1 package MojoFacets::Data;
6 use base 'Mojolicious::Controller';
8 use Data::Dump qw(dump);
15 use Time::HiRes qw(time);
23 my $path = $self->app->home->rel_dir('data');
24 die "no data dir $path" unless -d $path;
29 my $file = $File::Find::name;
30 if ( -f $file && $file =~ m/\.(js(on)?|txt)$/ ) {
31 $file =~ s/$path\/*//;
33 } elsif ( -f $file && $file =~ m/([^\/]+)\.edits\/(\d+\.\d+.+)/ ) {
34 push @{ $edits->{$1} }, $2
36 warn "IGNORE: $file\n";
40 @files = sort { lc $a cmp lc $b } @files;
42 $size->{$_} = -s "$path/$_" foreach @files;
44 if ( my $save_path = $self->session('save_path') ) {
45 $self->session( 'save_path' => 0 )
46 if ! defined $loaded->{$save_path};
54 dump_path => { map { $_ => $self->_dump_path($_) } @files },
60 my ( $self, $name ) = @_;
61 my $dir = $self->app->home->rel_dir('data');
64 return '/tmp/mojo_facets.' . $name . '.storable';
68 my ( $self, $path ) = @_;
70 my $dump_path = $self->_dump_path( $path );
71 my $first_load = ! -e $dump_path;
72 warn "save loaded to $dump_path";
73 my $info = $loaded->{$path};
74 store $info, $dump_path;
77 my $mtime = $loaded->{$path}->{mtime};
78 utime $mtime, $mtime, $dump_path;
79 warn "sync time to $path at $mtime\n";
82 warn $dump_path, ' ', -s $dump_path, " bytes\n";
91 my $nr_items = $#{ $_[0] } + 1;
93 warn "__stats $nr_items\n";
95 foreach my $e ( @{ $_[0] } ) {
96 foreach my $n ( keys %$e ) {
97 $stats->{$n}->{count}++;
99 if ( ref $e->{$n} eq 'ARRAY' ) {
100 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
106 foreach my $x ( @v ) {
107 $stats->{$n}->{numeric}++
108 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
109 $stats->{$n}->{empty}++
110 if length $x == 0; # faster than $x =~ m/^\s*$/;
116 foreach my $n ( keys %$stats ) {
117 my $s = $stats->{$n};
118 next unless defined $s->{array};
119 if ( $s->{array} == $s->{count} ) {
121 if ( $s->{count} == $nr_items ) {
122 warn "check $n for uniqeness\n";
124 foreach my $e ( @{ $_[0] } ) {
125 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
131 $stats->{$n}->{unique} = 1;
132 warn "# $n unique ",dump( $unique );
138 warn "# __stats ",dump($stats);
143 sub _param_or_session {
144 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
149 my $path = $self->_param_or_session('path');
150 warn "stats $path\n";
151 delete $loaded->{$path}->{stats};
152 $self->redirect_to( '/data/columns' );
157 my ( $self, $path ) = @_;
159 my $full_path = $self->app->home->rel_file( 'data/' . $path );
160 die "$full_path $!" unless -r $full_path;
162 my $dump_path = $self->_dump_path( $path );
164 if ( defined $loaded->{$path}->{data} ) {
165 my $mtime = (stat($full_path))[9];
166 return if $loaded->{$path}->{mtime} == $mtime;
167 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
168 } elsif ( -e $dump_path ) {
169 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
170 my $info = retrieve $dump_path;
171 $loaded->{ $path } = $info;
175 # we could use Mojo::JSON here, but it's too slow
176 # $data = from_json read_file $path;
177 my $data = read_file $full_path;
178 warn "# data snippet: ", substr($data,0,200);
180 if ( $path =~ m/\.js/ ) {
181 Encode::_utf8_on($data);
182 $data = from_json $data;
183 } elsif ( $path =~ m/\.txt/ ) {
184 my @lines = split(/\r?\n/, $data);
185 $data = { items => [] };
187 my $header_line = shift @lines;
188 my $multiline = $header_line =~ s/\^//g;
189 @header = split(/\|/, $header_line );
190 warn "# header ", dump( @header );
191 while ( my $line = shift @lines ) {
194 my @v = split(/\|/, $line);
195 while ( @lines && $#v < $#header ) {
199 my @more_v = split(/\|/, $line);
200 if ( $#v + $#more_v > $#header ) {
201 warn "short line: ",dump( @v );
205 $v[ $#v ] .= shift @more_v if @more_v;
206 push @v, @more_v if @more_v;
208 if ( $#v > $#header ) {
209 die "# splice $#header ", dump( @v );
210 @v = splice @v, 0, $#header;
214 foreach my $i ( 0 .. $#v ) {
215 $item->{ $header[$i] || "f_$i" } = [ $v[$i] ];
217 push @{ $data->{items} }, $item;
220 warn "file format unknown $path";
223 my $stats = __stats( $data->{items} );
226 if ( defined $data->{header} ) {
227 if ( ref $data->{header} eq 'ARRAY' ) {
228 @header = @{ $data->{header} };
230 warn "header not array ", dump( $data->{header} );
236 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
237 grep { defined $stats->{$_}->{count} } keys %$stats
241 header => [ @header ],
243 full_path => $full_path,
244 size => -s $full_path,
245 mtime => (stat($full_path))[9],
249 $loaded->{ $path } = $info;
250 $self->_save( $path );
