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);
18 use Statistics::Descriptive;
21 foreach my $module ( glob('lib/MojoFacets/Import/*.pm') ) {
22 $module =~ s{lib/(\w+)/(\w+)/(.*)\.pm}{$1::$2::$3};
24 die "$module: $!" if $!;
25 my ( $ext, $priority ) = $module->ext;
26 $imports->{$priority || 'file'}->{$ext} = $module;
27 warn "# import $ext $module\n";
30 warn "# import loaded ",dump( $imports );
33 my $full_path = shift;
35 # warn "# import_module $full_path\n";
37 foreach my $ext ( keys %{ $imports->{file} } ) {
38 if ( -f $full_path && $full_path =~ m/$ext/i ) {
39 return $imports->{file}->{$ext};
44 foreach my $ext ( keys %{ $imports->{directory} } ) {
45 if ( -f $full_path && $full_path =~ m/$ext/i ) {
46 return $imports->{directory}->{$ext};
58 my $data_dir = $self->app->home->rel_dir('data');
59 die "no data dir $data_dir" unless -d $data_dir;
65 my $file = $File::Find::name;
66 if ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
67 push @{ $changes->{$1} }, $2
68 } elsif ( import_module( $file ) ) {
69 $file =~ s/$data_dir\/*//;
72 #warn "IGNORE: $file\n";
76 no warnings qw(uninitialized); # mtime
77 @files = sort { $loaded->{$b}->{mtime} <=> $loaded->{$a}->{mtime} || lc $a cmp lc $b } @files,
78 grep { defined $loaded->{$_}->{generated} } keys %$loaded;
81 $size->{$_} = -s "$data_dir/$_" foreach @files;
88 dump_path => { map { $_ => $self->_dump_path($_) } @files },
94 my ( $self, $name ) = @_;
95 my $dir = $self->app->home->rel_dir('data');
98 return '/tmp/mojo_facets.' . $name . '.storable';
102 my ( $self, $path ) = @_;
104 my $dump_path = $self->_dump_path( $path );
105 my $first_load = ! -e $dump_path;
106 warn "save loaded to $dump_path";
107 my $info = $loaded->{$path};
108 store $info, $dump_path;
111 my $mtime = $loaded->{$path}->{mtime};
112 utime $mtime, $mtime, $dump_path;
113 warn "sync time to $path at $mtime\n";
116 warn $dump_path, ' ', -s $dump_path, " bytes\n";
125 my $nr_items = $#{ $_[0] } + 1;
127 warn "__stats $nr_items\n";
129 # foreach my $e ( @{ $_[0] } ) {
130 foreach my $i ( 0 .. $#{$_[0]} ) {
131 print STDERR " $i" if $i % 5000;
133 foreach my $n ( keys %$e ) {
134 $stats->{$n}->{count}++;
136 if ( ref $e->{$n} eq 'ARRAY' ) {
137 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
143 foreach my $x ( @v ) {
144 if ( ! defined $x ) { # FIXME really null
145 $stats->{$n}->{empty}++;
149 $stats->{$n}->{numeric}++
150 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
151 $stats->{$n}->{empty}++
152 if length $x == 0; # faster than $x =~ m/^\s*$/;
158 foreach my $n ( keys %$stats ) {
159 my $s = $stats->{$n};
160 next unless defined $s->{array};
161 if ( $s->{array} == $s->{count} ) {
163 if ( $s->{count} == $nr_items ) {
164 warn "check $n for uniqeness\n";
166 foreach my $e ( @{ $_[0] } ) {
167 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
173 $stats->{$n}->{unique} = 1;
174 #warn "# $n unique ",dump( $unique );
180 warn "# __stats ",dump($stats);
185 sub _param_or_session {
186 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
191 my $path = $self->_param_or_session('path');
192 warn "stats $path\n";
193 delete $loaded->{$path}->{stats};
194 return $self->redirect_to( '/data/columns' );
199 my ( $self, $path ) = @_;
201 return if defined $loaded->{$path}->{generated};
203 my $full_path = $self->app->home->rel_file( 'data/' . $path );
204 return $self->redirect_to('/data/index') unless -r $full_path;
