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 my $redirect_to = '/data/items';
266 $self->session( 'header' => $loaded->{$path}->{header} );
267 if ( ! defined $loaded->{$path}->{columns} ) {
268 my $columns_path = $self->_permanent_path( 'columns' );
269 if ( -e $columns_path ) {
270 my @columns = map { s/[\r\n]+$//; $_ } read_file $columns_path;
271 $loaded->{$path}->{columns} = [ @columns ];
272 warn "# columns_path $columns_path ",dump(@columns);
274 $loaded->{$path}->{columns} = $loaded->{$path}->{header}
277 $redirect_to = '/data/columns';
279 $self->session( 'columns' => $loaded->{$path}->{columns} );
280 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
281 $self->redirect_to( $redirect_to );
286 my ( $self, $name ) = @_;
287 my $path = $self->session('path') || $self->param('path');
288 $self->redirect_to('/data/index') unless $path;
289 if ( ! defined $loaded->{$path}->{$name} ) {
290 warn "$path $name isn't loaded\n";
291 $self->_load_path( $path );
292 $self->redirect_to('/data/index')
293 unless defined $loaded->{$path}->{$name};
295 return $loaded->{$path}->{$name};
302 $checked->{$_}++ foreach @_;
303 # warn "# _checked ",dump($checked);
307 sub _permanent_path {
309 my $path = $self->_param_or_session('path');
310 $self->app->home->rel_dir('data') . '/' . join('.', $path, @_);
315 my $path = $self->_param_or_session('path');
316 my $dir = $self->app->home->rel_dir('public') . '/export/';
317 mkdir $dir unless -e $dir;
319 mkdir $dir unless -e $dir;
320 $dir . '/' . join('.', @_);
326 if ( $self->param('columns') ) {
327 my @columns = $self->_param_array('columns');
328 write_file( $self->_permanent_path( 'columns' ), map { "$_\n" } @columns );
329 $self->redirect_to('/data/items');
332 my $stats = $self->_loaded( 'stats' );
335 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
337 foreach my $c ( @{ $self->session( 'header' ) } ) {
338 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
342 message => 'Select columns to display',
344 columns => \@columns,
345 checked => $self->_checked( $self->_param_array('columns') ),
350 my ($self,$name) = @_;
352 my @array = $self->param($name);
353 my $path = $self->session('path');
356 $self->session($name => [ @array ]);
357 $loaded->{$path}->{$name} = [ @array ];
358 } elsif ( my $session = $self->session($name) ) {
359 if ( ref $session eq 'ARRAY' ) {
362 die "$name not array ",dump($session);
365 #warn "# $name ",dump @array;
370 my ($self,$name,$default) = @_;
372 my $scalar = $self->param($name);
374 if ( defined $scalar ) {
375 $self->session($name => $scalar);
377 $scalar = $self->session($name);
380 if ( ! defined $scalar ) {
382 die "no default for $name" unless defined $scalar;
383 $self->session($name => $scalar);
386 warn "# _perm_scalar $name ",dump $scalar;
393 my $name = $self->param('filter_name') || die "name?";
394 my @vals = $self->param('filter_vals');
396 $self->_remove_filter( $name );
398 $self->_filter_on_data( $name, @vals );
399 if ( my $permanent = $self->param('_permanent') ) {
400 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
401 write_file $permanent_path, map { "$_\n" } @vals;
402 warn "permanent filter $permanent_path ", -s $permanent_path;
406 $self->session( 'offset' => 0 );
408 $self->redirect_to('/data/items');
411 sub _filter_on_data {
412 my ( $self, $name, @vals ) = @_;
414 my $path = $self->session('path');
416 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
417 warn "filter $name not found in data set";
421 $filters->{$name} = [ @vals ];
422 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
425 $filter_hash->{$_}++ foreach @vals;
427 #warn "# filter_hash ",dump( $filter_hash );
429 my $items = $self->_loaded('data')->{items};
431 my $include_missing = defined $filter_hash->{_missing};
434 foreach my $i ( 0 .. $#$items ) {
436 if ( defined $items->[$i]->{$name} ) {
437 foreach my $v ( @{ $items->[$i]->{$name} } ) {
438 if ( defined $filter_hash->{ $v } ) {
439 $filtered_items->{$i}++;
442 } elsif ( $include_missing ) {
443 $filtered_items->{$i}++;
447 #warn "# filter $name ",dump($filtered_items);
449 $loaded->{$path}->{filters}->{$name} = $filtered_items;
450 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
454 my ( $self, $all ) = @_;
455 my $data = $self->_loaded( 'data' );
457 return @{ $data->{items} } if $all == 1;
459 my $filters = $self->_current_filters;
461 foreach my $f ( keys %$filters ) {
462 foreach my $n ( @{ $filters->{$f} } ) {
463 $filter_value->{$f}->{$n} = 1;
466 my @items = @{ $data->{items} };
470 foreach my $n ( keys %$filter_value ) {
471 if ( ! exists $i->{$n} ) {
472 if ( defined $filter_value->{$n}->{_missing} ) {
480 # and match any of values in element
482 foreach my $v ( @{ $i->{$n} } ) { # FIXME not array?