258 my @paths = $self->param('paths');
259 warn "# paths ", dump @paths;
260 $self->_load_path( $_ ) foreach @paths;
262 my $path = $self->param('path') || $self->redirect_to( '/data/index' );
263 warn "# path $path\n";
264 $self->session('path' => $path);
265 $self->_load_path( $path );
267 my $redirect_to = '/data/items';
269 $self->session( 'header' => $loaded->{$path}->{header} );
270 if ( ! defined $loaded->{$path}->{columns} ) {
271 my $columns_path = $self->_permanent_path( 'columns' );
272 if ( -e $columns_path ) {
273 my @columns = map { s/[\r\n]+$//; $_ } read_file $columns_path;
274 $loaded->{$path}->{columns} = [ @columns ];
275 warn "# columns_path $columns_path ",dump(@columns);
277 $loaded->{$path}->{columns} = $loaded->{$path}->{header}
280 $redirect_to = '/data/columns';
282 $self->session( 'columns' => $loaded->{$path}->{columns} );
283 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
284 $self->redirect_to( $redirect_to );
289 my ( $self, $name ) = @_;
290 my $path = $self->session('path') || $self->param('path');
291 $self->redirect_to('/data/index') unless $path;
292 if ( ! defined $loaded->{$path}->{$name} ) {
293 warn "$path $name isn't loaded\n";
294 $self->_load_path( $path );
295 $self->redirect_to('/data/index')
296 unless defined $loaded->{$path}->{$name};
297 if ( ! defined $loaded->{$path}->{stats} ) {
298 warn "rebuild stats for $path\n";
299 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
302 return $loaded->{$path}->{$name};
309 $checked->{$_}++ foreach @_;
310 # warn "# _checked ",dump($checked);
314 sub _permanent_path {
316 my $path = $self->_param_or_session('path');
317 $self->app->home->rel_dir('data') . '/' . join('.', $path, @_);
322 my $path = $self->_param_or_session('path');
323 my $dir = $self->app->home->rel_dir('public') . '/export/';
324 mkdir $dir unless -e $dir;
326 mkdir $dir unless -e $dir;
327 $dir . '/' . join('.', @_);
333 if ( $self->param('columns') ) {
334 my @columns = $self->_param_array('columns');
335 write_file( $self->_permanent_path( 'columns' ), map { "$_\n" } @columns );
336 $self->redirect_to('/data/items');
339 my $stats = $self->_loaded( 'stats' );
342 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
344 foreach my $c ( @{ $self->session( 'header' ) } ) {
345 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
349 message => 'Select columns to display',
351 columns => \@columns,
352 checked => $self->_checked( $self->_param_array('columns') ),
357 my ($self,$name) = @_;
359 my @array = $self->param($name);
360 my $path = $self->session('path');
363 $self->session($name => [ @array ]);
364 $loaded->{$path}->{$name} = [ @array ];
365 } elsif ( my $session = $self->session($name) ) {
366 if ( ref $session eq 'ARRAY' ) {
369 die "$name not array ",dump($session);
372 #warn "# $name ",dump @array;
377 my ($self,$name,$default) = @_;
379 my $scalar = $self->param($name);
381 if ( defined $scalar ) {
382 $self->session($name => $scalar);
384 $scalar = $self->session($name);
387 if ( ! defined $scalar ) {
389 die "no default for $name" unless defined $scalar;
390 $self->session($name => $scalar);
393 warn "# _perm_scalar $name ",dump $scalar;
400 my $name = $self->param('filter_name') || die "name?";
401 my @vals = $self->param('filter_vals');
403 $self->_remove_filter( $name );
405 $self->_filter_on_data( $name, @vals );
406 if ( my $permanent = $self->param('_permanent') ) {
407 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
408 write_file $permanent_path, map { "$_\n" } @vals;
409 warn "permanent filter $permanent_path ", -s $permanent_path;
413 $self->session( 'offset' => 0 );
415 $self->redirect_to('/data/items');
418 sub _filter_on_data {
419 my ( $self, $name, @vals ) = @_;
421 my $path = $self->session('path');
423 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
424 warn "filter $name not found in data set";
428 $filters->{$name} = [ @vals ];
429 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
432 $filter_hash->{$_}++ foreach @vals;
434 #warn "# filter_hash ",dump( $filter_hash );
436 my $items = $self->_loaded('data')->{items};
438 my $include_missing = defined $filter_hash->{_missing};
441 foreach my $i ( 0 .. $#$items ) {
443 if ( defined $items->[$i]->{$name} ) {
444 foreach my $v ( @{ $items->[$i]->{$name} } ) {
445 if ( defined $filter_hash->{ $v } ) {
446 $filtered_items->{$i}++;
449 } elsif ( $include_missing ) {
450 $filtered_items->{$i}++;
454 #warn "# filter $name ",dump($filtered_items);
