1 package MojoFacets::Data;
6 use base 'Mojolicious::Controller';
8 #use Data::Dump qw(dump); # broken with Mojo::JSON, see https://rt.cpan.org/Public/Bug/Display.html?id=86592
11 sub dump { Dumper(@_) };
18 use Time::HiRes qw(time);
19 use File::Path qw(mkpath);
20 use Text::Unaccent::PurePerl;
22 use Statistics::Descriptive;
25 foreach my $module ( glob('lib/MojoFacets/Import/*.pm') ) {
26 $module =~ s{lib/(\w+)/(\w+)/(.*)\.pm}{$1::$2::$3};
28 die "$module: $!" if $!;
29 my ( $ext, $priority ) = $module->ext;
30 $imports->{$priority || 'file'}->{$ext} = $module;
31 warn "# import $ext $module\n";
34 warn "# import loaded ",dump( $imports );
37 my $full_path = shift;
39 # warn "# import_module $full_path\n";
41 return if $full_path =~ m/\.columns$/;
43 foreach my $ext ( keys %{ $imports->{file} } ) {
44 if ( -f $full_path && $full_path =~ m/$ext/i ) {
45 return $imports->{file}->{$ext};
50 foreach my $ext ( keys %{ $imports->{directory} } ) {
51 if ( -f $full_path && $full_path =~ m/$ext/i ) {
52 return $imports->{directory}->{$ext};
64 my $data_dir = $self->app->home->rel_file('data');
65 die "no data dir $data_dir" unless -d $data_dir;
71 my $file = $File::Find::name;
73 next if $file =~ m/.timefmt$/;
75 if ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+[\.,]\d+.+)/ ) {
76 push @{ $changes->{$1} }, $2
77 } elsif ( import_module( $file ) ) {
78 my $mtime = (stat($file))[9]; # mtime
79 $file =~ s/$data_dir\/*//;
81 $loaded->{$file}->{mtime} ||= $mtime;
83 #warn "IGNORE: $file\n";
87 no warnings qw(uninitialized); # mtime
88 @files = sort { $loaded->{$b}->{mtime} <=> $loaded->{$a}->{mtime} || lc $a cmp lc $b } @files,
89 grep { defined $loaded->{$_}->{generated} } keys %$loaded;
92 $size->{$_} = -s "$data_dir/$_" foreach @files;
99 dump_path => { map { $_ => $self->_dump_path($_) } @files },
105 my ( $self, $name ) = @_;
106 my $dir = $self->app->home->rel_file('data');
109 return '/tmp/mojo_facets.' . $name . '.storable';
113 my ( $self, $path ) = @_;
115 my $dump_path = $self->_dump_path( $path );
116 my $first_load = ! -e $dump_path;
117 warn "save loaded to $dump_path";
118 my $info = $loaded->{$path};
119 store $info, $dump_path;
122 my $mtime = $loaded->{$path}->{mtime};
123 utime $mtime, $mtime, $dump_path;
124 warn "sync time to $path at $mtime\n";
127 warn $dump_path, ' ', -s $dump_path, " bytes\n";
136 my $nr_items = $#{ $_[0] } + 1;
138 warn "__stats $nr_items\n";
140 # foreach my $e ( @{ $_[0] } ) {
141 foreach my $i ( 0 .. $#{$_[0]} ) {
142 print STDERR " $i" if $i % 5000;
144 foreach my $n ( keys %$e ) {
145 $stats->{$n}->{count}++;
147 if ( ref $e->{$n} eq 'ARRAY' ) {
148 $stats->{$n}->{array} += $#{ $e->{$n} } + 1;
154 foreach my $x ( @v ) {
155 if ( ! defined $x ) { # FIXME really null
156 $stats->{$n}->{empty}++;
160 $stats->{$n}->{numeric}++
161 if $x =~ m/^[-+]?([0-9]*\.[0-9]+|[0-9]+)$/;
162 $stats->{$n}->{empty}++
163 if length $x == 0; # faster than $x =~ m/^\s*$/;
169 foreach my $n ( keys %$stats ) {
170 my $s = $stats->{$n};
171 next unless defined $s->{array};
172 if ( $s->{array} == $s->{count} ) {
174 if ( $s->{count} == $nr_items ) {
175 warn "check $n for uniqeness\n";
177 foreach my $e ( @{ $_[0] } ) {
178 if ( ++$unique->{ $e->{$n}->[0] } == 2 ) {
184 $stats->{$n}->{unique} = 1;
185 #warn "# $n unique ",dump( $unique );
191 warn "# __stats ",dump($stats);
196 sub _param_or_session {
197 $_[0]->param( $_[1] ) || $_[0]->session( $_[1] )
202 my $path = $self->_param_or_session('path');
203 warn "stats $path\n";
204 delete $loaded->{$path}->{stats};
205 return $self->redirect_to( '/data/columns' );
210 my ( $self, $path ) = @_;
212 return if defined $loaded->{$path}->{generated};
214 my $full_path = $self->app->home->rel_file( 'data/' . $path );
215 return $self->redirect_to('/data/index') unless -r $full_path;
217 my $dump_path = $self->_dump_path( $path );
219 if ( defined $loaded->{$path}->{data} ) {
220 my $mtime = (stat($full_path))[9];
