1 package MojoFacets::Data;
6 use base 'Mojolicious::Controller';
8 use Data::Dump qw(dump);
14 use Time::HiRes qw(time);
15 use File::Path qw(mkpath);
16 use Text::Unaccent::PurePerl;
19 use MojoFacets::Import::File;
20 use MojoFacets::Import::HTMLTable;
21 use MojoFacets::Import::CSV;
22 use MojoFacets::Import::CouchDB;
23 use MojoFacets::Import::SQL;
31 my $data_dir = $self->app->home->rel_dir('data');
32 die "no data dir $data_dir" unless -d $data_dir;
37 my $file = $File::Find::name;
38 if ( -f $file && $file =~ m/\.(js(on)?|txt)$/ ) {
39 $file =~ s/$data_dir\/*//;
41 } elsif ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
42 push @{ $changes->{$1} }, $2
43 } elsif ( -d $file && $file =~ m/\.html$/ ) {
44 $file =~ s/$data_dir\/*//;
46 } elsif ( -f $file && $file =~ m/\.(csv|storabe|couchdb|sql)$/i ) {
47 $file =~ s/$data_dir\/*//;
50 #warn "IGNORE: $file\n";
54 no warnings qw(uninitialized); # mtime
55 @files = sort { $loaded->{$b}->{mtime} <=> $loaded->{$a}->{mtime} || lc $a cmp lc $b } @files,
56 grep { defined $loaded->{$_}->{generated} } keys %$loaded;
58 $size->{$_} = -s "$data_dir/$_" foreach @files;
65 dump_path => { map { $_ => $self->_dump_path($_) } @files },
71 my ( $self, $name ) = @_;
72 my $dir = $self->app->home->rel_dir('data');
75 return '/tmp/mojo_facets.' . $name . '.storable';
79 my ( $self, $path ) = @_;
81 my $dump_path = $self->_dump_path( $path );
82 my $first_load = ! -e $dump_path;
83 warn "save loaded to $dump_path";
84 my $info = $loaded->{$path};
85 store $info, $dump_path;
88 my $mtime = $loaded->{$path}->{mtime};
89 utime $mtime, $mtime, $dump_path;
90 warn "sync time to $path at $mtime\n";
93 warn $dump_path, ' ', -s $dump_path, " bytes\n";
102 my $nr_items = $#{ $_[0] } + 1;
104 warn "__stats $nr_items\n";
106 foreach my $e ( @{ $_[0] } ) {
107 foreach my $n ( keys %$e ) {
108 $stats->{$n}->{count}++;
110 if ( ref $e->{$n} eq 'ARRAY' ) {
111 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
117 foreach my $x ( @v ) {
118 $stats->{$n}->{numeric}++
119 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
120 $stats->{$n}->{empty}++
121 if length $x == 0; # faster than $x =~ m/^\s*$/;
127 foreach my $n ( keys %$stats ) {
128 my $s = $stats->{$n};
129 next unless defined $s->{array};
130 if ( $s->{array} == $s->{count} ) {
132 if ( $s->{count} == $nr_items ) {
133 warn "check $n for uniqeness\n";
135 foreach my $e ( @{ $_[0] } ) {
136 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
142 $stats->{$n}->{unique} = 1;
143 #warn "# $n unique ",dump( $unique );
149 warn "# __stats ",dump($stats);
154 sub _param_or_session {
155 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
160 my $path = $self->_param_or_session('path');
161 warn "stats $path\n";
162 delete $loaded->{$path}->{stats};
163 return $self->redirect_to( '/data/columns' );
168 my ( $self, $path ) = @_;
170 return if defined $loaded->{$path}->{generated};
172 my $full_path = $self->app->home->rel_file( 'data/' . $path );
173 return $self->redirect_to('/data/index') unless -r $full_path;
175 my $dump_path = $self->_dump_path( $path );
177 if ( defined $loaded->{$path}->{data} ) {
178 my $mtime = (stat($full_path))[9];
179 return if $loaded->{$path}->{mtime} == $mtime;
180 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
181 } elsif ( -e $dump_path ) {
182 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
183 my $info = retrieve $dump_path;
184 $loaded->{ $path } = $info;
189 if ( -f $full_path ) {
190 if ( $full_path =~ m/.csv/i ) {
