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/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
34 push @{ $changes->{$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 warn "save loaded to $dump_path";
72 my $info = $loaded->{$path};
73 store $info, $dump_path;
76 #my $mtime = $loaded->{$path}->{mtime};
77 #utime $mtime, $mtime, $dump_path;
79 warn $dump_path, ' ', -s $dump_path, " bytes\n";
84 my ( $self, $path ) = @_;
86 my $full_path = $self->app->home->rel_file( 'data/' . $path );
87 die "$full_path $!" unless -r $full_path;
89 my $dump_path = $self->_dump_path( $path );
91 if ( defined $loaded->{$path}->{data} ) {
92 my $mtime = (stat($full_path))[9];
93 return if $loaded->{$path}->{mtime} == $mtime;
94 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
95 } elsif ( -e $dump_path ) {
96 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
97 my $info = retrieve $dump_path;
98 $loaded->{ $path } = $info;
102 # we could use Mojo::JSON here, but it's too slow
103 # $data = from_json read_file $path;
104 my $data = read_file $full_path;
105 warn "# data snippet: ", substr($data,0,200);
107 if ( $path =~ m/\.js/ ) {
108 Encode::_utf8_on($data);
109 $data = from_json $data;
110 } elsif ( $path =~ m/\.txt/ ) {
111 my @lines = split(/\r?\n/, $data);
112 $data = { items => [] };
114 my $header_line = shift @lines;
115 my $multiline = $header_line =~ s/\^//g;
116 @header = split(/\|/, $header_line );
117 warn "# header ", dump( @header );
118 while ( my $line = shift @lines ) {
121 my @v = split(/\|/, $line);
122 while ( @lines && $#v < $#header ) {
126 my @more_v = split(/\|/, $line);
127 if ( $#v + $#more_v > $#header ) {
128 warn "short line: ",dump( @v );
132 $v[ $#v ] .= shift @more_v if @more_v;
133 push @v, @more_v if @more_v;
135 if ( $#v > $#header ) {
136 die "# splice $#header ", dump( @v );
137 @v = splice @v, 0, $#header;
141 foreach my $i ( 0 .. $#v ) {
142 $item->{ $header[$i] || "f_$i" } = [ $v[$i] ];
144 push @{ $data->{items} }, $item;
147 warn "file format unknown $path";
152 foreach my $e ( @{ $data->{items} } ) {
153 foreach my $n ( keys %$e ) {
154 $stats->{$n}->{count}++;
156 if ( ref $e->{$n} eq 'ARRAY' ) {
157 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
163 foreach my $x ( @v ) {
164 $stats->{$n}->{numeric}++
165 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
166 $stats->{$n}->{empty}++
167 if length $x == 0; # faster than $x =~ m/^\s*$/;
173 foreach my $n ( keys %$stats ) {
174 next unless defined $stats->{$n}->{array};
175 delete $stats->{$n}->{array}
176 if $stats->{$n}->{array} == $stats->{$n}->{count};
180 if ( defined $data->{header} ) {
181 if ( ref $data->{header} eq 'ARRAY' ) {
182 @header = @{ $data->{header} };
184 warn "header not array ", dump( $data->{header} );
190 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
191 grep { defined $stats->{$_}->{count} } keys %$stats
197 header => [ @header ],
199 full_path => $full_path,
200 size => -s $full_path,
201 mtime => (stat($full_path))[9],
205 $loaded->{ $path } = $info;
206 $self->_save( $path );
214 my @paths = $self->param('paths');
215 warn "# paths ", dump @paths;
216 $self->_load_path( $_ ) foreach @paths;
218 my $path = $self->param('path') || $self->redirect_to( '/data/index' );
219 warn "# path $path\n";
220 $self->session('path' => $path);
221 $self->_load_path( $path );
223 $self->session( 'header' => $loaded->{$path}->{header} );
224 if ( ! defined $loaded->{$path}->{columns} ) {
225 $self->session( 'columns' => $loaded->{$path}->{header} );
226 $self->session( 'order' => $loaded->{$path}->{header}->[0] );