221 return if $loaded->{$path}->{mtime} == $mtime;
222 warn "reload $full_path, modified ", time() - $mtime, " seconds ago\n";
223 } elsif ( -e $dump_path ) {
224 warn "dump_path $dump_path ", -s $dump_path, " bytes loading...\n";
225 my $info = retrieve $dump_path;
226 $loaded->{ $path } = $info;
231 if ( my $module = import_module( $full_path ) ) {
232 $data = $module->new( full_path => $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};
267 my $timefmt_path = $self->_permanent_path( 'timefmt' );
268 if ( -e $timefmt_path ) {
269 $self->session( 'timefmt', read_file $timefmt_path );
278 my @paths = @{ $self->every_param('paths') };
279 warn "# paths ", dump @paths;
281 foreach my $p ( keys %$loaded ) {
282 next if grep { /^\Q$p\E$/ } @paths;
283 warn "remove $p from memory\n";
284 delete $loaded->{$p};
287 $self->_load_path( $_ ) foreach @paths;
289 my $path = $self->param('path') || $self->session('path') || $paths[0] || $self->redirect_to('/data/index');
291 warn "# path $path\n";
292 $self->_load_path( $path );
294 $self->session( 'path' => $path );
296 my $redirect_to = '/data/items';
298 $self->session( 'header' => $loaded->{$path}->{header} );
299 if ( ! defined $loaded->{$path}->{columns} ) {
300 my $columns_path = $self->_permanent_path( 'columns' );
301 if ( -e $columns_path ) {
302 my @columns = map { s/[\r\n]+$//; $_ } read_file $columns_path, binmode => ':utf8';
303 $loaded->{$path}->{columns} = [ @columns ];
304 warn "# columns_path $columns_path ",dump(@columns);
306 $loaded->{$path}->{columns} = $loaded->{$path}->{header}
309 $redirect_to = '/data/columns';
311 $self->session( 'columns' => $loaded->{$path}->{columns} );
312 $self->session( 'order' => $loaded->{$path}->{columns}->[0] );
313 return $self->redirect_to( $redirect_to );
318 my ( $self, $name ) = @_;
319 my $path = $self->session('path') || $self->param('path');
320 return $self->redirect_to('/data/index') unless $path;
322 if ( defined $loaded->{$path}->{modified} && $loaded->{$path}->{modified} > 1 ) {
323 my $caller = (caller(1))[3];
324 if ( $caller =~ m/::edit/ ) {
325 warn "rebuild stats for $path ignored caller $caller\n";
327 warn "rebuild stats for $path FORCED by modified caller $caller\n";
328 # $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
329 $loaded->{$path}->{rebuild_stats} = 1;
330 $loaded->{$path}->{modified} = 1;
334 if ( defined $loaded->{$path}->{rebuild_stats} ) {
335 warn "rebuild_stats $path";
336 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
337 delete $loaded->{$path}->{rebuild_stats};
340 if ( ! defined $loaded->{$path}->{$name} ) {
341 warn "$path $name isn't loaded\n";
342 $self->_load_path( $path );
343 if ( ! defined $loaded->{$path}->{stats} ) {
344 warn "rebuild stats for $path\n";
345 $loaded->{$path}->{stats} = __stats( $loaded->{$path}->{data}->{items} );
347 if ( ! defined $loaded->{$path}->{$name} ) {
348 warn "MISSING $name for $path\n";
349 return $self->redirect_to('/data/index')
353 $self->session( 'modified' => $loaded->{$path}->{modified} );
355 return $loaded->{$path}->{$name};
362 $checked->{$_}++ foreach @_;
363 # warn "# _checked ",dump($checked);
367 sub _permanent_path {
369 my $path = $self->_param_or_session('path');
370 $self->app->home->rel_file('data') . '/' . join('.', $path, @_);
375 $n = unac_string('utf-8',$n);
380 sub _column_from_unac {
381 my ($self,$name) = @_;
383 my $stats = $self->_loaded('stats');
384 my $cols_norm = { map { __unac( $_ ) => $_ } keys %$stats };
386 $cols_norm->{$name} || die "can't find column $name in ", dump($cols_norm);
391 my $path = $self->_param_or_session('path');
393 warn "no path in param or session";
396 my $dir = $self->app->home->rel_file('public') . "/export/$path";
397 mkpath $dir unless -e $dir;
398 my $name = __export_path_name( $path, @_ );
399 my $full = $dir . '/' . $name;
400 $full =~ s/\/+$// if -d $full; # strip trailing slash for dirs
404 sub __export_path_name {