191 $data = MojoFacets::Import::CSV->new( full_path => $full_path )->data;
192 } elsif ( $full_path =~ m/.sql/i ) {
193 $data = MojoFacets::Import::SQL->new( full_path => $full_path )->data;
194 } elsif ( $full_path =~ m/.couchdb/i ) {
195 $data = MojoFacets::Import::CouchDB->new( full_path => $full_path )->data;
196 } elsif ( $full_path =~ m/.storable/ ) {
197 warn "open $full_path ", -s $full_path, " bytes";
198 open(my $pipe, "<", $full_path) || die $!;
199 while ( my $o = eval { Storable::fd_retrieve $pipe } ) {
200 if ( defined $o->{item} ) {
201 push @{ $data->{items} }, $o->{item};
203 warn "SKIP ",dump($o);
207 warn "loaded ", $#{ $data->{items} } + 1, " items from $full_path\n";
208 $data->{generated}++;
210 $data = MojoFacets::Import::File->new( full_path => $full_path, path => $path )->data;
212 } elsif ( -d $full_path && $full_path =~ m/.html/ ) {
213 $data = MojoFacets::Import::HTMLTable->new( dir => $full_path )->data;
215 die "can't load $full_path";
220 if ( defined $data->{header} ) {
221 if ( ref $data->{header} eq 'ARRAY' ) {
222 @header = @{ $data->{header} };
224 warn "header not array ", dump( $data->{header} );
228 my $stats = __stats( $data->{items} );
231 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
232 grep { defined $stats->{$_}->{count} } keys %$stats
236 header => [ @header ],
238 full_path => $full_path,
239 size => -s $full_path,
240 mtime => (stat($full_path))[9],
242 defined $data->{generated} ? ( generated => 1 ) : (),
245 $loaded->{ $path } = $info;
246 $self->_save( $path ) unless $info->{generated};
254 my $path = $self->param('path') || return $self->redirect_to( '/data/index' );
256 my @paths = $self->param('paths');
257 warn "# paths ", dump @paths;
259 foreach my $p ( keys %$loaded ) {
260 next if grep { /^\Q$p\E$/ } @paths;
261 warn "remove $p from memory\n";
262 delete $loaded->{$p};
265 $self->_load_path( $_ ) foreach @paths;
267 warn "# path $path\n";
268 $self->_load_path( $path );
270 $self->session( 'path' => $path );
272 my $redirect_to = '/data/items';
274 $self->session( 'header' => $loaded->{$path}->{header} );
275 if ( ! defined $loaded->{$path}->{columns} ) {
276 my $columns_path = $self->_permanent_path( 'columns' );
277 if ( -e $columns_path ) {
278 my @columns = map { s/[\r\n]+$//; $_ } read_file $columns_path, binmode => ':utf8';
279 $loaded->{$path}->{columns} = [ @columns ];
280 warn "# columns_path $columns_path ",dump(@columns);
282 $loaded->{$path}->{columns} = $loaded->{$path}->{header}
285 $redirect_to = '/data/columns';
287 $self->session( 'columns' => $loaded->{$path}->{columns} );
288 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
289 return $self->redirect_to( $redirect_to );
294 my ( $self, $name ) = @_;
295 my $path = $self->session('path') || $self->param('path');
296 return $self->redirect_to('/data/index') unless $path;
298 if ( defined $loaded->{$path}->{modified} && $loaded->{$path}->{modified} > 1 ) {
299 my $caller = (caller(1))[3];
300 if ( $caller =~ m/::edit/ ) {
301 warn "rebuild stats for $path ignored caller $caller\n";
303 warn "rebuild stats for $path FORCED by modified caller $caller\n";
304 # $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
305 $loaded->{$path}->{rebuild_stats} = 1;
306 $loaded->{$path}->{modified} = 1;
310 if ( defined $loaded->{$path}->{rebuild_stats} ) {
311 warn "rebuild_stats $path";
312 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
313 delete $loaded->{$path}->{rebuild_stats};
316 if ( ! defined $loaded->{$path}->{$name} ) {
317 warn "$path $name isn't loaded\n";
