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 foreach my $e ( @{ $_[0] } ) {
94 foreach my $n ( keys %$e ) {
95 $stats->{$n}->{count}++;
97 if ( ref $e->{$n} eq 'ARRAY' ) {
98 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
104 foreach my $x ( @v ) {
105 $stats->{$n}->{numeric}++
106 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
107 $stats->{$n}->{empty}++
108 if length $x == 0; # faster than $x =~ m/^\s*$/;
114 foreach my $n ( keys %$stats ) {
115 my $s = $stats->{$n};
116 next unless defined $s->{array};
117 if ( $s->{array} == $s->{count} ) {
119 if ( $s->{count} == $nr_items ) {
120 warn "check $n for uniqeness\n";
122 foreach my $e ( @{ $_[0] } ) {
123 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
129 $stats->{$n}->{unique} = 1;
130 warn "# $n unique ",dump( $unique );
136 warn "# __stats ",dump($stats);
144 my ( $self, $path ) = @_;
146 my $full_path = $self->app->home->rel_file( 'data/' . $path );
147 die "$full_path $!" unless -r $full_path;
149 my $dump_path = $self->_dump_path( $path );
151 if ( defined $loaded->{$path}->{data} ) {
152 my $mtime = (stat($full_path))[9];
153 return if $loaded->{$path}->{mtime} == $mtime;
154 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
155 } elsif ( -e $dump_path ) {
156 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
157 my $info = retrieve $dump_path;
158 $loaded->{ $path } = $info;
162 # we could use Mojo::JSON here, but it's too slow
163 # $data = from_json read_file $path;
164 my $data = read_file $full_path;
165 warn "# data snippet: ", substr($data,0,200);
167 if ( $path =~ m/\.js/ ) {
168 Encode::_utf8_on($data);
169 $data = from_json $data;
170 } elsif ( $path =~ m/\.txt/ ) {
171 my @lines = split(/\r?\n/, $data);
172 $data = { items => [] };
174 my $header_line = shift @lines;
175 my $multiline = $header_line =~ s/\^//g;
176 @header = split(/\|/, $header_line );
177 warn "# header ", dump( @header );
178 while ( my $line = shift @lines ) {
181 my @v = split(/\|/, $line);
182 while ( @lines && $#v < $#header ) {
186 my @more_v = split(/\|/, $line);
187 if ( $#v + $#more_v > $#header ) {
188 warn "short line: ",dump( @v );
192 $v[ $#v ] .= shift @more_v if @more_v;
193 push @v, @more_v if @more_v;
195 if ( $#v > $#header ) {
196 die "# splice $#header ", dump( @v );
197 @v = splice @v, 0, $#header;
201 foreach my $i ( 0 .. $#v ) {
202 $item->{ $header[$i] || "f_$i" } = [ $v[$i] ];
204 push @{ $data->{items} }, $item;
207 warn "file format unknown $path";
210 my $stats = __stats( $data->{items} );
213 if ( defined $data->{header} ) {
214 if ( ref $data->{header} eq 'ARRAY' ) {
215 @header = @{ $data->{header} };
217 warn "header not array ", dump( $data->{header} );
223 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
224 grep { defined $stats->{$_}->{count} } keys %$stats
228 header => [ @header ],
230 full_path => $full_path,
231 size => -s $full_path,
232 mtime => (stat($full_path))[9],
236 $loaded->{ $path } = $info;
237 $self->_save( $path );
245 my @paths = $self->param('paths');
246 warn "# paths ", dump @paths;
247 $self->_load_path( $_ ) foreach @paths;
249 my $path = $self->param('path') || $self->redirect_to( '/data/index' );
250 warn "# path $path\n";
251 $self->session('path' => $path);
252 $self->_load_path( $path );
254 $self->session( 'header' => $loaded->{$path}->{header} );
255 if ( ! defined $loaded->{$path}->{columns} ) {
256 $self->session( 'columns' => $loaded->{$path}->{header} );
257 $self->session( 'order' => $loaded->{$path}->{header}->[0] );
258 $self->redirect_to( '/data/columns' );
