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 return if $full_path =~ m/\.columns$/;
39 foreach my $ext ( keys %{ $imports->{file} } ) {
40 if ( -f $full_path && $full_path =~ m/$ext/i ) {
41 return $imports->{file}->{$ext};
46 foreach my $ext ( keys %{ $imports->{directory} } ) {
47 if ( -f $full_path && $full_path =~ m/$ext/i ) {
48 return $imports->{directory}->{$ext};
60 my $data_dir = $self->app->home->rel_dir('data');
61 die "no data dir $data_dir" unless -d $data_dir;
67 my $file = $File::Find::name;
68 if ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
69 push @{ $changes->{$1} }, $2
70 } elsif ( import_module( $file ) ) {
71 $file =~ s/$data_dir\/*//;
74 #warn "IGNORE: $file\n";
78 no warnings qw(uninitialized); # mtime
79 @files = sort { $loaded->{$b}->{mtime} <=> $loaded->{$a}->{mtime} || lc $a cmp lc $b } @files,
80 grep { defined $loaded->{$_}->{generated} } keys %$loaded;
83 $size->{$_} = -s "$data_dir/$_" foreach @files;
90 dump_path => { map { $_ => $self->_dump_path($_) } @files },
96 my ( $self, $name ) = @_;
97 my $dir = $self->app->home->rel_dir('data');
100 return '/tmp/mojo_facets.' . $name . '.storable';
104 my ( $self, $path ) = @_;
106 my $dump_path = $self->_dump_path( $path );
107 my $first_load = ! -e $dump_path;
108 warn "save loaded to $dump_path";
109 my $info = $loaded->{$path};
110 store $info, $dump_path;
113 my $mtime = $loaded->{$path}->{mtime};
114 utime $mtime, $mtime, $dump_path;
115 warn "sync time to $path at $mtime\n";
118 warn $dump_path, ' ', -s $dump_path, " bytes\n";
127 my $nr_items = $#{ $_[0] } + 1;
129 warn "__stats $nr_items\n";
131 # foreach my $e ( @{ $_[0] } ) {
132 foreach my $i ( 0 .. $#{$_[0]} ) {
133 print STDERR " $i" if $i % 5000;
135 foreach my $n ( keys %$e ) {
136 $stats->{$n}->{count}++;
138 if ( ref $e->{$n} eq 'ARRAY' ) {
139 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
145 foreach my $x ( @v ) {
146 if ( ! defined $x ) { # FIXME really null
147 $stats->{$n}->{empty}++;
151 $stats->{$n}->{numeric}++
152 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
153 $stats->{$n}->{empty}++
154 if length $x == 0; # faster than $x =~ m/^\s*$/;
160 foreach my $n ( keys %$stats ) {
161 my $s = $stats->{$n};
162 next unless defined $s->{array};
163 if ( $s->{array} == $s->{count} ) {
165 if ( $s->{count} == $nr_items ) {
166 warn "check $n for uniqeness\n";
168 foreach my $e ( @{ $_[0] } ) {
169 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
175 $stats->{$n}->{unique} = 1;
176 #warn "# $n unique ",dump( $unique );
182 warn "# __stats ",dump($stats);
187 sub _param_or_session {
188 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
193 my $path = $self->_param_or_session('path');
194 warn "stats $path\n";
195 delete $loaded->{$path}->{stats};
196 return $self->redirect_to( '/data/columns' );
201 my ( $self, $path ) = @_;
203 return if defined $loaded->{$path}->{generated};
205 my $full_path = $self->app->home->rel_file( 'data/' . $path );
206 return $self->redirect_to('/data/index') unless -r $full_path;
208 my $dump_path = $self->_dump_path( $path );
210 if ( defined $loaded->{$path}->{data} ) {
211 my $mtime = (stat($full_path))[9];
212 return if $loaded->{$path}->{mtime} == $mtime;
213 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
214 } elsif ( -e $dump_path ) {
215 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
216 my $info = retrieve $dump_path;
217 $loaded->{ $path } = $info;
222 if ( my $module = import_module( $full_path ) ) {
223 $data = $module->new( full_path => $full_path )->data;
225 die "can't load $full_path";
230 if ( defined $data->{header} ) {
231 if ( ref $data->{header} eq 'ARRAY' ) {
232 @header = @{ $data->{header} };
