1 package MojoFacets::Data;
6 use base 'Mojolicious::Controller';
8 use Data::Dump qw(dump);
14 use Time::HiRes qw(time);
15 use File::Path qw(mkpath);
16 use Text::Unaccent::PurePerl;
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;
32 my $data_dir = $self->app->home->rel_dir('data');
33 die "no data dir $data_dir" unless -d $data_dir;
38 my $file = $File::Find::name;
39 if ( -f $file && $file =~ m/\.(js(on)?|txt)$/ ) {
40 $file =~ s/$data_dir\/*//;
42 } elsif ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
43 push @{ $changes->{$1} }, $2
44 } elsif ( -d $file && $file =~ m/\.html$/ ) {
45 $file =~ s/$data_dir\/*//;
47 } elsif ( -f $file && $file =~ m/\.(csv|storable|couchdb|sql)$/i ) {
48 $file =~ s/$data_dir\/*//;
51 #warn "IGNORE: $file\n";
55 no warnings qw(uninitialized); # mtime
56 @files = sort { $loaded->{$b}->{mtime} <=> $loaded->{$a}->{mtime} || lc $a cmp lc $b } @files,
57 grep { defined $loaded->{$_}->{generated} } keys %$loaded;
60 $size->{$_} = -s "$data_dir/$_" foreach @files;
67 dump_path => { map { $_ => $self->_dump_path($_) } @files },
73 my ( $self, $name ) = @_;
74 my $dir = $self->app->home->rel_dir('data');
77 return '/tmp/mojo_facets.' . $name . '.storable';
81 my ( $self, $path ) = @_;
83 my $dump_path = $self->_dump_path( $path );
84 my $first_load = ! -e $dump_path;
85 warn "save loaded to $dump_path";
86 my $info = $loaded->{$path};
87 store $info, $dump_path;
90 my $mtime = $loaded->{$path}->{mtime};
91 utime $mtime, $mtime, $dump_path;
92 warn "sync time to $path at $mtime\n";
95 warn $dump_path, ' ', -s $dump_path, " bytes\n";
104 my $nr_items = $#{ $_[0] } + 1;
106 warn "__stats $nr_items\n";
108 # foreach my $e ( @{ $_[0] } ) {
109 foreach my $i ( 0 .. $#{$_[0]} ) {
110 print STDERR " $i" if $i % 5000;
112 foreach my $n ( keys %$e ) {
113 $stats->{$n}->{count}++;
115 if ( ref $e->{$n} eq 'ARRAY' ) {
116 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
122 foreach my $x ( @v ) {
123 if ( ! defined $x ) { # FIXME really null
124 $stats->{$n}->{empty}++;
128 $stats->{$n}->{numeric}++
129 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
130 $stats->{$n}->{empty}++
131 if length $x == 0; # faster than $x =~ m/^\s*$/;
137 foreach my $n ( keys %$stats ) {
138 my $s = $stats->{$n};
139 next unless defined $s->{array};
140 if ( $s->{array} == $s->{count} ) {
142 if ( $s->{count} == $nr_items ) {
143 warn "check $n for uniqeness\n";
145 foreach my $e ( @{ $_[0] } ) {
146 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
152 $stats->{$n}->{unique} = 1;
153 #warn "# $n unique ",dump( $unique );
159 warn "# __stats ",dump($stats);
164 sub _param_or_session {
165 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
170 my $path = $self->_param_or_session('path');
171 warn "stats $path\n";
172 delete $loaded->{$path}->{stats};
173 return $self->redirect_to( '/data/columns' );
178 my ( $self, $path ) = @_;
180 return if defined $loaded->{$path}->{generated};
182 my $full_path = $self->app->home->rel_file( 'data/' . $path );
183 return $self->redirect_to('/data/index') unless -r $full_path;
185 my $dump_path = $self->_dump_path( $path );
187 if ( defined $loaded->{$path}->{data} ) {
188 my $mtime = (stat($full_path))[9];
189 return if $loaded->{$path}->{mtime} == $mtime;
190 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
191 } elsif ( -e $dump_path ) {
192 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
193 my $info = retrieve $dump_path;
194 $loaded->{ $path } = $info;
199 if ( -f $full_path ) {
200 if ( $full_path =~ m/.storable$/ ) { # check storable first to catch files copied from /tmp/
201 $data->{generated}++;
202 warn "open $full_path ", -s $full_path, " bytes";
203 open(my $pipe, "<", $full_path) || die $!;
204 while ( my $o = eval { Storable::fd_retrieve $pipe } ) {
