Collapse po files

gettext-extract.rb

#!/usr/bin/ruby

# Collapses Gettext po files. for easy diff'ing
# Drops comments, and collapses quotes into one line.
# Example of what a po-file is: https://docs.transifex.com/formats/gettext
# Expanding the script by Raphaël Kolm at https://medium.com/@rkz_io/making-a-clutter-free-git-diff-on-gettext-translation-files-9c0c1bb1d8aa
# Sample output:
# 00001 msgid "is english def"
# 00001 msgstr "est anglais"
# 00002 msgid "good"
# 00002 msgstr "bien"
# 00003 msgstr "nice job"
# 00003 msgid "bon travail"
# Gettext: https://www.gnu.org/software/gettext/

lines = []
f = File.open(ARGV.first, "r:UTF-8") # you might be working with other kinds of encodings.
preamble_passed = false
f.readlines.each do |line|
  next unless preamble_passed || line =~ /^msg/
  preamble_passed = true
  lines << line if  line =~ /^\"|^msg/
end
f.close

collapsed_lines = []

collect_quoted = false

lines.each do |l|
  if l =~ /^msg/
    collect_quoted = false
    collapsed_lines << l.chomp
  else
    collect_quoted = true
    collapsed_lines.last << l.chomp
  end
end

translations = {}
(collapsed_lines.length/2).times do |n|
   key = collapsed_lines[2*n]
   value = collapsed_lines[2*n+1]
   translations[key] = value
end

index = 1
translations.sort.each do |k,v|
   puts '%05d %s' % [index, k]
   puts '%05d %s' % [index, v]
   index += 1
end