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;
20 use MojoFacets::Import::File;
21 use MojoFacets::Import::HTMLTable;
22 use MojoFacets::Import::CSV;
23 use MojoFacets::Import::CouchDB;
24 use MojoFacets::Import::SQL;
25 use MojoFacets::Import::Pairs;
33 my $data_dir = $self->app->home->rel_dir('data');
34 die "no data dir $data_dir" unless -d $data_dir;
39 my $file = $File::Find::name;
40 if ( -f $file && $file =~ m/\.(js(on)?|txt)$/ ) {
41 $file =~ s/$data_dir\/*//;
43 } elsif ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
44 push @{ $changes->{$1} }, $2
45 } elsif ( -d $file && $file =~ m/\.html$/ ) {
46 $file =~ s/$data_dir\/*//;
48 } elsif ( -f $file && $file =~ m/\.(csv|storable|couchdb|sql|pairs)$/i ) {
49 $file =~ s/$data_dir\/*//;
52 #warn "IGNORE: $file\n";
56 no warnings qw(uninitialized); # mtime
57 @files = sort { $loaded->{$b}->{mtime} <=> $loaded->{$a}->{mtime} || lc $a cmp lc $b } @files,
58 grep { defined $loaded->{$_}->{generated} } keys %$loaded;
61 $size->{$_} = -s "$data_dir/$_" foreach @files;
68 dump_path => { map { $_ => $self->_dump_path($_) } @files },
74 my ( $self, $name ) = @_;
75 my $dir = $self->app->home->rel_dir('data');
78 return '/tmp/mojo_facets.' . $name . '.storable';
82 my ( $self, $path ) = @_;
84 my $dump_path = $self->_dump_path( $path );
85 my $first_load = ! -e $dump_path;
86 warn "save loaded to $dump_path";
87 my $info = $loaded->{$path};
88 store $info, $dump_path;
91 my $mtime = $loaded->{$path}->{mtime};
92 utime $mtime, $mtime, $dump_path;
93 warn "sync time to $path at $mtime\n";
96 warn $dump_path, ' ', -s $dump_path, " bytes\n";
105 my $nr_items = $#{ $_[0] } + 1;
107 warn "__stats $nr_items\n";
109 # foreach my $e ( @{ $_[0] } ) {
110 foreach my $i ( 0 .. $#{$_[0]} ) {
111 print STDERR " $i" if $i % 5000;
113 foreach my $n ( keys %$e ) {
114 $stats->{$n}->{count}++;
116 if ( ref $e->{$n} eq 'ARRAY' ) {
117 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
123 foreach my $x ( @v ) {
124 if ( ! defined $x ) { # FIXME really null
125 $stats->{$n}->{empty}++;
129 $stats->{$n}->{numeric}++
130 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
131 $stats->{$n}->{empty}++
132 if length $x == 0; # faster than $x =~ m/^\s*$/;
138 foreach my $n ( keys %$stats ) {
139 my $s = $stats->{$n};
140 next unless defined $s->{array};
141 if ( $s->{array} == $s->{count} ) {
143 if ( $s->{count} == $nr_items ) {
144 warn "check $n for uniqeness\n";
146 foreach my $e ( @{ $_[0] } ) {
147 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
153 $stats->{$n}->{unique} = 1;
154 #warn "# $n unique ",dump( $unique );
160 warn "# __stats ",dump($stats);
165 sub _param_or_session {
166 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
171 my $path = $self->_param_or_session('path');
172 warn "stats $path\n";
173 delete $loaded->{$path}->{stats};
174 return $self->redirect_to( '/data/columns' );
179 my ( $self, $path ) = @_;
181 return if defined $loaded->{$path}->{generated};
183 my $full_path = $self->app->home->rel_file( 'data/' . $path );
184 return $self->redirect_to('/data/index') unless -r $full_path;
186 my $dump_path = $self->_dump_path( $path );
188 if ( defined $loaded->{$path}->{data} ) {
189 my $mtime = (stat($full_path))[9];
190 return if $loaded->{$path}->{mtime} == $mtime;
191 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
192 } elsif ( -e $dump_path ) {
193 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
194 my $info = retrieve $dump_path;
195 $loaded->{ $path } = $info;
200 if ( -f $full_path ) {
201 if ( $full_path =~ m/.storable$/ ) { # check storable first to catch files copied from /tmp/
202 $data->{generated}++;
203 warn "open $full_path ", -s $full_path, " bytes";
204 open(my $pipe, "<", $full_path) || die $!;
205 while ( my $o = eval { Storable::fd_retrieve $pipe } ) {
206 if ( exists $o->{item} ) {