205 if ( exists $o->{item} ) {
206 # stream of storable objects
207 push @{ $data->{items} }, $o->{item};
208 } elsif ( exists $o->{data}->{items} ) {
209 # /tmp/mojofacets.*.storable
210 $data->{items} = $o->{data}->{items};
211 $data->{header} = $o->{header};
212 delete $data->{generated};
214 warn "SKIP ",dump($o);
218 warn "loaded ", $#{ $data->{items} } + 1, " items from $full_path\n";
219 } elsif ( $full_path =~ m/.csv/i ) {
220 $data = MojoFacets::Import::CSV->new( full_path => $full_path )->data;
221 } elsif ( $full_path =~ m/.sql/i ) {
222 $data = MojoFacets::Import::SQL->new( full_path => $full_path )->data;
223 } elsif ( $full_path =~ m/.couchdb/i ) {
224 $data = MojoFacets::Import::CouchDB->new( full_path => $full_path )->data;
226 $data = MojoFacets::Import::File->new( full_path => $full_path, path => $path )->data;
228 } elsif ( -d $full_path && $full_path =~ m/.html/ ) {
229 $data = MojoFacets::Import::HTMLTable->new( dir => $full_path )->data;
231 die "can't load $full_path";
236 if ( defined $data->{header} ) {
237 if ( ref $data->{header} eq 'ARRAY' ) {
238 @header = @{ $data->{header} };
240 warn "header not array ", dump( $data->{header} );
244 my $stats = __stats( $data->{items} );
247 sort { $stats->{$b}->{count} <=> $stats->{$a}->{count} }
248 grep { defined $stats->{$_}->{count} } keys %$stats
252 header => [ @header ],
254 full_path => $full_path,
255 size => -s $full_path,
256 mtime => (stat($full_path))[9],
258 defined $data->{generated} ? ( generated => 1 ) : (),
261 $loaded->{ $path } = $info;
262 $self->_save( $path ) unless $info->{generated};
270 my $path = $self->param('path') || return $self->redirect_to( '/data/index' );
272 my @paths = $self->param('paths');
273 warn "# paths ", dump @paths;
275 foreach my $p ( keys %$loaded ) {
276 next if grep { /^\Q$p\E$/ } @paths;
277 warn "remove $p from memory\n";
278 delete $loaded->{$p};
281 $self->_load_path( $_ ) foreach @paths;
283 warn "# path $path\n";
284 $self->_load_path( $path );
286 $self->session( 'path' => $path );
288 my $redirect_to = '/data/items';
290 $self->session( 'header' => $loaded->{$path}->{header} );
291 if ( ! defined $loaded->{$path}->{columns} ) {
292 my $columns_path = $self->_permanent_path( 'columns' );
293 if ( -e $columns_path ) {
294 my @columns = map { s/[\r\n]+$//; $_ } read_file $columns_path, binmode => ':utf8';
295 $loaded->{$path}->{columns} = [ @columns ];
296 warn "# columns_path $columns_path ",dump(@columns);
298 $loaded->{$path}->{columns} = $loaded->{$path}->{header}
301 $redirect_to = '/data/columns';
303 $self->session( 'columns' => $loaded->{$path}->{columns} );
304 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
305 return $self->redirect_to( $redirect_to );
310 my ( $self, $name ) = @_;
311 my $path = $self->session('path') || $self->param('path');
312 return $self->redirect_to('/data/index') unless $path;
314 if ( defined $loaded->{$path}->{modified} && $loaded->{$path}->{modified} > 1 ) {
315 my $caller = (caller(1))[3];
316 if ( $caller =~ m/::edit/ ) {
317 warn "rebuild stats for $path ignored caller $caller\n";
319 warn "rebuild stats for $path FORCED by modified caller $caller\n";
320 # $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
321 $loaded->{$path}->{rebuild_stats} = 1;
322 $loaded->{$path}->{modified} = 1;
326 if ( defined $loaded->{$path}->{rebuild_stats} ) {
327 warn "rebuild_stats $path";
328 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
329 delete $loaded->{$path}->{rebuild_stats};
332 if ( ! defined $loaded->{$path}->{$name} ) {
333 warn "$path $name isn't loaded\n";
334 $self->_load_path( $path );
335 if ( ! defined $loaded->{$path}->{stats} ) {
336 warn "rebuild stats for $path\n";
337 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
