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;
28 my $file = $File::Find::name;
29 if ( -f $file && $file =~ m/\.(js(on)?|txt)$/ ) {
30 $file =~ s/$path\/*//;
32 } elsif ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
33 push @{ $changes->{$1} }, $2
35 warn "IGNORE: $file\n";
39 @files = sort { lc $a cmp lc $b } @files;
41 $size->{$_} = -s "$path/$_" 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 => { map { $_ => $self->_dump_path($_) } @files },
59 my ( $self, $name ) = @_;
60 my $dir = $self->app->home->rel_dir('data');
63 return '/tmp/mojo_facets.' . $name . '.storable';
67 my ( $self, $path ) = @_;
69 my $dump_path = $self->_dump_path( $path );
70 warn "save loaded to $dump_path";
71 my $info = $loaded->{$path};
72 store $info, $dump_path;
75 #my $mtime = $loaded->{$path}->{mtime};
76 #utime $mtime, $mtime, $dump_path;
78 warn $dump_path, ' ', -s $dump_path, " bytes\n";
83 my ( $self, $path ) = @_;
85 my $full_path = $self->app->home->rel_file( 'data/' . $path );
86 die "$full_path $!" unless -r $full_path;
88 my $dump_path = $self->_dump_path( $path );
90 if ( defined $loaded->{$path}->{data} ) {
91 my $mtime = (stat($full_path))[9];
92 return if $loaded->{$path}->{mtime} == $mtime;
93 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
94 } elsif ( -e $dump_path ) {
95 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
96 my $info = retrieve $dump_path;
97 $loaded->{ $path } = $info;
101 # we could use Mojo::JSON here, but it's too slow
102 # $data = from_json read_file $path;
103 my $data = read_file $full_path;
104 warn "# data snippet: ", substr($data,0,200);
106 if ( $path =~ m/\.js/ ) {
107 Encode::_utf8_on($data);
108 $data = from_json $data;
109 } elsif ( $path =~ m/\.txt/ ) {
110 my @lines = split(/\r?\n/, $data);
111 $data = { items => [] };
113 my $header_line = shift @lines;
114 my $multiline = $header_line =~ s/\^//g;
115 @header = split(/\|/, $header_line );
116 warn "# header ", dump( @header );
117 while ( my $line = shift @lines ) {
120 my @v = split(/\|/, $line);
121 while ( @lines && $#v < $#header ) {
125 my @more_v = split(/\|/, $line);
126 if ( $#v + $#more_v > $#header ) {
127 warn "short line: ",dump( @v );
131 $v[ $#v ] .= shift @more_v if @more_v;
132 push @v, @more_v if @more_v;
134 if ( $#v > $#header ) {
135 die "# splice $#header ", dump( @v );
136 @v = splice @v, 0, $#header;
140 foreach my $i ( 0 .. $#v ) {
141 $item->{ $header[$i] || "f_$i" } = [ $v[$i] ];
143 push @{ $data->{items} }, $item;
146 warn "file format unknown $path";
151 foreach my $e ( @{ $data->{items} } ) {
152 foreach my $n ( keys %$e ) {
153 $stats->{$n}->{count}++;
155 if ( ref $e->{$n} eq 'ARRAY' ) {
156 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
162 foreach my $x ( @v ) {
163 $stats->{$n}->{numeric}++
164 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
165 $stats->{$n}->{empty}++
166 if length $x == 0; # faster than $x =~ m/^\s*$/;
172 foreach my $n ( keys %$stats ) {
173 next unless defined $stats->{$n}->{array};
174 delete $stats->{$n}->{array}
175 if $stats->{$n}->{array} == $stats->{$n}->{count};
179 if ( defined $data->{header} ) {
180 if ( ref $data->{header} eq 'ARRAY' ) {
181 @header = @{ $data->{header} };
183 warn "header not array ", dump( $data->{header} );
189 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
190 grep { defined $stats->{$_}->{count} } keys %$stats
196 header => [ @header ],
198 full_path => $full_path,
199 size => -s $full_path,
200 mtime => (stat($full_path))[9],
204 $loaded->{ $path } = $info;
205 $self->_save( $path );
213 my @paths = $self->param('paths');
214 warn "# paths ", dump @paths;
215 $self->_load_path( $_ ) foreach @paths;
217 my $path = $self->param('path') || $self->redirect_to( '/data/index' );
218 warn "# path $path\n";
219 $self->session('path' => $path);
220 $self->_load_path( $path );