207 # stream of storable objects
208 push @{ $data->{items} }, $o->{item};
209 } elsif ( exists $o->{data}->{items} ) {
210 # /tmp/mojofacets.*.storable
211 $data->{items} = $o->{data}->{items};
212 $data->{header} = $o->{header};
213 delete $data->{generated};
215 warn "SKIP ",dump($o);
219 warn "loaded ", $#{ $data->{items} } + 1, " items from $full_path\n";
220 } elsif ( $full_path =~ m/.csv/i ) {
221 $data = MojoFacets::Import::CSV->new( full_path => $full_path )->data;
222 } elsif ( $full_path =~ m/.sql/i ) {
223 $data = MojoFacets::Import::SQL->new( full_path => $full_path )->data;
224 } elsif ( $full_path =~ m/.couchdb/i ) {
225 $data = MojoFacets::Import::CouchDB->new( full_path => $full_path )->data;
226 } elsif ( $full_path =~ m/.pairs/i ) {
227 $data = MojoFacets::Import::Pairs->new( full_path => $full_path )->data;
229 $data = MojoFacets::Import::File->new( full_path => $full_path, path => $path )->data;
231 } elsif ( -d $full_path && $full_path =~ m/.html/ ) {
232 $data = MojoFacets::Import::HTMLTable->new( dir => $full_path )->data;
234 die "can't load $full_path";
239 if ( defined $data->{header} ) {
240 if ( ref $data->{header} eq 'ARRAY' ) {
241 @header = @{ $data->{header} };
243 warn "header not array ", dump( $data->{header} );
247 my $stats = __stats( $data->{items} );
250 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
251 grep { defined $stats->{$_}->{count} } keys %$stats
255 header => [ @header ],
257 full_path => $full_path,
258 size => -s $full_path,
259 mtime => (stat($full_path))[9],
261 defined $data->{generated} ? ( generated => 1 ) : (),
264 $loaded->{ $path } = $info;
265 $self->_save( $path ) unless $info->{generated};
273 my @paths = $self->param('paths');
274 warn "# paths ", dump @paths;
276 foreach my $p ( keys %$loaded ) {
277 next if grep { /^\Q$p\E$/ } @paths;
278 warn "remove $p from memory\n";
279 delete $loaded->{$p};
282 $self->_load_path( $_ ) foreach @paths;
284 my $path = $self->param('path') || $self->session('path') || @paths[0] || $self->redirect_to('/data/index');
286 warn "# path $path\n";
287 $self->_load_path( $path );
289 $self->session( 'path' => $path );
291 my $redirect_to = '/data/items';
293 $self->session( 'header' => $loaded->{$path}->{header} );
294 if ( ! defined $loaded->{$path}->{columns} ) {
295 my $columns_path = $self->_permanent_path( 'columns' );
296 if ( -e $columns_path ) {
297 my @columns = map { s/[\r\n]+$//; $_ } read_file $columns_path, binmode => ':utf8';
298 $loaded->{$path}->{columns} = [ @columns ];
299 warn "# columns_path $columns_path ",dump(@columns);
301 $loaded->{$path}->{columns} = $loaded->{$path}->{header}
304 $redirect_to = '/data/columns';
306 $self->session( 'columns' => $loaded->{$path}->{columns} );
307 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
308 return $self->redirect_to( $redirect_to );
313 my ( $self, $name ) = @_;
314 my $path = $self->session('path') || $self->param('path');
315 return $self->redirect_to('/data/index') unless $path;
317 if ( defined $loaded->{$path}->{modified} && $loaded->{$path}->{modified} > 1 ) {
318 my $caller = (caller(1))[3];
319 if ( $caller =~ m/::edit/ ) {
320 warn "rebuild stats for $path ignored caller $caller\n";
322 warn "rebuild stats for $path FORCED by modified caller $caller\n";
323 # $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
324 $loaded->{$path}->{rebuild_stats} = 1;
325 $loaded->{$path}->{modified} = 1;
329 if ( defined $loaded->{$path}->{rebuild_stats} ) {
330 warn "rebuild_stats $path";
331 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
332 delete $loaded->{$path}->{rebuild_stats};
335 if ( ! defined $loaded->{$path}->{$name} ) {
336 warn "$path $name isn't loaded\n";
337 $self->_load_path( $path );
338 if ( ! defined $loaded->{$path}->{stats} ) {
339 warn "rebuild stats for $path\n";