408 my $name = join('.', map { __unac($_) } @_ );
409 if ( length($name) > $max_length ) {
410 $name = substr($name,0,$max_length) . Digest::MD5::md5_hex substr($name,$max_length);
418 my $view_path = $self->_permanent_path( 'views' );
420 if ( $self->param('columns') ) {
421 my @columns = $self->_param_array('columns');
422 write_file( $self->_permanent_path( 'columns' ), { binmode => ':utf8' }, map { "$_\n" } @columns );
423 if ( my $view = $self->param('view') ) {
424 mkdir $view_path unless -e $view_path;
425 write_file( "$view_path/$view", { binmode => ':utf8' }, map { "$_\n" } @columns );
426 warn "view $view_path/$view ", -s "$view_path/$view", " bytes\n";
429 return $self->redirect_to('/data/items');
431 } elsif ( ! $self->session('header') ) {
432 return $self->redirect_to('/data/load');
435 if ( my $id = $self->param('id') ) {
436 my $view_full = "$view_path/$id";
437 if ( -f $view_full ) {
438 my @columns = map { chomp; $_ } read_file $view_full, binmode => ':utf8';
439 warn "view $view_full loaded ", dump @columns;
440 $self->session( 'columns' => [ @columns ] );
441 return $self->redirect_to('/data/items');
445 my $stats = $self->_loaded( 'stats' );
448 @columns = grep { defined $stats->{$_}->{count} } @{ $self->session('columns') } if $self->session('columns');
450 foreach my $c ( @{ $self->session( 'header' ) } ) {
451 push @columns, $c unless grep { /^\Q$c\E$/ } @columns;
455 if ( -d $view_path ) {
456 @views = map { s{^\Q$view_path\E/*}{}; $_ } glob "$view_path/*";
457 warn "# views ",dump @views;
461 message => 'Select columns to display',
463 columns => \@columns,
464 checked => $self->_checked( $self->_param_array('columns') ),
470 my ($self,$name) = @_;
472 my @array = @{ $self->every_param($name) };
473 my $path = $self->session('path');
476 $self->session($name => [ @array ]);
477 $loaded->{$path}->{$name} = [ @array ];
478 } elsif ( my $session = $self->session($name) ) {
479 if ( ref $session eq 'ARRAY' ) {
482 die "$name not array ",dump($session);
485 #warn "# $name ",dump @array;
490 my ($self,$name,$default) = @_;
492 my $scalar = $self->param($name);
494 if ( defined $scalar ) {
495 $self->session($name => $scalar);
497 $scalar = $self->session($name);
500 if ( ! defined $scalar ) {
502 if ( defined $scalar ) {
503 $self->session($name => $scalar);
505 warn "no default for $name";
509 warn "# _param_scalar $name ",dump $scalar;
516 my $name = $self->param('filter_name') || die "name?";
517 my @vals = @{ $self->every_param('filter_vals') };
519 $self->_remove_filter( $name );
521 $self->_filter_on_data( $name, @vals );
522 if ( my $permanent = $self->param('_permanent') ) {
523 my $permanent_path = $self->_export_path( 'filter', $name, $permanent );
524 write_file $permanent_path, { binmode => ':utf8' }, map { "$_\n" } @vals;
525 warn "permanent filter $permanent_path ", -s $permanent_path;
529 $self->session( 'offset' => 0 );
531 return $self->redirect_to('/data/items');
534 sub _filter_on_data {
535 my ( $self, $name, @vals ) = @_;
537 my $path = $self->session('path');
539 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
540 warn "filter $name not found in data set";
544 $filters->{$name} = [ @vals ];
545 warn "_filter_on_data $name ", $#vals + 1, " values on $path\n";
548 $filter_hash->{$_}++ foreach @vals;
550 #warn "# filter_hash ",dump( $filter_hash );
552 my $items = $self->_loaded('data')->{items};
554 my $include_missing = defined $filter_hash->{_missing};
557 foreach my $i ( 0 .. $#$items ) {
559 if ( defined $items->[$i]->{$name} ) {
560 my $row = $items->[$i]->{$name};
561 $row = [ $row ] unless ref $row eq 'ARRAY'; # FIXME probably wrong place
562 foreach my $v ( @$row ) {
563 if ( defined $filter_hash->{ $v } ) {
564 $filtered_items->{$i}++;
567 } elsif ( $include_missing ) {
568 $filtered_items->{$i}++;
572 #warn "# filter $name ",dump($filtered_items);
574 # invalidate filters on other datasets
575 foreach my $dataset ( grep { exists $loaded->{$_}->{filters}->{$name} } keys %$loaded ) {