234 warn "header not array ", dump( $data->{header} );
238 my $stats = __stats( $data->{items} );
241 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
242 grep { defined $stats->{$_}->{count} } keys %$stats
246 header => [ @header ],
248 full_path => $full_path,
249 size => -s $full_path,
250 mtime => (stat($full_path))[9],
252 defined $data->{generated} ? ( generated => 1 ) : (),
255 $loaded->{ $path } = $info;
256 $self->_save( $path ) unless $info->{generated};
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 my $path = $self->param('path') || $self->session('path') || $paths[0] || $self->redirect_to('/data/index');
277 warn "# path $path\n";
278 $self->_load_path( $path );
280 $self->session( 'path' => $path );
282 my $redirect_to = '/data/items';
284 $self->session( 'header' => $loaded->{$path}->{header} );
285 if ( ! defined $loaded->{$path}->{columns} ) {
286 my $columns_path = $self->_permanent_path( 'columns' );
287 if ( -e $columns_path ) {
288 my @columns = map { s/[\r\n]+$//; $_ } read_file $columns_path, binmode => ':utf8';
289 $loaded->{$path}->{columns} = [ @columns ];
290 warn "# columns_path $columns_path ",dump(@columns);
292 $loaded->{$path}->{columns} = $loaded->{$path}->{header}
295 $redirect_to = '/data/columns';
297 $self->session( 'columns' => $loaded->{$path}->{columns} );
298 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
299 return $self->redirect_to( $redirect_to );
304 my ( $self, $name ) = @_;
305 my $path = $self->session('path') || $self->param('path');
306 return $self->redirect_to('/data/index') unless $path;
308 if ( defined $loaded->{$path}->{modified} && $loaded->{$path}->{modified} > 1 ) {
309 my $caller = (caller(1))[3];
310 if ( $caller =~ m/::edit/ ) {
311 warn "rebuild stats for $path ignored caller $caller\n";
313 warn "rebuild stats for $path FORCED by modified caller $caller\n";
314 # $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
315 $loaded->{$path}->{rebuild_stats} = 1;
316 $loaded->{$path}->{modified} = 1;
320 if ( defined $loaded->{$path}->{rebuild_stats} ) {
321 warn "rebuild_stats $path";
322 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
323 delete $loaded->{$path}->{rebuild_stats};
326 if ( ! defined $loaded->{$path}->{$name} ) {
327 warn "$path $name isn't loaded\n";
328 $self->_load_path( $path );
329 if ( ! defined $loaded->{$path}->{stats} ) {
330 warn "rebuild stats for $path\n";
331 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
333 if ( ! defined $loaded->{$path}->{$name} ) {
334 warn "MISSING $name for $path\n";
335 return $self->redirect_to('/data/index')
339 $self->session( 'modified' => $loaded->{$path}->{modified} );
341 return $loaded->{$path}->{$name};
348 $checked->{$_}++ foreach @_;
349 # warn "# _checked ",dump($checked);
353 sub _permanent_path {
355 my $path = $self->_param_or_session('path');
356 $self->app->home->rel_dir('data') . '/' . join('.', $path, @_);
361 $n = unac_string('utf-8',$n);
366 sub _column_from_unac {
367 my ($self,$name) = @_;
369 my $stats = $self->_loaded('stats');
370 my $cols_norm = { map { __unac( $_ ) => $_ } keys %$stats };
372 $cols_norm->{$name} || die "can't find column $name in ", dump($cols_norm);
377 my $path = $self->_param_or_session('path');
379 warn "no path in param or session";
382 my $dir = $self->app->home->rel_dir('public') . "/export/$path";
383 mkpath $dir unless -e $dir;
384 my $name = __export_path_name( $path, @_ );
385 my $full = $dir . '/' . $name;
386 $full =~ s/\/+$// if -d $full; # strip trailing slash for dirs
390 sub __export_path_name {
394 my $name = join('.', map { __unac($_) } @_ );
395 if ( length($name) > $max_length ) {
396 $name = substr($name,0,$max_length) . Digest::MD5::md5_hex substr($name,$max_length);