340 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
342 if ( ! defined $loaded->{$path}->{$name} ) {
343 warn "MISSING $name for $path\n";
344 return $self->redirect_to('/data/index')
348 $self->session( 'modified' => $loaded->{$path}->{modified} );
350 return $loaded->{$path}->{$name};
357 $checked->{$_}++ foreach @_;
358 # warn "# _checked ",dump($checked);
362 sub _permanent_path {
364 my $path = $self->_param_or_session('path');
365 $self->app->home->rel_dir('data') . '/' . join('.', $path, @_);
370 $n = unac_string('utf-8',$n);
375 sub _column_from_unac {
376 my ($self,$name) = @_;
378 my $stats = $self->_loaded('stats');
379 my $cols_norm = { map { __unac( $_ ) => $_ } keys %$stats };
381 $cols_norm->{$name} || die "can't find column $name in ", dump($cols_norm);
386 my $path = $self->_param_or_session('path');
388 warn "no path in param or session";
391 my $dir = $self->app->home->rel_dir('public') . "/export/$path";
392 mkpath $dir unless -e $dir;
393 my $name = __export_path_name( $path, @_ );
394 my $full = $dir . '/' . $name;
395 $full =~ s/\/+$// if -d $full; # strip trailing slash for dirs
399 sub __export_path_name {
403 my $name = join('.', map { __unac($_) } @_ );
404 if ( length($name) > $max_length ) {
405 $name = substr($name,0,$max_length) . Digest::MD5::md5_hex substr($name,$max_length);
413 my $view_path = $self->_permanent_path( 'views' );
415 if ( $self->param('columns') ) {
416 my @columns = $self->_param_array('columns');
417 write_file( $self->_permanent_path( 'columns' ), { binmode => ':utf8' }, map { "$_\n" } @columns );
418 if ( my $view = $self->param('view') ) {
419 mkdir $view_path unless -e $view_path;
420 write_file( "$view_path/$view", { binmode => ':utf8' }, map { "$_\n" } @columns );
421 warn "view $view_path/$view ", -s "$view_path/$view", " bytes\n";
424 return $self->redirect_to('/data/items');
426 } elsif ( ! $self->session('header') ) {
427 return $self->redirect_to('/data/load');
430 if ( my $id = $self->param('id') ) {
431 my $view_full = "$view_path/$id";
432 if ( -f $view_full ) {
433 my @columns = map { chomp; $_ } read_file $view_full, binmode => ':utf8';
434 warn "view $view_full loaded ", dump @columns;
435 $self->session( 'columns' => [ @columns ] );
436 return $self->redirect_to('/data/items');
440 my $stats = $self->_loaded( 'stats' );
443 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
445 foreach my $c ( @{ $self->session( 'header' ) } ) {
446 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
450 if ( -d $view_path ) {
451 @views = map { s{^\Q$view_path\E/*}{}; $_ } glob "$view_path/*";
452 warn "# views ",dump @views;
456 message => 'Select columns to display',
458 columns => \@columns,
459 checked => $self->_checked( $self->_param_array('columns') ),
465 my ($self,$name) = @_;
467 my @array = $self->param($name);
468 my $path = $self->session('path');
471 $self->session($name => [ @array ]);
472 $loaded->{$path}->{$name} = [ @array ];
473 } elsif ( my $session = $self->session($name) ) {
474 if ( ref $session eq 'ARRAY' ) {
477 die "$name not array ",dump($session);
480 #warn "# $name ",dump @array;
485 my ($self,$name,$default) = @_;
487 my $scalar = $self->param($name);
489 if ( defined $scalar ) {
490 $self->session($name => $scalar);
492 $scalar = $self->session($name);
495 if ( ! defined $scalar ) {
497 if ( defined $scalar ) {
498 $self->session($name => $scalar);
500 warn "no default for $name";
504 warn "# _param_scalar $name ",dump $scalar;
511 my $name = $self->param('filter_name') || die "name?";
512 my @vals = $self->param('filter_vals');
514 $self->_remove_filter( $name );
516 $self->_filter_on_data( $name, @vals );
517 if ( my $permanent = $self->param('_permanent') ) {
518 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
519 write_file $permanent_path, { binmode => ':utf8' }, map { "$_\n" } @vals;