206 my $dump_path = $self->_dump_path( $path );
208 if ( defined $loaded->{$path}->{data} ) {
209 my $mtime = (stat($full_path))[9];
210 return if $loaded->{$path}->{mtime} == $mtime;
211 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
212 } elsif ( -e $dump_path ) {
213 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
214 my $info = retrieve $dump_path;
215 $loaded->{ $path } = $info;
220 if ( my $module = import_module( $full_path ) ) {
221 $data = $module->new( full_path => $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 @paths = $self->param('paths');
263 warn "# paths ", dump @paths;
265 foreach my $p ( keys %$loaded ) {
266 next if grep { /^\Q$p\E$/ } @paths;
267 warn "remove $p from memory\n";
268 delete $loaded->{$p};
271 $self->_load_path( $_ ) foreach @paths;
273 my $path = $self->param('path') || $self->session('path') || $paths[0] || $self->redirect_to('/data/index');
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 $n = unac_string('utf-8',$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 if ( defined $scalar ) {
487 $self->session($name => $scalar);
489 warn "no default for $name";
493 warn "# _param_scalar $name ",dump $scalar;
500 my $name = $self->param('filter_name') || die "name?";
501 my @vals = $self->param('filter_vals');
503 $self->_remove_filter( $name );
505 $self->_filter_on_data( $name, @vals );
506 if ( my $permanent = $self->param('_permanent') ) {
507 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
508 write_file $permanent_path, { binmode => ':utf8' }, map { "$_\n" } @vals;
509 warn "permanent filter $permanent_path ", -s $permanent_path;
513 $self->session( 'offset' => 0 );
515 return $self->redirect_to('/data/items');
518 sub _filter_on_data {
519 my ( $self, $name, @vals ) = @_;
521 my $path = $self->session('path');
523 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
524 warn "filter $name not found in data set";
528 $filters->{$name} = [ @vals ];
529 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
532 $filter_hash->{$_}++ foreach @vals;
534 #warn "# filter_hash ",dump( $filter_hash );
536 my $items = $self->_loaded('data')->{items};
538 my $include_missing = defined $filter_hash->{_missing};
541 foreach my $i ( 0 .. $#$items ) {
543 if ( defined $items->[$i]->{$name} ) {
544 my $row = $items->[$i]->{$name};
545 $row = [ $row ] unless ref $row eq 'ARRAY'; # FIXME probably wrong place
546 foreach my $v ( @$row ) {
547 if ( defined $filter_hash->{ $v } ) {
548 $filtered_items->{$i}++;
551 } elsif ( $include_missing ) {
552 $filtered_items->{$i}++;
556 #warn "# filter $name ",dump($filtered_items);
558 # invalidate filters on other datasets
559 foreach my $dataset ( grep { exists $loaded->{$_}->{filters}->{$name} } keys %$loaded ) {
560 delete $loaded->{$dataset}->{filters}->{$name};
561 delete $loaded->{$dataset}->{filtered};
564 $loaded->{$path}->{filters}->{$name} = $filtered_items;
565 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
569 sub _current_filters {
572 my $stats = $self->_loaded('stats');
574 $current_filters->{ $_ } = $filters->{ $_ }
575 foreach ( grep { defined $filters->{ $_ } } keys %$stats )
577 warn "# _current_filters ",dump( keys %$current_filters );
578 return $current_filters;
581 sub _data_sorted_by {
582 my ( $self, $order ) = @_;
584 my $path = $self->session('path');
586 warn "_data_sorted_by $order from $path";