576 delete $loaded->{$dataset}->{filters}->{$name};
577 delete $loaded->{$dataset}->{filtered};
580 $loaded->{$path}->{filters}->{$name} = $filtered_items;
581 warn "filter $name with ", scalar keys %$filtered_items, " items created\n";
585 sub _current_filters {
588 my $stats = $self->_loaded('stats');
590 $current_filters->{ $_ } = $filters->{ $_ }
591 foreach ( grep { defined $filters->{ $_ } } keys %$stats )
593 warn "# _current_filters ",dump( keys %$current_filters );
594 return $current_filters;
597 sub _data_sorted_by {
598 my ( $self, $order ) = @_;
600 my $path = $self->session('path');
602 warn "_data_sorted_by $order from $path";
604 if ( defined $loaded->{$path}->{sorted}->{$order} ) {
605 return $loaded->{$path}->{sorted}->{$order};
608 my $data = $self->_loaded( 'data' );
609 my $numeric = $self->_is_numeric($order);
610 my $missing = $numeric ? 0 : '';
611 no warnings qw(numeric);
623 if ( ! exists $_->{$order} ) {
625 } elsif ( ref $_->{$order} eq 'ARRAY' ) {
626 $v = join('', @{$_->{$order}});
631 } @{ $data->{items} }
634 warn "sorted: $order numeric: $numeric items: ", $#sorted + 1, "\n";
635 #warn "# sorted ",dump( @sorted );
637 $loaded->{$path}->{sorted}->{$order} = [ @sorted ];
643 join(',', sort(@_), 'order', $order);
646 our ($out, $key,$value);
648 our $lookup_path_col;
652 warn "__commit_begin";
653 $lookup_path_col = undef;
659 $lookup_path_col = undef; # cleanup memory
664 warn "# lookup ",dump @_;
665 my ( $vals, $on_path, $on_col, $code, $stat_code ) = @_;
666 die "code is not sub{ ... } but ", dump $code unless ref $code eq 'CODE';
668 if ( ! exists $loaded->{$on_path} ) {
669 my @possible_paths = grep { /\Q$on_path\E/ } keys %$loaded;
670 die "more than one dataset available for '$on_path' ",dump @possible_paths if $#possible_paths > 0;
671 $on_path = shift @possible_paths;
672 warn "## fuzzy selected path $on_path";
675 my $items = $loaded->{$on_path}->{data}->{items} || die "no items for $on_path";
677 if ( ! exists $lookup_path_col->{$on_path}->{$on_col} ) {
678 warn "create lookup_path_col $on_path $on_col";
679 foreach my $i ( 0 .. $#$items ) {
680 my $item = $items->[$i];
681 if ( exists $item->{$on_col} ) {
682 if ( ref $item->{$on_col} eq 'ARRAY' ) {
683 foreach my $v ( @{ $item->{$on_col} } ) {
684 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
686 } elsif ( ! ref $item->{$on_col} ) { # scalar
687 my $v = $item->{$on_col};
688 push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
690 die "unknown type of ",dump $item->{$on_col};
694 warn "XXX ",dump $lookup_path_col->{$on_path}->{$on_col} if $ENV{DEBUG};
698 $stat = Statistics::Descriptive::Full->new() if $stat_code;
700 foreach my $v ( ref $vals eq 'ARRAY' ? @$vals : ( $vals ) ) {
701 foreach my $i ( @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} } ) {
703 warn "XXX lookup code $v $i ",dump $on if $ENV{DEBUG};
706 $stat_code->( $stat ) if $stat_code;
710 sub __commit_path_code {
711 my ( $path, $i, $code, $commit_changed ) = @_;
713 my $items = $loaded->{$path}->{data}->{items} || die "no items for $path";
714 my $row = $items->[$i];
717 foreach ( keys %$update ) {
718 $$commit_changed->{$_}++;
719 $loaded->{$path}->{data}->{items}->[$i]->{$_} = $update->{$_};
721 #warn "__commit_path_code $path $i ",dump( $update );
724 # uses templates/admin.html.ep
725 sub _switch_dataset {
730 foreach my $path ( keys %$loaded ) {
731 next unless exists $loaded->{$path}->{data};
732 push @$datasets, $path;
735 warn "# datasets ",dump($datasets);
737 $self->stash( 'datasets' => $datasets );
743 $self->_switch_dataset;
745 if ( my $timefmt = $self->param('timefmt') ) {
746 $self->session('timefmt', $timefmt);
747 warn "session store timefmt $timefmt\n";
748 my $timefmt_path = $self->_permanent_path( 'timefmt' );
749 write_file $timefmt_path, $timefmt;
750 warn "## $timefmt_path $timefmt"
753 if ( my $show = $self->param('id') ) {
754 $self->param('show', $show);