318 $self->_load_path( $path );
319 if ( ! defined $loaded->{$path}->{stats} ) {
320 warn "rebuild stats for $path\n";
321 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
323 if ( ! defined $loaded->{$path}->{$name} ) {
324 warn "MISSING $name for $path\n";
325 return $self->redirect_to('/data/index')
329 $self->session( 'modified' => $loaded->{$path}->{modified} );
331 return $loaded->{$path}->{$name};
338 $checked->{$_}++ foreach @_;
339 # warn "# _checked ",dump($checked);
343 sub _permanent_path {
345 my $path = $self->_param_or_session('path');
346 $self->app->home->rel_dir('data') . '/' . join('.', $path, @_);
351 $n = unac_string($n);
356 sub _column_from_unac {
357 my ($self,$name) = @_;
359 my $stats = $self->_loaded('stats');
360 my $cols_norm = { map { __unac( $_ ) => $_ } keys %$stats };
362 $cols_norm->{$name} || die "can't find column $name in ", dump($cols_norm);
367 my $path = $self->_param_or_session('path');
369 warn "no path in param or session";
372 my $dir = $self->app->home->rel_dir('public') . "/export/$path";
373 mkpath $dir unless -e $dir;
374 my $name = __export_path_name( $path, @_ );
375 my $full = $dir . '/' . $name;
376 $full =~ s/\/+$// if -d $full; # strip trailing slash for dirs
380 sub __export_path_name {
384 my $name = join('.', map { __unac($_) } @_ );
385 if ( length($name) > $max_length ) {
386 $name = substr($name,0,$max_length) . Digest::MD5::md5_hex substr($name,$max_length);
394 my $view_path = $self->_permanent_path( 'views' );
396 if ( $self->param('columns') ) {
397 my @columns = $self->_param_array('columns');
398 write_file( $self->_permanent_path( 'columns' ), { binmode => ':utf8' }, map { "$_\n" } @columns );
399 if ( my $view = $self->param('view') ) {
400 mkdir $view_path unless -e $view_path;
401 write_file( "$view_path/$view", { binmode => ':utf8' }, map { "$_\n" } @columns );
402 warn "view $view_path/$view ", -s "$view_path/$view", " bytes\n";
405 return $self->redirect_to('/data/items');
407 } elsif ( ! $self->session('header') ) {
408 return $self->redirect_to('/data/load');
411 if ( my $id = $self->param('id') ) {
412 my $view_full = "$view_path/$id";
413 if ( -f $view_full ) {
414 my @columns = map { chomp; $_ } read_file $view_full, binmode => ':utf8';
415 warn "view $view_full loaded ", dump @columns;
416 $self->session( 'columns' => [ @columns ] );
417 return $self->redirect_to('/data/items');
421 my $stats = $self->_loaded( 'stats' );
424 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
426 foreach my $c ( @{ $self->session( 'header' ) } ) {
427 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
431 if ( -d $view_path ) {
432 @views = map { s{^\Q$view_path\E/*}{}; $_ } glob "$view_path/*";
433 warn "# views ",dump @views;
437 message => 'Select columns to display',
439 columns => \@columns,
440 checked => $self->_checked( $self->_param_array('columns') ),
446 my ($self,$name) = @_;
448 my @array = $self->param($name);
449 my $path = $self->session('path');
452 $self->session($name => [ @array ]);
453 $loaded->{$path}->{$name} = [ @array ];
454 } elsif ( my $session = $self->session($name) ) {
455 if ( ref $session eq 'ARRAY' ) {
458 die "$name not array ",dump($session);
461 #warn "# $name ",dump @array;
466 my ($self,$name,$default) = @_;
468 my $scalar = $self->param($name);
470 if ( defined $scalar ) {
471 $self->session($name => $scalar);
473 $scalar = $self->session($name);
476 if ( ! defined $scalar ) {
478 die "no default for $name" unless defined $scalar;
479 $self->session($name => $scalar);