588 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
589 return $loaded->{$path}->{sorted}->{$order};
592 my $data = $self->_loaded( 'data' );
593 my $numeric = $self->_is_numeric($order);
594 my $missing = $numeric ? 0 : '';
595 no warnings qw(numeric);
607 if ( ! exists $_->{$order} ) {
609 } elsif ( ref $_->{$order} eq 'ARRAY' ) {
610 $v = join('', @{$_->{$order}});
615 } @{ $data->{items} }
618 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
619 #warn "# sorted ",dump( @sorted );
621 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
627 join(',', sort(@_), 'order', $order);
630 our ($out, $key,$value);
632 our $lookup_path_col;
636 warn "__commit_begin";
637 $lookup_path_col = undef;
643 $lookup_path_col = undef; # cleanup memory
648 warn "# lookup ",dump @_;
649 my ( $vals, $on_path, $on_col, $code, $stat_code ) = @_;
650 die "code is not sub{ ... } but ", dump $code unless ref $code eq 'CODE';
652 if ( ! exists $loaded->{$on_path} ) {
653 my @possible_paths = grep { /\Q$on_path\E/ } keys %$loaded;
654 die "more than one dataset available for '$on_path' ",dump @possible_paths if $#possible_paths > 0;
655 $on_path = shift @possible_paths;
656 warn "## fuzzy selected path $on_path";
659 my $items = $loaded->{$on_path}->{data}->{items} || die "no items for $on_path";
661 if ( ! exists $lookup_path_col->{$on_path}->{$on_col} ) {
662 warn "create lookup_path_col $on_path $on_col";
663 foreach my $i ( 0 .. $#$items ) {
664 my $item = $items->[$i];
665 if ( exists $item->{$on_col} ) {
666 if ( ref $item->{$on_col} eq 'ARRAY' ) {
667 foreach my $v ( @{ $item->{$on_col} } ) {
668 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
670 } elsif ( ! ref $item->{$on_col} ) { # scalar
671 my $v = $item->{$on_col};
672 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
674 die "unknown type of ",dump $item->{$on_col};
678 warn "XXX ",dump $lookup_path_col->{$on_path}->{$on_col} if $ENV{DEBUG};
682 $stat = Statistics::Descriptive::Full->new() if $stat_code;
684 foreach my $v ( ref $vals eq 'ARRAY' ? @$vals : ( $vals ) ) {
685 foreach my $i ( @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} } ) {
687 warn "XXX lookup code $v $i ",dump $on if $ENV{DEBUG};
690 $stat_code->( $stat ) if $stat_code;
694 sub __commit_path_code {
695 my ( $path, $i, $code, $commit_changed ) = @_;
697 my $items = $loaded->{$path}->{data}->{items} || die "no items for $path";
698 my $row = $items->[$i];
701 foreach ( keys %$update ) {
702 $$commit_changed->{$_}++;
703 $loaded->{$path}->{data}->{items}->[$i]->{$_} = $update->{$_};
705 #warn "__commit_path_code $path $i ",dump( $update );
708 # uses templates/admin.html.ep
709 sub _switch_dataset {
714 foreach my $path ( keys %$loaded ) {
715 next unless exists $loaded->{$path}->{data};
716 push @$datasets, $path;
719 warn "# datasets ",dump($datasets);
721 $self->stash( 'datasets' => $datasets );
727 $self->_switch_dataset;
729 if ( my $show = $self->param('id') ) {
730 $self->param('show', $show);
734 my $path = $self->_param_scalar('path');
736 my @columns = $self->_param_array('columns');
737 return $self->redirect_to('/data/columns') unless @columns;
738 my $order = $self->_param_scalar('order', $columns[0]);
739 my $sort = $self->_param_scalar('sort', 'a');
740 my $offset = $self->_param_scalar('offset', 0);
741 my $limit = $self->_param_scalar('limit', 20);
742 $self->_param_scalar('show', 'table');
744 # fix offset when changing limit
