Preprocess the source files for faster loading

lib/DDG/Goodie/NameDays.pm

@@ -27,80 +27,8 @@ triggers any => "name day", "name days", "nameday", "namedays", "imieniny", "jme
 
 
 # Load the data file
-my @names = (); # Names indexed by day
-my %dates = (); # Days indexed by name
-
-# File format: 366 lines (one for each day of a year).
-# Each line contains names separated with a space.
-# A line may contain the names in genitive case or variations of a name.
-# These variations are placed after vertical bar character (|); they are
-# not shown when searching for this day, but you can search for them.
-
-sub load_days_file {
-    my $file_name = shift();
-    
-    my @lines = share($file_name)->slurp(iomode => '<:encoding(UTF-8)');
-    
-    $file_name =~ s/\.txt$//;
-    $file_name =~ s/_/ /g;
-    
-    die "The text file must include 366 lines" unless scalar(@lines) == 366;
-
-    my $day_of_year = 1;
-
-    # Read names for each day and add them to the hash
-    for (@lines) {
-        # Add all names, including the names after vertical bar
-        my $names_for_date = lc($_);
-        $names_for_date =~ s/\|/ /;
-        for my $name (split(' ', $names_for_date)) {
-            push(@{$dates{$name}}, $file_name . '|' . $day_of_year);
-        }
-    
-        # Remove the names after vertical bar (|)
-        chomp;
-        s/\s*\|.*$//;
-        if ($_) {
-            $names[$day_of_year - 1] .= "; " if ($names[$day_of_year - 1]);
-            $names[$day_of_year - 1] .= $file_name . ': ' . $_;
-        }
-        
-        # Advance to the next day
-        $day_of_year++;
-    }
-}
-
-sub finish_loading {
-    # Convert the dates to string
-    for (keys %dates) {
-        # Group the dates by country
-        my %dates_by_country = ();
-        foreach (@{$dates{$_}}) {
-            die 'Internal error' unless /^(.*?)\|(\d+)$/;
-            # Any leap year here, because the text file includes February, 29
-            my $d = DateTime->from_day_of_year(year => 2000, day_of_year => $2);
-            if (exists $dates_by_country{$1}) {
-                $dates_by_country{$1} .= ', ';
-            }
-            $dates_by_country{$1} .= $d->strftime('%e %b');
-        }
-        
-        # Convert to string
-        my $res = '';
-        foreach (sort keys %dates_by_country) {
-            $res .= $_ . ': ' . $dates_by_country{$_} . "; ";
-        }
-        
-        $res =~ s/; $//;
-        $dates{$_} = $res;
-    }
-}
-
-
-load_days_file('Czech_Republic.txt');
-load_days_file('Hungary.txt');
-load_days_file('Poland.txt');
-finish_loading();
+my @names = share('preprocessed_names.txt')->slurp(iomode => '<:encoding(UTF-8)', chomp => 1); # Names indexed by day
+my %dates = share('preprocessed_dates.txt')->slurp(iomode => '<:encoding(UTF-8)', chomp => 1); # Days indexed by name
 
 
 sub parse_other_date_formats {

share/goodie/name_days/preprocess.pl

@@ -0,0 +1,101 @@
+#!/usr/bin/perl
+# Preprocess the source files and output data structures for slurping
+
+use strict;
+use DateTime;
+use Path::Class;
+
+# Load the data file
+my @names = (); # Names indexed by day
+my %dates = (); # Days indexed by name
+
+# File format: 366 lines (one for each day of a year).
+# Each line contains names separated with a space.
+# A line may contain the names in genitive case or variations of a name.
+# These variations are placed after vertical bar character (|); they are
+# not shown when searching for this day, but you can search for them.
+
+sub load_days_file {
+    my $file_name = shift();
+    
+    my @lines = file($file_name)->slurp(iomode => '<:encoding(UTF-8)');
+    
+    $file_name =~ s/\.txt$//;
+    $file_name =~ s/_/ /g;
+    
+    die "The text file must include 366 lines" unless scalar(@lines) == 366;
+
+    my $day_of_year = 1;
+
+    # Read names for each day and add them to the hash
+    for (@lines) {
+        # Add all names, including the names after vertical bar
+        my $names_for_date = lc($_);
+        $names_for_date =~ s/\|/ /;
+        for my $name (split(' ', $names_for_date)) {
+            push(@{$dates{$name}}, $file_name . '|' . $day_of_year);
+        }
+    
+        # Remove the names after vertical bar (|)
+        chomp;
+        s/\s*\|.*$//;
+        if ($_) {
+            $names[$day_of_year - 1] .= "; " if ($names[$day_of_year - 1]);
+            $names[$day_of_year - 1] .= $file_name . ': ' . $_;
+        }
+        
+        # Advance to the next day
+        $day_of_year++;
+    }
+}
+
+sub finish_loading {
+    # Convert the dates to string
+    for (keys %dates) {
+        # Group the dates by country
+        my %dates_by_country = ();
+        foreach (@{$dates{$_}}) {
+            die 'Internal error' unless /^(.*?)\|(\d+)$/;
+            # Any leap year here, because the text file includes February, 29
+            my $d = DateTime->from_day_of_year(year => 2000, day_of_year => $2);
+            if (exists $dates_by_country{$1}) {
+                $dates_by_country{$1} .= ', ';
+            }
+            $dates_by_country{$1} .= $d->strftime('%e %b');
+        }
+        
+        # Convert to string
+        my $res = '';
+        foreach (sort keys %dates_by_country) {
+            $res .= $_ . ': ' . $dates_by_country{$_} . "; ";
+        }
+        
+        $res =~ s/; $//;
+        $dates{$_} = $res;
+    }
+}
+
+
+# Load the source files into %dates and @names
+
+load_days_file('Czech_Republic.txt');
+load_days_file('Hungary.txt');
+load_days_file('Poland.txt');
+finish_loading();
+
+
+# Output the array and the hash.
+# Spew does not work for hashes, so use two loops here.
+
+open(my $fd, '>:encoding(UTF-8)', 'preprocessed_dates.txt') or die;
+
+for (keys %dates) {
+    print($fd $_ . "\n" . $dates{$_} . "\n");
+}
+
+
+open(my $fn, '>:encoding(UTF-8)', 'preprocessed_names.txt') or die;
+
+for (@names) {
+    print($fn $_ . "\n");
+}