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 $i ( 0 .. $#{$_[0]} ) {
108 print STDERR " $i" if $i % 5000;
110 foreach my $n ( keys %$e ) {
111 $stats->{$n}->{count}++;
113 if ( ref $e->{$n} eq 'ARRAY' ) {
114 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
120 foreach my $x ( @v ) {
121 if ( ! defined $x ) { # FIXME really null
122 $stats->{$n}->{empty}++;
126 $stats->{$n}->{numeric}++
127 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
128 $stats->{$n}->{empty}++
129 if length $x == 0; # faster than $x =~ m/^\s*$/;
135 foreach my $n ( keys %$stats ) {
136 my $s = $stats->{$n};
137 next unless defined $s->{array};
138 if ( $s->{array} == $s->{count} ) {
140 if ( $s->{count} == $nr_items ) {
141 warn "check $n for uniqeness\n";
143 foreach my $e ( @{ $_[0] } ) {
144 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
150 $stats->{$n}->{unique} = 1;
151 #warn "# $n unique ",dump( $unique );
157 warn "# __stats ",dump($stats);
162 sub _param_or_session {
163 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
168 my $path = $self->_param_or_session('path');
169 warn "stats $path\n";
170 delete $loaded->{$path}->{stats};
171 return $self->redirect_to( '/data/columns' );
176 my ( $self, $path ) = @_;
178 return if defined $loaded->{$path}->{generated};
180 my $full_path = $self->app->home->rel_file( 'data/' . $path );
181 return $self->redirect_to('/data/index') unless -r $full_path;
183 my $dump_path = $self->_dump_path( $path );
185 if ( defined $loaded->{$path}->{data} ) {
186 my $mtime = (stat($full_path))[9];
187 return if $loaded->{$path}->{mtime} == $mtime;
188 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
189 } elsif ( -e $dump_path ) {
190 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
191 my $info = retrieve $dump_path;
192 $loaded->{ $path } = $info;
197 if ( -f $full_path ) {
198 if ( $full_path =~ m/.csv/i ) {
199 $data = MojoFacets::Import::CSV->new( full_path => $full_path )->data;
200 } elsif ( $full_path =~ m/.sql/i ) {
201 $data = MojoFacets::Import::SQL->new( full_path => $full_path )->data;
202 } elsif ( $full_path =~ m/.couchdb/i ) {
203 $data = MojoFacets::Import::CouchDB->new( full_path => $full_path )->data;
204 } elsif ( $full_path =~ m/.storable/ ) {
205 warn "open $full_path ", -s $full_path, " bytes";
206 open(my $pipe, "<", $full_path) || die $!;
207 while ( my $o = eval { Storable::fd_retrieve $pipe } ) {
208 if ( defined $o->{item} ) {
209 push @{ $data->{items} }, $o->{item};
211 warn "SKIP ",dump($o);
215 warn "loaded ", $#{ $data->{items} } + 1, " items from $full_path\n";
216 $data->{generated}++;
218 $data = MojoFacets::Import::File->new( full_path => $full_path, path => $path )->data;
220 } elsif ( -d $full_path && $full_path =~ m/.html/ ) {
221 $data = MojoFacets::Import::HTMLTable->new( dir => $full_path )->data;
223 die "can't load $full_path";
228 if ( defined $data->{header} ) {
229 if ( ref $data->{header} eq 'ARRAY' ) {
230 @header = @{ $data->{header} };
232 warn "header not array ", dump( $data->{header} );
236 my $stats = __stats( $data->{items} );
239 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
240 grep { defined $stats->{$_}->{count} } keys %$stats
244 header => [ @header ],
246 full_path => $full_path,
247 size => -s $full_path,
248 mtime => (stat($full_path))[9],
250 defined $data->{generated} ? ( generated => 1 ) : (),
253 $loaded->{ $path } = $info;
254 $self->_save( $path ) unless $info->{generated};