482 warn "# _param_scalar $name ",dump $scalar;
489 my $name = $self->param('filter_name') || die "name?";
490 my @vals = $self->param('filter_vals');
492 $self->_remove_filter( $name );
494 $self->_filter_on_data( $name, @vals );
495 if ( my $permanent = $self->param('_permanent') ) {
496 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
497 write_file $permanent_path, { binmode => ':utf8' }, map { "$_\n" } @vals;
498 warn "permanent filter $permanent_path ", -s $permanent_path;
502 $self->session( 'offset' => 0 );
504 return $self->redirect_to('/data/items');
507 sub _filter_on_data {
508 my ( $self, $name, @vals ) = @_;
510 my $path = $self->session('path');
512 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
513 warn "filter $name not found in data set";
517 $filters->{$name} = [ @vals ];
518 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
521 $filter_hash->{$_}++ foreach @vals;
523 #warn "# filter_hash ",dump( $filter_hash );
525 my $items = $self->_loaded('data')->{items};
527 my $include_missing = defined $filter_hash->{_missing};
530 foreach my $i ( 0 .. $#$items ) {
532 if ( defined $items->[$i]->{$name} ) {
533 my $row = $items->[$i]->{$name};
534 $row = [ $row ] unless ref $row eq 'ARRAY'; # FIXME probably wrong place
535 foreach my $v ( @$row ) {
536 if ( defined $filter_hash->{ $v } ) {
537 $filtered_items->{$i}++;
540 } elsif ( $include_missing ) {
541 $filtered_items->{$i}++;
545 #warn "# filter $name ",dump($filtered_items);
547 $loaded->{$path}->{filters}->{$name} = $filtered_items;
548 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
552 sub _current_filters {
555 my $stats = $self->_loaded('stats');
557 $current_filters->{ $_ } = $filters->{ $_ }
558 foreach ( grep { defined $filters->{ $_ } } keys %$stats )
560 warn "# _current_filters ",dump( keys %$current_filters );
561 return $current_filters;
564 sub _data_sorted_by {
565 my ( $self, $order ) = @_;
567 my $path = $self->session('path');
569 warn "_data_sorted_by $order from $path";
571 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
572 return $loaded->{$path}->{sorted}->{$order};
575 my $data = $self->_loaded( 'data' );
576 my $numeric = $self->_is_numeric($order);
577 my $missing = $numeric ? 0 : '';
578 no warnings qw(numeric);
590 if ( ! exists $_->{$order} ) {
592 } elsif ( ref $_->{$order} eq 'ARRAY' ) {
593 $v = join('', @{$_->{$order}});
598 } @{ $data->{items} }
601 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
602 #warn "# sorted ",dump( @sorted );
604 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
610 join(',', sort(@_), 'order', $order);
613 our ($out, $key,$value);
615 sub __commit_path_code {
616 my ( $path, $i, $code, $commit_changed ) = @_;
618 my $items = $loaded->{$path}->{data}->{items} || die "no items for $path";
619 my $row = $items->[$i];
622 foreach ( keys %$update ) {
623 $$commit_changed->{$_}++;
624 $loaded->{$path}->{data}->{items}->[$i]->{$_} = $update->{$_};
626 #warn "__commit_path_code $path $i ",dump( $update );
629 # uses templates/admin.html.ep
630 sub _switch_dataset {
635 foreach my $path ( keys %$loaded ) {
636 next unless exists $loaded->{$path}->{data};
637 push @$datasets, $path;
640 warn "# datasets ",dump($datasets);
642 $self->stash( 'datasets' => $datasets );
648 $self->_switch_dataset;
650 if ( my $show = $self->param('id') ) {
651 $self->param('show', $show);
655 my $path = $self->_param_scalar('path');
657 my @columns = $self->_param_array('columns');
658 return $self->redirect_to('/data/columns') unless @columns;