339 if ( ! defined $loaded->{$path}->{$name} ) {
340 warn "MISSING $name for $path\n";
341 return $self->redirect_to('/data/index')
345 $self->session( 'modified' => $loaded->{$path}->{modified} );
347 return $loaded->{$path}->{$name};
354 $checked->{$_}++ foreach @_;
355 # warn "# _checked ",dump($checked);
359 sub _permanent_path {
361 my $path = $self->_param_or_session('path');
362 $self->app->home->rel_dir('data') . '/' . join('.', $path, @_);
367 $n = unac_string($n);
372 sub _column_from_unac {
373 my ($self,$name) = @_;
375 my $stats = $self->_loaded('stats');
376 my $cols_norm = { map { __unac( $_ ) => $_ } keys %$stats };
378 $cols_norm->{$name} || die "can't find column $name in ", dump($cols_norm);
383 my $path = $self->_param_or_session('path');
385 warn "no path in param or session";
388 my $dir = $self->app->home->rel_dir('public') . "/export/$path";
389 mkpath $dir unless -e $dir;
390 my $name = __export_path_name( $path, @_ );
391 my $full = $dir . '/' . $name;
392 $full =~ s/\/+$// if -d $full; # strip trailing slash for dirs
396 sub __export_path_name {
400 my $name = join('.', map { __unac($_) } @_ );
401 if ( length($name) > $max_length ) {
402 $name = substr($name,0,$max_length) . Digest::MD5::md5_hex substr($name,$max_length);
410 my $view_path = $self->_permanent_path( 'views' );
412 if ( $self->param('columns') ) {
413 my @columns = $self->_param_array('columns');
414 write_file( $self->_permanent_path( 'columns' ), { binmode => ':utf8' }, map { "$_\n" } @columns );
415 if ( my $view = $self->param('view') ) {
416 mkdir $view_path unless -e $view_path;
417 write_file( "$view_path/$view", { binmode => ':utf8' }, map { "$_\n" } @columns );
418 warn "view $view_path/$view ", -s "$view_path/$view", " bytes\n";
421 return $self->redirect_to('/data/items');
423 } elsif ( ! $self->session('header') ) {
424 return $self->redirect_to('/data/load');
427 if ( my $id = $self->param('id') ) {
428 my $view_full = "$view_path/$id";
429 if ( -f $view_full ) {
430 my @columns = map { chomp; $_ } read_file $view_full, binmode => ':utf8';
431 warn "view $view_full loaded ", dump @columns;
432 $self->session( 'columns' => [ @columns ] );
433 return $self->redirect_to('/data/items');
437 my $stats = $self->_loaded( 'stats' );
440 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
442 foreach my $c ( @{ $self->session( 'header' ) } ) {
443 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
447 if ( -d $view_path ) {
448 @views = map { s{^\Q$view_path\E/*}{}; $_ } glob "$view_path/*";
449 warn "# views ",dump @views;
453 message => 'Select columns to display',
455 columns => \@columns,
456 checked => $self->_checked( $self->_param_array('columns') ),
462 my ($self,$name) = @_;
464 my @array = $self->param($name);
465 my $path = $self->session('path');
468 $self->session($name => [ @array ]);
469 $loaded->{$path}->{$name} = [ @array ];
470 } elsif ( my $session = $self->session($name) ) {
471 if ( ref $session eq 'ARRAY' ) {
474 die "$name not array ",dump($session);
477 #warn "# $name ",dump @array;
482 my ($self,$name,$default) = @_;
484 my $scalar = $self->param($name);
486 if ( defined $scalar ) {
487 $self->session($name => $scalar);
489 $scalar = $self->session($name);
492 if ( ! defined $scalar ) {
494 die "no default for $name" unless defined $scalar;
495 $self->session($name => $scalar);
498 warn "# _param_scalar $name ",dump $scalar;
505 my $name = $self->param('filter_name') || die "name?";
506 my @vals = $self->param('filter_vals');
508 $self->_remove_filter( $name );
510 $self->_filter_on_data( $name, @vals );
511 if ( my $permanent = $self->param('_permanent') ) {
512 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
513 write_file $permanent_path, { binmode => ':utf8' }, map { "$_\n" } @vals;
