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;
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 $n = unac_string('utf-8',$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 if ( defined $scalar ) {
495 $self->session($name => $scalar);
497 warn "no default for $name";
501 warn "# _param_scalar $name ",dump $scalar;
508 my $name = $self->param('filter_name') || die "name?";
509 my @vals = $self->param('filter_vals');
511 $self->_remove_filter( $name );
513 $self->_filter_on_data( $name, @vals );
514 if ( my $permanent = $self->param('_permanent') ) {
515 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
516 write_file $permanent_path, { binmode => ':utf8' }, map { "$_\n" } @vals;
517 warn "permanent filter $permanent_path ", -s $permanent_path;
521 $self->session( 'offset' => 0 );
523 return $self->redirect_to('/data/items');
526 sub _filter_on_data {
527 my ( $self, $name, @vals ) = @_;
529 my $path = $self->session('path');
531 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
532 warn "filter $name not found in data set";
536 $filters->{$name} = [ @vals ];
537 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
540 $filter_hash->{$_}++ foreach @vals;
542 #warn "# filter_hash ",dump( $filter_hash );
544 my $items = $self->_loaded('data')->{items};
546 my $include_missing = defined $filter_hash->{_missing};
549 foreach my $i ( 0 .. $#$items ) {
551 if ( defined $items->[$i]->{$name} ) {
552 my $row = $items->[$i]->{$name};
553 $row = [ $row ] unless ref $row eq 'ARRAY'; # FIXME probably wrong place
554 foreach my $v ( @$row ) {
555 if ( defined $filter_hash->{ $v } ) {
556 $filtered_items->{$i}++;
559 } elsif ( $include_missing ) {
560 $filtered_items->{$i}++;
564 #warn "# filter $name ",dump($filtered_items);
566 # invalidate filters on other datasets
567 foreach my $dataset ( grep { exists $loaded->{$_}->{filters}->{$name} } keys %$loaded ) {
568 delete $loaded->{$dataset}->{filters}->{$name};
569 delete $loaded->{$dataset}->{filtered};
572 $loaded->{$path}->{filters}->{$name} = $filtered_items;
573 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
577 sub _current_filters {
580 my $stats = $self->_loaded('stats');
582 $current_filters->{ $_ } = $filters->{ $_ }
583 foreach ( grep { defined $filters->{ $_ } } keys %$stats )
585 warn "# _current_filters ",dump( keys %$current_filters );
586 return $current_filters;
589 sub _data_sorted_by {
590 my ( $self, $order ) = @_;
592 my $path = $self->session('path');
594 warn "_data_sorted_by $order from $path";
596 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
597 return $loaded->{$path}->{sorted}->{$order};
600 my $data = $self->_loaded( 'data' );
601 my $numeric = $self->_is_numeric($order);
602 my $missing = $numeric ? 0 : '';
603 no warnings qw(numeric);
615 if ( ! exists $_->{$order} ) {
617 } elsif ( ref $_->{$order} eq 'ARRAY' ) {
618 $v = join('', @{$_->{$order}});
623 } @{ $data->{items} }
626 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
627 #warn "# sorted ",dump( @sorted );
629 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
635 join(',', sort(@_), 'order', $order);
638 our ($out, $key,$value);
640 our $lookup_path_col;
644 warn "__commit_begin";
645 $lookup_path_col = undef;
651 $lookup_path_col = undef; # cleanup memory
656 warn "# lookup ",dump @_;
657 my ( $vals, $on_path, $on_col, $code, $stat_code ) = @_;
658 die "code is not sub{ ... } but ", dump $code unless ref $code eq 'CODE';
660 if ( ! exists $loaded->{$on_path} ) {
661 my @possible_paths = grep { /\Q$on_path\E/ } keys %$loaded;
662 die "more than one dataset available for '$on_path' ",dump @possible_paths if $#possible_paths > 0;
663 $on_path = shift @possible_paths;
664 warn "## fuzzy selected path $on_path";
667 my $items = $loaded->{$on_path}->{data}->{items} || die "no items for $on_path";
669 if ( ! exists $lookup_path_col->{$on_path}->{$on_col} ) {
670 warn "create lookup_path_col $on_path $on_col";
671 foreach my $i ( 0 .. $#$items ) {
672 my $item = $items->[$i];
673 if ( exists $item->{$on_col} ) {
674 if ( ref $item->{$on_col} eq 'ARRAY' ) {
675 foreach my $v ( @{ $item->{$on_col} } ) {