745 $offset = int( $offset / $limit ) * $limit;
747 if ( ! grep { /^\Q$order\E$/ } @columns ) {
748 $order = $columns[0];
749 $self->session( order => $order );
751 my $sorted = $self->_data_sorted_by( $order );
755 foreach my $name ( keys %$filters ) {
756 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
757 warn "skip filter $name not found in $path\n";
760 push @filter_names, $name;
762 warn "filter_names ",dump( @filter_names );
763 foreach my $name ( @filter_names ) {
764 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
765 $self->_filter_on_data( $name, @{ $filters->{$name} } );
769 my $all_filters = __all_filters( @filter_names,$order );
771 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
773 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
775 my $path_filters = $loaded->{$path}->{filters};
777 warn "create combined filter for $all_filters from ", $#$sorted + 1, " items\n";
780 foreach my $i ( 0 .. $#$sorted ) {
781 my $pos = $sorted->[$i];
783 if ( $#filter_names == -1 ) {
784 push @filtered, $pos;
789 foreach ( @filter_names ) {
790 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
794 push @filtered, $pos;
797 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
800 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
801 if defined $loaded->{$path}->{filtered}->{$all_filters};
803 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
805 my $data = $self->_loaded('data');
807 my $code = $self->_param_scalar('code','');
809 $code =~ s{\n+$}{\n}s;
811 # XXX convert @row->{foo} into @{$row->{foo}}
812 $code =~ s|\@(row->{[^}]+})|\@{\$$1}|gs;
814 my $commit = $self->param('commit');
815 my $test = $self->param('test');
820 if ( $code && ( $test || $commit ) ) {
821 # XXX find columns used in code snippet and show them to user
823 foreach my $column ( $code =~ m/\$row->{([^}]+)}/g ) {
824 if ( $column =~ s/^(['"])// ) {
827 next if $column =~ m/\$/; # hide columns with vars in them
828 $commit_changed->{$column} = 0;
832 my $code_path = $self->app->home->rel_dir('public') . "/code";
835 __path_modified( $path, 'commit' );
837 warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
838 ( $key, $value, $out ) = ( 'key', 'value' );
839 foreach ( 0 .. $#$filtered ) {
840 my $i = $filtered->[$_];
841 __commit_path_code( $path, $i, $code, \$commit_changed );
844 # this might move before $out to recalculate stats on source dataset?
845 __path_rebuild_stats( $path );
846 my $c = { map { $_ => 1 } @columns };
847 my @added_columns = sort grep { ! $c->{$_} } keys %$commit_changed;
848 warn "# added_columns ",dump( @added_columns );
849 unshift @columns, @added_columns;
851 $loaded->{$path}->{columns} = [ @columns ];
852 warn "# new columns ",dump( @columns );
854 __invalidate_path_column( $path, $_ ) foreach keys %$commit_changed;
856 $self->_save_change({
858 time => $self->param('time') || time(),
859 user => $self->param('user') || $ENV{'LOGNAME'},
861 commit_changed => $commit_changed,
864 if ( my $description = $self->param('code_description') ) {
865 my $depends = $self->param('code_depends') || die "no code_depends?";
866 my $path = "$code_path/$depends.$description.pl";
867 if ( -e $path && ! $self->param('overwrite') ) {
868 warn "# code $path not saved\n";
870 write_file( $path, { binmode => ':utf8' }, "$code\n" );
871 warn "code $path ", -s $path, " bytes saved\n";
878 my $commit_dataset = join('.'