514 warn "permanent filter $permanent_path ", -s $permanent_path;
518 $self->session( 'offset' => 0 );
520 return $self->redirect_to('/data/items');
523 sub _filter_on_data {
524 my ( $self, $name, @vals ) = @_;
526 my $path = $self->session('path');
528 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
529 warn "filter $name not found in data set";
533 $filters->{$name} = [ @vals ];
534 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
537 $filter_hash->{$_}++ foreach @vals;
539 #warn "# filter_hash ",dump( $filter_hash );
541 my $items = $self->_loaded('data')->{items};
543 my $include_missing = defined $filter_hash->{_missing};
546 foreach my $i ( 0 .. $#$items ) {
548 if ( defined $items->[$i]->{$name} ) {
549 my $row = $items->[$i]->{$name};
550 $row = [ $row ] unless ref $row eq 'ARRAY'; # FIXME probably wrong place
551 foreach my $v ( @$row ) {
552 if ( defined $filter_hash->{ $v } ) {
553 $filtered_items->{$i}++;
556 } elsif ( $include_missing ) {
557 $filtered_items->{$i}++;
561 #warn "# filter $name ",dump($filtered_items);
563 # invalidate filters on other datasets
564 foreach my $dataset ( grep { exists $loaded->{$_}->{filters}->{$name} } keys %$loaded ) {
565 delete $loaded->{$dataset}->{filters}->{$name};
566 delete $loaded->{$dataset}->{filtered};
569 $loaded->{$path}->{filters}->{$name} = $filtered_items;
570 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
574 sub _current_filters {
577 my $stats = $self->_loaded('stats');
579 $current_filters->{ $_ } = $filters->{ $_ }
580 foreach ( grep { defined $filters->{ $_ } } keys %$stats )
582 warn "# _current_filters ",dump( keys %$current_filters );
583 return $current_filters;
586 sub _data_sorted_by {
587 my ( $self, $order ) = @_;
589 my $path = $self->session('path');
591 warn "_data_sorted_by $order from $path";
593 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
594 return $loaded->{$path}->{sorted}->{$order};
597 my $data = $self->_loaded( 'data' );
598 my $numeric = $self->_is_numeric($order);
599 my $missing = $numeric ? 0 : '';
600 no warnings qw(numeric);
612 if ( ! exists $_->{$order} ) {
614 } elsif ( ref $_->{$order} eq 'ARRAY' ) {
615 $v = join('', @{$_->{$order}});
620 } @{ $data->{items} }
623 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
624 #warn "# sorted ",dump( @sorted );
626 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
632 join(',', sort(@_), 'order', $order);
635 our ($out, $key,$value);
637 our $lookup_path_col;
641 warn "__commit_begin";
642 $lookup_path_col = undef;
648 $lookup_path_col = undef; # cleanup memory
653 warn "# lookup ",dump @_;
654 my ( $vals, $on_path, $on_col, $code, $stat_code ) = @_;
655 die "code is not sub{ ... } but ", dump $code unless ref $code eq 'CODE';
657 if ( ! exists $loaded->{$on_path} ) {
658 my @possible_paths = grep { /\Q$on_path\E/ } keys %$loaded;
659 die "more than one dataset available for '$on_path' ",dump @possible_paths if $#possible_paths > 0;
660 $on_path = shift @possible_paths;
661 warn "## fuzzy selected path $on_path";
664 my $items = $loaded->{$on_path}->{data}->{items} || die "no items for $on_path";
666 if ( ! exists $lookup_path_col->{$on_path}->{$on_col} ) {
667 warn "create lookup_path_col $on_path $on_col";
668 foreach my $i ( 0 .. $#$items ) {
669 my $item = $items->[$i];
670 if ( exists $item->{$on_col} ) {
671 if ( ref $item->{$on_col} eq 'ARRAY' ) {
672 foreach my $v ( @{ $item->{$on_col} } ) {
673 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
675 } elsif ( ! ref $item->{$on_col} ) { # scalar
676 my $v = $item->{$on_col};
677 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
679 die "unknown type of ",dump $item->{$on_col};
683 warn "XXX ",dump $lookup_path_col->{$on_path}->{$on_col} if $ENV{DEBUG};
687 $stat = Statistics::Descriptive::Full->new() if $stat_code;
