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);
141 sub _param_or_session {
142 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
147 my $path = $self->_param_or_session('path');
148 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
149 $self->redirect_to( '/data/columns' );
154 my ( $self, $path ) = @_;
156 my $full_path = $self->app->home->rel_file( 'data/' . $path );
157 die "$full_path $!" unless -r $full_path;
159 my $dump_path = $self->_dump_path( $path );
161 if ( defined $loaded->{$path}->{data} ) {
162 my $mtime = (stat($full_path))[9];
163 return if $loaded->{$path}->{mtime} == $mtime;
164 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
165 } elsif ( -e $dump_path ) {
166 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
167 my $info = retrieve $dump_path;
168 $loaded->{ $path } = $info;
172 # we could use Mojo::JSON here, but it's too slow
173 # $data = from_json read_file $path;
174 my $data = read_file $full_path;
175 warn "# data snippet: ", substr($data,0,200);
177 if ( $path =~ m/\.js/ ) {
178 Encode::_utf8_on($data);
179 $data = from_json $data;
180 } elsif ( $path =~ m/\.txt/ ) {
181 my @lines = split(/\r?\n/, $data);
182 $data = { items => [] };
184 my $header_line = shift @lines;
185 my $multiline = $header_line =~ s/\^//g;
186 @header = split(/\|/, $header_line );
187 warn "# header ", dump( @header );
188 while ( my $line = shift @lines ) {
191 my @v = split(/\|/, $line);
192 while ( @lines && $#v < $#header ) {
196 my @more_v = split(/\|/, $line);
197 if ( $#v + $#more_v > $#header ) {
198 warn "short line: ",dump( @v );
202 $v[ $#v ] .= shift @more_v if @more_v;
203 push @v, @more_v if @more_v;
205 if ( $#v > $#header ) {
206 die "# splice $#header ", dump( @v );
207 @v = splice @v, 0, $#header;
211 foreach my $i ( 0 .. $#v ) {
212 $item->{ $header[$i] || "f_$i" } = [ $v[$i] ];
214 push @{ $data->{items} }, $item;
217 warn "file format unknown $path";
220 my $stats = __stats( $data->{items} );
223 if ( defined $data->{header} ) {
224 if ( ref $data->{header} eq 'ARRAY' ) {
225 @header = @{ $data->{header} };
227 warn "header not array ", dump( $data->{header} );
233 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
234 grep { defined $stats->{$_}->{count} } keys %$stats
238 header => [ @header ],
240 full_path => $full_path,
241 size => -s $full_path,
242 mtime => (stat($full_path))[9],
246 $loaded->{ $path } = $info;
247 $self->_save( $path );
255 my @paths = $self->param('paths');
256 warn "# paths ", dump @paths;
257 $self->_load_path( $_ ) foreach @paths;
259 my $path = $self->param('path') || $self->redirect_to( '/data/index' );
260 warn "# path $path\n";
261 $self->session('path' => $path);
262 $self->_load_path( $path );
264 $self->session( 'header' => $loaded->{$path}->{header} );
265 if ( ! defined $loaded->{$path}->{columns} ) {
266 $self->session( 'columns' => $loaded->{$path}->{header} );
267 $self->session( 'order' => $loaded->{$path}->{header}->[0] );
268 $self->redirect_to( '/data/columns' );
270 $self->session( 'columns' => $loaded->{$path}->{columns} );
271 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
272 $self->redirect_to( '/data/items' );
278 my ( $self, $name ) = @_;
279 my $path = $self->session('path') || $self->param('path');
280 $self->redirect_to('/data/index') unless $path;
281 if ( ! defined $loaded->{$path}->{$name} ) {
282 warn "$path $name isn't loaded\n";
283 $self->_load_path( $path );
284 $self->redirect_to('/data/index')
285 unless defined $loaded->{$path}->{$name};
287 return $loaded->{$path}->{$name};
294 $checked->{$_}++ foreach @_;
295 # warn "# _checked ",dump($checked);
303 if ( $self->param('columns') ) {
304 $self->_param_array('columns');
305 $self->redirect_to('/data/items');
308 my $stats = $self->_loaded( 'stats' );
311 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
313 foreach my $c ( @{ $self->session( 'header' ) } ) {
314 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
318 message => 'Select columns to display',
320 columns => \@columns,
321 checked => $self->_checked( $self->_param_array('columns') ),
326 my ($self,$name) = @_;
328 my @array = $self->param($name);
329 my $path = $self->session('path');
332 $self->session($name => [ @array ]);
333 $loaded->{$path}->{$name} = [ @array ];
334 } elsif ( my $session = $self->session($name) ) {
335 if ( ref $session eq 'ARRAY' ) {
338 die "$name not array ",dump($session);
341 #warn "# $name ",dump @array;
346 my ($self,$name,$default) = @_;
348 my $scalar = $self->param($name);
350 if ( defined $scalar ) {
351 $self->session($name => $scalar);
353 $scalar = $self->session($name);
356 if ( ! defined $scalar ) {
358 die "no default for $name" unless defined $scalar;
359 $self->session($name => $scalar);
362 warn "# _perm_scalar $name ",dump $scalar;
369 my $name = $self->param('filter_name') || die "name?";
370 my @vals = $self->param('filter_vals');