879 , $self->param('code_depends')
880 , $self->param('code_description')
885 warn "key $key value $value";
887 foreach my $n ( keys %$out ) {
888 my $i = { $key => [ $n ] };
889 my $ref = ref $out->{$n};
890 if ( $ref eq 'HASH' ) {
891 $i->{$_} = [ $out->{$n}->{$_} ] foreach keys %{ $out->{$n} };
892 } elsif ( $ref eq 'ARRAY' ) {
893 $i->{$_} = $out->{$n};
895 $i->{$value} = [ $out->{$n} ];
897 $i->{_error} = [ dump($out->{$n}) ];
902 my $stats = __stats( $items );
903 my @columns = grep { ! m/^\Q$key\E$/ } sort keys %$stats;
904 unshift @columns, $key;
906 $loaded->{$commit_dataset} = {
907 header => [ @columns ],
908 columns => [ @columns ],
910 data => { items => $items },
914 warn "# loaded out ", dump( $loaded->{$commit_dataset} );
915 $self->session('path', $commit_dataset);
916 $self->session('columns', [ @columns ]);
917 $self->session('order', $key);
918 return $self->redirect_to('/data/items');
921 $self->session('columns', [ @columns ]);
925 my $from_end = $sort eq 'd' ? $#$filtered : 0;
927 my ( $key, $value, $out ) = ( 'key', 'value' ); # XXX make local
928 foreach ( 0 .. $limit ) {
929 my $i = $_ + $offset;
930 last unless defined $filtered->[$i];
931 $i = $from_end - $i if $from_end;
932 my $id = $filtered->[$i];
933 my $row = Storable::dclone $data->{items}->[ $id ];
934 if ( $code && $test ) {
938 warn "ERROR evaling $@", dump($code);
939 $self->stash('eval_error', $@) if $@;
941 warn "EVAL ",dump($update);
942 foreach ( keys %$update ) {
943 $test_changed->{$_}++;
944 $row->{$_} = $update->{$_};
948 $row->{_row_id} ||= $id;
949 push @$sorted_items, $row;
952 if ( $self->param('export') ) {
953 my $export_path = $self->_export_path( 'items', @columns);
954 open(my $fh, '>', $export_path) || die "ERROR: can't open $export_path: $!";
955 print $fh "#",join("\t",@columns),"\n";
956 foreach my $f ( 0 .. $#$filtered ) {
957 print $fh join("\t", map {
958 my $i = $data->{items}->[ $filtered->[$f] ];
960 if ( ! defined $i->{$_} ) {
962 } elsif ( ref $i->{$_} eq 'ARRAY' ) {
963 $v =join(',', @{ $i->{$_} });
964 $v = '\N' if length($v) == 0;
965 } elsif ( ! ref $i->{$_} ) {
974 warn "export $export_path ", -s $export_path, " bytes\n";
977 my ( $code_depends, $code_description );
981 warn "# test_changed ",dump( $test_changed );
982 my $c = { map { $_ => 1 } @columns };
983 my @added_columns = sort grep { ! $c->{$_} } keys %$test_changed;
984 unshift @columns, @added_columns;
986 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
989 my $tmp = $code; $tmp =~ s/\$row->{(['"]?)([\w\s]+)\1/$depends_on->{$2}++/gse;
990 warn "# depends_on ",dump $depends_on;
992 my $test_added = Storable::dclone $test_changed;
993 delete $test_added->{$_} foreach keys %$depends_on;
995 $code_depends = $self->param('code_depends')
996 || join(',', keys %$depends_on);
998 $code_description = $self->param('code_description') ||
999 join(',', keys %$test_added);
1001 $code_depends ||= $code_description; # self-modifing
1002 if ( ! $code_depends && $out ) {
1003 $code_depends = $key;
1004 $code_description = $value;
1007 warn "# test_changed ",dump( $test_changed, $code_depends, $code_description );
1017 sorted => $sorted_items,
1018 columns => [ @columns ],
1019 rows => $#$filtered + 1,
1020 numeric => { map { $_, $self->_is_numeric($_) } @columns },
1021 unique => { map { $_, $self->_is_unique( $_) } @columns },
1022 filters => $self->_current_filters,
1024 cols_changed => $commit ? $commit_changed : $test_changed,
1025 code_depends => $code_depends,
1026 code_description => $code_description,
1027 code_path => $code_path,
1036 $self->session('order', $self->param('order'));
1037 $self->session('sort', $self->param('sort'));
1038 return $self->redirect_to('/data/items');
1042 my ( $self, $name ) = @_;