456 $loaded->{$path}->{filters}->{$name} = $filtered_items;
457 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
461 my ( $self, $all ) = @_;
462 my $data = $self->_loaded( 'data' );
464 return @{ $data->{items} } if $all == 1;
466 my $filters = $self->_current_filters;
468 foreach my $f ( keys %$filters ) {
469 foreach my $n ( @{ $filters->{$f} } ) {
470 $filter_value->{$f}->{$n} = 1;
473 my @items = @{ $data->{items} };
477 foreach my $n ( keys %$filter_value ) {
478 if ( ! exists $i->{$n} ) {
479 if ( defined $filter_value->{$n}->{_missing} ) {
487 # and match any of values in element
489 foreach my $v ( @{ $i->{$n} } ) { # FIXME not array?
490 $have_values ||= 1 if defined $filter_value->{$n}->{$v};
492 if ( ! $have_values ) {
498 } @items if $filter_value;
503 sub _current_filters {
506 $current_filters->{ $_ } = $filters->{ $_ }
508 grep { defined $filters->{ $_ } }
509 @{ $self->_loaded('header') }
511 #warn "# current_filters ",dump($current_filters);
512 return $current_filters;
515 sub _data_sorted_by {
516 my ( $self, $order ) = @_;
518 my $path = $self->session('path');
520 warn "_data_sorted_by $order from $path";
522 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
523 return $loaded->{$path}->{sorted}->{$order};
526 my $data = $self->_loaded( 'data' );
527 my $numeric = $self->_is_numeric($order);
528 my $missing = $numeric ? 0 : '';
529 no warnings qw(numeric);
540 [ $nr++, exists $_->{$order} ? join('', @{$_->{$order}}) : $missing ]
541 } @{ $data->{items} }
544 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
545 #warn "# sorted ",dump( @sorted );
547 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
554 my $path = $self->session('path');
555 $self->redirect_to('/data/index') unless defined $loaded->{ $path };
557 my @columns = $self->_param_array('columns');
558 $self->redirect_to('/data/columns') unless @columns;
559 my $order = $self->_param_scalar('order', $columns[0]);
560 my $sort = $self->_param_scalar('sort', 'a');
561 my $offset = $self->_param_scalar('offset', 0);
562 my $limit = $self->_param_scalar('limit', 20);
563 $self->_param_scalar('show', 'table');
565 # fix offset when changing limit
566 $offset = int( $offset / $limit ) * $limit;
568 my $sorted = $self->_data_sorted_by( $order );
572 foreach my $name ( keys %$filters ) {
573 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
574 warn "skip filter $name not found in $path\n";
577 push @filter_names, $name;
579 warn "filter_names ",dump( @filter_names );
580 foreach my $name ( @filter_names ) {
581 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
582 $self->_filter_on_data( $name, @{ $filters->{$name} } );
586 my $all_filters = join(' ',sort @filter_names,'order:',$order);
588 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
590 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
592 my $path_filters = $loaded->{$path}->{filters};
594 warn "create combined filter for $all_filters\n";
597 foreach my $i ( 0 .. $#$sorted ) {
598 my $pos = $sorted->[$i];
600 if ( $#filter_names == -1 ) {
601 push @filtered, $pos;
606 foreach ( @filter_names ) {
607 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
611 push @filtered, $pos;
614 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
617 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
618 if defined $loaded->{$path}->{filtered}->{$all_filters};
620 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
623 my $data = $self->_loaded('data');
624 my $from_end = $sort eq 'd' ? $#$filtered : 0;
625 foreach ( 0 .. $limit ) {
626 my $i = $_ + $offset;
627 last unless defined $filtered->[$i];
628 $i = $from_end - $i if $from_end;
629 my $id = $filtered->[$i];
631 my $item = $data->{items}->[ $id ];
632 $item->{_row_id} ||= $id;
635 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
641 sorted => $sorted_items,
642 columns => [ @columns ],
643 rows => $#$filtered + 1,
644 numeric => { map { $_, $self->_is_numeric($_) } @columns },
645 filters => $self->_current_filters,
653 $self->session('order', $self->param('order'));
654 $self->session('sort', $self->param('sort'));
655 $self->redirect_to('/data/items');
659 my ( $self, $name ) = @_;
661 my $stats = $self->_loaded( 'stats' );
663 # sort facet numerically if more >50% elements are numeric
664 my $count = $stats->{$name}->{count};
665 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
666 defined $stats->{$name}->{numeric} &&
667 $stats->{$name}->{numeric} > $count / 2;
671 my ($self,$name) = @_;