262 my $path = $self->param('path') || return $self->redirect_to( '/data/index' );
264 my @paths = $self->param('paths');
265 warn "# paths ", dump @paths;
267 foreach my $p ( keys %$loaded ) {
268 next if grep { /^\Q$p\E$/ } @paths;
269 warn "remove $p from memory\n";
270 delete $loaded->{$p};
273 $self->_load_path( $_ ) foreach @paths;
275 warn "# path $path\n";
276 $self->_load_path( $path );
278 $self->session( 'path' => $path );
280 my $redirect_to = '/data/items';
282 $self->session( 'header' => $loaded->{$path}->{header} );
283 if ( ! defined $loaded->{$path}->{columns} ) {
284 my $columns_path = $self->_permanent_path( 'columns' );
285 if ( -e $columns_path ) {
286 my @columns = map { s/[\r\n]+$//; $_ } read_file $columns_path, binmode => ':utf8';
287 $loaded->{$path}->{columns} = [ @columns ];
288 warn "# columns_path $columns_path ",dump(@columns);
290 $loaded->{$path}->{columns} = $loaded->{$path}->{header}
293 $redirect_to = '/data/columns';
295 $self->session( 'columns' => $loaded->{$path}->{columns} );
296 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
297 return $self->redirect_to( $redirect_to );
302 my ( $self, $name ) = @_;
303 my $path = $self->session('path') || $self->param('path');
304 return $self->redirect_to('/data/index') unless $path;
306 if ( defined $loaded->{$path}->{modified} && $loaded->{$path}->{modified} > 1 ) {
307 my $caller = (caller(1))[3];
308 if ( $caller =~ m/::edit/ ) {
309 warn "rebuild stats for $path ignored caller $caller\n";
311 warn "rebuild stats for $path FORCED by modified caller $caller\n";
312 # $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
313 $loaded->{$path}->{rebuild_stats} = 1;
314 $loaded->{$path}->{modified} = 1;
318 if ( defined $loaded->{$path}->{rebuild_stats} ) {
319 warn "rebuild_stats $path";
320 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
321 delete $loaded->{$path}->{rebuild_stats};
324 if ( ! defined $loaded->{$path}->{$name} ) {
325 warn "$path $name isn't loaded\n";
326 $self->_load_path( $path );
327 if ( ! defined $loaded->{$path}->{stats} ) {
328 warn "rebuild stats for $path\n";
329 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
331 if ( ! defined $loaded->{$path}->{$name} ) {
332 warn "MISSING $name for $path\n";
333 return $self->redirect_to('/data/index')
337 $self->session( 'modified' => $loaded->{$path}->{modified} );
339 return $loaded->{$path}->{$name};
346 $checked->{$_}++ foreach @_;
347 # warn "# _checked ",dump($checked);
351 sub _permanent_path {
353 my $path = $self->_param_or_session('path');
354 $self->app->home->rel_dir('data') . '/' . join('.', $path, @_);
359 $n = unac_string($n);
364 sub _column_from_unac {
365 my ($self,$name) = @_;
367 my $stats = $self->_loaded('stats');
368 my $cols_norm = { map { __unac( $_ ) => $_ } keys %$stats };
370 $cols_norm->{$name} || die "can't find column $name in ", dump($cols_norm);
375 my $path = $self->_param_or_session('path');
377 warn "no path in param or session";
380 my $dir = $self->app->home->rel_dir('public') . "/export/$path";
381 mkpath $dir unless -e $dir;
382 my $name = __export_path_name( $path, @_ );
383 my $full = $dir . '/' . $name;
384 $full =~ s/\/+$// if -d $full; # strip trailing slash for dirs
388 sub __export_path_name {
392 my $name = join('.', map { __unac($_) } @_ );
393 if ( length($name) > $max_length ) {
394 $name = substr($name,0,$max_length) . Digest::MD5::md5_hex substr($name,$max_length);
