use base 'Mojolicious::Controller';
-use Data::Dump qw(dump);
+#use Data::Dump qw(dump); # broken with Mojo::JSON, see https://rt.cpan.org/Public/Bug/Display.html?id=86592
+use Data::Dumper;
+use subs 'dump';
+sub dump { Dumper(@_) };
+
use File::Slurp;
use Encode;
use locale;
use Storable;
use Time::HiRes qw(time);
use File::Path qw(mkpath);
-use Text::Unaccent::PurePerl;
+use Text::Unaccent;
use Digest::MD5;
+use Statistics::Descriptive;
+
+our $imports;
+foreach my $module ( glob('lib/MojoFacets/Import/*.pm') ) {
+ $module =~ s{lib/(\w+)/(\w+)/(.*)\.pm}{$1::$2::$3};
+ eval "use $module";
+ die "$module: $!" if $!;
+ my ( $ext, $priority ) = $module->ext;
+ $imports->{$priority || 'file'}->{$ext} = $module;
+ warn "# import $ext $module\n";
+}
+
+warn "# import loaded ",dump( $imports );
+
+sub import_module {
+ my $full_path = shift;
+
+# warn "# import_module $full_path\n";
+
+ return if $full_path =~ m/\.columns$/;
+
+ foreach my $ext ( keys %{ $imports->{file} } ) {
+ if ( -f $full_path && $full_path =~ m/$ext/i ) {
+ return $imports->{file}->{$ext};
+ last;
+ }
+ }
-use MojoFacets::Import::File;
-use MojoFacets::Import::HTMLTable;
-use MojoFacets::Import::CSV;
-use MojoFacets::Import::CouchDB;
-use MojoFacets::Import::SQL;
+ foreach my $ext ( keys %{ $imports->{directory} } ) {
+ if ( -f $full_path && $full_path =~ m/$ext/i ) {
+ return $imports->{directory}->{$ext};
+ last;
+ }
+ }
+}
our $loaded;
our $filters;
my @files;
my $changes;
+
find( sub {
my $file = $File::Find::name;
- if ( -f $file && $file =~ m/\.(js(on)?|txt)$/ ) {
- $file =~ s/$data_dir\/*//;
- push @files, $file;
- } elsif ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
+ if ( -f $file && $file =~ m/([^\/]+)\.changes\/(\d+\.\d+.+)/ ) {
push @{ $changes->{$1} }, $2
- } elsif ( -d $file && $file =~ m/\.html$/ ) {
- $file =~ s/$data_dir\/*//;
- push @files, $file;
- } elsif ( -f $file && $file =~ m/\.(csv|storabe|couchdb|sql)$/i ) {
+ } elsif ( import_module( $file ) ) {
$file =~ s/$data_dir\/*//;
push @files, $file;
} else {
no warnings qw(uninitialized); # mtime
@files = sort { $loaded->{$b}->{mtime} <=> $loaded->{$a}->{mtime} || lc $a cmp lc $b } @files,
- grep { defined $loaded->{$_}->{generated} } keys %$loaded;
+ grep { defined $loaded->{$_}->{generated} } keys %$loaded;
+
my $size;
$size->{$_} = -s "$data_dir/$_" foreach @files;
}
my $data;
- if ( -f $full_path ) {
- if ( $full_path =~ m/.csv/i ) {
- $data = MojoFacets::Import::CSV->new( full_path => $full_path )->data;
- } elsif ( $full_path =~ m/.sql/i ) {
- $data = MojoFacets::Import::SQL->new( full_path => $full_path )->data;
- } elsif ( $full_path =~ m/.couchdb/i ) {
- $data = MojoFacets::Import::CouchDB->new( full_path => $full_path )->data;
- } elsif ( $full_path =~ m/.storable/ ) {
- warn "open $full_path ", -s $full_path, " bytes";
- open(my $pipe, "<", $full_path) || die $!;
- while ( my $o = eval { Storable::fd_retrieve $pipe } ) {
- if ( defined $o->{item} ) {
- push @{ $data->{items} }, $o->{item};
- } else {
- warn "SKIP ",dump($o);
- }
- }
- close($pipe);
- warn "loaded ", $#{ $data->{items} } + 1, " items from $full_path\n";
- $data->{generated}++;
- } else {
- $data = MojoFacets::Import::File->new( full_path => $full_path, path => $path )->data;
- }
- } elsif ( -d $full_path && $full_path =~ m/.html/ ) {
- $data = MojoFacets::Import::HTMLTable->new( dir => $full_path )->data;
+ if ( my $module = import_module( $full_path ) ) {
+ $data = $module->new( full_path => $full_path )->data;
} else {
die "can't load $full_path";
}
sub load {
my $self = shift;
- my $path = $self->param('path') || return $self->redirect_to( '/data/index' );
-
- my @paths = $self->param('paths');
+ my @paths = @{ $self->every_param('paths') };
warn "# paths ", dump @paths;
foreach my $p ( keys %$loaded ) {
$self->_load_path( $_ ) foreach @paths;
+ my $path = $self->param('path') || $self->session('path') || $paths[0] || $self->redirect_to('/data/index');