404 my $view_path = $self->_permanent_path( 'views' );
406 if ( $self->param('columns') ) {
407 my @columns = $self->_param_array('columns');
408 write_file( $self->_permanent_path( 'columns' ), { binmode => ':utf8' }, map { "$_\n" } @columns );
409 if ( my $view = $self->param('view') ) {
410 mkdir $view_path unless -e $view_path;
411 write_file( "$view_path/$view", { binmode => ':utf8' }, map { "$_\n" } @columns );
412 warn "view $view_path/$view ", -s "$view_path/$view", " bytes\n";
415 return $self->redirect_to('/data/items');
417 } elsif ( ! $self->session('header') ) {
418 return $self->redirect_to('/data/load');
421 if ( my $id = $self->param('id') ) {
422 my $view_full = "$view_path/$id";
423 if ( -f $view_full ) {
424 my @columns = map { chomp; $_ } read_file $view_full, binmode => ':utf8';
425 warn "view $view_full loaded ", dump @columns;
426 $self->session( 'columns' => [ @columns ] );
427 return $self->redirect_to('/data/items');
431 my $stats = $self->_loaded( 'stats' );
434 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
436 foreach my $c ( @{ $self->session( 'header' ) } ) {
437 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
441 if ( -d $view_path ) {
442 @views = map { s{^\Q$view_path\E/*}{}; $_ } glob "$view_path/*";
443 warn "# views ",dump @views;
447 message => 'Select columns to display',
449 columns => \@columns,
450 checked => $self->_checked( $self->_param_array('columns') ),
456 my ($self,$name) = @_;
458 my @array = $self->param($name);
459 my $path = $self->session('path');
462 $self->session($name => [ @array ]);
463 $loaded->{$path}->{$name} = [ @array ];
464 } elsif ( my $session = $self->session($name) ) {
465 if ( ref $session eq 'ARRAY' ) {
468 die "$name not array ",dump($session);
471 #warn "# $name ",dump @array;
476 my ($self,$name,$default) = @_;
478 my $scalar = $self->param($name);
480 if ( defined $scalar ) {
481 $self->session($name => $scalar);
483 $scalar = $self->session($name);
486 if ( ! defined $scalar ) {
488 if ( defined $scalar ) {
489 $self->session($name => $scalar);
491 warn "no default for $name";
495 warn "# _param_scalar $name ",dump $scalar;
502 my $name = $self->param('filter_name') || die "name?";
503 my @vals = $self->param('filter_vals');
505 $self->_remove_filter( $name );
507 $self->_filter_on_data( $name, @vals );
508 if ( my $permanent = $self->param('_permanent') ) {
509 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
510 write_file $permanent_path, { binmode => ':utf8' }, map { "$_\n" } @vals;
511 warn "permanent filter $permanent_path ", -s $permanent_path;
515 $self->session( 'offset' => 0 );
517 return $self->redirect_to('/data/items');
520 sub _filter_on_data {
521 my ( $self, $name, @vals ) = @_;
523 my $path = $self->session('path');
525 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
526 warn "filter $name not found in data set";
530 $filters->{$name} = [ @vals ];
531 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
534 $filter_hash->{$_}++ foreach @vals;
536 #warn "# filter_hash ",dump( $filter_hash );
538 my $items = $self->_loaded('data')->{items};
540 my $include_missing = defined $filter_hash->{_missing};
543 foreach my $i ( 0 .. $#$items ) {
545 if ( defined $items->[$i]->{$name} ) {
546 my $row = $items->[$i]->{$name};
547 $row = [ $row ] unless ref $row eq 'ARRAY'; # FIXME probably wrong place
548 foreach my $v ( @$row ) {
549 if ( defined $filter_hash->{ $v } ) {
550 $filtered_items->{$i}++;
553 } elsif ( $include_missing ) {
554 $filtered_items->{$i}++;
558 #warn "# filter $name ",dump($filtered_items);
560 # invalidate filters on other datasets
561 foreach my $dataset ( grep { exists $loaded->{$_}->{filters}->{$name} } keys %$loaded ) {