402 my $view_path = $self->_permanent_path( 'views' );
404 if ( $self->param('columns') ) {
405 my @columns = $self->_param_array('columns');
406 write_file( $self->_permanent_path( 'columns' ), { binmode => ':utf8' }, map { "$_\n" } @columns );
407 if ( my $view = $self->param('view') ) {
408 mkdir $view_path unless -e $view_path;
409 write_file( "$view_path/$view", { binmode => ':utf8' }, map { "$_\n" } @columns );
410 warn "view $view_path/$view ", -s "$view_path/$view", " bytes\n";
413 return $self->redirect_to('/data/items');
415 } elsif ( ! $self->session('header') ) {
416 return $self->redirect_to('/data/load');
419 if ( my $id = $self->param('id') ) {
420 my $view_full = "$view_path/$id";
421 if ( -f $view_full ) {
422 my @columns = map { chomp; $_ } read_file $view_full, binmode => ':utf8';
423 warn "view $view_full loaded ", dump @columns;
424 $self->session( 'columns' => [ @columns ] );
425 return $self->redirect_to('/data/items');
429 my $stats = $self->_loaded( 'stats' );
432 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
434 foreach my $c ( @{ $self->session( 'header' ) } ) {
435 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
439 if ( -d $view_path ) {
440 @views = map { s{^\Q$view_path\E/*}{}; $_ } glob "$view_path/*";
441 warn "# views ",dump @views;
445 message => 'Select columns to display',
447 columns => \@columns,
448 checked => $self->_checked( $self->_param_array('columns') ),
454 my ($self,$name) = @_;
456 my @array = $self->param($name);
457 my $path = $self->session('path');
460 $self->session($name => [ @array ]);
461 $loaded->{$path}->{$name} = [ @array ];
462 } elsif ( my $session = $self->session($name) ) {
463 if ( ref $session eq 'ARRAY' ) {
466 die "$name not array ",dump($session);
469 #warn "# $name ",dump @array;
474 my ($self,$name,$default) = @_;
476 my $scalar = $self->param($name);
478 if ( defined $scalar ) {
479 $self->session($name => $scalar);
481 $scalar = $self->session($name);
484 if ( ! defined $scalar ) {
486 die "no default for $name" unless defined $scalar;
487 $self->session($name => $scalar);
490 warn "# _param_scalar $name ",dump $scalar;
497 my $name = $self->param('filter_name') || die "name?";
498 my @vals = $self->param('filter_vals');
500 $self->_remove_filter( $name );
502 $self->_filter_on_data( $name, @vals );
503 if ( my $permanent = $self->param('_permanent') ) {
504 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
505 write_file $permanent_path, { binmode => ':utf8' }, map { "$_\n" } @vals;
506 warn "permanent filter $permanent_path ", -s $permanent_path;
510 $self->session( 'offset' => 0 );
512 return $self->redirect_to('/data/items');
515 sub _filter_on_data {
516 my ( $self, $name, @vals ) = @_;
518 my $path = $self->session('path');
520 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
521 warn "filter $name not found in data set";
525 $filters->{$name} = [ @vals ];
526 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
529 $filter_hash->{$_}++ foreach @vals;
531 #warn "# filter_hash ",dump( $filter_hash );
533 my $items = $self->_loaded('data')->{items};
535 my $include_missing = defined $filter_hash->{_missing};
538 foreach my $i ( 0 .. $#$items ) {
540 if ( defined $items->[$i]->{$name} ) {
541 my $row = $items->[$i]->{$name};
542 $row = [ $row ] unless ref $row eq 'ARRAY'; # FIXME probably wrong place
543 foreach my $v ( @$row ) {
544 if ( defined $filter_hash->{ $v } ) {
