1 package MojoFacets::Data;
6 use base 'Mojolicious::Controller';
8 use Data::Dump qw(dump);
22 my $path = $self->app->home->rel_dir('data');
23 die "no data dir $path" unless -d $path;
27 my $file = $File::Find::name;
28 if ( -f $file && $file =~ m/\.(js(on)?|txt)$/ ) {
29 $file =~ s/$path\/*//;
32 warn "IGNORE: $file\n";
36 @files = sort { lc $a cmp lc $b } @files;
38 $size->{$_} = -s "$path/$_" foreach @files;
41 $dump_path->{$_} = $self->_dump_path( $_ ) foreach @files;
43 if ( my $save_path = $self->session('save_path') ) {
44 $self->session( 'save_path' => 0 )
45 if ! defined $loaded->{$save_path};
53 dump_path => $dump_path,
58 my ( $self, $name ) = @_;
59 my $dir = $self->app->home->rel_dir('data');
62 return '/tmp/mojo_facets.' . $name . '.storable';
66 my ( $self, $path ) = @_;
68 my $dump_path = $self->_dump_path( $path );
69 warn "save loaded to $dump_path";
70 my $info = $loaded->{$path};
71 store $info, $dump_path;
74 #my $mtime = $loaded->{$path}->{mtime};
75 #utime $mtime, $mtime, $dump_path;
77 warn $dump_path, ' ', -s $dump_path, " bytes\n";
82 my ( $self, $path ) = @_;
84 my $full_path = $self->app->home->rel_file( 'data/' . $path );
85 die "$full_path $!" unless -r $full_path;
87 my $dump_path = $self->_dump_path( $path );
89 if ( defined $loaded->{$path}->{data} ) {
90 my $mtime = (stat($full_path))[9];
91 return if $loaded->{$path}->{mtime} == $mtime;
92 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
93 } elsif ( -e $dump_path ) {
94 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
95 my $info = retrieve $dump_path;
96 $loaded->{ $path } = $info;
100 # we could use Mojo::JSON here, but it's too slow
101 # $data = from_json read_file $path;
102 my $data = read_file $full_path;
103 warn "# data snippet: ", substr($data,0,200);
105 if ( $path =~ m/\.js/ ) {
106 Encode::_utf8_on($data);
107 $data = from_json $data;
108 } elsif ( $path =~ m/\.txt/ ) {
109 my @lines = split(/\r?\n/, $data);
110 $data = { items => [] };
112 my $header_line = shift @lines;
113 my $multiline = $header_line =~ s/\^//g;
114 @header = split(/\|/, $header_line );
115 warn "# header ", dump( @header );
116 while ( my $line = shift @lines ) {
119 my @v = split(/\|/, $line);
120 while ( @lines && $#v < $#header ) {
124 my @more_v = split(/\|/, $line);
125 if ( $#v + $#more_v > $#header ) {
126 warn "short line: ",dump( @v );
130 $v[ $#v ] .= shift @more_v if @more_v;
131 push @v, @more_v if @more_v;
133 if ( $#v > $#header ) {
134 die "# splice $#header ", dump( @v );
135 @v = splice @v, 0, $#header;
139 foreach my $i ( 0 .. $#v ) {
140 $item->{ $header[$i] || "f_$i" } = [ $v[$i] ];
142 push @{ $data->{items} }, $item;
145 warn "file format unknown $path";
150 foreach my $e ( @{ $data->{items} } ) {
151 foreach my $n ( keys %$e ) {
152 $stats->{$n}->{count}++;
154 if ( ref $e->{$n} eq 'ARRAY' ) {
155 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
161 foreach my $x ( @v ) {
162 $stats->{$n}->{numeric}++
163 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
164 $stats->{$n}->{empty}++
165 if length $x == 0; # faster than $x =~ m/^\s*$/;
171 foreach my $n ( keys %$stats ) {
172 next unless defined $stats->{$n}->{array};
173 delete $stats->{$n}->{array}
174 if $stats->{$n}->{array} == $stats->{$n}->{count};
178 if ( defined $data->{header} ) {
179 if ( ref $data->{header} eq 'ARRAY' ) {
180 @header = @{ $data->{header} };
182 warn "header not array ", dump( $data->{header} );
188 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
189 grep { defined $stats->{$_}->{count} } keys %$stats
195 header => [ @header ],
197 full_path => $full_path,
198 size => -s $full_path,
199 mtime => (stat($full_path))[9],
203 $loaded->{ $path } = $info;
204 $self->_save( $path );
212 my @paths = $self->param('paths');
213 warn "# paths ", dump @paths;
214 $self->_load_path( $_ ) foreach @paths;
216 my $path = $self->param('path') || $self->redirect_to( '/data/index' );