758 my $path = $self->_param_scalar('path');
760 my @columns = $self->_param_array('columns');
761 return $self->redirect_to('/data/columns') unless @columns;
762 my $order = $self->_param_scalar('order', $columns[0]);
763 my $sort = $self->_param_scalar('sort', 'a');
764 my $offset = $self->_param_scalar('offset', 0);
765 my $limit = $self->_param_scalar('limit', 20);
766 $self->_param_scalar('show', 'table');
768 # fix offset when changing limit
769 $offset = int( $offset / $limit ) * $limit;
771 if ( ! grep { /^\Q$order\E$/ } @columns ) {
772 $order = $columns[0];
773 $self->session( order => $order );
775 my $sorted = $self->_data_sorted_by( $order );
779 foreach my $name ( keys %$filters ) {
780 if ( ! defined $loaded->{$path}->{stats}->{ $name } ) {
781 warn "skip filter $name not found in $path\n";
784 push @filter_names, $name;
786 warn "filter_names ",dump( @filter_names );
787 foreach my $name ( @filter_names ) {
788 next if ref $loaded->{$path}->{filters}->{$name} eq 'ARRAY';
789 $self->_filter_on_data( $name, @{ $filters->{$name} } );
793 my $all_filters = __all_filters( @filter_names,$order );
795 # warn "# all_filters $all_filters ", dump( $loaded->{$path}->{filtered}->{$all_filters} );
797 if ( ! defined $loaded->{$path}->{filtered}->{$all_filters} ) {
799 my $path_filters = $loaded->{$path}->{filters};
801 warn "create combined filter for $all_filters from ", $#$sorted + 1, " items\n";
804 foreach my $i ( 0 .. $#$sorted ) {
805 my $pos = $sorted->[$i];
807 if ( $#filter_names == -1 ) {
808 push @filtered, $pos;
813 foreach ( @filter_names ) {
814 $skip ||= 1 if ! defined $path_filters->{$_}->{$pos};
818 push @filtered, $pos;
821 $loaded->{$path}->{filtered}->{$all_filters} = [ @filtered ];
824 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
825 if defined $loaded->{$path}->{filtered}->{$all_filters};
827 warn "all_filters $all_filters produced ", $#$filtered + 1, " items\n" if $filtered;
829 my $data = $self->_loaded('data');
831 my $code = $self->_param_scalar('code','');
833 $code =~ s{\n+$}{\n}s;
835 # XXX convert @row->{foo} into @{$row->{foo}}
836 $code =~ s|\@(row->\{[^}]+\})|\@{\$$1}|gs;
838 my $commit = $self->param('commit');
839 my $test = $self->param('test');
844 if ( $code && ( $test || $commit ) ) {
845 # XXX find columns used in code snippet and show them to user
847 foreach my $column ( $code =~ m/\$row->\{([^}]+)\}/g ) {
848 if ( $column =~ s/^(['"])// ) {
851 next if $column =~ m/\$/; # hide columns with vars in them
852 $commit_changed->{$column} = 0;
856 my $code_path = $self->app->home->rel_file('public') . "/code";
859 __path_modified( $path, 'commit' );
861 warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
862 ( $key, $value, $out ) = ( 'key', 'value' );
863 foreach ( 0 .. $#$filtered ) {
864 my $i = $filtered->[$_];
865 __commit_path_code( $path, $i, $code, \$commit_changed );
868 # this might move before $out to recalculate stats on source dataset?
869 __path_rebuild_stats( $path );
870 my $c = { map { $_ => 1 } @columns };
871 my @added_columns = sort grep { ! $c->{$_} } keys %$commit_changed;
872 warn "# added_columns ",dump( @added_columns );
873 unshift @columns, @added_columns;
875 $loaded->{$path}->{columns} = [ @columns ];
876 warn "# new columns ",dump( @columns );
878 __invalidate_path_column( $path, $_ ) foreach keys %$commit_changed;
880 $self->_save_change({
882 time => $self->param('time') || time(),
883 user => $self->param('user') || $ENV{'LOGNAME'},
885 commit_changed => $commit_changed,
888 if ( my $description = $self->param('code_description') ) {
889 my $depends = $self->param('code_depends') || die "no code_depends?";
890 my $path = "$code_path/$depends.$description.pl";
891 if ( -e $path && ! $self->param('overwrite') ) {
892 warn "# code $path not saved\n";
894 write_file( $path, { binmode => ':utf8' }, "$code\n" );
895 warn "code $path ", -s $path, " bytes saved\n";
902 my $commit_dataset = join('.'