689 foreach my $v ( ref $vals eq 'ARRAY' ? @$vals : ( $vals ) ) {
690 foreach my $i ( @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} } ) {
692 warn "XXX lookup code $v $i ",dump $on if $ENV{DEBUG};
695 $stat_code->( $stat ) if $stat_code;
699 sub __commit_path_code {
700 my ( $path, $i, $code, $commit_changed ) = @_;
702 my $items = $loaded->{$path}->{data}->{items} || die "no items for $path";
703 my $row = $items->[$i];
706 foreach ( keys %$update ) {
707 $$commit_changed->{$_}++;
708 $loaded->{$path}->{data}->{items}->[$i]->{$_} = $update->{$_};
710 #warn "__commit_path_code $path $i ",dump( $update );
713 # uses templates/admin.html.ep
714 sub _switch_dataset {
719 foreach my $path ( keys %$loaded ) {
720 next unless exists $loaded->{$path}->{data};
721 push @$datasets, $path;
724 warn "# datasets ",dump($datasets);
726 $self->stash( 'datasets' => $datasets );
732 $self->_switch_dataset;
734 if ( my $show = $self->param('id') ) {
735 $self->param('show', $show);
739 my $path = $self->_param_scalar('path');
741 my @columns = $self->_param_array('columns');
742 return $self->redirect_to('/data/columns') unless @columns;
743 my $order = $self->_param_scalar('order', $columns[0]);
744 my $sort = $self->_param_scalar('sort', 'a');
745 my $offset = $self->_param_scalar('offset', 0);
746 my $limit = $self->_param_scalar('limit', 20);
747 $self->_param_scalar('show', 'table');
749 # fix offset when changing limit
750 $offset = int( $offset / $limit ) * $limit;
752 if ( ! grep { /^\Q$order\E$/ } @columns ) {
753 $order = $columns[0];
754 $self->session( order => $order );
756 my $sorted = $self->_data_sorted_by( $order );
760 foreach my $name ( keys %$filters ) {
761 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
762 warn "skip filter $name not found in $path\n";
765 push @filter_names, $name;
767 warn "filter_names ",dump( @filter_names );
768 foreach my $name ( @filter_names ) {
769 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
770 $self->_filter_on_data( $name, @{ $filters->{$name} } );
774 my $all_filters = __all_filters( @filter_names,$order );
776 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
778 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
780 my $path_filters = $loaded->{$path}->{filters};
782 warn "create combined filter for $all_filters from ", $#$sorted + 1, " items\n";
785 foreach my $i ( 0 .. $#$sorted ) {
786 my $pos = $sorted->[$i];
788 if ( $#filter_names == -1 ) {
789 push @filtered, $pos;
794 foreach ( @filter_names ) {
795 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
799 push @filtered, $pos;
802 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
805 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
806 if defined $loaded->{$path}->{filtered}->{$all_filters};
808 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
810 my $data = $self->_loaded('data');
812 my $code = $self->_param_scalar('code','');
814 $code =~ s{\n+$}{\n}s;
816 # XXX convert @row->{foo} into @{$row->{foo}}
817 $code =~ s|\@(row->{[^}]+})|\@{\$$1}|gs;
819 my $commit = $self->param('commit');
820 my $test = $self->param('test');
825 if ( $code && ( $test || $commit ) ) {
826 # XXX find columns used in code snippet and show them to user
828 foreach my $column ( $code =~ m/\$row->{([^}]+)}/g ) {
829 if ( $column =~ s/^(['"])// ) {
832 next if $column =~ m/\$/; # hide columns with vars in them
833 $commit_changed->{$column} = 0;
837 my $code_path = $self->app->home->rel_dir('public') . "/code";
840 warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
841 ( $key, $value, $out ) = ( 'key', 'value' );
842 foreach ( 0 .. $#$filtered ) {
843 my $i = $filtered->[$_];
844 __commit_path_code( $path, $i, $code, \$commit_changed );
847 # this might move before $out to recalculate stats on source dataset?