222 $self->session( 'header' => $loaded->{$path}->{header} );
223 if ( ! defined $loaded->{$path}->{columns} ) {
224 $self->session( 'columns' => $loaded->{$path}->{header} );
225 $self->session( 'order' => $loaded->{$path}->{header}->[0] );
226 $self->redirect_to( '/data/columns' );
228 $self->session( 'columns' => $loaded->{$path}->{columns} );
229 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
230 $self->redirect_to( '/data/items' );
236 my ( $self, $name ) = @_;
237 my $path = $self->session('path');
238 $self->redirect_to('/data/index') unless $path;
239 if ( ! defined $loaded->{$path}->{$name} ) {
240 warn "$path $name isn't loaded\n";
241 $self->_load_path( $path );
242 $self->redirect_to('/data/index')
243 unless defined $loaded->{$path}->{$name};
245 return $loaded->{$path}->{$name};
252 $checked->{$_}++ foreach @_;
253 # warn "# _checked ",dump($checked);
261 if ( $self->param('columns') ) {
262 $self->_perm_array('columns');
263 $self->redirect_to('/data/items');
266 my $stats = $self->_loaded( 'stats' ); # || $self->redirect_to( '/data/index' );
269 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
271 foreach my $c ( @{ $self->session( 'header' ) } ) {
272 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
276 message => 'Select columns to display',
278 columns => \@columns,
279 checked => $self->_checked( $self->_perm_array('columns') ),
284 my ($self,$name) = @_;
286 my @array = $self->param($name);
287 my $path = $self->session('path');
290 $self->session($name => [ @array ]);
291 $loaded->{$path}->{$name} = [ @array ];
292 } elsif ( my $session = $self->session($name) ) {
293 if ( ref $session eq 'ARRAY' ) {
296 die "$name not array ",dump($session);
299 #warn "# $name ",dump @array;
304 my ($self,$name,$default) = @_;
306 my $scalar = $self->param($name);
308 if ( defined $scalar ) {
309 $self->session($name => $scalar);
311 $scalar = $self->session($name);
314 if ( ! defined $scalar ) {
316 die "no default for $name" unless defined $scalar;
317 $self->session($name => $scalar);
320 warn "# _perm_scalar $name ",dump $scalar;
327 my $name = $self->param('filter_name') || die "name?";
328 my @vals = $self->param('filter_vals');
330 $self->_remove_filter( $name );
331 $self->_filter_on_data( $name, @vals ) if @vals;
333 $self->session( 'offset' => 0 );
335 $self->redirect_to('/data/items');
338 sub _filter_on_data {
339 my ( $self, $name, @vals ) = @_;
341 my $path = $self->session('path');
343 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
344 warn "filter $name not found in data set";
348 $filters->{$name} = [ @vals ];
349 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
352 $filter_hash->{$_}++ foreach @vals;
354 #warn "# filter_hash ",dump( $filter_hash );
356 my $items = $self->_loaded('data')->{items};
358 my $include_missing = defined $filter_hash->{_missing};
361 foreach my $i ( 0 .. $#$items ) {
363 if ( defined $items->[$i]->{$name} ) {
364 foreach my $v ( @{ $items->[$i]->{$name} } ) {
365 if ( defined $filter_hash->{ $v } ) {
366 $filtered_items->{$i}++;
369 } elsif ( $include_missing ) {
370 $filtered_items->{$i}++;
374 #warn "# filter $name ",dump($filtered_items);
376 $loaded->{$path}->{filters}->{$name} = $filtered_items;
377 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
381 my ( $self, $all ) = @_;
382 my $data = $self->_loaded( 'data' );
384 return @{ $data->{items} } if $all == 1;
386 my $filters = $self->_current_filters;
388 foreach my $f ( keys %$filters ) {
389 foreach my $n ( @{ $filters->{$f} } ) {
390 $filter_value->{$f}->{$n} = 1;
393 my @items = @{ $data->{items} };
397 foreach my $n ( keys %$filter_value ) {
398 if ( ! exists $i->{$n} ) {
399 if ( defined $filter_value->{$n}->{_missing} ) {
407 # and match any of values in element
409 foreach my $v ( @{ $i->{$n} } ) { # FIXME not array?