+
warn "# path $path\n";
$self->_load_path( $path );
sub __unac {
my $n = shift;
- $n = unac_string($n);
+ $n = unac_string('utf-8',$n);
$n =~ s/\W+/_/g;
return $n;
}
sub _param_array {
my ($self,$name) = @_;
- my @array = $self->param($name);
+ my @array = @{ $self->every_param($name) };
my $path = $self->session('path');
if ( @array ) {
if ( ! defined $scalar ) {
$scalar = $default;
- die "no default for $name" unless defined $scalar;
- $self->session($name => $scalar);
+ if ( defined $scalar ) {
+ $self->session($name => $scalar);
+ } else {
+ warn "no default for $name";
+ }
}
warn "# _param_scalar $name ",dump $scalar;
my $self = shift;
my $name = $self->param('filter_name') || die "name?";
- my @vals = $self->param('filter_vals');
+ my @vals = @{ $self->every_param('filter_vals') };
$self->_remove_filter( $name );
if ( @vals ) {
our ($out, $key,$value);
+our $lookup_path_col;
+our $on;
+
+sub __commit_begin {
+ warn "__commit_begin";
+ $lookup_path_col = undef;
+ $on = undef;
+}
+
+sub __commit_end {
+ warn "__commit_end";
+ $lookup_path_col = undef; # cleanup memory
+ $on = undef;
+}
+
+sub lookup {
+ warn "# lookup ",dump @_;
+ my ( $vals, $on_path, $on_col, $code, $stat_code ) = @_;
+ die "code is not sub{ ... } but ", dump $code unless ref $code eq 'CODE';
+
+ if ( ! exists $loaded->{$on_path} ) {
+ my @possible_paths = grep { /\Q$on_path\E/ } keys %$loaded;
+ die "more than one dataset available for '$on_path' ",dump @possible_paths if $#possible_paths > 0;
+ $on_path = shift @possible_paths;
+ warn "## fuzzy selected path $on_path";
+ }
+
+ my $items = $loaded->{$on_path}->{data}->{items} || die "no items for $on_path";
+
+ if ( ! exists $lookup_path_col->{$on_path}->{$on_col} ) {
+ warn "create lookup_path_col $on_path $on_col";
+ foreach my $i ( 0 .. $#$items ) {
+ my $item = $items->[$i];
+ if ( exists $item->{$on_col} ) {
+ if ( ref $item->{$on_col} eq 'ARRAY' ) {
+ foreach my $v ( @{ $item->{$on_col} } ) {
+ push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
+ }
+ } elsif ( ! ref $item->{$on_col} ) { # scalar
+ my $v = $item->{$on_col};
+ push @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} }, $i;
+ } else {
+ die "unknown type of ",dump $item->{$on_col};
+ }
+ }
+ }
+ warn "XXX ",dump $lookup_path_col->{$on_path}->{$on_col} if $ENV{DEBUG};
+ }
+
+ my $stat;
+ $stat = Statistics::Descriptive::Full->new() if $stat_code;
+
+ foreach my $v ( ref $vals eq 'ARRAY' ? @$vals : ( $vals ) ) {
+ foreach my $i ( @{ $lookup_path_col->{$on_path}->{$on_col}->{$v} } ) {
+ $on = $items->[$i];
+ warn "XXX lookup code $v $i ",dump $on if $ENV{DEBUG};
+ $code->($stat);
+ }
+ $stat_code->( $stat ) if $stat_code;
+ }
+}
+
sub __commit_path_code {
my ( $path, $i, $code, $commit_changed ) = @_;
$code =~ s{\r}{}gs;
$code =~ s{\n+$}{\n}s;
+ # XXX convert @row->{foo} into @{$row->{foo}}
+ $code =~ s|\@(row->{[^}]+})|\@{\$$1}|gs;
+
my $commit = $self->param('commit');
my $test = $self->param('test');
my $commit_changed;
+ __commit_begin;
if ( $code && ( $test || $commit ) ) {
# XXX find columns used in code snippet and show them to user
my $code_path = $self->app->home->rel_dir('public') . "/code";
if ( $commit ) {
+ __path_modified( $path, 'commit' );
+
warn "# commit on ", $#$filtered + 1, " items:\n$code\n";
( $key, $value, $out ) = ( 'key', 'value' );
foreach ( 0 .. $#$filtered ) {
} # test?
+ __commit_end;
+
$self->render(
order => $order,
offset => $offset,
keys %$loaded;
}
+sub reload {
+ my $self = shift;
+ $self->stash( reload => 1 );
+ $self->remove;
+# $self->_load_path( $self->param('path') );
+ $self->redirect_to('/data/load?path=' . $self->param('path') );
+}
+
sub remove {
my $self = shift;
my $path = $self->param('path');
} else {
warn "WARNING: $path unlink ignored";
}
+ return if $self->stash('reload');
return $self->redirect_to( '/data/load' );
}