545 $filtered_items->{$i}++;
548 } elsif ( $include_missing ) {
549 $filtered_items->{$i}++;
553 #warn "# filter $name ",dump($filtered_items);
555 # invalidate filters on other datasets
556 foreach my $dataset ( grep { exists $loaded->{$_}->{filters}->{$name} } keys %$loaded ) {
557 delete $loaded->{$dataset}->{filters}->{$name};
558 delete $loaded->{$dataset}->{filtered};
561 $loaded->{$path}->{filters}->{$name} = $filtered_items;
562 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
566 sub _current_filters {
569 my $stats = $self->_loaded('stats');
571 $current_filters->{ $_ } = $filters->{ $_ }
572 foreach ( grep { defined $filters->{ $_ } } keys %$stats )
574 warn "# _current_filters ",dump( keys %$current_filters );
575 return $current_filters;
578 sub _data_sorted_by {
579 my ( $self, $order ) = @_;
581 my $path = $self->session('path');
583 warn "_data_sorted_by $order from $path";
585 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
586 return $loaded->{$path}->{sorted}->{$order};
589 my $data = $self->_loaded( 'data' );
590 my $numeric = $self->_is_numeric($order);
591 my $missing = $numeric ? 0 : '';
592 no warnings qw(numeric);
604 if ( ! exists $_->{$order} ) {
606 } elsif ( ref $_->{$order} eq 'ARRAY' ) {
607 $v = join('', @{$_->{$order}});
612 } @{ $data->{items} }
615 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
616 #warn "# sorted ",dump( @sorted );
618 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
624 join(',', sort(@_), 'order', $order);
627 our ($out, $key,$value);
629 our $lookup_path_col;
633 warn "__commit_begin";
634 $lookup_path_col = undef;
640 $lookup_path_col = undef; # cleanup memory
645 warn "# lookup ",dump @_;
646 my ( $vals, $on_path, $on_col, $code ) = @_;
647 die "code is not sub{ ... } but ", dump $code unless ref $code eq 'CODE';
648 my $items = $loaded->{$on_path}->{data}->{items} || die "no items for $on_path";
650 if ( ! exists $lookup_path_col->{$on_path}->{$on_col} ) {
651 warn "create lookup_path_col $on_path $on_col";
652 foreach my $i ( 0 .. $#$items ) {
653 my $item = $items->[$i];
654 if ( exists $item->{$on_col} ) {
655 foreach my $v ( @{ $item->{$on_col} } ) {
656 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
660 warn dump $lookup_path_col->{$on_path}->{$on_col};
663 foreach my $v ( @$vals ) {
664 foreach my $i ( @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} } ) {
666 warn "# lookup code $v $i ",dump $on;
672 sub __commit_path_code {
673 my ( $path, $i, $code, $commit_changed ) = @_;
675 my $items = $loaded->{$path}->{data}->{items} || die "no items for $path";
676 my $row = $items->[$i];
679 foreach ( keys %$update ) {
680 $$commit_changed->{$_}++;
681 $loaded->{$path}->{data}->{items}->[$i]->{$_} = $update->{$_};
683 #warn "__commit_path_code $path $i ",dump( $update );
686 # uses templates/admin.html.ep
687 sub _switch_dataset {
692 foreach my $path ( keys %$loaded ) {
693 next unless exists $loaded->{$path}->{data};
694 push @$datasets, $path;
697 warn "# datasets ",dump($datasets);
699 $self->stash( 'datasets' => $datasets );
705 $self->_switch_dataset;
707 if ( my $show = $self->param('id') ) {
708 $self->param('show', $show);
712 my $path = $self->_param_scalar('path');
714 my @columns = $self->_param_array('columns');
715 return $self->redirect_to('/data/columns') unless @columns;
716 my $order = $self->_param_scalar('order', $columns[0]);
717 my $sort = $self->_param_scalar('sort', 'a');