562 delete $loaded->{$dataset}->{filters}->{$name};
563 delete $loaded->{$dataset}->{filtered};
566 $loaded->{$path}->{filters}->{$name} = $filtered_items;
567 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
571 sub _current_filters {
574 my $stats = $self->_loaded('stats');
576 $current_filters->{ $_ } = $filters->{ $_ }
577 foreach ( grep { defined $filters->{ $_ } } keys %$stats )
579 warn "# _current_filters ",dump( keys %$current_filters );
580 return $current_filters;
583 sub _data_sorted_by {
584 my ( $self, $order ) = @_;
586 my $path = $self->session('path');
588 warn "_data_sorted_by $order from $path";
590 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
591 return $loaded->{$path}->{sorted}->{$order};
594 my $data = $self->_loaded( 'data' );
595 my $numeric = $self->_is_numeric($order);
596 my $missing = $numeric ? 0 : '';
597 no warnings qw(numeric);
609 if ( ! exists $_->{$order} ) {
611 } elsif ( ref $_->{$order} eq 'ARRAY' ) {
612 $v = join('', @{$_->{$order}});
617 } @{ $data->{items} }
620 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
621 #warn "# sorted ",dump( @sorted );
623 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
629 join(',', sort(@_), 'order', $order);
632 our ($out, $key,$value);
634 our $lookup_path_col;
638 warn "__commit_begin";
639 $lookup_path_col = undef;
645 $lookup_path_col = undef; # cleanup memory
650 warn "# lookup ",dump @_;
651 my ( $vals, $on_path, $on_col, $code, $stat_code ) = @_;
652 die "code is not sub{ ... } but ", dump $code unless ref $code eq 'CODE';
654 if ( ! exists $loaded->{$on_path} ) {
655 my @possible_paths = grep { /\Q$on_path\E/ } keys %$loaded;
656 die "more than one dataset available for '$on_path' ",dump @possible_paths if $#possible_paths > 0;
657 $on_path = shift @possible_paths;
658 warn "## fuzzy selected path $on_path";
661 my $items = $loaded->{$on_path}->{data}->{items} || die "no items for $on_path";
663 if ( ! exists $lookup_path_col->{$on_path}->{$on_col} ) {
664 warn "create lookup_path_col $on_path $on_col";
665 foreach my $i ( 0 .. $#$items ) {
666 my $item = $items->[$i];
667 if ( exists $item->{$on_col} ) {
668 if ( ref $item->{$on_col} eq 'ARRAY' ) {
669 foreach my $v ( @{ $item->{$on_col} } ) {
670 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
672 } elsif ( ! ref $item->{$on_col} ) { # scalar
673 my $v = $item->{$on_col};
674 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
676 die "unknown type of ",dump $item->{$on_col};
680 warn "XXX ",dump $lookup_path_col->{$on_path}->{$on_col} if $ENV{DEBUG};
684 $stat = Statistics::Descriptive::Full->new() if $stat_code;
686 foreach my $v ( ref $vals eq 'ARRAY' ? @$vals : ( $vals ) ) {
687 foreach my $i ( @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} } ) {
689 warn "XXX lookup code $v $i ",dump $on if $ENV{DEBUG};
692 $stat_code->( $stat ) if $stat_code;
696 sub __commit_path_code {
697 my ( $path, $i, $code, $commit_changed ) = @_;
699 my $items = $loaded->{$path}->{data}->{items} || die "no items for $path";
700 my $row = $items->[$i];
703 foreach ( keys %$update ) {
704 $$commit_changed->{$_}++;
705 $loaded->{$path}->{data}->{items}->[$i]->{$_} = $update->{$_};
707 #warn "__commit_path_code $path $i ",dump( $update );
710 # uses templates/admin.html.ep
711 sub _switch_dataset {
716 foreach my $path ( keys %$loaded ) {
717 next unless exists $loaded->{$path}->{data};
718 push @$datasets, $path;
721 warn "# datasets ",dump($datasets);
723 $self->stash( 'datasets' => $datasets );
729 $self->_switch_dataset;
731 if ( my $show = $self->param('id') ) {
732 $self->param('show', $show);
736 my $path = $self->_param_scalar('path');
738 my @columns = $self->_param_array('columns');