410 $have_values ||= 1 if defined $filter_value->{$n}->{$v};
412 if ( ! $have_values ) {
418 } @items if $filter_value;
423 sub _current_filters {
426 $current_filters->{ $_ } = $filters->{ $_ }
428 grep { defined $filters->{ $_ } }
429 @{ $self->_loaded('header') }
431 #warn "# current_filters ",dump($current_filters);
432 return $current_filters;
435 sub _data_sorted_by {
436 my ( $self, $order ) = @_;
438 my $path = $self->session('path');
440 warn "_data_sorted_by $order from $path";
442 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
443 return $loaded->{$path}->{sorted}->{$order};
446 my $data = $self->_loaded( 'data' );
447 my $numeric = $self->_is_numeric($order);
448 my $missing = $numeric ? 0 : '';
449 no warnings qw(numeric);
460 [ $nr++, exists $_->{$order} ? join('', @{$_->{$order}}) : $missing ]
461 } @{ $data->{items} }
464 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
465 #warn "# sorted ",dump( @sorted );
467 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
474 my $path = $self->session('path');
475 $self->redirect_to('/data/index') unless defined $loaded->{ $path };
477 my @columns = $self->_perm_array('columns');
478 $self->redirect_to('/data/columns') unless @columns;
479 my $order = $self->_perm_scalar('order', $columns[0]);
480 my $sort = $self->_perm_scalar('sort', 'a');
481 my $offset = $self->_perm_scalar('offset', 0);
482 my $limit = $self->_perm_scalar('limit', 20);
483 $self->_perm_scalar('show', 'table');
485 # fix offset when changing limit
486 $offset = int( $offset / $limit ) * $limit;
488 my $sorted = $self->_data_sorted_by( $order );
492 foreach my $name ( keys %$filters ) {
493 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
494 warn "skip filter $name not found in $path\n";
497 push @filter_names, $name;
499 warn "filter_names ",dump( @filter_names );
500 foreach my $name ( @filter_names ) {
501 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
502 $self->_filter_on_data( $name, @{ $filters->{$name} } );
506 my $all_filters = join(' ',sort @filter_names,'order:',$order);
508 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
510 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
512 my $path_filters = $loaded->{$path}->{filters};
514 warn "create combined filter for $all_filters\n";
517 foreach my $i ( 0 .. $#$sorted ) {
518 my $pos = $sorted->[$i];
520 if ( $#filter_names == -1 ) {
521 push @filtered, $pos;
526 foreach ( @filter_names ) {
527 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
531 push @filtered, $pos;
534 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
537 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
538 if defined $loaded->{$path}->{filtered}->{$all_filters};
540 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
543 my $data = $self->_loaded('data');
544 my $from_end = $sort eq 'd' ? $#$filtered : 0;
545 foreach ( 0 .. $limit ) {
546 my $i = $_ + $offset;
547 last unless defined $filtered->[$i];
548 $i = $from_end - $i if $from_end;
549 my $id = $filtered->[$i];
551 my $item = $data->{items}->[ $id ];
552 $item->{_row_id} ||= $id;
555 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
561 sorted => $sorted_items,
562 columns => [ @columns ],
563 rows => $#$filtered + 1,
564 numeric => { map { $_, $self->_is_numeric($_) } @columns },
565 filters => $self->_current_filters,
573 $self->session('order', $self->param('order'));
574 $self->session('sort', $self->param('sort'));
575 $self->redirect_to('/data/items');
579 my ( $self, $name ) = @_;
581 my $stats = $self->_loaded( 'stats' );
583 # sort facet numerically if more >50% elements are numeric
584 my $count = $stats->{$name}->{count};
585 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
586 defined $stats->{$name}->{numeric} &&