217 warn "# path $path\n";
218 $self->session('path' => $path);
219 $self->_load_path( $path );
221 $self->session( 'header' => $loaded->{$path}->{header} );
222 if ( ! defined $loaded->{$path}->{columns} ) {
223 $self->session( 'columns' => $loaded->{$path}->{header} );
224 $self->session( 'order' => $loaded->{$path}->{header}->[0] );
225 $self->redirect_to( '/data/columns' );
227 $self->session( 'columns' => $loaded->{$path}->{columns} );
228 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
229 $self->redirect_to( '/data/items' );
235 my ( $self, $name ) = @_;
236 my $path = $self->session('path');
237 $self->redirect_to('/data/index') unless $path;
238 if ( ! defined $loaded->{$path}->{$name} ) {
239 warn "$path $name isn't loaded\n";
240 $self->_load_path( $path );
241 $self->redirect_to('/data/index')
242 unless defined $loaded->{$path}->{$name};
244 return $loaded->{$path}->{$name};
251 $checked->{$_}++ foreach @_;
252 # warn "# _checked ",dump($checked);
260 if ( $self->param('columns') ) {
261 $self->_perm_array('columns');
262 $self->redirect_to('/data/items');
265 my $stats = $self->_loaded( 'stats' ); # || $self->redirect_to( '/data/index' );
268 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
270 foreach my $c ( @{ $self->session( 'header' ) } ) {
271 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
275 message => 'Select columns to display',
277 columns => \@columns,
278 checked => $self->_checked( $self->_perm_array('columns') ),
283 my ($self,$name) = @_;
285 my @array = $self->param($name);
286 my $path = $self->session('path');
289 $self->session($name => [ @array ]);
290 $loaded->{$path}->{$name} = [ @array ];
291 } elsif ( my $session = $self->session($name) ) {
292 if ( ref $session eq 'ARRAY' ) {
295 die "$name not array ",dump($session);
298 #warn "# $name ",dump @array;
303 my ($self,$name,$default) = @_;
305 my $scalar = $self->param($name);
307 if ( defined $scalar ) {
308 $self->session($name => $scalar);
310 $scalar = $self->session($name);
313 if ( ! defined $scalar ) {
315 die "no default for $name" unless defined $scalar;
316 $self->session($name => $scalar);
319 warn "# _perm_scalar $name ",dump $scalar;
326 my $name = $self->param('filter_name') || die "name?";
327 my @vals = $self->param('filter_vals');
329 $self->_remove_filter( $name );
330 $self->_filter_on_data( $name, @vals ) if @vals;
332 $self->session( 'offset' => 0 );
334 $self->redirect_to('/data/items');
337 sub _filter_on_data {
338 my ( $self, $name, @vals ) = @_;
340 my $path = $self->session('path');
342 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
343 warn "filter $name not found in data set";
347 $filters->{$name} = [ @vals ];
348 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
351 $filter_hash->{$_}++ foreach @vals;
353 #warn "# filter_hash ",dump( $filter_hash );
355 my $items = $self->_loaded('data')->{items};
357 my $include_missing = defined $filter_hash->{_missing};
360 foreach my $i ( 0 .. $#$items ) {
362 if ( defined $items->[$i]->{$name} ) {
363 foreach my $v ( @{ $items->[$i]->{$name} } ) {
364 if ( defined $filter_hash->{ $v } ) {
365 $filtered_items->{$i}++;
368 } elsif ( $include_missing ) {
369 $filtered_items->{$i}++;
373 #warn "# filter $name ",dump($filtered_items);
375 $loaded->{$path}->{filters}->{$name} = $filtered_items;
376 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
380 my ( $self, $all ) = @_;
381 my $data = $self->_loaded( 'data' );
383 return @{ $data->{items} } if $all == 1;
385 my $filters = $self->_current_filters;
387 foreach my $f ( keys %$filters ) {
388 foreach my $n ( @{ $filters->{$f} } ) {
389 $filter_value->{$f}->{$n} = 1;
392 my @items = @{ $data->{items} };
396 foreach my $n ( keys %$filter_value ) {
397 if ( ! exists $i->{$n} ) {
398 if ( defined $filter_value->{$n}->{_missing} ) {
406 # and match any of values in element
408 foreach my $v ( @{ $i->{$n} } ) { # FIXME not array?