739 return $self->redirect_to('/data/columns') unless @columns;
740 my $order = $self->_param_scalar('order', $columns[0]);
741 my $sort = $self->_param_scalar('sort', 'a');
742 my $offset = $self->_param_scalar('offset', 0);
743 my $limit = $self->_param_scalar('limit', 20);
744 $self->_param_scalar('show', 'table');
746 # fix offset when changing limit
747 $offset = int( $offset / $limit ) * $limit;
749 if ( ! grep { /^\Q$order\E$/ } @columns ) {
750 $order = $columns[0];
751 $self->session( order => $order );
753 my $sorted = $self->_data_sorted_by( $order );
757 foreach my $name ( keys %$filters ) {
758 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
759 warn "skip filter $name not found in $path\n";
762 push @filter_names, $name;
764 warn "filter_names ",dump( @filter_names );
765 foreach my $name ( @filter_names ) {
766 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
767 $self->_filter_on_data( $name, @{ $filters->{$name} } );
771 my $all_filters = __all_filters( @filter_names,$order );
773 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
775 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
777 my $path_filters = $loaded->{$path}->{filters};
779 warn "create combined filter for $all_filters from ", $#$sorted + 1, " items\n";
782 foreach my $i ( 0 .. $#$sorted ) {
783 my $pos = $sorted->[$i];
785 if ( $#filter_names == -1 ) {
786 push @filtered, $pos;
791 foreach ( @filter_names ) {
792 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
796 push @filtered, $pos;
799 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
802 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
803 if defined $loaded->{$path}->{filtered}->{$all_filters};
805 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
807 my $data = $self->_loaded('data');
809 my $code = $self->_param_scalar('code','');
811 $code =~ s{\n+$}{\n}s;
813 # XXX convert @row->{foo} into @{$row->{foo}}
814 $code =~ s|\@(row->{[^}]+})|\@{\$$1}|gs;
816 my $commit = $self->param('commit');
817 my $test = $self->param('test');
822 if ( $code && ( $test || $commit ) ) {
823 # XXX find columns used in code snippet and show them to user
825 foreach my $column ( $code =~ m/\$row->{([^}]+)}/g ) {
826 if ( $column =~ s/^(['"])// ) {
829 next if $column =~ m/\$/; # hide columns with vars in them
830 $commit_changed->{$column} = 0;
834 my $code_path = $self->app->home->rel_dir('public') . "/code";
837 __path_modified( $path, 'commit' );
839 warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
840 ( $key, $value, $out ) = ( 'key', 'value' );
841 foreach ( 0 .. $#$filtered ) {
842 my $i = $filtered->[$_];
843 __commit_path_code( $path, $i, $code, \$commit_changed );
846 # this might move before $out to recalculate stats on source dataset?
847 __path_rebuild_stats( $path );
848 my $c = { map { $_ => 1 } @columns };
849 my @added_columns = sort grep { ! $c->{$_} } keys %$commit_changed;
850 warn "# added_columns ",dump( @added_columns );
851 unshift @columns, @added_columns;
853 $loaded->{$path}->{columns} = [ @columns ];
854 warn "# new columns ",dump( @columns );
856 __invalidate_path_column( $path, $_ ) foreach keys %$commit_changed;
858 $self->_save_change({
860 time => $self->param('time') || time(),
861 user => $self->param('user') || $ENV{'LOGNAME'},
863 commit_changed => $commit_changed,
866 if ( my $description = $self->param('code_description') ) {
867 my $depends = $self->param('code_depends') || die "no code_depends?";
868 my $path = "$code_path/$depends.$description.pl";
869 if ( -e $path && ! $self->param('overwrite') ) {
870 warn "# code $path not saved\n";
872 write_file( $path, { binmode => ':utf8' }, "$code\n" );
873 warn "code $path ", -s $path, " bytes saved\n";
880 my $commit_dataset = join('.'