848 __path_rebuild_stats( $path );
849 my $c = { map { $_ => 1 } @columns };
850 my @added_columns = sort grep { ! $c->{$_} } keys %$commit_changed;
851 warn "# added_columns ",dump( @added_columns );
852 unshift @columns, @added_columns;
854 $loaded->{$path}->{columns} = [ @columns ];
855 warn "# new columns ",dump( @columns );
857 __invalidate_path_column( $path, $_ ) foreach keys %$commit_changed;
859 $self->_save_change({
861 time => $self->param('time') || time(),
862 user => $self->param('user') || $ENV{'LOGNAME'},
864 commit_changed => $commit_changed,
867 if ( my $description = $self->param('code_description') ) {
868 my $depends = $self->param('code_depends') || die "no code_depends?";
869 my $path = "$code_path/$depends.$description.pl";
870 if ( -e $path && ! $self->param('overwrite') ) {
871 warn "# code $path not saved\n";
873 write_file( $path, { binmode => ':utf8' }, "$code\n" );
874 warn "code $path ", -s $path, " bytes saved\n";
881 my $commit_dataset = join('.'
882 , $self->param('code_depends')
883 , $self->param('code_description')
888 warn "key $key value $value";
890 foreach my $n ( keys %$out ) {
891 my $i = { $key => [ $n ] };
892 my $ref = ref $out->{$n};
893 if ( $ref eq 'HASH' ) {
894 $i->{$_} = [ $out->{$n}->{$_} ] foreach keys %{ $out->{$n} };
895 } elsif ( $ref eq 'ARRAY' ) {
896 $i->{$_} = $out->{$n};
898 $i->{$value} = [ $out->{$n} ];
900 $i->{_error} = [ dump($out->{$n}) ];
905 my $stats = __stats( $items );
906 my @columns = grep { ! m/^\Q$key\E$/ } sort keys %$stats;
907 unshift @columns, $key;
909 $loaded->{$commit_dataset} = {
910 header => [ @columns ],
911 columns => [ @columns ],
913 data => { items => $items },
917 warn "# loaded out ", dump( $loaded->{$commit_dataset} );
918 $self->session('path', $commit_dataset);
919 $self->session('columns', [ @columns ]);
920 $self->session('order', $key);
921 return $self->redirect_to('/data/items');
924 $self->session('columns', [ @columns ]);
928 my $from_end = $sort eq 'd' ? $#$filtered : 0;
930 my ( $key, $value, $out ) = ( 'key', 'value' ); # XXX make local
931 foreach ( 0 .. $limit ) {
932 my $i = $_ + $offset;
933 last unless defined $filtered->[$i];
934 $i = $from_end - $i if $from_end;
935 my $id = $filtered->[$i];
936 my $row = Storable::dclone $data->{items}->[ $id ];
937 if ( $code && $test ) {
941 warn "ERROR evaling $@", dump($code);
942 $self->stash('eval_error', $@) if $@;
944 warn "EVAL ",dump($update);
945 foreach ( keys %$update ) {
946 $test_changed->{$_}++;
947 $row->{$_} = $update->{$_};
951 $row->{_row_id} ||= $id;
952 push @$sorted_items, $row;
955 if ( $self->param('export') ) {
956 my $export_path = $self->_export_path( 'items', @columns);
957 open(my $fh, '>', $export_path) || die "ERROR: can't open $export_path: $!";
958 print $fh "#",join("\t",@columns),"\n";
959 foreach my $f ( 0 .. $#$filtered ) {
960 print $fh join("\t", map {
961 my $i = $data->{items}->[ $filtered->[$f] ];
963 if ( ! defined $i->{$_} ) {
965 } elsif ( ref $i->{$_} eq 'ARRAY' ) {
966 $v =join(',', @{ $i->{$_} });
967 $v = '\N' if length($v) == 0;
968 } elsif ( ! ref $i->{$_} ) {
977 warn "export $export_path ", -s $export_path, " bytes\n";
980 my ( $code_depends, $code_description );
984 warn "# test_changed ",dump( $test_changed );
985 my $c = { map { $_ => 1 } @columns };
986 my @added_columns = sort grep { ! $c->{$_} } keys %$test_changed;
987 unshift @columns, @added_columns;
989 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
992 my $tmp = $code; $tmp =~ s/\$row->{(['"]?)([\w\s]+)\1/$depends_on->{$2}++/gse;
993 warn "# depends_on ",dump $depends_on;
995 my $test_added = Storable::dclone $test_changed;
996 delete $test_added->{$_} foreach keys %$depends_on;
998 $code_depends = $self->param('code_depends')
999 || join(',', keys %$depends_on);
1001 $code_description = $self->param('code_description') ||
1002 join(',', keys %$test_added);
1004 $code_depends ||= $code_description; # self-modifing
1005 if ( ! $code_depends && $out ) {
1006 $code_depends = $key;