659 my $order = $self->_param_scalar('order', $columns[0]);
660 my $sort = $self->_param_scalar('sort', 'a');
661 my $offset = $self->_param_scalar('offset', 0);
662 my $limit = $self->_param_scalar('limit', 20);
663 $self->_param_scalar('show', 'table');
665 # fix offset when changing limit
666 $offset = int( $offset / $limit ) * $limit;
668 if ( ! grep { /^\Q$order\E$/ } @columns ) {
669 $order = $columns[0];
670 $self->session( order => $order );
672 my $sorted = $self->_data_sorted_by( $order );
676 foreach my $name ( keys %$filters ) {
677 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
678 warn "skip filter $name not found in $path\n";
681 push @filter_names, $name;
683 warn "filter_names ",dump( @filter_names );
684 foreach my $name ( @filter_names ) {
685 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
686 $self->_filter_on_data( $name, @{ $filters->{$name} } );
690 my $all_filters = __all_filters( @filter_names,$order );
692 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
694 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
696 my $path_filters = $loaded->{$path}->{filters};
698 warn "create combined filter for $all_filters from ", $#$sorted + 1, " items\n";
701 foreach my $i ( 0 .. $#$sorted ) {
702 my $pos = $sorted->[$i];
704 if ( $#filter_names == -1 ) {
705 push @filtered, $pos;
710 foreach ( @filter_names ) {
711 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
715 push @filtered, $pos;
718 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
721 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
722 if defined $loaded->{$path}->{filtered}->{$all_filters};
724 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
726 my $data = $self->_loaded('data');
728 my $code = $self->_param_scalar('code','');
730 $code =~ s{\n+$}{\n}s;
732 my $commit = $self->param('commit');
733 my $test = $self->param('test');
737 if ( $code && ( $test || $commit ) ) {
738 # XXX find columns used in code snippet and show them to user
740 foreach my $column ( $code =~ m/\$row->{([^}]+)}/g ) {
741 if ( $column =~ s/^(['"])// ) {
744 next if $column =~ m/\$/; # hide columns with vars in them
745 $commit_changed->{$column} = 0;
749 my $code_path = $self->app->home->rel_dir('public') . "/code";
752 warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
753 ( $key, $value, $out ) = ( 'key', 'value' );
754 foreach ( 0 .. $#$filtered ) {
755 my $i = $filtered->[$_];
756 __commit_path_code( $path, $i, $code, \$commit_changed );
759 # this might move before $out to recalculate stats on source dataset?
760 __path_rebuild_stats( $path );
761 my $c = { map { $_ => 1 } @columns };
762 my @added_columns = sort grep { ! $c->{$_} } keys %$commit_changed;
763 warn "# added_columns ",dump( @added_columns );
764 unshift @columns, @added_columns;
766 $loaded->{$path}->{columns} = [ @columns ];
767 warn "# new columns ",dump( @columns );
769 __invalidate_path_column( $path, $_ ) foreach keys %$commit_changed;
771 $self->_save_change({
773 time => $self->param('time') || time(),
774 user => $self->param('user') || $ENV{'LOGNAME'},
776 commit_changed => $commit_changed,
779 if ( my $description = $self->param('code_description') ) {
780 my $depends = $self->param('code_depends') || die "no code_depends?";
781 my $path = "$code_path/$depends.$description.pl";
782 if ( -e $path && ! $self->param('overwrite') ) {
783 warn "# code $path not saved\n";
785 write_file( $path, { binmode => ':utf8' }, "$code\n" );
786 warn "code $path ", -s $path, " bytes saved\n";
793 my $commit_dataset = join('.'