260 $self->session( 'columns' => $loaded->{$path}->{columns} );
261 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
262 $self->redirect_to( '/data/items' );
268 my ( $self, $name ) = @_;
269 my $path = $self->session('path') || $self->param('path');
270 $self->redirect_to('/data/index') unless $path;
271 if ( ! defined $loaded->{$path}->{$name} ) {
272 warn "$path $name isn't loaded\n";
273 $self->_load_path( $path );
274 $self->redirect_to('/data/index')
275 unless defined $loaded->{$path}->{$name};
277 return $loaded->{$path}->{$name};
284 $checked->{$_}++ foreach @_;
285 # warn "# _checked ",dump($checked);
293 if ( $self->param('columns') ) {
294 $self->_perm_array('columns');
295 $self->redirect_to('/data/items');
298 my $stats = $self->_loaded( 'stats' ); # || $self->redirect_to( '/data/index' );
301 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
303 foreach my $c ( @{ $self->session( 'header' ) } ) {
304 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
308 message => 'Select columns to display',
310 columns => \@columns,
311 checked => $self->_checked( $self->_perm_array('columns') ),
316 my ($self,$name) = @_;
318 my @array = $self->param($name);
319 my $path = $self->session('path');
322 $self->session($name => [ @array ]);
323 $loaded->{$path}->{$name} = [ @array ];
324 } elsif ( my $session = $self->session($name) ) {
325 if ( ref $session eq 'ARRAY' ) {
328 die "$name not array ",dump($session);
331 #warn "# $name ",dump @array;
336 my ($self,$name,$default) = @_;
338 my $scalar = $self->param($name);
340 if ( defined $scalar ) {
341 $self->session($name => $scalar);
343 $scalar = $self->session($name);
346 if ( ! defined $scalar ) {
348 die "no default for $name" unless defined $scalar;
349 $self->session($name => $scalar);
352 warn "# _perm_scalar $name ",dump $scalar;
359 my $name = $self->param('filter_name') || die "name?";
360 my @vals = $self->param('filter_vals');
362 $self->_remove_filter( $name );
363 $self->_filter_on_data( $name, @vals ) if @vals;
365 $self->session( 'offset' => 0 );
367 $self->redirect_to('/data/items');
370 sub _filter_on_data {
371 my ( $self, $name, @vals ) = @_;
373 my $path = $self->session('path');
375 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
376 warn "filter $name not found in data set";
380 $filters->{$name} = [ @vals ];
381 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
384 $filter_hash->{$_}++ foreach @vals;
386 #warn "# filter_hash ",dump( $filter_hash );
388 my $items = $self->_loaded('data')->{items};
390 my $include_missing = defined $filter_hash->{_missing};
393 foreach my $i ( 0 .. $#$items ) {
395 if ( defined $items->[$i]->{$name} ) {
396 foreach my $v ( @{ $items->[$i]->{$name} } ) {
397 if ( defined $filter_hash->{ $v } ) {
398 $filtered_items->{$i}++;
401 } elsif ( $include_missing ) {
402 $filtered_items->{$i}++;
406 #warn "# filter $name ",dump($filtered_items);
408 $loaded->{$path}->{filters}->{$name} = $filtered_items;
409 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
413 my ( $self, $all ) = @_;
414 my $data = $self->_loaded( 'data' );
416 return @{ $data->{items} } if $all == 1;
418 my $filters = $self->_current_filters;
420 foreach my $f ( keys %$filters ) {
421 foreach my $n ( @{ $filters->{$f} } ) {
422 $filter_value->{$f}->{$n} = 1;
425 my @items = @{ $data->{items} };
429 foreach my $n ( keys %$filter_value ) {
430 if ( ! exists $i->{$n} ) {
431 if ( defined $filter_value->{$n}->{_missing} ) {
439 # and match any of values in element
441 foreach my $v ( @{ $i->{$n} } ) { # FIXME not array?