718 my $offset = $self->_param_scalar('offset', 0);
719 my $limit = $self->_param_scalar('limit', 20);
720 $self->_param_scalar('show', 'table');
722 # fix offset when changing limit
723 $offset = int( $offset / $limit ) * $limit;
725 if ( ! grep { /^\Q$order\E$/ } @columns ) {
726 $order = $columns[0];
727 $self->session( order => $order );
729 my $sorted = $self->_data_sorted_by( $order );
733 foreach my $name ( keys %$filters ) {
734 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
735 warn "skip filter $name not found in $path\n";
738 push @filter_names, $name;
740 warn "filter_names ",dump( @filter_names );
741 foreach my $name ( @filter_names ) {
742 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
743 $self->_filter_on_data( $name, @{ $filters->{$name} } );
747 my $all_filters = __all_filters( @filter_names,$order );
749 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
751 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
753 my $path_filters = $loaded->{$path}->{filters};
755 warn "create combined filter for $all_filters from ", $#$sorted + 1, " items\n";
758 foreach my $i ( 0 .. $#$sorted ) {
759 my $pos = $sorted->[$i];
761 if ( $#filter_names == -1 ) {
762 push @filtered, $pos;
767 foreach ( @filter_names ) {
768 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
772 push @filtered, $pos;
775 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
778 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
779 if defined $loaded->{$path}->{filtered}->{$all_filters};
781 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
783 my $data = $self->_loaded('data');
785 my $code = $self->_param_scalar('code','');
787 $code =~ s{\n+$}{\n}s;
789 my $commit = $self->param('commit');
790 my $test = $self->param('test');
795 if ( $code && ( $test || $commit ) ) {
796 # XXX find columns used in code snippet and show them to user
798 foreach my $column ( $code =~ m/\$row->{([^}]+)}/g ) {
799 if ( $column =~ s/^(['"])// ) {
802 next if $column =~ m/\$/; # hide columns with vars in them
803 $commit_changed->{$column} = 0;
807 my $code_path = $self->app->home->rel_dir('public') . "/code";
810 warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
811 ( $key, $value, $out ) = ( 'key', 'value' );
812 foreach ( 0 .. $#$filtered ) {
813 my $i = $filtered->[$_];
814 __commit_path_code( $path, $i, $code, \$commit_changed );
817 # this might move before $out to recalculate stats on source dataset?
818 __path_rebuild_stats( $path );
819 my $c = { map { $_ => 1 } @columns };
820 my @added_columns = sort grep { ! $c->{$_} } keys %$commit_changed;
821 warn "# added_columns ",dump( @added_columns );
822 unshift @columns, @added_columns;
824 $loaded->{$path}->{columns} = [ @columns ];
825 warn "# new columns ",dump( @columns );
827 __invalidate_path_column( $path, $_ ) foreach keys %$commit_changed;
829 $self->_save_change({
831 time => $self->param('time') || time(),
832 user => $self->param('user') || $ENV{'LOGNAME'},
834 commit_changed => $commit_changed,
837 if ( my $description = $self->param('code_description') ) {
838 my $depends = $self->param('code_depends') || die "no code_depends?";
839 my $path = "$code_path/$depends.$description.pl";
840 if ( -e $path && ! $self->param('overwrite') ) {
841 warn "# code $path not saved\n";
843 write_file( $path, { binmode => ':utf8' }, "$code\n" );
844 warn "code $path ", -s $path, " bytes saved\n";
851 my $commit_dataset = join('.'