903 , $self->param('code_depends')
904 , $self->param('code_description')
909 warn "key $key value $value";
911 foreach my $n ( keys %$out ) {
912 my $i = { $key => [ $n ] };
913 my $ref = ref $out->{$n};
914 if ( $ref eq 'HASH' ) {
915 $i->{$_} = [ $out->{$n}->{$_} ] foreach keys %{ $out->{$n} };
916 } elsif ( $ref eq 'ARRAY' ) {
917 $i->{$_} = $out->{$n};
919 $i->{$value} = [ $out->{$n} ];
921 $i->{_error} = [ dump($out->{$n}) ];
926 my $stats = __stats( $items );
927 my @columns = grep { ! m/^\Q$key\E$/ } sort keys %$stats;
928 unshift @columns, $key;
930 $loaded->{$commit_dataset} = {
931 header => [ @columns ],
932 columns => [ @columns ],
934 data => { items => $items },
938 warn "# loaded out ", dump( $loaded->{$commit_dataset} );
939 $self->session('path', $commit_dataset);
940 $self->session('columns', [ @columns ]);
941 $self->session('order', $key);
942 return $self->redirect_to('/data/items');
945 $self->session('columns', [ @columns ]);
949 my $from_end = $sort eq 'd' ? $#$filtered : 0;
951 my ( $key, $value, $out ) = ( 'key', 'value' ); # XXX make local
952 foreach ( 0 .. $limit ) {
953 my $i = $_ + $offset;
954 last unless defined $filtered->[$i];
955 $i = $from_end - $i if $from_end;
956 my $id = $filtered->[$i];
957 my $row = Storable::dclone $data->{items}->[ $id ];
958 if ( $code && $test ) {
962 warn "ERROR evaling $@", dump($code);
963 $self->stash('eval_error', $@) if $@;
965 warn "EVAL ",dump($update);
966 foreach ( keys %$update ) {
967 $test_changed->{$_}++;
968 $row->{$_} = $update->{$_};
972 $row->{_row_id} ||= $id;
973 push @$sorted_items, $row;
976 if ( $self->param('export') ) {
977 my $export_path = $self->_export_path( 'items', @columns);
978 open(my $fh, '>', $export_path) || die "ERROR: can't open $export_path: $!";
979 print $fh "#",join("\t",@columns),"\n";
980 foreach my $f ( 0 .. $#$filtered ) {
981 print $fh join("\t", map {
982 my $i = $data->{items}->[ $filtered->[$f] ];
984 if ( ! defined $i->{$_} ) {
986 } elsif ( ref $i->{$_} eq 'ARRAY' ) {
987 $v =join(',', @{ $i->{$_} });
988 $v = '\N' if length($v) == 0;
989 } elsif ( ! ref $i->{$_} ) {
998 warn "export $export_path ", -s $export_path, " bytes\n";
1001 my ( $code_depends, $code_description );
1005 warn "# test_changed ",dump( $test_changed );
1006 my $c = { map { $_ => 1 } @columns };
1007 my @added_columns = sort grep { ! $c->{$_} } keys %$test_changed;
1008 unshift @columns, @added_columns;
1010 warn "# sorted_items ", $#$sorted_items + 1, " offset $offset limit $limit order $sort";
1013 my $tmp = $code; $tmp =~ s/\$row->\{(['"]?)([\w\s]+)\1/$depends_on->{$2}++/gse;
1014 warn "# depends_on ",dump $depends_on;
1016 my $test_added = Storable::dclone $test_changed;
1017 delete $test_added->{$_} foreach keys %$depends_on;
1019 $code_depends = $self->param('code_depends')
1020 || join(',', keys %$depends_on);
1022 $code_description = $self->param('code_description') ||
1023 join(',', keys %$test_added);
1025 $code_depends ||= $code_description; # self-modifing
1026 if ( ! $code_depends && $out ) {
1027 $code_depends = $key;
1028 $code_description = $value;
1031 warn "# test_changed ",dump( $test_changed, $code_depends, $code_description );
1041 sorted => $sorted_items,
1042 columns => [ @columns ],
1043 rows => $#$filtered + 1,
1044 numeric => { map { $_, $self->_is_numeric($_) } @columns },
1045 unique => { map { $_, $self->_is_unique( $_) } @columns },
1046 filters => $self->_current_filters,
1048 cols_changed => $commit ? $commit_changed : $test_changed,
1049 code_depends => $code_depends,
1050 code_description => $code_description,
1051 code_path => $code_path,
1060 $self->session('order', $self->param('order'));
1061 $self->session('sort', $self->param('sort'));
1062 return $self->redirect_to('/data/items');
1066 my ( $self, $name ) = @_;
1068 my $stats = $self->_loaded( 'stats' );