520 warn "permanent filter $permanent_path ", -s $permanent_path;
524 $self->session( 'offset' => 0 );
526 return $self->redirect_to('/data/items');
529 sub _filter_on_data {
530 my ( $self, $name, @vals ) = @_;
532 my $path = $self->session('path');
534 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
535 warn "filter $name not found in data set";
539 $filters->{$name} = [ @vals ];
540 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
543 $filter_hash->{$_}++ foreach @vals;
545 #warn "# filter_hash ",dump( $filter_hash );
547 my $items = $self->_loaded('data')->{items};
549 my $include_missing = defined $filter_hash->{_missing};
552 foreach my $i ( 0 .. $#$items ) {
554 if ( defined $items->[$i]->{$name} ) {
555 my $row = $items->[$i]->{$name};
556 $row = [ $row ] unless ref $row eq 'ARRAY'; # FIXME probably wrong place
557 foreach my $v ( @$row ) {
558 if ( defined $filter_hash->{ $v } ) {
559 $filtered_items->{$i}++;
562 } elsif ( $include_missing ) {
563 $filtered_items->{$i}++;
567 #warn "# filter $name ",dump($filtered_items);
569 # invalidate filters on other datasets
570 foreach my $dataset ( grep { exists $loaded->{$_}->{filters}->{$name} } keys %$loaded ) {
571 delete $loaded->{$dataset}->{filters}->{$name};
572 delete $loaded->{$dataset}->{filtered};
575 $loaded->{$path}->{filters}->{$name} = $filtered_items;
576 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
580 sub _current_filters {
583 my $stats = $self->_loaded('stats');
585 $current_filters->{ $_ } = $filters->{ $_ }
586 foreach ( grep { defined $filters->{ $_ } } keys %$stats )
588 warn "# _current_filters ",dump( keys %$current_filters );
589 return $current_filters;
592 sub _data_sorted_by {
593 my ( $self, $order ) = @_;
595 my $path = $self->session('path');
597 warn "_data_sorted_by $order from $path";
599 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
600 return $loaded->{$path}->{sorted}->{$order};
603 my $data = $self->_loaded( 'data' );
604 my $numeric = $self->_is_numeric($order);
605 my $missing = $numeric ? 0 : '';
606 no warnings qw(numeric);
618 if ( ! exists $_->{$order} ) {
620 } elsif ( ref $_->{$order} eq 'ARRAY' ) {
621 $v = join('', @{$_->{$order}});
626 } @{ $data->{items} }
629 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
630 #warn "# sorted ",dump( @sorted );
632 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
638 join(',', sort(@_), 'order', $order);
641 our ($out, $key,$value);
643 our $lookup_path_col;
647 warn "__commit_begin";
648 $lookup_path_col = undef;
654 $lookup_path_col = undef; # cleanup memory
659 warn "# lookup ",dump @_;
660 my ( $vals, $on_path, $on_col, $code, $stat_code ) = @_;
661 die "code is not sub{ ... } but ", dump $code unless ref $code eq 'CODE';
663 if ( ! exists $loaded->{$on_path} ) {
664 my @possible_paths = grep { /\Q$on_path\E/ } keys %$loaded;
665 die "more than one dataset available for '$on_path' ",dump @possible_paths if $#possible_paths > 0;
666 $on_path = shift @possible_paths;
667 warn "## fuzzy selected path $on_path";
670 my $items = $loaded->{$on_path}->{data}->{items} || die "no items for $on_path";
672 if ( ! exists $lookup_path_col->{$on_path}->{$on_col} ) {
673 warn "create lookup_path_col $on_path $on_col";
674 foreach my $i ( 0 .. $#$items ) {
675 my $item = $items->[$i];
676 if ( exists $item->{$on_col} ) {
677 if ( ref $item->{$on_col} eq 'ARRAY' ) {
678 foreach my $v ( @{ $item->{$on_col} } ) {
679 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
681 } elsif ( ! ref $item->{$on_col} ) { # scalar
682 my $v = $item->{$on_col};
683 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
685 die "unknown type of ",dump $item->{$on_col};
689 warn "XXX ",dump $lookup_path_col->{$on_path}->{$on_col} if $ENV{DEBUG};
693 $stat = Statistics::Descriptive::Full->new() if $stat_code;