676 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
678 } elsif ( ! ref $item->{$on_col} ) { # scalar
679 my $v = $item->{$on_col};
680 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
682 die "unknown type of ",dump $item->{$on_col};
686 warn "XXX ",dump $lookup_path_col->{$on_path}->{$on_col} if $ENV{DEBUG};
690 $stat = Statistics::Descriptive::Full->new() if $stat_code;
692 foreach my $v ( ref $vals eq 'ARRAY' ? @$vals : ( $vals ) ) {
693 foreach my $i ( @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} } ) {
695 warn "XXX lookup code $v $i ",dump $on if $ENV{DEBUG};
698 $stat_code->( $stat ) if $stat_code;
702 sub __commit_path_code {
703 my ( $path, $i, $code, $commit_changed ) = @_;
705 my $items = $loaded->{$path}->{data}->{items} || die "no items for $path";
706 my $row = $items->[$i];
709 foreach ( keys %$update ) {
710 $$commit_changed->{$_}++;
711 $loaded->{$path}->{data}->{items}->[$i]->{$_} = $update->{$_};
713 #warn "__commit_path_code $path $i ",dump( $update );
716 # uses templates/admin.html.ep
717 sub _switch_dataset {
722 foreach my $path ( keys %$loaded ) {
723 next unless exists $loaded->{$path}->{data};
724 push @$datasets, $path;
727 warn "# datasets ",dump($datasets);
729 $self->stash( 'datasets' => $datasets );
735 $self->_switch_dataset;
737 if ( my $show = $self->param('id') ) {
738 $self->param('show', $show);
742 my $path = $self->_param_scalar('path');
744 my @columns = $self->_param_array('columns');
745 return $self->redirect_to('/data/columns') unless @columns;
746 my $order = $self->_param_scalar('order', $columns[0]);
747 my $sort = $self->_param_scalar('sort', 'a');
748 my $offset = $self->_param_scalar('offset', 0);
749 my $limit = $self->_param_scalar('limit', 20);
750 $self->_param_scalar('show', 'table');
752 # fix offset when changing limit
753 $offset = int( $offset / $limit ) * $limit;
755 if ( ! grep { /^\Q$order\E$/ } @columns ) {
756 $order = $columns[0];
757 $self->session( order => $order );
759 my $sorted = $self->_data_sorted_by( $order );
763 foreach my $name ( keys %$filters ) {
764 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
765 warn "skip filter $name not found in $path\n";
768 push @filter_names, $name;
770 warn "filter_names ",dump( @filter_names );
771 foreach my $name ( @filter_names ) {
772 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
773 $self->_filter_on_data( $name, @{ $filters->{$name} } );
777 my $all_filters = __all_filters( @filter_names,$order );
779 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
781 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
783 my $path_filters = $loaded->{$path}->{filters};
785 warn "create combined filter for $all_filters from ", $#$sorted + 1, " items\n";
788 foreach my $i ( 0 .. $#$sorted ) {
789 my $pos = $sorted->[$i];
791 if ( $#filter_names == -1 ) {
792 push @filtered, $pos;
797 foreach ( @filter_names ) {
798 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
802 push @filtered, $pos;
805 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
808 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
809 if defined $loaded->{$path}->{filtered}->{$all_filters};
811 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
813 my $data = $self->_loaded('data');
815 my $code = $self->_param_scalar('code','');
817 $code =~ s{\n+$}{\n}s;
819 # XXX convert @row->{foo} into @{$row->{foo}}
820 $code =~ s|\@(row->{[^}]+})|\@{\$$1}|gs;
822 my $commit = $self->param('commit');
823 my $test = $self->param('test');
828 if ( $code && ( $test || $commit ) ) {
829 # XXX find columns used in code snippet and show them to user
831 foreach my $column ( $code =~ m/\$row->{([^}]+)}/g ) {
832 if ( $column =~ s/^(['"])// ) {
835 next if $column =~ m/\$/; # hide columns with vars in them
836 $commit_changed->{$column} = 0;
840 my $code_path = $self->app->home->rel_dir('public') . "/code";
843 warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
844 ( $key, $value, $out ) = ( 'key', 'value' );
845 foreach ( 0 .. $#$filtered ) {
846 my $i = $filtered->[$_];
847 __commit_path_code( $path, $i, $code, \$commit_changed );
850 # this might move before $out to recalculate stats on source dataset?