881 , $self->param('code_depends')
882 , $self->param('code_description')
887 warn "key $key value $value";
889 foreach my $n ( keys %$out ) {
890 my $i = { $key => [ $n ] };
891 my $ref = ref $out->{$n};
892 if ( $ref eq 'HASH' ) {
893 $i->{$_} = [ $out->{$n}->{$_} ] foreach keys %{ $out->{$n} };
894 } elsif ( $ref eq 'ARRAY' ) {
895 $i->{$_} = $out->{$n};
897 $i->{$value} = [ $out->{$n} ];
899 $i->{_error} = [ dump($out->{$n}) ];
904 my $stats = __stats( $items );
905 my @columns = grep { ! m/^\Q$key\E$/ } sort keys %$stats;
906 unshift @columns, $key;
908 $loaded->{$commit_dataset} = {
909 header => [ @columns ],
910 columns => [ @columns ],
912 data => { items => $items },
916 warn "# loaded out ", dump( $loaded->{$commit_dataset} );
917 $self->session('path', $commit_dataset);
918 $self->session('columns', [ @columns ]);
919 $self->session('order', $key);
920 return $self->redirect_to('/data/items');
923 $self->session('columns', [ @columns ]);
927 my $from_end = $sort eq 'd' ? $#$filtered : 0;
929 my ( $key, $value, $out ) = ( 'key', 'value' ); # XXX make local
930 foreach ( 0 .. $limit ) {
931 my $i = $_ + $offset;
932 last unless defined $filtered->[$i];
933 $i = $from_end - $i if $from_end;
934 my $id = $filtered->[$i];
935 my $row = Storable::dclone $data->{items}->[ $id ];
936 if ( $code && $test ) {
940 warn "ERROR evaling $@", dump($code);
941 $self->stash('eval_error', $@) if $@;
943 warn "EVAL ",dump($update);
944 foreach ( keys %$update ) {
945 $test_changed->{$_}++;
946 $row->{$_} = $update->{$_};
950 $row->{_row_id} ||= $id;
951 push @$sorted_items, $row;
954 if ( $self->param('export') ) {
955 my $export_path = $self->_export_path( 'items', @columns);
956 open(my $fh, '>', $export_path) || die "ERROR: can't open $export_path: $!";
957 print $fh "#",join("\t",@columns),"\n";
958 foreach my $f ( 0 .. $#$filtered ) {
959 print $fh join("\t", map {
960 my $i = $data->{items}->[ $filtered->[$f] ];
962 if ( ! defined $i->{$_} ) {
964 } elsif ( ref $i->{$_} eq 'ARRAY' ) {
965 $v =join(',', @{ $i->{$_} });
966 $v = '\N' if length($v) == 0;
967 } elsif ( ! ref $i->{$_} ) {
976 warn "export $export_path ", -s $export_path, " bytes\n";
979 my ( $code_depends, $code_description );
983 warn "# test_changed ",dump( $test_changed );
984 my $c = { map { $_ => 1 } @columns };
985 my @added_columns = sort grep { ! $c->{$_} } keys %$test_changed;
986 unshift @columns, @added_columns;
988 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
991 my $tmp = $code; $tmp =~ s/\$row->{(['"]?)([\w\s]+)\1/$depends_on->{$2}++/gse;
992 warn "# depends_on ",dump $depends_on;
994 my $test_added = Storable::dclone $test_changed;
995 delete $test_added->{$_} foreach keys %$depends_on;
997 $code_depends = $self->param('code_depends')
998 || join(',', keys %$depends_on);
1000 $code_description = $self->param('code_description') ||
1001 join(',', keys %$test_added);
1003 $code_depends ||= $code_description; # self-modifing
1004 if ( ! $code_depends && $out ) {
1005 $code_depends = $key;
1006 $code_description = $value;
1009 warn "# test_changed ",dump( $test_changed, $code_depends, $code_description );
1019 sorted => $sorted_items,
1020 columns => [ @columns ],
1021 rows => $#$filtered + 1,
1022 numeric => { map { $_, $self->_is_numeric($_) } @columns },
1023 unique => { map { $_, $self->_is_unique( $_) } @columns },
1024 filters => $self->_current_filters,
1026 cols_changed => $commit ? $commit_changed : $test_changed,
1027 code_depends => $code_depends,
1028 code_description => $code_description,
1029 code_path => $code_path,
1038 $self->session('order', $self->param('order'));
1039 $self->session('sort', $self->param('sort'));
1040 return $self->redirect_to('/data/items');
1044 my ( $self, $name ) = @_;