695 foreach my $v ( ref $vals eq 'ARRAY' ? @$vals : ( $vals ) ) {
696 foreach my $i ( @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} } ) {
698 warn "XXX lookup code $v $i ",dump $on if $ENV{DEBUG};
701 $stat_code->( $stat ) if $stat_code;
705 sub __commit_path_code {
706 my ( $path, $i, $code, $commit_changed ) = @_;
708 my $items = $loaded->{$path}->{data}->{items} || die "no items for $path";
709 my $row = $items->[$i];
712 foreach ( keys %$update ) {
713 $$commit_changed->{$_}++;
714 $loaded->{$path}->{data}->{items}->[$i]->{$_} = $update->{$_};
716 #warn "__commit_path_code $path $i ",dump( $update );
719 # uses templates/admin.html.ep
720 sub _switch_dataset {
725 foreach my $path ( keys %$loaded ) {
726 next unless exists $loaded->{$path}->{data};
727 push @$datasets, $path;
730 warn "# datasets ",dump($datasets);
732 $self->stash( 'datasets' => $datasets );
738 $self->_switch_dataset;
740 if ( my $show = $self->param('id') ) {
741 $self->param('show', $show);
745 my $path = $self->_param_scalar('path');
747 my @columns = $self->_param_array('columns');
748 return $self->redirect_to('/data/columns') unless @columns;
749 my $order = $self->_param_scalar('order', $columns[0]);
750 my $sort = $self->_param_scalar('sort', 'a');
751 my $offset = $self->_param_scalar('offset', 0);
752 my $limit = $self->_param_scalar('limit', 20);
753 $self->_param_scalar('show', 'table');
755 # fix offset when changing limit
756 $offset = int( $offset / $limit ) * $limit;
758 if ( ! grep { /^\Q$order\E$/ } @columns ) {
759 $order = $columns[0];
760 $self->session( order => $order );
762 my $sorted = $self->_data_sorted_by( $order );
766 foreach my $name ( keys %$filters ) {
767 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
768 warn "skip filter $name not found in $path\n";
771 push @filter_names, $name;
773 warn "filter_names ",dump( @filter_names );
774 foreach my $name ( @filter_names ) {
775 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
776 $self->_filter_on_data( $name, @{ $filters->{$name} } );
780 my $all_filters = __all_filters( @filter_names,$order );
782 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
784 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
786 my $path_filters = $loaded->{$path}->{filters};
788 warn "create combined filter for $all_filters from ", $#$sorted + 1, " items\n";
791 foreach my $i ( 0 .. $#$sorted ) {
792 my $pos = $sorted->[$i];
794 if ( $#filter_names == -1 ) {
795 push @filtered, $pos;
800 foreach ( @filter_names ) {
801 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
805 push @filtered, $pos;
808 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
811 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
812 if defined $loaded->{$path}->{filtered}->{$all_filters};
814 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
816 my $data = $self->_loaded('data');
818 my $code = $self->_param_scalar('code','');
820 $code =~ s{\n+$}{\n}s;
822 # XXX convert @row->{foo} into @{$row->{foo}}
823 $code =~ s|\@(row->{[^}]+})|\@{\$$1}|gs;
825 my $commit = $self->param('commit');
826 my $test = $self->param('test');
831 if ( $code && ( $test || $commit ) ) {
832 # XXX find columns used in code snippet and show them to user
834 foreach my $column ( $code =~ m/\$row->{([^}]+)}/g ) {
835 if ( $column =~ s/^(['"])// ) {
838 next if $column =~ m/\$/; # hide columns with vars in them
839 $commit_changed->{$column} = 0;
843 my $code_path = $self->app->home->rel_dir('public') . "/code";
846 __path_modified( $path, 'commit' );
848 warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
849 ( $key, $value, $out ) = ( 'key', 'value' );
850 foreach ( 0 .. $#$filtered ) {
851 my $i = $filtered->[$_];
852 __commit_path_code( $path, $i, $code, \$commit_changed );
855 # this might move before $out to recalculate stats on source dataset?