852 , $self->param('code_depends')
853 , $self->param('code_description')
858 warn "key $key value $value";
860 foreach my $n ( keys %$out ) {
861 my $i = { $key => [ $n ] };
862 my $ref = ref $out->{$n};
863 if ( $ref eq 'HASH' ) {
864 $i->{$_} = [ $out->{$n}->{$_} ] foreach keys %{ $out->{$n} };
865 } elsif ( $ref eq 'ARRAY' ) {
866 $i->{$_} = $out->{$n};
868 $i->{$value} = [ $out->{$n} ];
870 $i->{_error} = [ dump($out->{$n}) ];
875 my $stats = __stats( $items );
876 my @columns = grep { ! m/^\Q$key\E$/ } sort keys %$stats;
877 unshift @columns, $key;
879 $loaded->{$commit_dataset} = {
880 header => [ @columns ],
881 columns => [ @columns ],
883 data => { items => $items },
887 warn "# loaded out ", dump( $loaded->{$commit_dataset} );
888 $self->session('path', $commit_dataset);
889 $self->session('columns', [ @columns ]);
890 $self->session('order', $key);
891 return $self->redirect_to('/data/items');
894 $self->session('columns', [ @columns ]);
898 my $from_end = $sort eq 'd' ? $#$filtered : 0;
900 my ( $key, $value, $out ) = ( 'key', 'value' ); # XXX make local
901 foreach ( 0 .. $limit ) {
902 my $i = $_ + $offset;
903 last unless defined $filtered->[$i];
904 $i = $from_end - $i if $from_end;
905 my $id = $filtered->[$i];
906 my $row = Storable::dclone $data->{items}->[ $id ];
907 if ( $code && $test ) {
911 warn "ERROR evaling $@", dump($code);
912 $self->stash('eval_error', $@) if $@;
914 warn "EVAL ",dump($update);
915 foreach ( keys %$update ) {
916 $test_changed->{$_}++;
917 $row->{$_} = $update->{$_};
921 $row->{_row_id} ||= $id;
922 push @$sorted_items, $row;
925 if ( $self->param('export') ) {
926 my $export_path = $self->_export_path( 'items', @columns);
927 open(my $fh, '>', $export_path) || die "ERROR: can't open $export_path: $!";
928 print $fh "#",join("\t",@columns),"\n";
929 foreach my $f ( 0 .. $#$filtered ) {
930 print $fh join("\t", map {
931 my $i = $data->{items}->[ $filtered->[$f] ];
933 if ( ! defined $i->{$_} ) {
935 } elsif ( ref $i->{$_} eq 'ARRAY' ) {
936 $v =join(',', @{ $i->{$_} });
937 $v = '\N' if length($v) == 0;
938 } elsif ( ! ref $i->{$_} ) {
947 warn "export $export_path ", -s $export_path, " bytes\n";
950 my ( $code_depends, $code_description );
954 warn "# test_changed ",dump( $test_changed );
955 my $c = { map { $_ => 1 } @columns };
956 my @added_columns = sort grep { ! $c->{$_} } keys %$test_changed;
957 unshift @columns, @added_columns;
959 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
962 my $tmp = $code; $tmp =~ s/\$row->{(['"]?)([\w\s]+)\1/$depends_on->{$2}++/gse;
963 warn "# depends_on ",dump $depends_on;
965 my $test_added = Storable::dclone $test_changed;
966 delete $test_added->{$_} foreach keys %$depends_on;
968 $code_depends = $self->param('code_depends')
969 || join(',', keys %$depends_on);
971 $code_description = $self->param('code_description') ||
972 join(',', keys %$test_added);
974 $code_depends ||= $code_description; # self-modifing
975 if ( ! $code_depends && $out ) {
976 $code_depends = $key;
977 $code_description = $value;
980 warn "# test_changed ",dump( $test_changed, $code_depends, $code_description );
990 sorted => $sorted_items,
991 columns => [ @columns ],
992 rows => $#$filtered + 1,
993 numeric => { map { $_, $self->_is_numeric($_) } @columns },
994 unique => { map { $_, $self->_is_unique( $_) } @columns },
995 filters => $self->_current_filters,
997 cols_changed => $commit ? $commit_changed : $test_changed,
998 code_depends => $code_depends,
999 code_description => $code_description,
1000 code_path => $code_path,
1009 $self->session('order', $self->param('order'));
1010 $self->session('sort', $self->param('sort'));
1011 return $self->redirect_to('/data/items');
