#! /usr/bin/python2 -O
# -*- coding: iso-8859-1 -*-

# This script iterates through the html, lib, and template directories
# looking for php scripts that contain __() functions.
# It creates/appends to the corresponding 'xxx.po' file in the
# 'lang' subdirectory and places the i18n strings into the file
# in the proper format.
#
# usage: translation_tool [-v] [-f]
#        -v: Verbose, print duplicate terms that could be moved to common_po
#        -f: Force, overwrite existing translated files, otherwise append
#

import getopt
import os
import re
import sys

scriptdirs = ['html', 'lib', 'template']

try:
	opts, args =getopt.getopt(sys.argv[1:], 'vf')
except getopt.GetoptError, e:
	print 'error: %s' % str(e)
	raise SystemExit

translator_name = raw_input("What is your full name? ")
translator_email = raw_input("What is your email address? ")
trans_native = raw_input("What is the native name of the language? ")
trans_eng = raw_input("What is the English name of the language? ")
trans_abbrv = raw_input("What is the ISO 639-1 Alpha-2 abbreviation for the language? ")

if len(trans_abbrv) != 2:
	print "Must use 2 character abbreviation"
	raise SystemExit

INC_HEADER = """\
<?php
# INSTRUCTIONS TO TRANSLATORS
#
# This file contains the i18n translations for a subset of the
# Arch Linux User Community Repository (AUR).  This is a PHP
# script, and as such, you MUST pay great attention to the syntax.
# If your text contains any double-quotes ("), you MUST escape
# them with a backslash (\).
#
# %s (%s) translation
# Translator: %s <%s>

global $_t;


""" % (trans_eng, trans_native, translator_name, translator_email)


print_dupes = '-v' in sys.argv
force = '-f' in sys.argv

up = re.compile('_\(\s*"(([^"]|(?<=\\\\)["])+)"')

lang = {}

current_dir = os.getcwd()

pofile = '%s.po' % trans_abbrv

# Iterate through various places where the php files might be.
#
for dir in scriptdirs:
	dir = '../%s' % dir

	if os.path.exists(dir):
		# Find all the PHP files in the current directory.
		files = [x for x in os.listdir(dir)
					if (x[-4:] == '.inc' and x[-7:] != '.po')
					or x[-6:] == '.class'
					or x[-4:] == '.php'
					or x[-6:] == '.phtml'
				]
		os.chdir(dir)

		for file in files:
			f = open(file,'r')
			lines = f.readlines()
			f.close()

			# Parse files.
			print "Parsing %s..." % file
			for line in lines:
				match = up.search(line)
				while match:
					term = match.group(1).replace('\\"','"')
					if print_dupes:
						if term in lang.keys():
							print 'Multiple use of "%s"' % term

					lang[term] = 1
					line = line[match.end(1):]
					match = up.search(line)

		os.chdir(current_dir)

# Now generate the .po file if it doesn't already exist.
# If it does exist, only append new stuff to the end.
# If the 'force' option is passed, just overwrite.

print """
You will now be prompted for all needed translations.
Please translate the requested lines, hitting [enter]
goes to the next one. You may stop at any time using
ctrl+c, and pick up where you left off by running
translation_tool again.

If there are escapes in the original English, you may
need to include them in your translation. The
following is a list of escapes and what they do:
%h - HTML code inserted at run-time
%s - Nontranslated string inserted at run-time (such as username)
\\" - A double quote (")

When you have finished your translation, make a tarball
of the .po file and send it to aur-dev@archlinux.org
for inclusion in the AUR.

By submitting a translation, you are implying that you
are also willing to maintain it. When there are
new strings to be translated, you will be contacted.
****************************************************
"""

os.chdir(current_dir)
if force:
	# Just overwrite any existing files.
	# NOT RECOMMENDED! OVERWRITES ALL OTHER LANGUAGE SUPPORT
	#
	print "Generating %s..." % pofile

	f = open(pofile,'w')
	f.write(INC_HEADER)

	for term in lang:
		f.write("\n")
		trans = raw_input("\n%s\n= " % term)
		f.write('$_t["%s"] = "%s";\n' % (term, trans))
	f.write("\n");
	f.close()
else:
	# Leave existing file intact. Only append on new terms.
	mapre = re.compile('^\$_t\["(.*)"\].*$')

	got_match = False
	print "Updating %s..." % pofile
	try:
		f = open(pofile, 'r')
		new_file = 0
	except:
		new_file = 1
	if not new_file:
		contents = f.readlines()
		f.close()

		# Strip beginning/ending empty lines
		while contents[0] == '':
			del contents[0]
		while contents[-1] in ['', "\n", "?>", "?>\n", "\n?>"]:
			del contents[-1]
					
		f = open(pofile,'w')
		f.write("".join(contents))
		f.write("\n");
		f.close()
	else:
		f = open(pofile,'w')
		f.write(INC_HEADER)
		f.write('\n')
		f.close()

	# Read in file contents so we can hash what already exists
	try:
		f = open(pofile, 'r')
		new_file = 0
	except:
		new_file = 1

	existing_terms = []
	if not new_file:
		contents = f.readlines()
		f.close()

		# Strip beginning/ending empty lines
		while contents[0] == '':
			del contents[0]
		while contents[-1] in ['', "\n", "?>", "?>\n", "\n?>"]:
			del contents[-1]

		# Collect existing terms
		for line in contents:
			match = mapre.search(line)
			if match:
				existing_terms.append(match.group(1))

	# Append any new terms to EOF
	f = open(pofile, 'w')
	if not new_file:
		f.write("".join(contents))
	else:
		f.write(INC_HEADER)

	for term in lang.keys():
		if term not in existing_terms:
			f.write("\n");
			trans = raw_input("\n%s\n= " % term)
			f.write('$_t["%s"] = "%s";\n' % (term, trans))
	f.write("\n");
	f.close()