC using configuration specified by C. my $webpac = new WebPAC( config_file => 'name.conf', - [code_page => 'ISO-8859-2',] + code_page => 'ISO-8859-2', + low_mem => 1, ); Default C is C. -It will also read configuration files +Default is not to use C options (see L below). + +This method will also read configuration files C (used by indexer and Web font-end) and configuration file specified by C which describes databases to be indexed. @@ -59,6 +63,8 @@ my $self = {@_}; bless($self, $class); + $self->{'start_t'} = time(); + my $log_file = $self->{'log'} || "log.conf"; Log::Log4perl->init($log_file); @@ -106,6 +112,32 @@ EVAL_PERL => 1, ); + # running with low_mem flag? well, use DBM::Deep then. + if ($self->{'low_mem'}) { + $log->info("running with low_mem which impacts performance (<32 Mb memory usage)"); + + my $db_file = "data.db"; + + if (-e $db_file) { + unlink $db_file or $log->logdie("can't remove '$db_file' from last run"); + $log->debug("removed '$db_file' from last run"); + } + + require DBM::Deep; + + my $db = new DBM::Deep $db_file; + + $log->logdie("DBM::Deep error: $!") unless ($db); + + if ($db->error()) { + $log->logdie("can't open '$db_file' under low_mem: ",$db->error()); + } else { + $log->debug("using file '$db_file' for DBM::Deep"); + } + + $self->{'db'} = $db; + } + return $self; } @@ -116,12 +148,16 @@ $webpac->open_isis( filename => '/data/ISIS/ISIS', code_page => '852', - limit_mfn => '500', + limit_mfn => 500, + start_mfn => 6000, lookup => [ ... ], ); By default, ISIS code page is assumed to be C<852>. +If optional parametar C is set, this will be first MFN to read +from database (so you can skip beginning of your database if you need to). + If optional parametar C is set, it will read just 500 records from database in example above. @@ -149,6 +185,12 @@ $log->logcroak("need filename") if (! $arg->{'filename'}); my $code_page = $arg->{'code_page'} || '852'; + $log->logdie("can't find database ",$arg->{'filename'}) unless (glob($arg->{'filename'}.'.*')); + + # store data in object + $self->{'isis_filename'} = $arg->{'filename'}; + $self->{'isis_code_page'} = $code_page; + use OpenIsis; #$self->{'isis_code_page'} = $code_page; @@ -157,17 +199,31 @@ my $cp = Text::Iconv->new($code_page,$self->{'code_page'}); $log->info("reading ISIS database '",$arg->{'filename'},"'"); + $log->debug("isis code page: $code_page"); my $isis_db = OpenIsis::open($arg->{'filename'}); my $maxmfn = OpenIsis::maxRowid( $isis_db ) || 1; + my $startmfn = 1; - $maxmfn = $self->{limit_mfn} if ($self->{limit_mfn}); + if (my $s = $self->{'start_mfn'}) { + $log->info("skipping to MFN $s"); + $startmfn = $s; + } else { + $self->{'start_mfn'} = $startmfn; + } - $log->info("processing $maxmfn records..."); + $maxmfn = $startmfn + $self->{limit_mfn} if ($self->{limit_mfn}); + + $log->info("processing ",($maxmfn-$startmfn)." records..."); # read database - for (my $mfn = 1; $mfn <= $maxmfn; $mfn++) { + for (my $mfn = $startmfn; $mfn <= $maxmfn; $mfn++) { + + + $log->debug("mfn: $mfn\n"); + + my $rec; # read record my $row = OpenIsis::read( $isis_db, $mfn ); @@ -186,19 +242,34 @@ $val = $l; } - push @{$self->{'data'}->{$mfn}->{$k}}, $val; + push @{$rec->{$k}}, $val; } + } else { + push @{$rec->{'000'}}, $mfn; } } + $log->confess("record $mfn empty?") unless ($rec); + + # store + if ($self->{'low_mem'}) { + $self->{'db'}->put($mfn, $rec); + } else { + $self->{'data'}->{$mfn} = $rec; + } + # create lookup - my $rec = $self->{'data'}->{$mfn}; $self->create_lookup($rec, @{$arg->{'lookup'}}); + $self->progress_bar($mfn,$maxmfn); + } - $self->{'current_mfn'} = 1; + $self->{'current_mfn'} = -1; + $self->{'last_pcnt'} = 0; + + $log->debug("max mfn: $maxmfn"); # store max mfn and return it. return $self->{'max_mfn'} = $maxmfn; @@ -218,7 +289,15 @@ my $log = $self->_get_logger(); - my $mfn = $self->{'current_mfn'}++ || $log->logconfess("it seems that you didn't load database!"); + $log->logconfess("it seems that you didn't load