1070 # sort facet numerically if more >50% elements are numeric
1071 my $count = $stats->{$name}->{count};
1072 $count -= $stats->{$name}->{empty} if defined $stats->{$name}->{empty};
1073 defined $stats->{$name}->{numeric} &&
1074 $stats->{$name}->{numeric} > $count / 2;
1078 my ( $self, $name ) = @_;
1079 my $stats = $self->_loaded( 'stats' );
1080 defined $stats->{$name}->{unique};
1083 sub _remove_filter {
1084 my ($self,$name) = @_;
1085 warn "_remove_filter $name\n";
1087 my $path = $self->session('path');
1089 delete $filters->{$name};
1090 delete $loaded->{$path}->{filters}->{$name};
1091 warn "filters left: ", keys %{ $loaded->{$path}->{filters} };
1094 grep { /\Q$name\E/ }
1095 keys %{ $loaded->{$path}->{filtered} }
1097 delete $loaded->{$path}->{filtered}->{$_};
1098 warn "remove filtered cache $_";
1105 my $path = $self->session('path') || return $self->redirect_to( '/data/index' );
1107 if ( my $name = $self->param('remove') ) {
1108 $self->_remove_filter( $name );
1109 return $self->redirect_to( '/data/items' );
1113 my $name = $self->param('name') || die "no name";
1115 my $all = $self->_param_scalar('all', 1);
1116 my $data = $self->_loaded('data');
1118 my $filters = $self->_current_filters;
1119 my $all_filters = __all_filters( keys %$filters,$self->session('order') );
1120 my $filtered = $loaded->{$path}->{filtered}->{$all_filters}
1121 if defined $loaded->{$path}->{filtered}->{$all_filters};
1123 if ( ! $filtered || $all ) {
1124 $filtered = [ 0 .. $#{ $data->{items} } ];
1125 warn "filter all values\n";
1127 warn "filter using $all_filters\n";
1130 foreach my $i ( @$filtered ) {
1131 my $item = $data->{items}->[$i];
1132 if ( ! exists $item->{$name} || ! defined $item->{$name} ) {
1133 $facet->{ _missing }++;
1134 } elsif ( ref $item->{$name} eq 'ARRAY' ) {
1135 $facet->{$_}++ foreach @{ $item->{$name} };
1137 $facet->{ $item->{$name} }++;
1141 my $checked_values = $self->_checked( @{ $filters->{$name} } ) if defined $filters->{$name};
1143 if ( my $code = $self->param('code') ) {
1145 foreach my $value ( keys %$facet ) {
1146 my $count = $facet->{$value};
1147 my $checked = $checked_values->{$value};
1151 warn "ERROR: $@\n$code\n";
1153 } elsif ( $checked != $checked_values->{$value} ) {
1154 warn "checked $value $count -> $checked\n";
1155 $checked_values->{$value} = $checked;
1158 warn "out ",dump( $out );
1159 $self->stash( out => $out );
1162 # warn "# facet $name ",dump $facet;
1166 : defined $filters->{$name} ? @{ $filters->{$name} }
1169 my $numeric = $self->_is_numeric($name);
1171 my $sort = $self->param('sort');
1172 # sort numeric facets with more than 5 values ascending
1173 $sort ||= $numeric && $#facet_names > 4 ? 'a' : 'c';
1175 @facet_names = sort {
1177 if ( $sort eq 'a' ) {
1178 $result = $numeric ? $a <=> $b : lc $a cmp lc $b;
1179 } elsif ( $sort eq 'd' ) {
1180 $result = $numeric ? $b <=> $a : lc $b cmp lc $a;
1181 } elsif ( $sort eq 'c' ) {
1182 $result = ( $facet->{$b} || -1 ) <=> ( $facet->{$a} || -1 )
1184 warn "unknown sort: $sort";
1185 $result = $a cmp $b;
1187 $result = $a cmp $b unless defined $result; # FIXME cludge for numeric facets with invalid data
1191 $self->render( name => $name, facet => $facet, checked => $checked_values,
1192 facet_names => \@facet_names, sort => $sort, numeric => $numeric,
1197 sub __invalidate_path_column {
1198 my ( $path, $name ) = @_;
1200 if ( defined $loaded->{$path}->{sorted}->{$name} ) {
1201 delete $loaded->{$path}->{sorted}->{$name};
1202 warn "# invalidate $path sorted $name\n";
1205 foreach ( grep { m/\Q$name\E/ } keys %{ $loaded->{$path}->{filtered} } ) {
1206 delete $loaded->{$path}->{filtered}->{$_};
1207 warn "# invalidate $path filtered $_\n";
1211 sub __path_modified {
1212 my ( $path, $value ) = @_;
1213 $value = 1 unless defined $value;