372 $self->_remove_filter( $name );
373 $self->_filter_on_data( $name, @vals ) if @vals;
375 $self->session( 'offset' => 0 );
377 $self->redirect_to('/data/items');
380 sub _filter_on_data {
381 my ( $self, $name, @vals ) = @_;
383 my $path = $self->session('path');
385 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
386 warn "filter $name not found in data set";
390 $filters->{$name} = [ @vals ];
391 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
394 $filter_hash->{$_}++ foreach @vals;
396 #warn "# filter_hash ",dump( $filter_hash );
398 my $items = $self->_loaded('data')->{items};
400 my $include_missing = defined $filter_hash->{_missing};
403 foreach my $i ( 0 .. $#$items ) {
405 if ( defined $items->[$i]->{$name} ) {
406 foreach my $v ( @{ $items->[$i]->{$name} } ) {
407 if ( defined $filter_hash->{ $v } ) {
408 $filtered_items->{$i}++;
411 } elsif ( $include_missing ) {
412 $filtered_items->{$i}++;
416 #warn "# filter $name ",dump($filtered_items);
418 $loaded->{$path}->{filters}->{$name} = $filtered_items;
419 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
423 my ( $self, $all ) = @_;
424 my $data = $self->_loaded( 'data' );
426 return @{ $data->{items} } if $all == 1;
428 my $filters = $self->_current_filters;
430 foreach my $f ( keys %$filters ) {
431 foreach my $n ( @{ $filters->{$f} } ) {
432 $filter_value->{$f}->{$n} = 1;
435 my @items = @{ $data->{items} };
439 foreach my $n ( keys %$filter_value ) {
440 if ( ! exists $i->{$n} ) {
441 if ( defined $filter_value->{$n}->{_missing} ) {
449 # and match any of values in element
451 foreach my $v ( @{ $i->{$n} } ) { # FIXME not array?
452 $have_values ||= 1 if defined $filter_value->{$n}->{$v};
454 if ( ! $have_values ) {
460 } @items if $filter_value;
465 sub _current_filters {
468 $current_filters->{ $_ } = $filters->{ $_ }
470 grep { defined $filters->{ $_ } }
471 @{ $self->_loaded('header') }
473 #warn "# current_filters ",dump($current_filters);
474 return $current_filters;
477 sub _data_sorted_by {
478 my ( $self, $order ) = @_;
480 my $path = $self->session('path');
482 warn "_data_sorted_by $order from $path";
484 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
485 return $loaded->{$path}->{sorted}->{$order};
488 my $data = $self->_loaded( 'data' );
489 my $numeric = $self->_is_numeric($order);
490 my $missing = $numeric ? 0 : '';
491 no warnings qw(numeric);
502 [ $nr++, exists $_->{$order} ? join('', @{$_->{$order}}) : $missing ]
503 } @{ $data->{items} }
506 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
507 #warn "# sorted ",dump( @sorted );
509 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
516 my $path = $self->session('path');
517 $self->redirect_to('/data/index') unless defined $loaded->{ $path };
519 my @columns = $self->_param_array('columns');
520 $self->redirect_to('/data/columns') unless @columns;
521 my $order = $self->_param_scalar('order', $columns[0]);
522 my $sort = $self->_param_scalar('sort', 'a');
523 my $offset = $self->_param_scalar('offset', 0);
524 my $limit = $self->_param_scalar('limit', 20);
525 $self->_param_scalar('show', 'table');
527 # fix offset when changing limit
528 $offset = int( $offset / $limit ) * $limit;
530 my $sorted = $self->_data_sorted_by( $order );
534 foreach my $name ( keys %$filters ) {
535 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
536 warn "skip filter $name not found in $path\n";
539 push @filter_names, $name;
541 warn "filter_names ",dump( @filter_names );
542 foreach my $name ( @filter_names ) {
543 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
544 $self->_filter_on_data( $name, @{ $filters->{$name} } );
548 my $all_filters = join(' ',sort @filter_names,'order:',$order);
550 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
552 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
554 my $path_filters = $loaded->{$path}->{filters};
556 warn "create combined filter for $all_filters\n";
559 foreach my $i ( 0 .. $#$sorted ) {
560 my $pos = $sorted->[$i];
562 if ( $#filter_names == -1 ) {
563 push @filtered, $pos;
568 foreach ( @filter_names ) {
569 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
573 push @filtered, $pos;
576 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
579 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
580 if defined $loaded->{$path}->{filtered}->{$all_filters};
582 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
585 my $data = $self->_loaded('data');
586 my $from_end = $sort eq 'd' ? $#$filtered : 0;
587 foreach ( 0 .. $limit ) {
588 my $i = $_ + $offset;
589 last unless defined $filtered->[$i];
590 $i = $from_end - $i if $from_end;
591 my $id = $filtered->[$i];
593 my $item = $data->{items}->[ $id ];
594 $item->{_row_id} ||= $id;
597 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
603 sorted => $sorted_items,
604 columns => [ @columns ],
605 rows => $#$filtered + 1,