1044 my $stats = $self->_loaded( 'stats' );
1046 # sort facet numerically if more >50% elements are numeric
1047 my $count = $stats->{$name}->{count};
1048 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
1049 defined $stats->{$name}->{numeric} &&
1050 $stats->{$name}->{numeric} > $count / 2;
1054 my ( $self, $name ) = @_;
1055 my $stats = $self->_loaded( 'stats' );
1056 defined $stats->{$name}->{unique};
1059 sub _remove_filter {
1060 my ($self,$name) = @_;
1061 warn "_remove_filter $name\n";
1063 my $path = $self->session('path');
1065 delete $filters->{$name};
1066 delete $loaded->{$path}->{filters}->{$name};
1067 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
1070 grep { /\Q$name\E/ }
1071 keys %{ $loaded->{$path}->{filtered} }
1073 delete $loaded->{$path}->{filtered}->{$_};
1074 warn "remove filtered cache $_";
1081 my $path = $self->session('path') || return $self->redirect_to( '/data/index' );
1083 if ( my $name = $self->param('remove') ) {
1084 $self->_remove_filter( $name );
1085 return $self->redirect_to( '/data/items' );
1089 my $name = $self->param('name') || die "no name";
1091 my $all = $self->_param_scalar('all', 1);
1092 my $data = $self->_loaded('data');
1094 my $filters = $self->_current_filters;
1095 my $all_filters = __all_filters( keys %$filters,$self->session('order') );
1096 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
1097 if defined $loaded->{$path}->{filtered}->{$all_filters};
1099 if ( ! $filtered || $all ) {
1100 $filtered = [ 0 .. $#{ $data->{items} } ];
1101 warn "filter all values\n";
1103 warn "filter using $all_filters\n";
1106 foreach my $i ( @$filtered ) {
1107 my $item = $data->{items}->[$i];
1108 if ( ! exists $item->{$name} || ! defined $item->{$name} ) {
1109 $facet->{ _missing }++;
1110 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
1111 $facet->{$_}++ foreach @{ $item->{$name} };
1113 $facet->{ $item->{$name} }++;
1117 my $checked_values = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
1119 if ( my $code = $self->param('code') ) {
1121 foreach my $value ( keys %$facet ) {
1122 my $count = $facet->{$value};
1123 my $checked = $checked_values->{$value};
1127 warn "ERROR: $@\n$code\n";
1129 } elsif ( $checked != $checked_values->{$value} ) {
1130 warn "checked $value $count -> $checked\n";
1131 $checked_values->{$value} = $checked;
1134 warn "out ",dump( $out );
1135 $self->stash( out => $out );
1138 # warn "# facet $name ",dump $facet;
1142 : defined $filters->{$name} ? @{ $filters->{$name} }
1145 my $numeric = $self->_is_numeric($name);
1147 my $sort = $self->param('sort');
1148 # sort numeric facets with more than 5 values ascending
1149 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
1151 @facet_names = sort {
1153 if ( $sort eq 'a' ) {
1154 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
1155 } elsif ( $sort eq 'd' ) {
1156 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
1157 } elsif ( $sort eq 'c' ) {
1158 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
1160 warn "unknown sort: $sort";
1161 $result = $a cmp $b;
1163 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
1167 $self->render( name => $name, facet => $facet, checked => $checked_values,
1168 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
1173 sub __invalidate_path_column {
1174 my ( $path, $name ) = @_;
1176 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
1177 delete $loaded->{$path}->{sorted}->{$name};
1178 warn "# invalidate $path sorted $name\n";
1181 foreach ( grep { m/\Q$name\E/ } keys %{ $loaded->{$path}->{filtered} } ) {
1182 delete $loaded->{$path}->{filtered}->{$_};
1183 warn "# invalidate $path filtered $_\n";
1187 sub __path_modified {
1188 my ( $path, $value ) = @_;
1189 $value = 1 unless defined $value;