1015 my ( $self, $name ) = @_;
1017 my $stats = $self->_loaded( 'stats' );
1019 # sort facet numerically if more >50% elements are numeric
1020 my $count = $stats->{$name}->{count};
1021 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
1022 defined $stats->{$name}->{numeric} &&
1023 $stats->{$name}->{numeric} > $count / 2;
1027 my ( $self, $name ) = @_;
1028 my $stats = $self->_loaded( 'stats' );
1029 defined $stats->{$name}->{unique};
1032 sub _remove_filter {
1033 my ($self,$name) = @_;
1034 warn "_remove_filter $name\n";
1036 my $path = $self->session('path');
1038 delete $filters->{$name};
1039 delete $loaded->{$path}->{filters}->{$name};
1040 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
1043 grep { /\Q$name\E/ }
1044 keys %{ $loaded->{$path}->{filtered} }
1046 delete $loaded->{$path}->{filtered}->{$_};
1047 warn "remove filtered cache $_";
1054 my $path = $self->session('path') || return $self->redirect_to( '/data/index' );
1056 if ( my $name = $self->param('remove') ) {
1057 $self->_remove_filter( $name );
1058 return $self->redirect_to( '/data/items' );
1062 my $name = $self->param('name') || die "no name";
1064 my $all = $self->_param_scalar('all', 1);
1065 my $data = $self->_loaded('data');
1067 my $filters = $self->_current_filters;
1068 my $all_filters = __all_filters( keys %$filters,$self->session('order') );
1069 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
1070 if defined $loaded->{$path}->{filtered}->{$all_filters};
1072 if ( ! $filtered || $all ) {
1073 $filtered = [ 0 .. $#{ $data->{items} } ];
1074 warn "filter all values\n";
1076 warn "filter using $all_filters\n";
1079 foreach my $i ( @$filtered ) {
1080 my $item = $data->{items}->[$i];
1081 if ( ! exists $item->{$name} || ! defined $item->{$name} ) {
1082 $facet->{ _missing }++;
1083 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
1084 $facet->{$_}++ foreach @{ $item->{$name} };
1086 $facet->{ $item->{$name} }++;
1090 my $checked_values = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
1092 if ( my $code = $self->param('code') ) {
1094 foreach my $value ( keys %$facet ) {
1095 my $count = $facet->{$value};
1096 my $checked = $checked_values->{$value};
1100 warn "ERROR: $@\n$code\n";
1102 } elsif ( $checked != $checked_values->{$value} ) {
1103 warn "checked $value $count -> $checked\n";
1104 $checked_values->{$value} = $checked;
1107 warn "out ",dump( $out );
1108 $self->stash( out => $out );
1111 # warn "# facet $name ",dump $facet;
1115 : defined $filters->{$name} ? @{ $filters->{$name} }
1118 my $numeric = $self->_is_numeric($name);
1120 my $sort = $self->param('sort');
1121 # sort numeric facets with more than 5 values ascending
1122 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
1124 @facet_names = sort {
1126 if ( $sort eq 'a' ) {
1127 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
1128 } elsif ( $sort eq 'd' ) {
1129 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
1130 } elsif ( $sort eq 'c' ) {
1131 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
1133 warn "unknown sort: $sort";
1134 $result = $a cmp $b;
1136 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
1140 $self->render( name => $name, facet => $facet, checked => $checked_values,
1141 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
1146 sub __invalidate_path_column {
1147 my ( $path, $name ) = @_;
1149 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
1150 delete $loaded->{$path}->{sorted}->{$name};
1151 warn "# invalidate $path sorted $name\n";
1154 foreach ( grep { m/\Q$name\E/ } keys %{ $loaded->{$path}->{filtered} } ) {
1155 delete $loaded->{$path}->{filtered}->{$_};
1156 warn "# invalidate $path filtered $_\n";