409 $have_values ||= 1 if defined $filter_value->{$n}->{$v};
411 if ( ! $have_values ) {
417 } @items if $filter_value;
422 sub _current_filters {
425 $current_filters->{ $_ } = $filters->{ $_ }
427 grep { defined $filters->{ $_ } }
428 @{ $self->_loaded('header') }
430 #warn "# current_filters ",dump($current_filters);
431 return $current_filters;
434 sub _data_sorted_by {
435 my ( $self, $order ) = @_;
437 my $path = $self->session('path');
439 warn "_data_sorted_by $order from $path";
441 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
442 return $loaded->{$path}->{sorted}->{$order};
445 my $data = $self->_loaded( 'data' );
446 my $numeric = $self->_is_numeric($order);
447 my $missing = $numeric ? 0 : '';
448 no warnings qw(numeric);
459 [ $nr++, exists $_->{$order} ? join('', @{$_->{$order}}) : $missing ]
460 } @{ $data->{items} }
463 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
464 #warn "# sorted ",dump( @sorted );
466 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
473 my $path = $self->session('path');
474 $self->redirect_to('/data/index') unless defined $loaded->{ $path };
476 my @columns = $self->_perm_array('columns');
477 $self->redirect_to('/data/columns') unless @columns;
478 my $order = $self->_perm_scalar('order', $columns[0]);
479 my $sort = $self->_perm_scalar('sort', 'a');
480 my $offset = $self->_perm_scalar('offset', 0);
481 my $limit = $self->_perm_scalar('limit', 20);
482 $self->_perm_scalar('show', 'table');
484 # fix offset when changing limit
485 $offset = int( $offset / $limit ) * $limit;
487 my $sorted = $self->_data_sorted_by( $order );
491 foreach my $name ( keys %$filters ) {
492 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
493 warn "skip filter $name not found in $path\n";
496 push @filter_names, $name;
498 warn "filter_names ",dump( @filter_names );
499 foreach my $name ( @filter_names ) {
500 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
501 $self->_filter_on_data( $name, @{ $filters->{$name} } );
505 my $all_filters = join(' ',sort @filter_names,'order:',$order);
507 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
509 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
511 my $path_filters = $loaded->{$path}->{filters};
513 warn "create combined filter for $all_filters\n";
516 foreach my $i ( 0 .. $#$sorted ) {
517 my $pos = $sorted->[$i];
519 if ( $#filter_names == -1 ) {
520 push @filtered, $pos;
525 foreach ( @filter_names ) {
526 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
530 push @filtered, $pos;
533 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
536 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
537 if defined $loaded->{$path}->{filtered}->{$all_filters};
539 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
542 my $data = $self->_loaded('data');
543 my $from_end = $sort eq 'd' ? $#$filtered : 0;
544 foreach ( 0 .. $limit ) {
545 my $i = $_ + $offset;
546 last unless defined $filtered->[$i];
547 $i = $from_end - $i if $from_end;
548 my $id = $filtered->[$i];
550 my $item = $data->{items}->[ $id ];
551 $item->{_row_id} ||= $id;
554 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
560 sorted => $sorted_items,
561 columns => [ @columns ],
562 rows => $#$filtered + 1,
563 numeric => { map { $_, $self->_is_numeric($_) } @columns },
564 filters => $self->_current_filters,
572 $self->session('order', $self->param('order'));
573 $self->session('sort', $self->param('sort'));
574 $self->redirect_to('/data/items');
578 my ( $self, $name ) = @_;
580 my $stats = $self->_loaded( 'stats' );
582 # sort facet numerically if more >50% elements are numeric
583 my $count = $stats->{$name}->{count};
584 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
585 defined $stats->{$name}->{numeric} &&