851 __path_rebuild_stats( $path );
852 my $c = { map { $_ => 1 } @columns };
853 my @added_columns = sort grep { ! $c->{$_} } keys %$commit_changed;
854 warn "# added_columns ",dump( @added_columns );
855 unshift @columns, @added_columns;
857 $loaded->{$path}->{columns} = [ @columns ];
858 warn "# new columns ",dump( @columns );
860 __invalidate_path_column( $path, $_ ) foreach keys %$commit_changed;
862 $self->_save_change({
864 time => $self->param('time') || time(),
865 user => $self->param('user') || $ENV{'LOGNAME'},
867 commit_changed => $commit_changed,
870 if ( my $description = $self->param('code_description') ) {
871 my $depends = $self->param('code_depends') || die "no code_depends?";
872 my $path = "$code_path/$depends.$description.pl";
873 if ( -e $path && ! $self->param('overwrite') ) {
874 warn "# code $path not saved\n";
876 write_file( $path, { binmode => ':utf8' }, "$code\n" );
877 warn "code $path ", -s $path, " bytes saved\n";
884 my $commit_dataset = join('.'
885 , $self->param('code_depends')
886 , $self->param('code_description')
891 warn "key $key value $value";
893 foreach my $n ( keys %$out ) {
894 my $i = { $key => [ $n ] };
895 my $ref = ref $out->{$n};
896 if ( $ref eq 'HASH' ) {
897 $i->{$_} = [ $out->{$n}->{$_} ] foreach keys %{ $out->{$n} };
898 } elsif ( $ref eq 'ARRAY' ) {
899 $i->{$_} = $out->{$n};
901 $i->{$value} = [ $out->{$n} ];
903 $i->{_error} = [ dump($out->{$n}) ];
908 my $stats = __stats( $items );
909 my @columns = grep { ! m/^\Q$key\E$/ } sort keys %$stats;
910 unshift @columns, $key;
912 $loaded->{$commit_dataset} = {
913 header => [ @columns ],
914 columns => [ @columns ],
916 data => { items => $items },
920 warn "# loaded out ", dump( $loaded->{$commit_dataset} );
921 $self->session('path', $commit_dataset);
922 $self->session('columns', [ @columns ]);
923 $self->session('order', $key);
924 return $self->redirect_to('/data/items');
927 $self->session('columns', [ @columns ]);
931 my $from_end = $sort eq 'd' ? $#$filtered : 0;
933 my ( $key, $value, $out ) = ( 'key', 'value' ); # XXX make local
934 foreach ( 0 .. $limit ) {
935 my $i = $_ + $offset;
936 last unless defined $filtered->[$i];
937 $i = $from_end - $i if $from_end;
938 my $id = $filtered->[$i];
939 my $row = Storable::dclone $data->{items}->[ $id ];
940 if ( $code && $test ) {
944 warn "ERROR evaling $@", dump($code);
945 $self->stash('eval_error', $@) if $@;
947 warn "EVAL ",dump($update);
948 foreach ( keys %$update ) {
949 $test_changed->{$_}++;
950 $row->{$_} = $update->{$_};
954 $row->{_row_id} ||= $id;
955 push @$sorted_items, $row;
958 if ( $self->param('export') ) {
959 my $export_path = $self->_export_path( 'items', @columns);
960 open(my $fh, '>', $export_path) || die "ERROR: can't open $export_path: $!";
961 print $fh "#",join("\t",@columns),"\n";
962 foreach my $f ( 0 .. $#$filtered ) {
963 print $fh join("\t", map {
964 my $i = $data->{items}->[ $filtered->[$f] ];
966 if ( ! defined $i->{$_} ) {
968 } elsif ( ref $i->{$_} eq 'ARRAY' ) {
969 $v =join(',', @{ $i->{$_} });
970 $v = '\N' if length($v) == 0;
971 } elsif ( ! ref $i->{$_} ) {
980 warn "export $export_path ", -s $export_path, " bytes\n";
983 my ( $code_depends, $code_description );
987 warn "# test_changed ",dump( $test_changed );
988 my $c = { map { $_ => 1 } @columns };
989 my @added_columns = sort grep { ! $c->{$_} } keys %$test_changed;
990 unshift @columns, @added_columns;
992 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
995 my $tmp = $code; $tmp =~ s/\$row->{(['"]?)([\w\s]+)\1/$depends_on->{$2}++/gse;
996 warn "# depends_on ",dump $depends_on;
998 my $test_added = Storable::dclone $test_changed;
999 delete $test_added->{$_} foreach keys %$depends_on;
1001 $code_depends = $self->param('code_depends')
1002 || join(',', keys %$depends_on);
1004 $code_description = $self->param('code_description') ||
1005 join(',', keys %$test_added);
1007 $code_depends ||= $code_description; # self-modifing
1008 if ( ! $code_depends && $out ) {
1009 $code_depends = $key;
1010 $code_description = $value;