856 __path_rebuild_stats( $path );
857 my $c = { map { $_ => 1 } @columns };
858 my @added_columns = sort grep { ! $c->{$_} } keys %$commit_changed;
859 warn "# added_columns ",dump( @added_columns );
860 unshift @columns, @added_columns;
862 $loaded->{$path}->{columns} = [ @columns ];
863 warn "# new columns ",dump( @columns );
865 __invalidate_path_column( $path, $_ ) foreach keys %$commit_changed;
867 $self->_save_change({
869 time => $self->param('time') || time(),
870 user => $self->param('user') || $ENV{'LOGNAME'},
872 commit_changed => $commit_changed,
875 if ( my $description = $self->param('code_description') ) {
876 my $depends = $self->param('code_depends') || die "no code_depends?";
877 my $path = "$code_path/$depends.$description.pl";
878 if ( -e $path && ! $self->param('overwrite') ) {
879 warn "# code $path not saved\n";
881 write_file( $path, { binmode => ':utf8' }, "$code\n" );
882 warn "code $path ", -s $path, " bytes saved\n";
889 my $commit_dataset = join('.'
890 , $self->param('code_depends')
891 , $self->param('code_description')
896 warn "key $key value $value";
898 foreach my $n ( keys %$out ) {
899 my $i = { $key => [ $n ] };
900 my $ref = ref $out->{$n};
901 if ( $ref eq 'HASH' ) {
902 $i->{$_} = [ $out->{$n}->{$_} ] foreach keys %{ $out->{$n} };
903 } elsif ( $ref eq 'ARRAY' ) {
904 $i->{$_} = $out->{$n};
906 $i->{$value} = [ $out->{$n} ];
908 $i->{_error} = [ dump($out->{$n}) ];
913 my $stats = __stats( $items );
914 my @columns = grep { ! m/^\Q$key\E$/ } sort keys %$stats;
915 unshift @columns, $key;
917 $loaded->{$commit_dataset} = {
918 header => [ @columns ],
919 columns => [ @columns ],
921 data => { items => $items },
925 warn "# loaded out ", dump( $loaded->{$commit_dataset} );
926 $self->session('path', $commit_dataset);
927 $self->session('columns', [ @columns ]);
928 $self->session('order', $key);
929 return $self->redirect_to('/data/items');
932 $self->session('columns', [ @columns ]);
936 my $from_end = $sort eq 'd' ? $#$filtered : 0;
938 my ( $key, $value, $out ) = ( 'key', 'value' ); # XXX make local
939 foreach ( 0 .. $limit ) {
940 my $i = $_ + $offset;
941 last unless defined $filtered->[$i];
942 $i = $from_end - $i if $from_end;
943 my $id = $filtered->[$i];
944 my $row = Storable::dclone $data->{items}->[ $id ];
945 if ( $code && $test ) {
949 warn "ERROR evaling $@", dump($code);
950 $self->stash('eval_error', $@) if $@;
952 warn "EVAL ",dump($update);
953 foreach ( keys %$update ) {
954 $test_changed->{$_}++;
955 $row->{$_} = $update->{$_};
959 $row->{_row_id} ||= $id;
960 push @$sorted_items, $row;
963 if ( $self->param('export') ) {
964 my $export_path = $self->_export_path( 'items', @columns);
965 open(my $fh, '>', $export_path) || die "ERROR: can't open $export_path: $!";
966 print $fh "#",join("\t",@columns),"\n";
967 foreach my $f ( 0 .. $#$filtered ) {
968 print $fh join("\t", map {
969 my $i = $data->{items}->[ $filtered->[$f] ];
971 if ( ! defined $i->{$_} ) {
973 } elsif ( ref $i->{$_} eq 'ARRAY' ) {
974 $v =join(',', @{ $i->{$_} });
975 $v = '\N' if length($v) == 0;
976 } elsif ( ! ref $i->{$_} ) {
985 warn "export $export_path ", -s $export_path, " bytes\n";
988 my ( $code_depends, $code_description );
992 warn "# test_changed ",dump( $test_changed );
993 my $c = { map { $_ => 1 } @columns };
994 my @added_columns = sort grep { ! $c->{$_} } keys %$test_changed;
995 unshift @columns, @added_columns;
997 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
1000 my $tmp = $code; $tmp =~ s/\$row->{(['"]?)([\w\s]+)\1/$depends_on->{$2}++/gse;
1001 warn "# depends_on ",dump $depends_on;
1003 my $test_added = Storable::dclone $test_changed;
1004 delete $test_added->{$_} foreach keys %$depends_on;
1006 $code_depends = $self->param('code_depends')
1007 || join(',', keys %$depends_on);
1009 $code_description = $self->param('code_description') ||
1010 join(',', keys %$test_added);
1012 $code_depends ||= $code_description; # self-modifing
1013 if ( ! $code_depends && $out ) {
1014 $code_depends = $key;
1015 $code_description = $value;