1215 $loaded->{$path}->{modified} = $value;
1217 warn "# __path_modified $path $value\n";
1220 sub __path_rebuild_stats { $loaded->{ $_[0] }->{rebuild_stats} = 1 };
1223 my ($self,$change) = @_;
1225 my $change_path = $self->_permanent_path( 'changes' );
1226 mkdir $change_path unless -d $change_path;
1227 $change_path .= '/' . $change->{time};
1228 store $change, $change_path;
1229 utime $change->{time}, $change->{time}, $change_path;
1230 warn "_save_change $change_path ", dump($change);
1235 my $new_content = $self->param('new_content');
1236 $new_content ||= $self->param('content'); # backward compatibility with old actions
1238 my $i = $self->param('_row_id');
1239 die "invalid _row_id ",dump($i) unless $i =~ m/^\d+$/;
1240 my $path = $self->param('path') || die "no path";
1241 my $name = $self->param('name') || die "no name";
1242 my $status = 200; # 200 = OK, 201 = Created
1244 my $data = $self->_loaded('data');
1246 if ( defined $loaded->{$path}->{data}->{items}->[$i] ) {
1247 $new_content =~ s/^\s+//s;
1248 $new_content =~ s/\s+$//s;
1250 if ( $new_content =~ /\xB6/ ) { # para
1251 $v = [ split(/\s*\xB6\s*/, $new_content) ];
1253 $v = [ $new_content ];
1256 my $old = dump $loaded->{$path}->{data}->{items}->[$i]->{$name};
1259 && ! ( $old eq 'undef' && length($new_content) == 0 ) # new value empty, previous undef
1261 $self->_save_change({
1265 old => $loaded->{$path}->{data}->{items}->[$i]->{$name},
1267 time => $self->param('time') || time(),
1268 user => $self->param('user') || $ENV{'LOGNAME'},
1270 map { $_ => $loaded->{$path}->{data}->{items}->[$i]->{$_}->[0] }
1271 grep { defined $loaded->{$path}->{stats}->{$_}->{unique} }
1272 keys %{ $loaded->{$path}->{stats} }
1276 warn "# change $path $i $old -> $new\n";
1277 $loaded->{$path}->{data}->{items}->[$i]->{$name} = $v;
1279 __invalidate_path_column( $path, $name );
1281 $status = 201; # created
1282 __path_rebuild_stats( $path );
1284 $new_content = join("\xB6",@$v);
1287 warn "# unchanged $path $i $old\n";
1291 $new_content = "$path $i $name doesn't exist\n";
1295 warn "# edit $status ", dump $new_content;
1299 new_content => scalar $new_content,
1306 my $path = $self->_param_or_session('path');
1307 my $dump_path = $self->_save( $path );
1308 __path_modified( $path, 0 );
1310 return $self->redirect_to( '/data/items' );
1316 my $dir = $self->app->home->rel_file('public');
1318 if ( my $import = $self->param('import') ) {
1320 if ( $import =~ m{/filter\.(.+?)\..+} ) {
1321 my $name = $self->_column_from_unac( $1 );
1323 my @vals = map { chomp; $_ }
1324 read_file "$dir/export/$import", binmode => ':utf8';
1326 $self->_remove_filter( $name );
1327 $self->_filter_on_data( $name, @vals );
1328 $self->session( 'offset' => 0 );
1329 return $self->redirect_to('/data/items');
1331 warn "UNKNOWN IMPORT $import";
1335 if ( my $remove = $self->param('remove') ) {
1336 my $path = "$dir/export/$remove";
1337 unlink $path if -e $path;
1339 unlink $path if -e $path;
1342 my $path = $self->_export_path || return $self->redirect_to('/data/index');
1344 my @files = grep { ! /\.png$/ } glob "$path/*";
1345 my $mtime = { map { $_ => (stat($_))[9] } @files };
1346 @files = sort { $mtime->{$b} <=> $mtime->{$a} } @files;
1347 $self->render( export => [ @files ] );
1350 sub __loaded_paths {
1352 grep { defined $loaded->{$_}->{data} }
1358 $self->stash( reload => 1 );
1360 # $self->_load_path( $self->param('path') );
1361 $self->redirect_to('/data/load?path=' . $self->param('path') );
1366 my $path = $self->param('path');
1367 if ( $path =~ m{^/tmp/mojo_facets\.} ) {
1369 warn "# unlink $path";
1370 if ( my $name = $self->param('name') ) {
1371 delete $loaded->{$name};
1372 warn "# remove $name from memory";
1375 warn "WARNING: $path unlink ignored";
1377 return if $self->stash('reload');
1378 return $self->redirect_to( '/data/load' );