442 $have_values ||= 1 if defined $filter_value->{$n}->{$v};
444 if ( ! $have_values ) {
450 } @items if $filter_value;
455 sub _current_filters {
458 $current_filters->{ $_ } = $filters->{ $_ }
460 grep { defined $filters->{ $_ } }
461 @{ $self->_loaded('header') }
463 #warn "# current_filters ",dump($current_filters);
464 return $current_filters;
467 sub _data_sorted_by {
468 my ( $self, $order ) = @_;
470 my $path = $self->session('path');
472 warn "_data_sorted_by $order from $path";
474 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
475 return $loaded->{$path}->{sorted}->{$order};
478 my $data = $self->_loaded( 'data' );
479 my $numeric = $self->_is_numeric($order);
480 my $missing = $numeric ? 0 : '';
481 no warnings qw(numeric);
492 [ $nr++, exists $_->{$order} ? join('', @{$_->{$order}}) : $missing ]
493 } @{ $data->{items} }
496 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
497 #warn "# sorted ",dump( @sorted );
499 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
506 my $path = $self->session('path');
507 $self->redirect_to('/data/index') unless defined $loaded->{ $path };
509 my @columns = $self->_perm_array('columns');
510 $self->redirect_to('/data/columns') unless @columns;
511 my $order = $self->_perm_scalar('order', $columns[0]);
512 my $sort = $self->_perm_scalar('sort', 'a');
513 my $offset = $self->_perm_scalar('offset', 0);
514 my $limit = $self->_perm_scalar('limit', 20);
515 $self->_perm_scalar('show', 'table');
517 # fix offset when changing limit
518 $offset = int( $offset / $limit ) * $limit;
520 my $sorted = $self->_data_sorted_by( $order );
524 foreach my $name ( keys %$filters ) {
525 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
526 warn "skip filter $name not found in $path\n";
529 push @filter_names, $name;
531 warn "filter_names ",dump( @filter_names );
532 foreach my $name ( @filter_names ) {
533 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
534 $self->_filter_on_data( $name, @{ $filters->{$name} } );
538 my $all_filters = join(' ',sort @filter_names,'order:',$order);
540 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
542 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
544 my $path_filters = $loaded->{$path}->{filters};
546 warn "create combined filter for $all_filters\n";
549 foreach my $i ( 0 .. $#$sorted ) {
550 my $pos = $sorted->[$i];
552 if ( $#filter_names == -1 ) {
553 push @filtered, $pos;
558 foreach ( @filter_names ) {
559 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
563 push @filtered, $pos;
566 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
569 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
570 if defined $loaded->{$path}->{filtered}->{$all_filters};
572 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
575 my $data = $self->_loaded('data');
576 my $from_end = $sort eq 'd' ? $#$filtered : 0;
577 foreach ( 0 .. $limit ) {
578 my $i = $_ + $offset;
579 last unless defined $filtered->[$i];
580 $i = $from_end - $i if $from_end;
581 my $id = $filtered->[$i];
583 my $item = $data->{items}->[ $id ];
584 $item->{_row_id} ||= $id;
587 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
593 sorted => $sorted_items,
594 columns => [ @columns ],
595 rows => $#$filtered + 1,
596 numeric => { map { $_, $self->_is_numeric($_) } @columns },
597 filters => $self->_current_filters,
605 $self->session('order', $self->param('order'));
606 $self->session('sort', $self->param('sort'));
607 $self->redirect_to('/data/items');
611 my ( $self, $name ) = @_;
613 my $stats = $self->_loaded( 'stats' );
615 # sort facet numerically if more >50% elements are numeric
616 my $count = $stats->{$name}->{count};
617 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
618 defined $stats->{$name}->{numeric} &&
619 $stats->{$name}->{numeric} > $count / 2;
623 my ($self,$name) = @_;
624 warn "_remove_filter $name\n";
626 my $path = $self->session('path');
628 delete $filters->{$name};
629 delete $loaded->{$path}->{filters}->{$name};
630 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
634 keys %{ $loaded->{$path}->{filtered} }