794 , $self->param('code_depends')
795 , $self->param('code_description')
800 warn "key $key value $value";
802 foreach my $n ( keys %$out ) {
803 my $i = { $key => [ $n ] };
804 my $ref = ref $out->{$n};
805 if ( $ref eq 'HASH' ) {
806 $i->{$_} = [ $out->{$n}->{$_} ] foreach keys %{ $out->{$n} };
807 } elsif ( $ref eq 'ARRAY' ) {
808 $i->{$_} = $out->{$n};
810 $i->{$value} = [ $out->{$n} ];
812 $i->{_error} = [ dump($out->{$n}) ];
817 my $stats = __stats( $items );
818 my @columns = grep { ! m/^\Q$key\E$/ } sort keys %$stats;
819 unshift @columns, $key;
821 $loaded->{$commit_dataset} = {
822 header => [ @columns ],
823 columns => [ @columns ],
825 data => { items => $items },
829 warn "# loaded out ", dump( $loaded->{$commit_dataset} );
830 $self->session('path', $commit_dataset);
831 $self->session('columns', [ @columns ]);
832 $self->session('order', $key);
833 return $self->redirect_to('/data/items');
836 $self->session('columns', [ @columns ]);
840 my $from_end = $sort eq 'd' ? $#$filtered : 0;
842 my ( $key, $value, $out ) = ( 'key', 'value' ); # XXX make local
843 foreach ( 0 .. $limit ) {
844 my $i = $_ + $offset;
845 last unless defined $filtered->[$i];
846 $i = $from_end - $i if $from_end;
847 my $id = $filtered->[$i];
848 my $row = Storable::dclone $data->{items}->[ $id ];
849 if ( $code && $test ) {
853 warn "ERROR evaling $@", dump($code);
854 $self->stash('eval_error', $@) if $@;
856 warn "EVAL ",dump($update);
857 foreach ( keys %$update ) {
858 $test_changed->{$_}++;
859 $row->{$_} = $update->{$_};
863 $row->{_row_id} ||= $id;
864 push @$sorted_items, $row;
867 if ( $self->param('export') ) {
868 my $export_path = $self->_export_path( 'items', @columns);
869 open(my $fh, '>', $export_path) || die "ERROR: can't open $export_path: $!";
870 print $fh "#",join("\t",@columns),"\n";
871 foreach my $f ( 0 .. $#$filtered ) {
872 print $fh join("\t", map {
873 my $i = $data->{items}->[ $filtered->[$f] ];
875 if ( ! defined $i->{$_} ) {
877 } elsif ( ref $i->{$_} eq 'ARRAY' ) {
878 $v =join(',', @{ $i->{$_} });
879 $v = '\N' if length($v) == 0;
880 } elsif ( ! ref $i->{$_} ) {
889 warn "export $export_path ", -s $export_path, " bytes\n";
892 warn "# test_changed ",dump( $test_changed );
893 my $c = { map { $_ => 1 } @columns };
894 my @added_columns = sort grep { ! $c->{$_} } keys %$test_changed;
895 unshift @columns, @added_columns;
897 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
900 my $tmp = $code; $tmp =~ s/\$row->{(['"]?)(\w+)\1/$depends_on->{$2}++/gse;
901 warn "# depends_on ",dump $depends_on;
903 my $test_added = Storable::dclone $test_changed;
904 delete $test_added->{$_} foreach keys %$depends_on;
906 my $code_depends = $self->param('code_depends')
907 || join(',', keys %$depends_on);
909 my $code_description = $self->param('code_description') ||
910 join(',', keys %$test_added);
912 $code_depends ||= $code_description; # self-modifing
913 if ( ! $code_depends && $out ) {
914 $code_depends = $key;
915 $code_description = $value;
918 warn "# test_changed ",dump( $test_changed, $code_depends, $code_description );
924 sorted => $sorted_items,
925 columns => [ @columns ],
926 rows => $#$filtered + 1,
927 numeric => { map { $_, $self->_is_numeric($_) } @columns },
928 unique => { map { $_, $self->_is_unique( $_) } @columns },
929 filters => $self->_current_filters,
931 cols_changed => $commit ? $commit_changed : $test_changed,
932 code_depends => $code_depends,
933 code_description => $code_description,
934 code_path => $code_path,
943 $self->session('order', $self->param('order'));
944 $self->session('sort', $self->param('sort'));
945 return $self->redirect_to('/data/items');