672 warn "_remove_filter $name\n";
674 my $path = $self->session('path');
676 delete $filters->{$name};
677 delete $loaded->{$path}->{filters}->{$name};
678 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
682 keys %{ $loaded->{$path}->{filtered} }
684 delete $loaded->{$path}->{filtered}->{$_};
685 warn "remove filtered cache $_";
692 my $path = $self->session('path') || $self->redirect_to( '/data/index' );
694 if ( my $name = $self->param('remove') ) {
695 $self->_remove_filter( $name );
696 $self->redirect_to( '/data/items' );
700 my $name = $self->param('name') || die "no name";
702 my $all = $self->_param_scalar('all', 1);
703 my $data = $self->_loaded('data');
705 my $filters = $self->_current_filters;
706 my $all_filters = join(' ',sort keys %$filters,'order:',$self->session('order'));
707 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
708 if defined $loaded->{$path}->{filtered}->{$all_filters};
710 if ( ! $filtered || $all ) {
711 $filtered = [ 0 .. $#{ $data->{items} } ];
712 warn "filter all values\n";
714 warn "filter using $all_filters\n";
717 foreach my $i ( @$filtered ) {
718 my $item = $data->{items}->[$i];
719 if ( ! exists $item->{$name} ) {
720 $facet->{ _missing }++;
721 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
722 $facet->{$_}++ foreach @{ $item->{$name} };
724 $facet->{ $item->{$name} }++;
728 # warn "# facet $name ",dump $facet;
733 : defined $filters->{$name} ? @{ $filters->{$name} }
736 $checked = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
738 my $numeric = $self->_is_numeric($name);
740 my $sort = $self->param('sort');
741 # sort numeric facets with more than 5 values ascending
742 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
744 @facet_names = sort {
746 if ( $sort eq 'a' ) {
747 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
748 } elsif ( $sort eq 'd' ) {
749 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
750 } elsif ( $sort eq 'c' ) {
751 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
753 warn "unknown sort: $sort";
756 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
760 $self->render( name => $name, facet => $facet, checked => $checked,
761 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
767 my $new_content = $self->param('new_content');
769 my $i = $self->param('_row_id');
770 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
771 my $path = $self->param('path') || die "no path";
772 my $name = $self->param('name') || die "no name";
773 my $status = 200; # 200 = OK, 201 = Created
775 my $data = $self->_loaded('data');
777 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
778 $new_content =~ s/^\s+//s;
779 $new_content =~ s/\s+$//s;
781 if ( $new_content =~ /\xB6/ ) { # para
782 $v = [ split(/\s*\xB6\s*/, $new_content) ];
784 $v = [ $new_content ];
787 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
790 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
796 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
798 time => $self->param('time') || time(),
799 user => $self->param('user') || $ENV{'LOGNAME'},
801 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
802 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
803 keys %{ $loaded->{$path}->{stats} }
806 my $edit_path = $self->_permanent_path( 'edits' );
807 mkdir $edit_path unless -d $edit_path;
808 $edit_path .= '/' . $edit->{time};
809 store $edit, $edit_path;
810 utime $edit->{time}, $edit->{time}, $edit_path;
811 warn "# $edit_path ", dump($edit);
813 warn "# edit $path $i $old -> $new\n";
814 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
816 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
817 delete $loaded->{$path}->{sorted}->{$name};
818 warn "# invalidate $path sorted $name\n";
821 foreach ( grep { m/$name/ } keys %{ $loaded->{$path}->{filtered} } ) {
822 delete $loaded->{$path}->{filtered}->{$_};
823 warn "# invalidate $path filtered $_\n";
826 $status = 201; # created
827 $self->session('save_path' => $path);
829 $new_content = join("\xB6",@$v);
832 warn "# unchanged $path $i $old\n";
836 $new_content = "$path $i $name doesn't exist\n";
840 warn "# edit $status ", dump $new_content;
844 new_content => scalar $new_content,
851 my $path = $self->_param_or_session('path');
852 my $dump_path = $self->_save( $path );
853 $self->session('save_path' => 0);
855 $self->redirect_to( '/data/items' );
860 $self->render( export => [
861 map { s{^.+/public/export/}{}; $_ }
862 glob( $self->_export_path . '*' )
868 grep { defined $loaded->{$_}->{data} }