bc-zdump2abbrev.pl

#!/bin/perl

# For each time zone in the output of:
# find /usr/share/zoneinfo -type f | xargs -n 1 zdump -v
# find the latest (and thus most likely current) abbreviation

# Note that abbreviations are not unique (tzdata uses EDT for both the
# United States and Australia's eastern daylight timezones), but this
# should help me canonize time zone names (maybe)

require "/usr/local/lib/bclib.pl";

my(%latest,%abbrev,%geonames);

# limit to those timezones actually found in geonames
my(%geonames) = list2hash(split(/\n/,read_file("$bclib{githome}/ASTRO/tz-names-in-geonames.txt")));

open(A,"bzcat $bclib{githome}/ASTRO/zdump.txt.srt.bz2|");

while (<A>) {

  # we need: filename, UTC year, abbrev
  my(@data) = split(/\s+/,$_);

  my($fname, $utcy, $abbrev) = @data[0,5,13];
  $fname=~s%/usr/share/zoneinfo/%%;

  unless ($geonames{$fname}) {next;}

  # if I already have later abbrev for this tz, ignore
  if ($latest{$fname} > $utcy) {next;}

  $latest{$fname} = $utcy;
  $abbrev{$fname} = $abbrev;
}

for $i (keys %abbrev) {print "$i $abbrev{$i}\n";}