227 $self->redirect_to( '/data/columns' );
229 $self->session( 'columns' => $loaded->{$path}->{columns} );
230 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
231 $self->redirect_to( '/data/items' );
237 my ( $self, $name ) = @_;
238 my $path = $self->session('path') || $self->param('path');
239 $self->redirect_to('/data/index') unless $path;
240 if ( ! defined $loaded->{$path}->{$name} ) {
241 warn "$path $name isn't loaded\n";
242 $self->_load_path( $path );
243 $self->redirect_to('/data/index')
244 unless defined $loaded->{$path}->{$name};
246 return $loaded->{$path}->{$name};
253 $checked->{$_}++ foreach @_;
254 # warn "# _checked ",dump($checked);
262 if ( $self->param('columns') ) {
263 $self->_perm_array('columns');
264 $self->redirect_to('/data/items');
267 my $stats = $self->_loaded( 'stats' ); # || $self->redirect_to( '/data/index' );
270 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
272 foreach my $c ( @{ $self->session( 'header' ) } ) {
273 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
277 message => 'Select columns to display',
279 columns => \@columns,
280 checked => $self->_checked( $self->_perm_array('columns') ),
285 my ($self,$name) = @_;
287 my @array = $self->param($name);
288 my $path = $self->session('path');
291 $self->session($name => [ @array ]);
292 $loaded->{$path}->{$name} = [ @array ];
293 } elsif ( my $session = $self->session($name) ) {
294 if ( ref $session eq 'ARRAY' ) {
297 die "$name not array ",dump($session);
300 #warn "# $name ",dump @array;
305 my ($self,$name,$default) = @_;
307 my $scalar = $self->param($name);
309 if ( defined $scalar ) {
310 $self->session($name => $scalar);
312 $scalar = $self->session($name);
315 if ( ! defined $scalar ) {
317 die "no default for $name" unless defined $scalar;
318 $self->session($name => $scalar);
321 warn "# _perm_scalar $name ",dump $scalar;
328 my $name = $self->param('filter_name') || die "name?";
329 my @vals = $self->param('filter_vals');
331 $self->_remove_filter( $name );
332 $self->_filter_on_data( $name, @vals ) if @vals;
334 $self->session( 'offset' => 0 );
336 $self->redirect_to('/data/items');
339 sub _filter_on_data {
340 my ( $self, $name, @vals ) = @_;
342 my $path = $self->session('path');
344 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
345 warn "filter $name not found in data set";
349 $filters->{$name} = [ @vals ];
350 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
353 $filter_hash->{$_}++ foreach @vals;
355 #warn "# filter_hash ",dump( $filter_hash );
357 my $items = $self->_loaded('data')->{items};
359 my $include_missing = defined $filter_hash->{_missing};
362 foreach my $i ( 0 .. $#$items ) {
364 if ( defined $items->[$i]->{$name} ) {
365 foreach my $v ( @{ $items->[$i]->{$name} } ) {
366 if ( defined $filter_hash->{ $v } ) {
367 $filtered_items->{$i}++;
370 } elsif ( $include_missing ) {
371 $filtered_items->{$i}++;
375 #warn "# filter $name ",dump($filtered_items);
377 $loaded->{$path}->{filters}->{$name} = $filtered_items;
378 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
382 my ( $self, $all ) = @_;
383 my $data = $self->_loaded( 'data' );
385 return @{ $data->{items} } if $all == 1;
387 my $filters = $self->_current_filters;
389 foreach my $f ( keys %$filters ) {
390 foreach my $n ( @{ $filters->{$f} } ) {
391 $filter_value->{$f}->{$n} = 1;
394 my @items = @{ $data->{items} };
398 foreach my $n ( keys %$filter_value ) {
399 if ( ! exists $i->{$n} ) {
400 if ( defined $filter_value->{$n}->{_missing} ) {
408 # and match any of values in element
410 foreach my $v ( @{ $i->{$n} } ) { # FIXME not array?