1013 warn "# test_changed ",dump( $test_changed, $code_depends, $code_description );
1023 sorted => $sorted_items,
1024 columns => [ @columns ],
1025 rows => $#$filtered + 1,
1026 numeric => { map { $_, $self->_is_numeric($_) } @columns },
1027 unique => { map { $_, $self->_is_unique( $_) } @columns },
1028 filters => $self->_current_filters,
1030 cols_changed => $commit ? $commit_changed : $test_changed,
1031 code_depends => $code_depends,
1032 code_description => $code_description,
1033 code_path => $code_path,
1042 $self->session('order', $self->param('order'));
1043 $self->session('sort', $self->param('sort'));
1044 return $self->redirect_to('/data/items');
1048 my ( $self, $name ) = @_;
1050 my $stats = $self->_loaded( 'stats' );
1052 # sort facet numerically if more >50% elements are numeric
1053 my $count = $stats->{$name}->{count};
1054 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
1055 defined $stats->{$name}->{numeric} &&
1056 $stats->{$name}->{numeric} > $count / 2;
1060 my ( $self, $name ) = @_;
1061 my $stats = $self->_loaded( 'stats' );
1062 defined $stats->{$name}->{unique};
1065 sub _remove_filter {
1066 my ($self,$name) = @_;
1067 warn "_remove_filter $name\n";
1069 my $path = $self->session('path');
1071 delete $filters->{$name};
1072 delete $loaded->{$path}->{filters}->{$name};
1073 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
1076 grep { /\Q$name\E/ }
1077 keys %{ $loaded->{$path}->{filtered} }
1079 delete $loaded->{$path}->{filtered}->{$_};
1080 warn "remove filtered cache $_";
1087 my $path = $self->session('path') || return $self->redirect_to( '/data/index' );
1089 if ( my $name = $self->param('remove') ) {
1090 $self->_remove_filter( $name );
1091 return $self->redirect_to( '/data/items' );
1095 my $name = $self->param('name') || die "no name";
1097 my $all = $self->_param_scalar('all', 1);
1098 my $data = $self->_loaded('data');
1100 my $filters = $self->_current_filters;
1101 my $all_filters = __all_filters( keys %$filters,$self->session('order') );
1102 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
1103 if defined $loaded->{$path}->{filtered}->{$all_filters};
1105 if ( ! $filtered || $all ) {
1106 $filtered = [ 0 .. $#{ $data->{items} } ];
1107 warn "filter all values\n";
1109 warn "filter using $all_filters\n";
1112 foreach my $i ( @$filtered ) {
1113 my $item = $data->{items}->[$i];
1114 if ( ! exists $item->{$name} || ! defined $item->{$name} ) {
1115 $facet->{ _missing }++;
1116 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
1117 $facet->{$_}++ foreach @{ $item->{$name} };
1119 $facet->{ $item->{$name} }++;
1123 my $checked_values = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
1125 if ( my $code = $self->param('code') ) {
1127 foreach my $value ( keys %$facet ) {
1128 my $count = $facet->{$value};
1129 my $checked = $checked_values->{$value};
1133 warn "ERROR: $@\n$code\n";
1135 } elsif ( $checked != $checked_values->{$value} ) {
1136 warn "checked $value $count -> $checked\n";
1137 $checked_values->{$value} = $checked;
1140 warn "out ",dump( $out );
1141 $self->stash( out => $out );
1144 # warn "# facet $name ",dump $facet;
1148 : defined $filters->{$name} ? @{ $filters->{$name} }
1151 my $numeric = $self->_is_numeric($name);
1153 my $sort = $self->param('sort');
1154 # sort numeric facets with more than 5 values ascending
1155 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
1157 @facet_names = sort {
1159 if ( $sort eq 'a' ) {
1160 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
1161 } elsif ( $sort eq 'd' ) {
1162 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
1163 } elsif ( $sort eq 'c' ) {
1164 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
1166 warn "unknown sort: $sort";
1167 $result = $a cmp $b;
1169 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
1173 $self->render( name => $name, facet => $facet, checked => $checked_values,
1174 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
1179 sub __invalidate_path_column {
1180 my ( $path, $name ) = @_;
1182 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
1183 delete $loaded->{$path}->{sorted}->{$name};