949 my ( $self, $name ) = @_;
951 my $stats = $self->_loaded( 'stats' );
953 # sort facet numerically if more >50% elements are numeric
954 my $count = $stats->{$name}->{count};
955 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
956 defined $stats->{$name}->{numeric} &&
957 $stats->{$name}->{numeric} > $count / 2;
961 my ( $self, $name ) = @_;
962 my $stats = $self->_loaded( 'stats' );
963 defined $stats->{$name}->{unique};
967 my ($self,$name) = @_;
968 warn "_remove_filter $name\n";
970 my $path = $self->session('path');
972 delete $filters->{$name};
973 delete $loaded->{$path}->{filters}->{$name};
974 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
978 keys %{ $loaded->{$path}->{filtered} }
980 delete $loaded->{$path}->{filtered}->{$_};
981 warn "remove filtered cache $_";
988 my $path = $self->session('path') || return $self->redirect_to( '/data/index' );
990 if ( my $name = $self->param('remove') ) {
991 $self->_remove_filter( $name );
992 return $self->redirect_to( '/data/items' );
996 my $name = $self->param('name') || die "no name";
998 my $all = $self->_param_scalar('all', 1);
999 my $data = $self->_loaded('data');
1001 my $filters = $self->_current_filters;
1002 my $all_filters = __all_filters( keys %$filters,$self->session('order') );
1003 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
1004 if defined $loaded->{$path}->{filtered}->{$all_filters};
1006 if ( ! $filtered || $all ) {
1007 $filtered = [ 0 .. $#{ $data->{items} } ];
1008 warn "filter all values\n";
1010 warn "filter using $all_filters\n";
1013 foreach my $i ( @$filtered ) {
1014 my $item = $data->{items}->[$i];
1015 if ( ! exists $item->{$name} ) {
1016 $facet->{ _missing }++;
1017 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
1018 $facet->{$_}++ foreach @{ $item->{$name} };
1020 $facet->{ $item->{$name} }++;
1024 my $checked_values = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
1026 if ( my $code = $self->param('code') ) {
1028 foreach my $value ( keys %$facet ) {
1029 my $count = $facet->{$value};
1030 my $checked = $checked_values->{$value};
1034 warn "ERROR: $@\n$code\n";
1036 } elsif ( $checked != $checked_values->{$value} ) {
1037 warn "checked $value $count -> $checked\n";
1038 $checked_values->{$value} = $checked;
1041 warn "out ",dump( $out );
1042 $self->stash( out => $out );
1045 # warn "# facet $name ",dump $facet;
1049 : defined $filters->{$name} ? @{ $filters->{$name} }
1052 my $numeric = $self->_is_numeric($name);
1054 my $sort = $self->param('sort');
1055 # sort numeric facets with more than 5 values ascending
1056 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
1058 @facet_names = sort {
1060 if ( $sort eq 'a' ) {
1061 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
1062 } elsif ( $sort eq 'd' ) {
1063 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
1064 } elsif ( $sort eq 'c' ) {
1065 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
1067 warn "unknown sort: $sort";
1068 $result = $a cmp $b;
1070 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
1074 $self->render( name => $name, facet => $facet, checked => $checked_values,
1075 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
1080 sub __invalidate_path_column {
1081 my ( $path, $name ) = @_;
1083 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
1084 delete $loaded->{$path}->{sorted}->{$name};
1085 warn "# invalidate $path sorted $name\n";
1088 foreach ( grep { m/\Q$name\E/ } keys %{ $loaded->{$path}->{filtered} } ) {
1089 delete $loaded->{$path}->{filtered}->{$_};
1090 warn "# invalidate $path filtered $_\n";
1094 sub __path_modified {