636 delete $loaded->{$path}->{filtered}->{$_};
637 warn "remove filtered cache $_";
644 my $path = $self->session('path') || $self->redirect_to( '/data/index' );
646 if ( my $name = $self->param('remove') ) {
647 $self->_remove_filter( $name );
648 $self->redirect_to( '/data/items' );
652 my $name = $self->param('name') || die "no name";
654 my $all = $self->_perm_scalar('all', 1);
655 my $data = $self->_loaded('data');
657 my $filters = $self->_current_filters;
658 my $all_filters = join(' ',sort keys %$filters,'order:',$self->session('order'));
659 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
660 if defined $loaded->{$path}->{filtered}->{$all_filters};
662 if ( ! $filtered || $all ) {
663 $filtered = [ 0 .. $#{ $data->{items} } ];
664 warn "filter all values\n";
666 warn "filter using $all_filters\n";
669 foreach my $i ( @$filtered ) {
670 my $item = $data->{items}->[$i];
671 if ( ! exists $item->{$name} ) {
672 $facet->{ _missing }++;
673 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
674 $facet->{$_}++ foreach @{ $item->{$name} };
676 $facet->{ $item->{$name} }++;
680 # warn "# facet $name ",dump $facet;
685 : defined $filters->{$name} ? @{ $filters->{$name} }
688 $checked = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
690 my $numeric = $self->_is_numeric($name);
692 my $sort = $self->param('sort');
693 # sort numeric facets with more than 5 values ascending
694 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
696 @facet_names = sort {
698 if ( $sort eq 'a' ) {
699 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
700 } elsif ( $sort eq 'd' ) {
701 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
702 } elsif ( $sort eq 'c' ) {
703 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
705 warn "unknown sort: $sort";
708 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
712 $self->render( name => $name, facet => $facet, checked => $checked,
713 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
719 my $content = $self->param('content');
721 my $i = $self->param('_row_id');
722 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
723 my $path = $self->param('path') || die "no path";
724 my $name = $self->param('name') || die "no name";
725 my $status = 200; # 200 = OK, 201 = Created
727 my $data = $self->_loaded('data');
729 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
730 $content =~ s/^\s+//s;
731 $content =~ s/\s+$//s;
733 if ( $content =~ /\xB6/ ) { # para
734 $v = [ split(/\s*\xB6\s*/, $content) ];
739 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
742 && ! ( $old eq 'undef' && length($content) == 0 ) # new value empty, previous undef
748 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
750 time => $self->param('time') || time(),
751 user => $self->param('user') || $ENV{'LOGNAME'},
753 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
754 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
755 keys %{ $loaded->{$path}->{stats} }
758 my $edit_path = $self->app->home->rel_dir('data') . '/' . $path . '.edits';
759 mkdir $edit_path unless -d $edit_path;
760 $edit_path .= '/' . $edit->{time};
761 store $edit, $edit_path;
762 utime $edit->{time}, $edit->{time}, $edit_path;
763 warn "# $edit_path ", dump($edit);
765 warn "# edit $path $i $old -> $new\n";
766 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
768 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
769 delete $loaded->{$path}->{sorted}->{$name};
770 warn "# invalidate $path sorted $name\n";
773 foreach ( grep { m/$name/ } keys %{ $loaded->{$path}->{filtered} } ) {
774 delete $loaded->{$path}->{filtered}->{$_};
775 warn "# invalidate $path filtered $_\n";
778 $status = 201; # created
779 $self->session('save_path' => $path);
782 warn "# unchanged $path $i $old\n";
786 $content = "$path $i $name doesn't exist\n";
790 warn "# edit $status $content";
801 my $path = $self->param('path');
802 $path ||= $self->session('path');
804 my $dump_path = $self->_save( $path );
805 $self->session('save_path' => 0);
807 $self->redirect_to( '/data/items' );