1184 warn "# invalidate $path sorted $name\n";
1187 foreach ( grep { m/\Q$name\E/ } keys %{ $loaded->{$path}->{filtered} } ) {
1188 delete $loaded->{$path}->{filtered}->{$_};
1189 warn "# invalidate $path filtered $_\n";
1193 sub __path_modified {
1194 my ( $path, $value ) = @_;
1195 $value = 1 unless defined $value;
1197 $loaded->{$path}->{modified} = $value;
1199 warn "# __path_modified $path $value\n";
1202 sub __path_rebuild_stats { $loaded->{ $_[0] }->{rebuild_stats} = 1 };
1205 my ($self,$change) = @_;
1207 my $change_path = $self->_permanent_path( 'changes' );
1208 mkdir $change_path unless -d $change_path;
1209 $change_path .= '/' . $change->{time};
1210 store $change, $change_path;
1211 utime $change->{time}, $change->{time}, $change_path;
1212 warn "_save_change $change_path ", dump($change);
1217 my $new_content = $self->param('new_content');
1218 $new_content ||= $self->param('content'); # backward compatibility with old actions
1220 my $i = $self->param('_row_id');
1221 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
1222 my $path = $self->param('path') || die "no path";
1223 my $name = $self->param('name') || die "no name";
1224 my $status = 200; # 200 = OK, 201 = Created
1226 my $data = $self->_loaded('data');
1228 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
1229 $new_content =~ s/^\s+//s;
1230 $new_content =~ s/\s+$//s;
1232 if ( $new_content =~ /\xB6/ ) { # para
1233 $v = [ split(/\s*\xB6\s*/, $new_content) ];
1235 $v = [ $new_content ];
1238 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
1241 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
1243 $self->_save_change({
1247 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
1249 time => $self->param('time') || time(),
1250 user => $self->param('user') || $ENV{'LOGNAME'},
1252 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
1253 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
1254 keys %{ $loaded->{$path}->{stats} }
1258 warn "# change $path $i $old -> $new\n";
1259 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
1261 __invalidate_path_column( $path, $name );
1263 $status = 201; # created
1264 __path_rebuild_stats( $path );
1266 $new_content = join("\xB6",@$v);
1269 warn "# unchanged $path $i $old\n";
1273 $new_content = "$path $i $name doesn't exist\n";
1277 warn "# edit $status ", dump $new_content;
1281 new_content => scalar $new_content,
1288 my $path = $self->_param_or_session('path');
1289 my $dump_path = $self->_save( $path );
1290 __path_modified( $path, 0 );
1292 return $self->redirect_to( '/data/items' );
1298 my $dir = $self->app->home->rel_dir('public');
1300 if ( my $import = $self->param('import') ) {
1302 if ( $import =~ m{/filter\.(.+?)\..+} ) {
1303 my $name = $self->_column_from_unac( $1 );
1305 my @vals = map { chomp; $_ }
1306 read_file "$dir/export/$import", binmode => ':utf8';
1308 $self->_remove_filter( $name );
1309 $self->_filter_on_data( $name, @vals );
1310 $self->session( 'offset' => 0 );
1311 return $self->redirect_to('/data/items');
1313 warn "UNKNOWN IMPORT $import";
1317 if ( my $remove = $self->param('remove') ) {
1318 my $path = "$dir/export/$remove";
1319 unlink $path if -e $path;
1321 unlink $path if -e $path;
1324 my $path = $self->_export_path || return $self->redirect_to('/data/index');
1326 my @files = grep { ! /\.png$/ } glob "$path/*";
1327 my $mtime = { map { $_ => (stat($_))[9] } @files };
1328 @files = sort { $mtime->{$b} <=> $mtime->{$a} } @files;
1329 $self->render( export => [ @files ] );
1332 sub __loaded_paths {
1334 grep { defined $loaded->{$_}->{data} }
1340 $self->stash( reload => 1 );
1342 # $self->_load_path( $self->param('path') );
1343 $self->redirect_to('/data/load?path=' . $self->param('path') );
1348 my $path = $self->param('path');
1349 if ( $path =~ m{^/tmp/mojo_facets\.} ) {
1351 warn "# unlink $path";
1352 if ( my $name = $self->param('name') ) {
1353 delete $loaded->{$name};
1354 warn "# remove $name from memory";
1357 warn "WARNING: $path unlink ignored";
1359 return if $self->stash('reload');
1360 return $self->redirect_to( '/data/load' );