1095 my ( $path, $value ) = @_;
1096 $value = 1 unless defined $value;
1098 $loaded->{$path}->{modified} = $value;
1100 warn "# __path_modified $path $value\n";
1103 sub __path_rebuild_stats { $loaded->{ $_[0] }->{rebuild_stats} = 1 };
1106 my ($self,$change) = @_;
1108 my $change_path = $self->_permanent_path( 'changes' );
1109 mkdir $change_path unless -d $change_path;
1110 $change_path .= '/' . $change->{time};
1111 store $change, $change_path;
1112 utime $change->{time}, $change->{time}, $change_path;
1113 warn "_save_change $change_path ", dump($change);
1118 my $new_content = $self->param('new_content');
1119 $new_content ||= $self->param('content'); # backward compatibility with old actions
1121 my $i = $self->param('_row_id');
1122 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
1123 my $path = $self->param('path') || die "no path";
1124 my $name = $self->param('name') || die "no name";
1125 my $status = 200; # 200 = OK, 201 = Created
1127 my $data = $self->_loaded('data');
1129 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
1130 $new_content =~ s/^\s+//s;
1131 $new_content =~ s/\s+$//s;
1133 if ( $new_content =~ /\xB6/ ) { # para
1134 $v = [ split(/\s*\xB6\s*/, $new_content) ];
1136 $v = [ $new_content ];
1139 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
1142 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
1144 $self->_save_change({
1148 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
1150 time => $self->param('time') || time(),
1151 user => $self->param('user') || $ENV{'LOGNAME'},
1153 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
1154 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
1155 keys %{ $loaded->{$path}->{stats} }
1159 warn "# change $path $i $old -> $new\n";
1160 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
1162 __invalidate_path_column( $path, $name );
1164 $status = 201; # created
1165 __path_rebuild_stats( $path );
1167 $new_content = join("\xB6",@$v);
1170 warn "# unchanged $path $i $old\n";
1174 $new_content = "$path $i $name doesn't exist\n";
1178 warn "# edit $status ", dump $new_content;
1182 new_content => scalar $new_content,
1189 my $path = $self->_param_or_session('path');
1190 my $dump_path = $self->_save( $path );
1191 __path_modified( $path, 0 );
1193 return $self->redirect_to( '/data/items' );
1199 my $dir = $self->app->home->rel_dir('public');
1201 if ( my $import = $self->param('import') ) {
1203 if ( $import =~ m{/filter\.(.+?)\..+} ) {
1204 my $name = $self->_column_from_unac( $1 );
1206 my @vals = map { chomp; $_ }
1207 read_file "$dir/export/$import", binmode => ':utf8';
1209 $self->_remove_filter( $name );
1210 $self->_filter_on_data( $name, @vals );
1211 $self->session( 'offset' => 0 );
1212 return $self->redirect_to('/data/items');
1214 warn "UNKNOWN IMPORT $import";
1218 if ( my $remove = $self->param('remove') ) {
1219 my $path = "$dir/export/$remove";
1220 unlink $path if -e $path;
1222 unlink $path if -e $path;
1225 my $path = $self->_export_path || return $self->redirect_to('/data/index');
1227 my @files = grep { ! /\.png$/ } glob "$path/*";
1228 my $mtime = { map { $_ => (stat($_))[9] } @files };
1229 @files = sort { $mtime->{$b} <=> $mtime->{$a} } @files;
1230 $self->render( export => [ @files ] );
1233 sub __loaded_paths {
1235 grep { defined $loaded->{$_}->{data} }
1241 my $path = $self->param('path');
1242 if ( $path =~ m{^/tmp/mojo_facets\.} ) {
1244 warn "# unlink $path";
1245 if ( my $name = $self->param('name') ) {
1246 delete $loaded->{$name};
1247 warn "# remove $name from memory";
1250 warn "WARNING: $path unlink ignored";
1252 return $self->redirect_to( '/data/load' );