1046 my $stats = $self->_loaded( 'stats' );
1048 # sort facet numerically if more >50% elements are numeric
1049 my $count = $stats->{$name}->{count};
1050 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
1051 defined $stats->{$name}->{numeric} &&
1052 $stats->{$name}->{numeric} > $count / 2;
1056 my ( $self, $name ) = @_;
1057 my $stats = $self->_loaded( 'stats' );
1058 defined $stats->{$name}->{unique};
1061 sub _remove_filter {
1062 my ($self,$name) = @_;
1063 warn "_remove_filter $name\n";
1065 my $path = $self->session('path');
1067 delete $filters->{$name};
1068 delete $loaded->{$path}->{filters}->{$name};
1069 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
1072 grep { /\Q$name\E/ }
1073 keys %{ $loaded->{$path}->{filtered} }
1075 delete $loaded->{$path}->{filtered}->{$_};
1076 warn "remove filtered cache $_";
1083 my $path = $self->session('path') || return $self->redirect_to( '/data/index' );
1085 if ( my $name = $self->param('remove') ) {
1086 $self->_remove_filter( $name );
1087 return $self->redirect_to( '/data/items' );
1091 my $name = $self->param('name') || die "no name";
1093 my $all = $self->_param_scalar('all', 1);
1094 my $data = $self->_loaded('data');
1096 my $filters = $self->_current_filters;
1097 my $all_filters = __all_filters( keys %$filters,$self->session('order') );
1098 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
1099 if defined $loaded->{$path}->{filtered}->{$all_filters};
1101 if ( ! $filtered || $all ) {
1102 $filtered = [ 0 .. $#{ $data->{items} } ];
1103 warn "filter all values\n";
1105 warn "filter using $all_filters\n";
1108 foreach my $i ( @$filtered ) {
1109 my $item = $data->{items}->[$i];
1110 if ( ! exists $item->{$name} || ! defined $item->{$name} ) {
1111 $facet->{ _missing }++;
1112 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
1113 $facet->{$_}++ foreach @{ $item->{$name} };
1115 $facet->{ $item->{$name} }++;
1119 my $checked_values = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
1121 if ( my $code = $self->param('code') ) {
1123 foreach my $value ( keys %$facet ) {
1124 my $count = $facet->{$value};
1125 my $checked = $checked_values->{$value};
1129 warn "ERROR: $@\n$code\n";
1131 } elsif ( $checked != $checked_values->{$value} ) {
1132 warn "checked $value $count -> $checked\n";
1133 $checked_values->{$value} = $checked;
1136 warn "out ",dump( $out );
1137 $self->stash( out => $out );
1140 # warn "# facet $name ",dump $facet;
1144 : defined $filters->{$name} ? @{ $filters->{$name} }
1147 my $numeric = $self->_is_numeric($name);
1149 my $sort = $self->param('sort');
1150 # sort numeric facets with more than 5 values ascending
1151 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
1153 @facet_names = sort {
1155 if ( $sort eq 'a' ) {
1156 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
1157 } elsif ( $sort eq 'd' ) {
1158 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
1159 } elsif ( $sort eq 'c' ) {
1160 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
1162 warn "unknown sort: $sort";
1163 $result = $a cmp $b;
1165 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
1169 $self->render( name => $name, facet => $facet, checked => $checked_values,
1170 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
1175 sub __invalidate_path_column {
1176 my ( $path, $name ) = @_;
1178 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
1179 delete $loaded->{$path}->{sorted}->{$name};
1180 warn "# invalidate $path sorted $name\n";
1183 foreach ( grep { m/\Q$name\E/ } keys %{ $loaded->{$path}->{filtered} } ) {
1184 delete $loaded->{$path}->{filtered}->{$_};
1185 warn "# invalidate $path filtered $_\n";
1189 sub __path_modified {
1190 my ( $path, $value ) = @_;
1191 $value = 1 unless defined $value;