1007 $code_description = $value;
1010 warn "# test_changed ",dump( $test_changed, $code_depends, $code_description );
1020 sorted => $sorted_items,
1021 columns => [ @columns ],
1022 rows => $#$filtered + 1,
1023 numeric => { map { $_, $self->_is_numeric($_) } @columns },
1024 unique => { map { $_, $self->_is_unique( $_) } @columns },
1025 filters => $self->_current_filters,
1027 cols_changed => $commit ? $commit_changed : $test_changed,
1028 code_depends => $code_depends,
1029 code_description => $code_description,
1030 code_path => $code_path,
1039 $self->session('order', $self->param('order'));
1040 $self->session('sort', $self->param('sort'));
1041 return $self->redirect_to('/data/items');
1045 my ( $self, $name ) = @_;
1047 my $stats = $self->_loaded( 'stats' );
1049 # sort facet numerically if more >50% elements are numeric
1050 my $count = $stats->{$name}->{count};
1051 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
1052 defined $stats->{$name}->{numeric} &&
1053 $stats->{$name}->{numeric} > $count / 2;
1057 my ( $self, $name ) = @_;
1058 my $stats = $self->_loaded( 'stats' );
1059 defined $stats->{$name}->{unique};
1062 sub _remove_filter {
1063 my ($self,$name) = @_;
1064 warn "_remove_filter $name\n";
1066 my $path = $self->session('path');
1068 delete $filters->{$name};
1069 delete $loaded->{$path}->{filters}->{$name};
1070 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
1073 grep { /\Q$name\E/ }
1074 keys %{ $loaded->{$path}->{filtered} }
1076 delete $loaded->{$path}->{filtered}->{$_};
1077 warn "remove filtered cache $_";
1084 my $path = $self->session('path') || return $self->redirect_to( '/data/index' );
1086 if ( my $name = $self->param('remove') ) {
1087 $self->_remove_filter( $name );
1088 return $self->redirect_to( '/data/items' );
1092 my $name = $self->param('name') || die "no name";
1094 my $all = $self->_param_scalar('all', 1);
1095 my $data = $self->_loaded('data');
1097 my $filters = $self->_current_filters;
1098 my $all_filters = __all_filters( keys %$filters,$self->session('order') );
1099 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
1100 if defined $loaded->{$path}->{filtered}->{$all_filters};
1102 if ( ! $filtered || $all ) {
1103 $filtered = [ 0 .. $#{ $data->{items} } ];
1104 warn "filter all values\n";
1106 warn "filter using $all_filters\n";
1109 foreach my $i ( @$filtered ) {
1110 my $item = $data->{items}->[$i];
1111 if ( ! exists $item->{$name} || ! defined $item->{$name} ) {
1112 $facet->{ _missing }++;
1113 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
1114 $facet->{$_}++ foreach @{ $item->{$name} };
1116 $facet->{ $item->{$name} }++;
1120 my $checked_values = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
1122 if ( my $code = $self->param('code') ) {
1124 foreach my $value ( keys %$facet ) {
1125 my $count = $facet->{$value};
1126 my $checked = $checked_values->{$value};
1130 warn "ERROR: $@\n$code\n";
1132 } elsif ( $checked != $checked_values->{$value} ) {
1133 warn "checked $value $count -> $checked\n";
1134 $checked_values->{$value} = $checked;
1137 warn "out ",dump( $out );
1138 $self->stash( out => $out );
1141 # warn "# facet $name ",dump $facet;
1145 : defined $filters->{$name} ? @{ $filters->{$name} }
1148 my $numeric = $self->_is_numeric($name);
1150 my $sort = $self->param('sort');
1151 # sort numeric facets with more than 5 values ascending
1152 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
1154 @facet_names = sort {
1156 if ( $sort eq 'a' ) {
1157 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
1158 } elsif ( $sort eq 'd' ) {
1159 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
1160 } elsif ( $sort eq 'c' ) {
1161 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
1163 warn "unknown sort: $sort";
1164 $result = $a cmp $b;
1166 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
1170 $self->render( name => $name, facet => $facet, checked => $checked_values,
1171 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
1176 sub __invalidate_path_column {