database!") unless ($self->{'current_mfn'}); + + if ($self->{'current_mfn'} == -1) { + $self->{'current_mfn'} = $self->{'start_mfn'}; + } else { + $self->{'current_mfn'}++; + } + + my $mfn = $self->{'current_mfn'}; if ($mfn > $self->{'max_mfn'}) { $self->{'current_mfn'} = $self->{'max_mfn'}; @@ -226,7 +305,99 @@ return; } - return $self->{'data'}->{$mfn}; + $self->progress_bar($mfn,$self->{'max_mfn'}); + + if ($self->{'low_mem'}) { + return $self->{'db'}->get($mfn); + } else { + return $self->{'data'}->{$mfn}; + } +} + +=head2 mfn + +Returns current record number (MFN). + + print $webpac->mfn; + +=cut + +sub mfn { + my $self = shift; + return $self->{'current_mfn'}; +} + +=head2 progress_bar + +Draw progress bar on STDERR. + + $webpac->progress_bar($current, $max); + +=cut + +sub progress_bar { + my $self = shift; + + my ($curr,$max) = @_; + + my $log = $self->_get_logger(); + + $log->logconfess("no current value!") if (! $curr); + $log->logconfess("no maximum value!") if (! $max); + + if ($curr > $max) { + $max = $curr; + $log->debug("overflow to $curr"); + } + + $self->{'last_pcnt'} ||= 1; + + my $p = int($curr * 100 / $max) || 1; + + # reset on re-run + if ($p < $self->{'last_pcnt'}) { + $self->{'last_pcnt'} = $p; + $self->{'last_t'} = time(); + $self->{'last_curr'} = undef; + } + + $self->{'last_t'} ||= time(); + + if ($p != $self->{'last_pcnt'}) { + + my $last_curr = $self->{'last_curr'} || $curr; + my $t = time(); + my $rate = ($curr - $last_curr) / (($t - $self->{'last_t'} || 1)); + my $eta = ($max-$curr) / ($rate || 1); + printf STDERR ("%5d [%-38s] %-5d %0.1f/s %s\r",$curr,"=" x ($p/3)."$p%>", $max, $rate, $self->fmt_time($eta)); + $self->{'last_pcnt'} = $p; + $self->{'last_t'} = time(); + $self->{'last_curr'} = $curr; + } + print STDERR "\n" if ($p == 100); +} + +=head2 fmt_time + +Format time (in seconds) for display. + + print $webpac->fmt_time(time()); + +This method is called by L to display remaining time. + +=cut + +sub fmt_time { + my $self = shift; + + my $t = shift || 0; + my $out = ""; + + my ($ss,$mm,$hh) = gmtime($t); + $out .= "${hh}h" if ($hh); + $out .= sprintf("%02d:%02d", $mm,$ss); + $out .= " " if ($hh == 0); + return $out; } =head2 open_import_xml @@ -252,17 +423,21 @@ $self->{'tag'} = $type2tag{$type_base}; - $log->debug("using type '",$self->{'type'},"' tag <",$self->{'tag'},">") if ($self->{'debug'}); + $log->info("using type '",$self->{'type'},"' tag <",$self->{'tag'},">"); my $f = "./import_xml/".$self->{'type'}.".xml"; $log->logconfess("import_xml file '$f' doesn't exist!") if (! -e "$f"); - $log->debug("reading '$f'") if ($self->{'debug'}); + $log->info("reading '$f'"); + + $self->{'import_xml_file'} = $f; $self->{'import_xml'} = XMLin($f, ForceArray => [ $self->{'tag'}, 'config', 'format' ], ); + $log->debug("import xml is ",sub { Dumper($self->{'import_xml'}) }); + } =head2 create_lookup @@ -284,21 +459,23 @@ $log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); foreach my $i (@_) { - if ($i->{'eval'}) { - my $eval = $self->fill_in($rec,$i->{'eval'}); - my $key = $self->fill_in($rec,$i->{'key'}); - my @val = $self->fill_in($rec,$i->{'val'}); - if ($key && @val && eval $eval) { + $log->logconfess("need key") unless defined($i->{'key'}); + $log->logconfess("need val") unless defined($i->{'val'}); + + if (defined($i->{'eval'})) { + # eval first, so we can skip fill_in for key and val + my $eval = $self->fill_in($rec,$i->{'eval'}) || next; + if ($self->_eval($eval)) { + my $key = $self->fill_in($rec,$i->{'key'}) || next; + my @val = $self->fill_in($rec,$i->{'val'}) || next; $log->debug("stored $key = ",sub { join(" | ",@val) }); push @{$self->{'lookup'}->{$key}}, @val; } } else { - my $key = $self->fill_in($rec,$i->{'key'}); - my @val = $self->fill_in($rec,$i->{'val'}); - if ($key && @val) { - $log->debug("stored $key = ",sub { join(" | ",@val) }); - push @{$self->{'lookup'}->{$key}}, @val; - } + my $key = $self->fill_in($rec,$i->{'key'}) || next; + my @val = $self->fill_in($rec,$i->{'val'}) || next; + $log->debug("stored $key = ",sub { join(" | ",@val) }); + push @{$self->{'lookup'}->{$key}}, @val; } } } @@ -329,6 +506,7 @@ if ($$rec->{$f}) { return '' if (! $$rec->{$f}->[$i]); + no strict 'refs'; if ($sf && $$rec->{$f}->[$i]->{$sf}) { $$found++ if (defined($$found)); return $$rec->{$f}->[$i]->{$sf}; @@ -369,6 +547,9 @@ This function B perform parsing of format to inteligenty skip delimiters before fields which aren't used. +This method will automatically decode UTF-8 string to local code page +if needed. + =cut sub fill_in { @@ -384,6 +565,10 @@ # FIXME remove for speedup? $log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); + if (utf8::is_utf8($format)) { + $format = $self->_x($format); + } + my $found = 0; my $eval_code; @@ -481,7 +666,7 @@ $i = 0 if (! $i); - my $format = $self->{'utf2cp'}->convert($format_utf8) || $log->logconfess("can't convert '$format_utf8' from UTF-8 to ",$self->{'code_page'}); + my $format = $self->_x($format_utf8) || $log->logconfess("can't convert '$format_utf8' from UTF-8 to ",$self->{'code_page'}); my @out; @@ -594,6 +779,31 @@ return @arr; } +=head2 sort_arr + +Sort array ignoring case and html in data + + my @sorted = $webpac->sort_arr(@unsorted); + +=cut + +sub sort_arr { + my $self = shift; + + my $log = $self->_get_logger(); + + # FIXME add Schwartzian Transformation? + + my @sorted = sort { + $a =~ s#<[^>]+/*>##; + $b =~ s#<[^>]+/*>##; + lc($b) cmp lc($a) + } @_; + $log->debug("sorted values: ",sub { join(", ",@sorted) }); + + return @sorted; +} + =head2 data_structure @@ -602,6 +812,10 @@ my @ds = $webpac->data_structure($rec); +This method will also set C<$webpac->{'currnet_filename'}> if there is + tag in C and C<$webpac->{'headline'}> if there is + tag. + =cut sub data_structure { @@ -612,6 +826,9 @@ my $rec = shift; $log->logconfess("need HASH as first argument!") if ($rec !~ /HASH/o); + undef $self->{'currnet_filename'}; + undef $self->{'headline'}; + my @sorted_tags; if ($self->{tags_by_order}) { @sorted_tags = @{$self->{tags_by_order}}; @@ -643,19 +860,73 @@ } next if (! @v); - # does tag have type? - if ($tag->{'type'}) { - push @{$row->{$tag->{'type'}}}, @v; - } else { - push @{$row->{'display'}}, @v; - push @{$row->{'swish'}}, @v; + if ($tag->{'sort'}) { + @v = $self->sort_arr(@v); + $log->warn("sort within tag is usually not what you want!"); + } + + # use format? + if ($tag->{'format_name'}) { + @v = map { $self->apply_format($tag->{'format_name'},$tag->{'format_delimiter'},$_) } @v; + } + + if ($field eq 'filename') { + $self->{'current_filename'} = join('',@v); + $log->debug("filename: ",$self->{'current_filename'}); + } elsif ($field eq 'headline') { + $self->{'headline'} .= join('',@v); + $log->debug("headline: ",$self->{'headline'}); + next; # don't return headline in data_structure! + } + + # delimiter will join repeatable fields + if ($tag->{'delimiter'}) { + @v = ( join($tag->{'delimiter'}, @v) ); + } + + # default types + my @types = qw(display swish); + # override by type attribute + @types = ( $tag->{'type'} ) if ($tag->{'type'}); + + foreach my $type (@types) { + # append to previous line? + $log->debug("type: $type ",sub { join(" ",@v) }, $row->{'append'} || 'no append'); + if ($tag->{'append'}) { + + # I will delimit appended part with + # delimiter (or ,) + my $d = $tag->{'delimiter'}; + # default delimiter + $d ||= ", "; + + my $last = pop @{$row->{$type}}; + $d = "" if (! $last); + $last .