1193 $loaded->{$path}->{modified} = $value;
1195 warn "# __path_modified $path $value\n";
1198 sub __path_rebuild_stats { $loaded->{ $_[0] }->{rebuild_stats} = 1 };
1201 my ($self,$change) = @_;
1203 my $change_path = $self->_permanent_path( 'changes' );
1204 mkdir $change_path unless -d $change_path;
1205 $change_path .= '/' . $change->{time};
1206 store $change, $change_path;
1207 utime $change->{time}, $change->{time}, $change_path;
1208 warn "_save_change $change_path ", dump($change);
1213 my $new_content = $self->param('new_content');
1214 $new_content ||= $self->param('content'); # backward compatibility with old actions
1216 my $i = $self->param('_row_id');
1217 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
1218 my $path = $self->param('path') || die "no path";
1219 my $name = $self->param('name') || die "no name";
1220 my $status = 200; # 200 = OK, 201 = Created
1222 my $data = $self->_loaded('data');
1224 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
1225 $new_content =~ s/^\s+//s;
1226 $new_content =~ s/\s+$//s;
1228 if ( $new_content =~ /\xB6/ ) { # para
1229 $v = [ split(/\s*\xB6\s*/, $new_content) ];
1231 $v = [ $new_content ];
1234 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
1237 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
1239 $self->_save_change({
1243 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
1245 time => $self->param('time') || time(),
1246 user => $self->param('user') || $ENV{'LOGNAME'},
1248 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
1249 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
1250 keys %{ $loaded->{$path}->{stats} }
1254 warn "# change $path $i $old -> $new\n";
1255 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
1257 __invalidate_path_column( $path, $name );
1259 $status = 201; # created
1260 __path_rebuild_stats( $path );
1262 $new_content = join("\xB6",@$v);
1265 warn "# unchanged $path $i $old\n";
1269 $new_content = "$path $i $name doesn't exist\n";
1273 warn "# edit $status ", dump $new_content;
1277 new_content => scalar $new_content,
1284 my $path = $self->_param_or_session('path');
1285 my $dump_path = $self->_save( $path );
1286 __path_modified( $path, 0 );
1288 return $self->redirect_to( '/data/items' );
1294 my $dir = $self->app->home->rel_dir('public');
1296 if ( my $import = $self->param('import') ) {
1298 if ( $import =~ m{/filter\.(.+?)\..+} ) {
1299 my $name = $self->_column_from_unac( $1 );
1301 my @vals = map { chomp; $_ }
1302 read_file "$dir/export/$import", binmode => ':utf8';
1304 $self->_remove_filter( $name );
1305 $self->_filter_on_data( $name, @vals );
1306 $self->session( 'offset' => 0 );
1307 return $self->redirect_to('/data/items');
1309 warn "UNKNOWN IMPORT $import";
1313 if ( my $remove = $self->param('remove') ) {
1314 my $path = "$dir/export/$remove";
1315 unlink $path if -e $path;
1317 unlink $path if -e $path;
1320 my $path = $self->_export_path || return $self->redirect_to('/data/index');
1322 my @files = grep { ! /\.png$/ } glob "$path/*";
1323 my $mtime = { map { $_ => (stat($_))[9] } @files };
1324 @files = sort { $mtime->{$b} <=> $mtime->{$a} } @files;
1325 $self->render( export => [ @files ] );
1328 sub __loaded_paths {
1330 grep { defined $loaded->{$_}->{data} }
1336 $self->stash( reload => 1 );
1338 # $self->_load_path( $self->param('path') );
1339 $self->redirect_to('/data/load?path=' . $self->param('path') );
1344 my $path = $self->param('path');
1345 if ( $path =~ m{^/tmp/mojo_facets\.} ) {
1347 warn "# unlink $path";
1348 if ( my $name = $self->param('name') ) {
1349 delete $loaded->{$name};
1350 warn "# remove $name from memory";
1353 warn "WARNING: $path unlink ignored";
1355 return if $self->stash('reload');
1356 return $self->redirect_to( '/data/load' );