483 $have_values ||= 1 if defined $filter_value->{$n}->{$v};
485 if ( ! $have_values ) {
491 } @items if $filter_value;
496 sub _current_filters {
499 $current_filters->{ $_ } = $filters->{ $_ }
501 grep { defined $filters->{ $_ } }
502 @{ $self->_loaded('header') }
504 #warn "# current_filters ",dump($current_filters);
505 return $current_filters;
508 sub _data_sorted_by {
509 my ( $self, $order ) = @_;
511 my $path = $self->session('path');
513 warn "_data_sorted_by $order from $path";
515 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
516 return $loaded->{$path}->{sorted}->{$order};
519 my $data = $self->_loaded( 'data' );
520 my $numeric = $self->_is_numeric($order);
521 my $missing = $numeric ? 0 : '';
522 no warnings qw(numeric);
533 [ $nr++, exists $_->{$order} ? join('', @{$_->{$order}}) : $missing ]
534 } @{ $data->{items} }
537 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
538 #warn "# sorted ",dump( @sorted );
540 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
547 my $path = $self->session('path');
548 $self->redirect_to('/data/index') unless defined $loaded->{ $path };
550 my @columns = $self->_param_array('columns');
551 $self->redirect_to('/data/columns') unless @columns;
552 my $order = $self->_param_scalar('order', $columns[0]);
553 my $sort = $self->_param_scalar('sort', 'a');
554 my $offset = $self->_param_scalar('offset', 0);
555 my $limit = $self->_param_scalar('limit', 20);
556 $self->_param_scalar('show', 'table');
558 # fix offset when changing limit
559 $offset = int( $offset / $limit ) * $limit;
561 my $sorted = $self->_data_sorted_by( $order );
565 foreach my $name ( keys %$filters ) {
566 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
567 warn "skip filter $name not found in $path\n";
570 push @filter_names, $name;
572 warn "filter_names ",dump( @filter_names );
573 foreach my $name ( @filter_names ) {
574 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
575 $self->_filter_on_data( $name, @{ $filters->{$name} } );
579 my $all_filters = join(' ',sort @filter_names,'order:',$order);
581 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
583 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
585 my $path_filters = $loaded->{$path}->{filters};
587 warn "create combined filter for $all_filters\n";
590 foreach my $i ( 0 .. $#$sorted ) {
591 my $pos = $sorted->[$i];
593 if ( $#filter_names == -1 ) {
594 push @filtered, $pos;
599 foreach ( @filter_names ) {
600 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
604 push @filtered, $pos;
607 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
610 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
611 if defined $loaded->{$path}->{filtered}->{$all_filters};
613 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
616 my $data = $self->_loaded('data');
617 my $from_end = $sort eq 'd' ? $#$filtered : 0;
618 foreach ( 0 .. $limit ) {
619 my $i = $_ + $offset;
620 last unless defined $filtered->[$i];
621 $i = $from_end - $i if $from_end;
622 my $id = $filtered->[$i];
624 my $item = $data->{items}->[ $id ];
625 $item->{_row_id} ||= $id;
628 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
634 sorted => $sorted_items,
635 columns => [ @columns ],
636 rows => $#$filtered + 1,
637 numeric => { map { $_, $self->_is_numeric($_) } @columns },
638 filters => $self->_current_filters,
646 $self->session('order', $self->param('order'));
647 $self->session('sort', $self->param('sort'));
648 $self->redirect_to('/data/items');
652 my ( $self, $name ) = @_;
654 my $stats = $self->_loaded( 'stats' );
656 # sort facet numerically if more >50% elements are numeric
657 my $count = $stats->{$name}->{count};
658 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
659 defined $stats->{$name}->{numeric} &&
660 $stats->{$name}->{numeric} > $count / 2;
664 my ($self,$name) = @_;
665 warn "_remove_filter $name\n";
667 my $path = $self->session('path');
669 delete $filters->{$name};
670 delete $loaded->{$path}->{filters}->{$name};
671 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