1191 $loaded->{$path}->{modified} = $value;
1193 warn "# __path_modified $path $value\n";
1196 sub __path_rebuild_stats { $loaded->{ $_[0] }->{rebuild_stats} = 1 };
1199 my ($self,$change) = @_;
1201 my $change_path = $self->_permanent_path( 'changes' );
1202 mkdir $change_path unless -d $change_path;
1203 $change_path .= '/' . $change->{time};
1204 store $change, $change_path;
1205 utime $change->{time}, $change->{time}, $change_path;
1206 warn "_save_change $change_path ", dump($change);
1211 my $new_content = $self->param('new_content');
1212 $new_content ||= $self->param('content'); # backward compatibility with old actions
1214 my $i = $self->param('_row_id');
1215 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
1216 my $path = $self->param('path') || die "no path";
1217 my $name = $self->param('name') || die "no name";
1218 my $status = 200; # 200 = OK, 201 = Created
1220 my $data = $self->_loaded('data');
1222 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
1223 $new_content =~ s/^\s+//s;
1224 $new_content =~ s/\s+$//s;
1226 if ( $new_content =~ /\xB6/ ) { # para
1227 $v = [ split(/\s*\xB6\s*/, $new_content) ];
1229 $v = [ $new_content ];
1232 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
1235 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
1237 $self->_save_change({
1241 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
1243 time => $self->param('time') || time(),
1244 user => $self->param('user') || $ENV{'LOGNAME'},
1246 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
1247 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
1248 keys %{ $loaded->{$path}->{stats} }
1252 warn "# change $path $i $old -> $new\n";
1253 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
1255 __invalidate_path_column( $path, $name );
1257 $status = 201; # created
1258 __path_rebuild_stats( $path );
1260 $new_content = join("\xB6",@$v);
1263 warn "# unchanged $path $i $old\n";
1267 $new_content = "$path $i $name doesn't exist\n";
1271 warn "# edit $status ", dump $new_content;
1275 new_content => scalar $new_content,
1282 my $path = $self->_param_or_session('path');
1283 my $dump_path = $self->_save( $path );
1284 __path_modified( $path, 0 );
1286 return $self->redirect_to( '/data/items' );
1292 my $dir = $self->app->home->rel_dir('public');
1294 if ( my $import = $self->param('import') ) {
1296 if ( $import =~ m{/filter\.(.+?)\..+} ) {
1297 my $name = $self->_column_from_unac( $1 );
1299 my @vals = map { chomp; $_ }
1300 read_file "$dir/export/$import", binmode => ':utf8';
1302 $self->_remove_filter( $name );
1303 $self->_filter_on_data( $name, @vals );
1304 $self->session( 'offset' => 0 );
1305 return $self->redirect_to('/data/items');
1307 warn "UNKNOWN IMPORT $import";
1311 if ( my $remove = $self->param('remove') ) {
1312 my $path = "$dir/export/$remove";
1313 unlink $path if -e $path;
1315 unlink $path if -e $path;
1318 my $path = $self->_export_path || return $self->redirect_to('/data/index');
1320 my @files = grep { ! /\.png$/ } glob "$path/*";
1321 my $mtime = { map { $_ => (stat($_))[9] } @files };
1322 @files = sort { $mtime->{$b} <=> $mtime->{$a} } @files;
1323 $self->render( export => [ @files ] );
1326 sub __loaded_paths {
1328 grep { defined $loaded->{$_}->{data} }
1334 $self->stash( reload => 1 );
1336 # $self->_load_path( $self->param('path') );
1337 $self->redirect_to('/data/load?path=' . $self->param('path') );
1342 my $path = $self->param('path');
1343 if ( $path =~ m{^/tmp/mojo_facets\.} ) {
1345 warn "# unlink $path";
1346 if ( my $name = $self->param('name') ) {
1347 delete $loaded->{$name};
1348 warn "# remove $name from memory";
1351 warn "WARNING: $path unlink ignored";
1353 return if $self->stash('reload');
1354 return $self->redirect_to( '/data/load' );