586 $stats->{$name}->{numeric} > $count / 2;
590 my ($self,$name) = @_;
591 warn "_remove_filter $name\n";
593 my $path = $self->session('path');
595 delete $filters->{$name};
596 delete $loaded->{$path}->{filters}->{$name};
597 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
601 keys %{ $loaded->{$path}->{filtered} }
603 delete $loaded->{$path}->{filtered}->{$_};
604 warn "remove filtered cache $_";
611 my $path = $self->session('path') || $self->redirect_to( '/data/index' );
613 if ( my $name = $self->param('remove') ) {
614 $self->_remove_filter( $name );
615 $self->redirect_to( '/data/items' );
619 my $name = $self->param('name') || die "no name";
621 my $all = $self->_perm_scalar('all', 1);
622 my $data = $self->_loaded('data');
624 my $filters = $self->_current_filters;
625 my $all_filters = join(' ',sort keys %$filters,'order:',$self->session('order'));
626 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
627 if defined $loaded->{$path}->{filtered}->{$all_filters};
629 if ( ! $filtered || $all ) {
630 $filtered = [ 0 .. $#{ $data->{items} } ];
631 warn "filter all values\n";
633 warn "filter using $all_filters\n";
636 foreach my $i ( @$filtered ) {
637 my $item = $data->{items}->[$i];
638 if ( ! exists $item->{$name} ) {
639 $facet->{ _missing }++;
640 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
641 $facet->{$_}++ foreach @{ $item->{$name} };
643 $facet->{ $item->{$name} }++;
647 # warn "# facet $name ",dump $facet;
652 : defined $filters->{$name} ? @{ $filters->{$name} }
655 $checked = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
657 my $numeric = $self->_is_numeric($name);
659 my $sort = $self->param('sort');
660 # sort numeric facets with more than 5 values ascending
661 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
663 @facet_names = sort {
665 if ( $sort eq 'a' ) {
666 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
667 } elsif ( $sort eq 'd' ) {
668 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
669 } elsif ( $sort eq 'c' ) {
670 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
672 warn "unknown sort: $sort";
675 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
679 $self->render( name => $name, facet => $facet, checked => $checked,
680 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
686 my $content = $self->param('content');
688 my $i = $self->param('_row_id');
689 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
690 my $path = $self->param('path') || die "no path";
691 my $name = $self->param('name') || die "no name";
692 my $status = 200; # 200 = OK, 201 = Created
694 my $data = $self->_loaded('data');
696 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
697 $content =~ s/^\s+//s;
698 $content =~ s/\s+$//s;
700 if ( $content =~ /\xB6/ ) { # para
701 $v = [ split(/\s*\xB6\s*/, $content) ];
706 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
709 && ! ( $old eq 'undef' && length($content) == 0 ) # new value empty, previous undef
711 warn "# update $path $i $old -> $new\n";
712 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
714 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
715 delete $loaded->{$path}->{sorted}->{$name};
716 warn "# invalidate $path sorted $name\n";
719 foreach ( grep { m/$name/ } keys %{ $loaded->{$path}->{filtered} } ) {
720 delete $loaded->{$path}->{filtered}->{$_};
721 warn "# invalidate $path filtered $_\n";
724 $status = 201; # created
725 $self->session('save_path' => $path);
728 warn "# unchanged $path $i $old\n";
732 $content = "$path $i $name doesn't exist\n";
736 warn "# edit $status $content";
747 my $path = $self->param('path');
748 $path ||= $self->session('path');
750 my $dump_path = $self->_save( $path );
751 $self->session('save_path' => 0);
753 $self->redirect_to( '/data/items' );