675 keys %{ $loaded->{$path}->{filtered} }
677 delete $loaded->{$path}->{filtered}->{$_};
678 warn "remove filtered cache $_";
685 my $path = $self->session('path') || $self->redirect_to( '/data/index' );
687 if ( my $name = $self->param('remove') ) {
688 $self->_remove_filter( $name );
689 $self->redirect_to( '/data/items' );
693 my $name = $self->param('name') || die "no name";
695 my $all = $self->_param_scalar('all', 1);
696 my $data = $self->_loaded('data');
698 my $filters = $self->_current_filters;
699 my $all_filters = join(' ',sort keys %$filters,'order:',$self->session('order'));
700 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
701 if defined $loaded->{$path}->{filtered}->{$all_filters};
703 if ( ! $filtered || $all ) {
704 $filtered = [ 0 .. $#{ $data->{items} } ];
705 warn "filter all values\n";
707 warn "filter using $all_filters\n";
710 foreach my $i ( @$filtered ) {
711 my $item = $data->{items}->[$i];
712 if ( ! exists $item->{$name} ) {
713 $facet->{ _missing }++;
714 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
715 $facet->{$_}++ foreach @{ $item->{$name} };
717 $facet->{ $item->{$name} }++;
721 # warn "# facet $name ",dump $facet;
726 : defined $filters->{$name} ? @{ $filters->{$name} }
729 $checked = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
731 my $numeric = $self->_is_numeric($name);
733 my $sort = $self->param('sort');
734 # sort numeric facets with more than 5 values ascending
735 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
737 @facet_names = sort {
739 if ( $sort eq 'a' ) {
740 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
741 } elsif ( $sort eq 'd' ) {
742 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
743 } elsif ( $sort eq 'c' ) {
744 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
746 warn "unknown sort: $sort";
749 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
753 $self->render( name => $name, facet => $facet, checked => $checked,
754 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
760 my $content = $self->param('content');
762 my $i = $self->param('_row_id');
763 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
764 my $path = $self->param('path') || die "no path";
765 my $name = $self->param('name') || die "no name";
766 my $status = 200; # 200 = OK, 201 = Created
768 my $data = $self->_loaded('data');
770 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
771 $content =~ s/^\s+//s;
772 $content =~ s/\s+$//s;
774 if ( $content =~ /\xB6/ ) { # para
775 $v = [ split(/\s*\xB6\s*/, $content) ];
780 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
783 && ! ( $old eq 'undef' && length($content) == 0 ) # new value empty, previous undef
789 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
791 time => $self->param('time') || time(),
792 user => $self->param('user') || $ENV{'LOGNAME'},
794 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
795 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
796 keys %{ $loaded->{$path}->{stats} }
799 my $edit_path = $self->_permanent_path( 'edits' );
800 mkdir $edit_path unless -d $edit_path;
801 $edit_path .= '/' . $edit->{time};
802 store $edit, $edit_path;
803 utime $edit->{time}, $edit->{time}, $edit_path;
804 warn "# $edit_path ", dump($edit);
806 warn "# edit $path $i $old -> $new\n";
807 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
809 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
810 delete $loaded->{$path}->{sorted}->{$name};
811 warn "# invalidate $path sorted $name\n";
814 foreach ( grep { m/$name/ } keys %{ $loaded->{$path}->{filtered} } ) {
815 delete $loaded->{$path}->{filtered}->{$_};
816 warn "# invalidate $path filtered $_\n";
819 $status = 201; # created
820 $self->session('save_path' => $path);
823 warn "# unchanged $path $i $old\n";
827 $content = "$path $i $name doesn't exist\n";
831 warn "# edit $status $content";
842 my $path = $self->_param_or_session('path');
843 my $dump_path = $self->_save( $path );
844 $self->session('save_path' => 0);
846 $self->redirect_to( '/data/items' );