1018 warn "# test_changed ",dump( $test_changed, $code_depends, $code_description );
1028 sorted => $sorted_items,
1029 columns => [ @columns ],
1030 rows => $#$filtered + 1,
1031 numeric => { map { $_, $self->_is_numeric($_) } @columns },
1032 unique => { map { $_, $self->_is_unique( $_) } @columns },
1033 filters => $self->_current_filters,
1035 cols_changed => $commit ? $commit_changed : $test_changed,
1036 code_depends => $code_depends,
1037 code_description => $code_description,
1038 code_path => $code_path,
1047 $self->session('order', $self->param('order'));
1048 $self->session('sort', $self->param('sort'));
1049 return $self->redirect_to('/data/items');
1053 my ( $self, $name ) = @_;
1055 my $stats = $self->_loaded( 'stats' );
1057 # sort facet numerically if more >50% elements are numeric
1058 my $count = $stats->{$name}->{count};
1059 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
1060 defined $stats->{$name}->{numeric} &&
1061 $stats->{$name}->{numeric} > $count / 2;
1065 my ( $self, $name ) = @_;
1066 my $stats = $self->_loaded( 'stats' );
1067 defined $stats->{$name}->{unique};
1070 sub _remove_filter {
1071 my ($self,$name) = @_;
1072 warn "_remove_filter $name\n";
1074 my $path = $self->session('path');
1076 delete $filters->{$name};
1077 delete $loaded->{$path}->{filters}->{$name};
1078 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
1081 grep { /\Q$name\E/ }
1082 keys %{ $loaded->{$path}->{filtered} }
1084 delete $loaded->{$path}->{filtered}->{$_};
1085 warn "remove filtered cache $_";
1092 my $path = $self->session('path') || return $self->redirect_to( '/data/index' );
1094 if ( my $name = $self->param('remove') ) {
1095 $self->_remove_filter( $name );
1096 return $self->redirect_to( '/data/items' );
1100 my $name = $self->param('name') || die "no name";
1102 my $all = $self->_param_scalar('all', 1);
1103 my $data = $self->_loaded('data');
1105 my $filters = $self->_current_filters;
1106 my $all_filters = __all_filters( keys %$filters,$self->session('order') );
1107 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
1108 if defined $loaded->{$path}->{filtered}->{$all_filters};
1110 if ( ! $filtered || $all ) {
1111 $filtered = [ 0 .. $#{ $data->{items} } ];
1112 warn "filter all values\n";
1114 warn "filter using $all_filters\n";
1117 foreach my $i ( @$filtered ) {
1118 my $item = $data->{items}->[$i];
1119 if ( ! exists $item->{$name} || ! defined $item->{$name} ) {
1120 $facet->{ _missing }++;
1121 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
1122 $facet->{$_}++ foreach @{ $item->{$name} };
1124 $facet->{ $item->{$name} }++;
1128 my $checked_values = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
1130 if ( my $code = $self->param('code') ) {
1132 foreach my $value ( keys %$facet ) {
1133 my $count = $facet->{$value};
1134 my $checked = $checked_values->{$value};
1138 warn "ERROR: $@\n$code\n";
1140 } elsif ( $checked != $checked_values->{$value} ) {
1141 warn "checked $value $count -> $checked\n";
1142 $checked_values->{$value} = $checked;
1145 warn "out ",dump( $out );
1146 $self->stash( out => $out );
1149 # warn "# facet $name ",dump $facet;
1153 : defined $filters->{$name} ? @{ $filters->{$name} }
1156 my $numeric = $self->_is_numeric($name);
1158 my $sort = $self->param('sort');
1159 # sort numeric facets with more than 5 values ascending
1160 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
1162 @facet_names = sort {
1164 if ( $sort eq 'a' ) {
1165 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
1166 } elsif ( $sort eq 'd' ) {
1167 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
1168 } elsif ( $sort eq 'c' ) {
1169 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
1171 warn "unknown sort: $sort";
1172 $result = $a cmp $b;
1174 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
1178 $self->render( name => $name, facet => $facet, checked => $checked_values,
1179 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
1184 sub __invalidate_path_column {
1185 my ( $path, $name ) = @_;
1187 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
1188 delete $loaded->{$path}->{sorted}->{$name};