606 numeric => { map { $_, $self->_is_numeric($_) } @columns },
607 filters => $self->_current_filters,
615 $self->session('order', $self->param('order'));
616 $self->session('sort', $self->param('sort'));
617 $self->redirect_to('/data/items');
621 my ( $self, $name ) = @_;
623 my $stats = $self->_loaded( 'stats' );
625 # sort facet numerically if more >50% elements are numeric
626 my $count = $stats->{$name}->{count};
627 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
628 defined $stats->{$name}->{numeric} &&
629 $stats->{$name}->{numeric} > $count / 2;
633 my ($self,$name) = @_;
634 warn "_remove_filter $name\n";
636 my $path = $self->session('path');
638 delete $filters->{$name};
639 delete $loaded->{$path}->{filters}->{$name};
640 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
644 keys %{ $loaded->{$path}->{filtered} }
646 delete $loaded->{$path}->{filtered}->{$_};
647 warn "remove filtered cache $_";
654 my $path = $self->session('path') || $self->redirect_to( '/data/index' );
656 if ( my $name = $self->param('remove') ) {
657 $self->_remove_filter( $name );
658 $self->redirect_to( '/data/items' );
662 my $name = $self->param('name') || die "no name";
664 my $all = $self->_param_scalar('all', 1);
665 my $data = $self->_loaded('data');
667 my $filters = $self->_current_filters;
668 my $all_filters = join(' ',sort keys %$filters,'order:',$self->session('order'));
669 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
670 if defined $loaded->{$path}->{filtered}->{$all_filters};
672 if ( ! $filtered || $all ) {
673 $filtered = [ 0 .. $#{ $data->{items} } ];
674 warn "filter all values\n";
676 warn "filter using $all_filters\n";
679 foreach my $i ( @$filtered ) {
680 my $item = $data->{items}->[$i];
681 if ( ! exists $item->{$name} ) {
682 $facet->{ _missing }++;
683 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
684 $facet->{$_}++ foreach @{ $item->{$name} };
686 $facet->{ $item->{$name} }++;
690 # warn "# facet $name ",dump $facet;
695 : defined $filters->{$name} ? @{ $filters->{$name} }
698 $checked = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
700 my $numeric = $self->_is_numeric($name);
702 my $sort = $self->param('sort');
703 # sort numeric facets with more than 5 values ascending
704 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
706 @facet_names = sort {
708 if ( $sort eq 'a' ) {
709 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
710 } elsif ( $sort eq 'd' ) {
711 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
712 } elsif ( $sort eq 'c' ) {
713 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
715 warn "unknown sort: $sort";
718 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
722 $self->render( name => $name, facet => $facet, checked => $checked,
723 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
729 my $content = $self->param('content');
731 my $i = $self->param('_row_id');
732 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
733 my $path = $self->param('path') || die "no path";
734 my $name = $self->param('name') || die "no name";
735 my $status = 200; # 200 = OK, 201 = Created
737 my $data = $self->_loaded('data');
739 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
740 $content =~ s/^\s+//s;
741 $content =~ s/\s+$//s;
743 if ( $content =~ /\xB6/ ) { # para
744 $v = [ split(/\s*\xB6\s*/, $content) ];
749 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
752 && ! ( $old eq 'undef' && length($content) == 0 ) # new value empty, previous undef
758 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
760 time => $self->param('time') || time(),
761 user => $self->param('user') || $ENV{'LOGNAME'},
763 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
764 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
765 keys %{ $loaded->{$path}->{stats} }
768 my $edit_path = $self->app->home->rel_dir('data') . '/' . $path . '.edits';
769 mkdir $edit_path unless -d $edit_path;
770 $edit_path .= '/' . $edit->{time};
771 store $edit, $edit_path;
772 utime $edit->{time}, $edit->{time}, $edit_path;
773 warn "# $edit_path ", dump($edit);
775 warn "# edit $path $i $old -> $new\n";
776 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
778 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
779 delete $loaded->{$path}->{sorted}->{$name};
780 warn "# invalidate $path sorted $name\n";
783 foreach ( grep { m/$name/ } keys %{ $loaded->{$path}->{filtered} } ) {
784 delete $loaded->{$path}->{filtered}->{$_};
785 warn "# invalidate $path filtered $_\n";
788 $status = 201; # created
789 $self->session('save_path' => $path);
792 warn "# unchanged $path $i $old\n";
796 $content = "$path $i $name doesn't exist\n";
800 warn "# edit $status $content";
811 my $path = $self->_param_or_session('path');
812 my $dump_path = $self->_save( $path );
813 $self->session('save_path' => 0);
815 $self->redirect_to( '/data/items' );