1160 sub __path_modified {
1161 my ( $path, $value ) = @_;
1162 $value = 1 unless defined $value;
1164 $loaded->{$path}->{modified} = $value;
1166 warn "# __path_modified $path $value\n";
1169 sub __path_rebuild_stats { $loaded->{ $_[0] }->{rebuild_stats} = 1 };
1172 my ($self,$change) = @_;
1174 my $change_path = $self->_permanent_path( 'changes' );
1175 mkdir $change_path unless -d $change_path;
1176 $change_path .= '/' . $change->{time};
1177 store $change, $change_path;
1178 utime $change->{time}, $change->{time}, $change_path;
1179 warn "_save_change $change_path ", dump($change);
1184 my $new_content = $self->param('new_content');
1185 $new_content ||= $self->param('content'); # backward compatibility with old actions
1187 my $i = $self->param('_row_id');
1188 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
1189 my $path = $self->param('path') || die "no path";
1190 my $name = $self->param('name') || die "no name";
1191 my $status = 200; # 200 = OK, 201 = Created
1193 my $data = $self->_loaded('data');
1195 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
1196 $new_content =~ s/^\s+//s;
1197 $new_content =~ s/\s+$//s;
1199 if ( $new_content =~ /\xB6/ ) { # para
1200 $v = [ split(/\s*\xB6\s*/, $new_content) ];
1202 $v = [ $new_content ];
1205 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
1208 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
1210 $self->_save_change({
1214 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
1216 time => $self->param('time') || time(),
1217 user => $self->param('user') || $ENV{'LOGNAME'},
1219 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
1220 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
1221 keys %{ $loaded->{$path}->{stats} }
1225 warn "# change $path $i $old -> $new\n";
1226 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
1228 __invalidate_path_column( $path, $name );
1230 $status = 201; # created
1231 __path_rebuild_stats( $path );
1233 $new_content = join("\xB6",@$v);
1236 warn "# unchanged $path $i $old\n";
1240 $new_content = "$path $i $name doesn't exist\n";
1244 warn "# edit $status ", dump $new_content;
1248 new_content => scalar $new_content,
1255 my $path = $self->_param_or_session('path');
1256 my $dump_path = $self->_save( $path );
1257 __path_modified( $path, 0 );
1259 return $self->redirect_to( '/data/items' );
1265 my $dir = $self->app->home->rel_dir('public');
1267 if ( my $import = $self->param('import') ) {
1269 if ( $import =~ m{/filter\.(.+?)\..+} ) {
1270 my $name = $self->_column_from_unac( $1 );
1272 my @vals = map { chomp; $_ }
1273 read_file "$dir/export/$import", binmode => ':utf8';
1275 $self->_remove_filter( $name );
1276 $self->_filter_on_data( $name, @vals );
1277 $self->session( 'offset' => 0 );
1278 return $self->redirect_to('/data/items');
1280 warn "UNKNOWN IMPORT $import";
1284 if ( my $remove = $self->param('remove') ) {
1285 my $path = "$dir/export/$remove";
1286 unlink $path if -e $path;
1288 unlink $path if -e $path;
1291 my $path = $self->_export_path || return $self->redirect_to('/data/index');
1293 my @files = grep { ! /\.png$/ } glob "$path/*";
1294 my $mtime = { map { $_ => (stat($_))[9] } @files };
1295 @files = sort { $mtime->{$b} <=> $mtime->{$a} } @files;
1296 $self->render( export => [ @files ] );
1299 sub __loaded_paths {
1301 grep { defined $loaded->{$_}->{data} }
1307 my $path = $self->param('path');
1308 if ( $path =~ m{^/tmp/mojo_facets\.} ) {
1310 warn "# unlink $path";
1311 if ( my $name = $self->param('name') ) {
1312 delete $loaded->{$name};
1313 warn "# remove $name from memory";
1316 warn "WARNING: $path unlink ignored";
1318 return $self->redirect_to( '/data/load' );