1189 warn "# invalidate $path sorted $name\n";
1192 foreach ( grep { m/\Q$name\E/ } keys %{ $loaded->{$path}->{filtered} } ) {
1193 delete $loaded->{$path}->{filtered}->{$_};
1194 warn "# invalidate $path filtered $_\n";
1198 sub __path_modified {
1199 my ( $path, $value ) = @_;
1200 $value = 1 unless defined $value;
1202 $loaded->{$path}->{modified} = $value;
1204 warn "# __path_modified $path $value\n";
1207 sub __path_rebuild_stats { $loaded->{ $_[0] }->{rebuild_stats} = 1 };
1210 my ($self,$change) = @_;
1212 my $change_path = $self->_permanent_path( 'changes' );
1213 mkdir $change_path unless -d $change_path;
1214 $change_path .= '/' . $change->{time};
1215 store $change, $change_path;
1216 utime $change->{time}, $change->{time}, $change_path;
1217 warn "_save_change $change_path ", dump($change);
1222 my $new_content = $self->param('new_content');
1223 $new_content ||= $self->param('content'); # backward compatibility with old actions
1225 my $i = $self->param('_row_id');
1226 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
1227 my $path = $self->param('path') || die "no path";
1228 my $name = $self->param('name') || die "no name";
1229 my $status = 200; # 200 = OK, 201 = Created
1231 my $data = $self->_loaded('data');
1233 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
1234 $new_content =~ s/^\s+//s;
1235 $new_content =~ s/\s+$//s;
1237 if ( $new_content =~ /\xB6/ ) { # para
1238 $v = [ split(/\s*\xB6\s*/, $new_content) ];
1240 $v = [ $new_content ];
1243 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
1246 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
1248 $self->_save_change({
1252 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
1254 time => $self->param('time') || time(),
1255 user => $self->param('user') || $ENV{'LOGNAME'},
1257 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
1258 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
1259 keys %{ $loaded->{$path}->{stats} }
1263 warn "# change $path $i $old -> $new\n";
1264 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
1266 __invalidate_path_column( $path, $name );
1268 $status = 201; # created
1269 __path_rebuild_stats( $path );
1271 $new_content = join("\xB6",@$v);
1274 warn "# unchanged $path $i $old\n";
1278 $new_content = "$path $i $name doesn't exist\n";
1282 warn "# edit $status ", dump $new_content;
1286 new_content => scalar $new_content,
1293 my $path = $self->_param_or_session('path');
1294 my $dump_path = $self->_save( $path );
1295 __path_modified( $path, 0 );
1297 return $self->redirect_to( '/data/items' );
1303 my $dir = $self->app->home->rel_dir('public');
1305 if ( my $import = $self->param('import') ) {
1307 if ( $import =~ m{/filter\.(.+?)\..+} ) {
1308 my $name = $self->_column_from_unac( $1 );
1310 my @vals = map { chomp; $_ }
1311 read_file "$dir/export/$import", binmode => ':utf8';
1313 $self->_remove_filter( $name );
1314 $self->_filter_on_data( $name, @vals );
1315 $self->session( 'offset' => 0 );
1316 return $self->redirect_to('/data/items');
1318 warn "UNKNOWN IMPORT $import";
1322 if ( my $remove = $self->param('remove') ) {
1323 my $path = "$dir/export/$remove";
1324 unlink $path if -e $path;
1326 unlink $path if -e $path;
1329 my $path = $self->_export_path || return $self->redirect_to('/data/index');
1331 my @files = grep { ! /\.png$/ } glob "$path/*";
1332 my $mtime = { map { $_ => (stat($_))[9] } @files };
1333 @files = sort { $mtime->{$b} <=> $mtime->{$a} } @files;
1334 $self->render( export => [ @files ] );
1337 sub __loaded_paths {
1339 grep { defined $loaded->{$_}->{data} }
1345 $self->stash( reload => 1 );
1347 # $self->_load_path( $self->param('path') );
1348 $self->redirect_to('/data/load?path=' . $self->param('path') );
1353 my $path = $self->param('path');
1354 if ( $path =~ m{^/tmp/mojo_facets\.} ) {
1356 warn "# unlink $path";
1357 if ( my $name = $self->param('name') ) {
1358 delete $loaded->{$name};
1359 warn "# remove $name from memory";
1362 warn "WARNING: $path unlink ignored";
1364 return if $self->stash('reload');
1365 return $self->redirect_to( '/data/load' );