411 $have_values ||= 1 if defined $filter_value->{$n}->{$v};
413 if ( ! $have_values ) {
419 } @items if $filter_value;
424 sub _current_filters {
427 $current_filters->{ $_ } = $filters->{ $_ }
429 grep { defined $filters->{ $_ } }
430 @{ $self->_loaded('header') }
432 #warn "# current_filters ",dump($current_filters);
433 return $current_filters;
436 sub _data_sorted_by {
437 my ( $self, $order ) = @_;
439 my $path = $self->session('path');
441 warn "_data_sorted_by $order from $path";
443 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
444 return $loaded->{$path}->{sorted}->{$order};
447 my $data = $self->_loaded( 'data' );
448 my $numeric = $self->_is_numeric($order);
449 my $missing = $numeric ? 0 : '';
450 no warnings qw(numeric);
461 [ $nr++, exists $_->{$order} ? join('', @{$_->{$order}}) : $missing ]
462 } @{ $data->{items} }
465 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
466 #warn "# sorted ",dump( @sorted );
468 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
475 my $path = $self->session('path');
476 $self->redirect_to('/data/index') unless defined $loaded->{ $path };
478 my @columns = $self->_perm_array('columns');
479 $self->redirect_to('/data/columns') unless @columns;
480 my $order = $self->_perm_scalar('order', $columns[0]);
481 my $sort = $self->_perm_scalar('sort', 'a');
482 my $offset = $self->_perm_scalar('offset', 0);
483 my $limit = $self->_perm_scalar('limit', 20);
484 $self->_perm_scalar('show', 'table');
486 # fix offset when changing limit
487 $offset = int( $offset / $limit ) * $limit;
489 my $sorted = $self->_data_sorted_by( $order );
493 foreach my $name ( keys %$filters ) {
494 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
495 warn "skip filter $name not found in $path\n";
498 push @filter_names, $name;
500 warn "filter_names ",dump( @filter_names );
501 foreach my $name ( @filter_names ) {
502 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
503 $self->_filter_on_data( $name, @{ $filters->{$name} } );
507 my $all_filters = join(' ',sort @filter_names,'order:',$order);
509 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
511 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
513 my $path_filters = $loaded->{$path}->{filters};
515 warn "create combined filter for $all_filters\n";
518 foreach my $i ( 0 .. $#$sorted ) {
519 my $pos = $sorted->[$i];
521 if ( $#filter_names == -1 ) {
522 push @filtered, $pos;
527 foreach ( @filter_names ) {
528 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
532 push @filtered, $pos;
535 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
538 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
539 if defined $loaded->{$path}->{filtered}->{$all_filters};
541 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
544 my $data = $self->_loaded('data');
545 my $from_end = $sort eq 'd' ? $#$filtered : 0;
546 foreach ( 0 .. $limit ) {
547 my $i = $_ + $offset;
548 last unless defined $filtered->[$i];
549 $i = $from_end - $i if $from_end;
550 my $id = $filtered->[$i];
552 my $item = $data->{items}->[ $id ];
553 $item->{_row_id} ||= $id;
556 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
562 sorted => $sorted_items,
563 columns => [ @columns ],
564 rows => $#$filtered + 1,
565 numeric => { map { $_, $self->_is_numeric($_) } @columns },
566 filters => $self->_current_filters,
574 $self->session('order', $self->param('order'));
575 $self->session('sort', $self->param('sort'));
576 $self->redirect_to('/data/items');
580 my ( $self, $name ) = @_;
582 my $stats = $self->_loaded( 'stats' );
584 # sort facet numerically if more >50% elements are numeric
585 my $count = $stats->{$name}->{count};
586 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
587 defined $stats->{$name}->{numeric} &&
588 $stats->{$name}->{numeric} > $count / 2;
592 my ($self,$name) = @_;
593 warn "_remove_filter $name\n";
595 my $path = $self->session('path');
597 delete $filters->{$name};
598 delete $loaded->{$path}->{filters}->{$name};