1177 my ( $path, $name ) = @_;
1179 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
1180 delete $loaded->{$path}->{sorted}->{$name};
1181 warn "# invalidate $path sorted $name\n";
1184 foreach ( grep { m/\Q$name\E/ } keys %{ $loaded->{$path}->{filtered} } ) {
1185 delete $loaded->{$path}->{filtered}->{$_};
1186 warn "# invalidate $path filtered $_\n";
1190 sub __path_modified {
1191 my ( $path, $value ) = @_;
1192 $value = 1 unless defined $value;
1194 $loaded->{$path}->{modified} = $value;
1196 warn "# __path_modified $path $value\n";
1199 sub __path_rebuild_stats { $loaded->{ $_[0] }->{rebuild_stats} = 1 };
1202 my ($self,$change) = @_;
1204 my $change_path = $self->_permanent_path( 'changes' );
1205 mkdir $change_path unless -d $change_path;
1206 $change_path .= '/' . $change->{time};
1207 store $change, $change_path;
1208 utime $change->{time}, $change->{time}, $change_path;
1209 warn "_save_change $change_path ", dump($change);
1214 my $new_content = $self->param('new_content');
1215 $new_content ||= $self->param('content'); # backward compatibility with old actions
1217 my $i = $self->param('_row_id');
1218 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
1219 my $path = $self->param('path') || die "no path";
1220 my $name = $self->param('name') || die "no name";
1221 my $status = 200; # 200 = OK, 201 = Created
1223 my $data = $self->_loaded('data');
1225 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
1226 $new_content =~ s/^\s+//s;
1227 $new_content =~ s/\s+$//s;
1229 if ( $new_content =~ /\xB6/ ) { # para
1230 $v = [ split(/\s*\xB6\s*/, $new_content) ];
1232 $v = [ $new_content ];
1235 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
1238 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
1240 $self->_save_change({
1244 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
1246 time => $self->param('time') || time(),
1247 user => $self->param('user') || $ENV{'LOGNAME'},
1249 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
1250 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
1251 keys %{ $loaded->{$path}->{stats} }
1255 warn "# change $path $i $old -> $new\n";
1256 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
1258 __invalidate_path_column( $path, $name );
1260 $status = 201; # created
1261 __path_rebuild_stats( $path );
1263 $new_content = join("\xB6",@$v);
1266 warn "# unchanged $path $i $old\n";
1270 $new_content = "$path $i $name doesn't exist\n";
1274 warn "# edit $status ", dump $new_content;
1278 new_content => scalar $new_content,
1285 my $path = $self->_param_or_session('path');
1286 my $dump_path = $self->_save( $path );
1287 __path_modified( $path, 0 );
1289 return $self->redirect_to( '/data/items' );
1295 my $dir = $self->app->home->rel_dir('public');
1297 if ( my $import = $self->param('import') ) {
1299 if ( $import =~ m{/filter\.(.+?)\..+} ) {
1300 my $name = $self->_column_from_unac( $1 );
1302 my @vals = map { chomp; $_ }
1303 read_file "$dir/export/$import", binmode => ':utf8';
1305 $self->_remove_filter( $name );
1306 $self->_filter_on_data( $name, @vals );
1307 $self->session( 'offset' => 0 );
1308 return $self->redirect_to('/data/items');
1310 warn "UNKNOWN IMPORT $import";
1314 if ( my $remove = $self->param('remove') ) {
1315 my $path = "$dir/export/$remove";
1316 unlink $path if -e $path;
1318 unlink $path if -e $path;
1321 my $path = $self->_export_path || return $self->redirect_to('/data/index');
1323 my @files = grep { ! /\.png$/ } glob "$path/*";
1324 my $mtime = { map { $_ => (stat($_))[9] } @files };
1325 @files = sort { $mtime->{$b} <=> $mtime->{$a} } @files;
1326 $self->render( export => [ @files ] );
1329 sub __loaded_paths {
1331 grep { defined $loaded->{$_}->{data} }
1337 my $path = $self->param('path');
1338 if ( $path =~ m{^/tmp/mojo_facets\.} ) {
1340 warn "# unlink $path";
1341 if ( my $name = $self->param('name') ) {
1342 delete $loaded->{$name};
1343 warn "# remove $name from memory";
1346 warn "WARNING: $path unlink ignored";
1348 return $self->redirect_to( '/data/load' );