gergely/astrognome
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Redesign geodata.xml so it also stores country names

Browse Source
This commit is contained in:
Gergely Polonkai 2014-08-16 23:48:04 +02:00
parent 4a52e037af
commit 15737de254
1 changed files with 16 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
data/geonames/geonames_process.pl
View File

@ -16,25 +16,34 @@ while (<TIMEZONES>) {
} }
close(TIMEZONES); close(TIMEZONES);
my $xml_file = IO::File->new('>geodata.xml');
my $writer = XML::Writer->new(OUTPUT => $xml_file, NEWLINES => 0);
$writer->xmlDecl('utf-8');
$writer->startTag('geodata');
$writer->startTag('countries');
open(COUNTRIES, 'countryInfo.txt') or die("Cannot open countryInfo.txt: $!\n"); open(COUNTRIES, 'countryInfo.txt') or die("Cannot open countryInfo.txt: $!\n");
while (<COUNTRIES>) { while (<COUNTRIES>) {
my ($country_code, $iso3, $iso_numeric, $fips, $name, $capital, $area, $population, $continent, $tld, $currency_code, $currency_name, $phone, $postal_code_format, $postal_code_regex, $languages, $geonameid, $neighbours, $equivalent_fips_code) = split(/\t/, $_); my ($country_code, $iso3, $iso_numeric, $fips, $name, $capital, $area, $population, $continent, $tld, $currency_code, $currency_name, $phone, $postal_code_format, $postal_code_regex, $languages, $geonameid, $neighbours, $equivalent_fips_code) = split(/\t/, $_);
next if ($country_code !~ /^[A-Z]{2}$/); next if ($country_code !~ /^[A-Z]{2}$/);
$writer->emptyTag('c',
'n' => $name,
'c' => $country_code,
);
if ($country_code =~ /^[A-Z]{2}$/) { if ($country_code =~ /^[A-Z]{2}$/) {
$countries{$country_code} = $name; $countries{$country_code} = $name;
} }
} }
close(COUNTRIES); close(COUNTRIES);
$writer->endTag('countries');
$writer->startTag('places');
open(GEONAMES, "cities.txt") or die("Cannot open cities.txt: $!\n"); open(GEONAMES, "cities.txt") or die("Cannot open cities.txt: $!\n");
my $xml_file = IO::File->new('>geodata.xml');
my $writer = XML::Writer->new(OUTPUT => $xml_file, NEWLINES => 0);
$writer->xmlDecl('utf-8');
$writer->startTag('geodata');
while (<GEONAMES>) { while (<GEONAMES>) {
chomp($_); chomp($_);
my ($country_code, $name, $latitude, $longitude, $elevation, $timezone) = split(/\t/, $_); my ($country_code, $name, $latitude, $longitude, $elevation, $timezone) = split(/\t/, $_);
@ -62,6 +71,7 @@ while (<GEONAMES>) {
print $., "\n" if ($. % 19083 == 0); print $., "\n" if ($. % 19083 == 0);
} }
$writer->endTag('places');
$writer->endTag('geodata'); $writer->endTag('geodata');
$writer->end(); $writer->end();
$xml_file->close(); $xml_file->close();