587 $stats->{$name}->{numeric} > $count / 2;
591 my ($self,$name) = @_;
592 warn "_remove_filter $name\n";
594 my $path = $self->session('path');
596 delete $filters->{$name};
597 delete $loaded->{$path}->{filters}->{$name};
598 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
602 keys %{ $loaded->{$path}->{filtered} }
604 delete $loaded->{$path}->{filtered}->{$_};
605 warn "remove filtered cache $_";
612 my $path = $self->session('path') || $self->redirect_to( '/data/index' );
614 if ( my $name = $self->param('remove') ) {
615 $self->_remove_filter( $name );
616 $self->redirect_to( '/data/items' );
620 my $name = $self->param('name') || die "no name";
622 my $all = $self->_perm_scalar('all', 1);
623 my $data = $self->_loaded('data');
625 my $filters = $self->_current_filters;
626 my $all_filters = join(' ',sort keys %$filters,'order:',$self->session('order'));
627 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
628 if defined $loaded->{$path}->{filtered}->{$all_filters};
630 if ( ! $filtered || $all ) {
631 $filtered = [ 0 .. $#{ $data->{items} } ];
632 warn "filter all values\n";
634 warn "filter using $all_filters\n";
637 foreach my $i ( @$filtered ) {
638 my $item = $data->{items}->[$i];
639 if ( ! exists $item->{$name} ) {
640 $facet->{ _missing }++;
641 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
642 $facet->{$_}++ foreach @{ $item->{$name} };
644 $facet->{ $item->{$name} }++;
648 # warn "# facet $name ",dump $facet;
653 : defined $filters->{$name} ? @{ $filters->{$name} }
656 $checked = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
658 my $numeric = $self->_is_numeric($name);
660 my $sort = $self->param('sort');
661 # sort numeric facets with more than 5 values ascending
662 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
664 @facet_names = sort {
666 if ( $sort eq 'a' ) {
667 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
668 } elsif ( $sort eq 'd' ) {
669 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
670 } elsif ( $sort eq 'c' ) {
671 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
673 warn "unknown sort: $sort";
676 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
680 $self->render( name => $name, facet => $facet, checked => $checked,
681 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
687 my $content = $self->param('content');
689 my $i = $self->param('_row_id');
690 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
691 my $path = $self->param('path') || die "no path";
692 my $name = $self->param('name') || die "no name";
693 my $status = 200; # 200 = OK, 201 = Created
695 my $data = $self->_loaded('data');
697 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
698 $content =~ s/^\s+//s;
699 $content =~ s/\s+$//s;
701 if ( $content =~ /\xB6/ ) { # para
702 $v = [ split(/\s*\xB6\s*/, $content) ];
707 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
710 && ! ( $old eq 'undef' && length($content) == 0 ) # new value empty, previous undef
712 warn "# update $path $i $old -> $new\n";
713 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
715 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
716 delete $loaded->{$path}->{sorted}->{$name};
717 warn "# invalidate $path sorted $name\n";
720 foreach ( grep { m/$name/ } keys %{ $loaded->{$path}->{filtered} } ) {
721 delete $loaded->{$path}->{filtered}->{$_};
722 warn "# invalidate $path filtered $_\n";
725 $status = 201; # created
726 $self->session('save_path' => $path);
729 warn "# unchanged $path $i $old\n";
733 $content = "$path $i $name doesn't exist\n";
737 warn "# edit $status $content";
748 my $path = $self->param('path');
749 $path ||= $self->session('path');
751 my $dump_path = $self->_save( $path );
752 $self->session('save_path' => 0);
754 $self->redirect_to( '/data/items' );