= $d . join($d, @v); + push @{$row->{$type}}, $last; + + } else { + push @{$row->{$type}}, @v; + } } + } if ($row) { $row->{'tag'} = $field; + + # TODO: name_sigular, name_plural + my $name = $self->{'import_xml'}->{'indexer'}->{$field}->{'name'}; + $row->{'name'} = $name ? $self->_x($name) : $field; + + # post-sort all values in field + if ($self->{'import_xml'}->{'indexer'}->{$field}->{'sort'}) { + $log->warn("sort at field tag not implemented"); + } + push @ds, $row; + $log->debug("row $field: ",sub { Dumper($row) }); } @@ -694,6 +965,79 @@ return $out; } +=head2 output_file + +Create output from in-memory data structure using Template Toolkit template +to a file. + + $webpac->output_file( + file => 'out.txt', + template => 'text.tt', + data => @ds + ); + +=cut + +sub output_file { + my $self = shift; + + my $args = {@_}; + + my $log = $self->_get_logger(); + + my $file = $args->{'file'} || $log->logconfess("need file name"); + + $log->debug("creating file ",$file); + + open(my $fh, ">", $file) || $log->logdie("can't open output file '$file': $!"); + print $fh $self->output( + template => $args->{'template'}, + data => $args->{'data'}, + ) || $log->logdie("print: $!"); + close($fh) || $log->logdie("close: $!"); +} + +=head2 apply_format + +Apply format specified in tag with C and +C. + + my $text = $webpac->apply_format($format_name,$format_delimiter,$data); + +Formats can contain C if you need them. + +=cut + +sub apply_format { + my $self = shift; + + my ($name,$delimiter,$data) = @_; + + my $log = $self->_get_logger(); + + if (! $self->{'import_xml'}->{'format'}->{$name}) { + $log->warn(" is not defined in ",$self->{'import_xml_file'}); + return $data; + } + + $log->warn("no delimiter for format $name") if (! $delimiter); + + my $format = $self->_x($self->{'import_xml'}->{'format'}->{$name}->{'content'}) || $log->logdie("can't find format '$name'"); + + my @data = split(/\Q$delimiter\E/, $data); + + my $out = sprintf($format, @data); + $log->debug("using format $name [$format] on $data to produce: $out"); + + if ($out =~ m/$LOOKUP_REGEX/o) { + return $self->lookup($out); + } else { + return $out; + } + +} + + # # # @@ -746,11 +1090,36 @@ return $va <=> $vb; } +=head2 _get_logger + +Get C object with a twist: domains are defined for each +method + + my $log = $webpac->_get_logger(); + +=cut + sub _get_logger { my $self = shift; - my @c = caller(1); - return get_logger($c[3]); + my $name = (caller(1))[3] || caller; + return get_logger($name); +} + +=head2 _x + +Convert string from UTF-8 to code page defined in C. + + my $text = $webpac->_x('utf8 text'); + +=cut + +sub _x { + my $self = shift; + my $utf8 = shift || return; + + return $self->{'utf2cp'}->convert($utf8) || + $self->_get_logger()->logwarn("can't convert '$utf8'"); } # @@ -771,6 +1140,37 @@ also use method names, and not only classes (which are just few) to filter logging. + +=head1 MEMORY USAGE + +C options is double-edged sword. If enabled, WebPAC +will run on memory constraint machines (which doesn't have enough +physical RAM to create memory structure for whole source database). + +If your machine has 512Mb or more of RAM and database is around 10000 records, +memory shouldn't be an issue. If you don't have enough physical RAM, you +might consider using virtual memory (if your operating system is handling it +well, like on FreeBSD or Linux) instead of dropping to L to handle +parsed structure of ISIS database (this is what C option does). + +Hitting swap at end of reading source database is probably o.k. However, +hitting swap before 90% will dramatically decrease performance and you will +be better off with C and using rest of availble memory for +operating system disk cache (Linux is particuallary good about this). +However, every access to database record will require disk access, so +generation phase will be slower 10-100 times. + +Parsed structures are essential - you just have option to trade RAM memory +(which is fast) for disk space (which is slow). Be sure to have planty of +disk space if you are using C and thus L. + +However, when WebPAC is running on desktop machines (or laptops :-), it's +highly undesireable for system to start swapping. Using C option can +reduce WecPAC memory usage to around 64Mb for same database with lookup +fields and sorted indexes which stay in RAM. Performance will suffer, but +memory usage will really be minimal. It might be also more confortable to +run WebPAC reniced on those machines. + =cut 1;