599 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
603 keys %{ $loaded->{$path}->{filtered} }
605 delete $loaded->{$path}->{filtered}->{$_};
606 warn "remove filtered cache $_";
613 my $path = $self->session('path') || $self->redirect_to( '/data/index' );
615 if ( my $name = $self->param('remove') ) {
616 $self->_remove_filter( $name );
617 $self->redirect_to( '/data/items' );
621 my $name = $self->param('name') || die "no name";
623 my $all = $self->_perm_scalar('all', 1);
624 my $data = $self->_loaded('data');
626 my $filters = $self->_current_filters;
627 my $all_filters = join(' ',sort keys %$filters,'order:',$self->session('order'));
628 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
629 if defined $loaded->{$path}->{filtered}->{$all_filters};
631 if ( ! $filtered || $all ) {
632 $filtered = [ 0 .. $#{ $data->{items} } ];
633 warn "filter all values\n";
635 warn "filter using $all_filters\n";
638 foreach my $i ( @$filtered ) {
639 my $item = $data->{items}->[$i];
640 if ( ! exists $item->{$name} ) {
641 $facet->{ _missing }++;
642 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
643 $facet->{$_}++ foreach @{ $item->{$name} };
645 $facet->{ $item->{$name} }++;
649 # warn "# facet $name ",dump $facet;
654 : defined $filters->{$name} ? @{ $filters->{$name} }
657 $checked = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
659 my $numeric = $self->_is_numeric($name);
661 my $sort = $self->param('sort');
662 # sort numeric facets with more than 5 values ascending
663 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
665 @facet_names = sort {
667 if ( $sort eq 'a' ) {
668 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
669 } elsif ( $sort eq 'd' ) {
670 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
671 } elsif ( $sort eq 'c' ) {
672 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
674 warn "unknown sort: $sort";
677 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
681 $self->render( name => $name, facet => $facet, checked => $checked,
682 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
688 my $content = $self->param('content');
690 my $i = $self->param('_row_id');
691 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
692 my $path = $self->param('path') || die "no path";
693 my $name = $self->param('name') || die "no name";
694 my $status = 200; # 200 = OK, 201 = Created
696 my $data = $self->_loaded('data');
698 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
699 $content =~ s/^\s+//s;
700 $content =~ s/\s+$//s;
702 if ( $content =~ /\xB6/ ) { # para
703 $v = [ split(/\s*\xB6\s*/, $content) ];
708 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
711 && ! ( $old eq 'undef' && length($content) == 0 ) # new value empty, previous undef
717 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
719 time => $self->param('time') || time(),
720 user => $self->param('user') || $ENV{'LOGNAME'},
722 my $change_path = $self->app->home->rel_dir('data') . '/' . $path . '.changes';
723 mkdir $change_path unless -d $change_path;
724 $change_path .= '/' . $change->{time};
725 store $change, $change_path;
726 warn "# $change_path ", dump($change);
728 warn "# change $path $i $old -> $new\n";
729 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
731 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
732 delete $loaded->{$path}->{sorted}->{$name};
733 warn "# invalidate $path sorted $name\n";
736 foreach ( grep { m/$name/ } keys %{ $loaded->{$path}->{filtered} } ) {
737 delete $loaded->{$path}->{filtered}->{$_};
738 warn "# invalidate $path filtered $_\n";
741 $status = 201; # created
742 $self->session('save_path' => $path);
745 warn "# unchanged $path $i $old\n";
749 $content = "$path $i $name doesn't exist\n";
753 warn "# edit $status $content";
764 my $path = $self->param('path');
765 $path ||= $self->session('path');
767 my $dump_path = $self->_save( $path );
768 $self->session('save_path' => 0);
770 $self->redirect_to( '/data/items' );