gramps/po/check_po

#! /usr/bin/env python
#
# check_po - a gramps tool to check validity of po files
#
# Copyright (C) 2006-2006  Kees Bakker
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA

# $Id:$

import sys
import re

all_total = {}
all_fuzzy = {}
all_untranslated = {}
all_percent_s = {}
all_named_s = {}
all_bnamed_s = {}
all_context = {}
all_coverage = {}
all_template_coverage = {}

def strip_quotes(st):
	st = st.strip()
	if len(st) >= 2 and st[0] == '"' and st[len(st)-1] == '"':
		st = st.strip()[1:-1]
	return st

class Msgid:
	fuzzy_pat = re.compile( 'fuzzy' )
	tips_xml_pat = re.compile( r'tips\.xml' )
	def __init__( self, msgnr, lineno ):
		self._msgid = []
		self._msgstr = []
		self._cmnt = []
		self.nr = msgnr
		self.lineno = lineno
		self.is_fuzzy = 0
		self.has_sfmt_mismatch = 0
		self.has_named_sfmt_mismatch = 0
		self.has_fmt_missing_sd = 0
		self.has_context_error = 0
		self.has_named_fmt_mismatch = 0
		self.has_xml_error = 0
		self.has_lastchar_error = 0

	def diag( self ):
		if 1:
			print
			print "msg nr: %d, lineno: %d" % ( self.nr, self.lineno )
			sys.stdout.write( ''.join( self._msgid ) )
			sys.stdout.write( ''.join( self._msgstr ) )
		else:
			# Compatible with the old check_po
			print "%d '%s' : '%s'" % ( self.lineno, self.msgid(), self.msgstr() )

	def msgid( self ):
		if not self._msgid:
			return None
		txt = ''
		for l in self._msgid:
			l = re.sub( r'msgid\s+', '', l )
			l = strip_quotes( l )
			txt += l
		return txt

	def add_msgid( self, line ):
		self._msgid.append( line )

	def msgstr( self ):
		if not self._msgstr:
			return None
		txt = ''
		for l in self._msgstr:
			l = re.sub( r'msgstr\s+', '', l )
			l = strip_quotes( l )
			txt += l
		return txt

	def add_msgstr( self, line ):
		self._msgstr.append( line )

	def add_cmnt( self, line ):
		self._cmnt.append( line )
		if not self.is_fuzzy and self.fuzzy_pat.search( line ):
			self.is_fuzzy = 1

	def is_tips_xml( self ):
		for c in self._cmnt:
			if self.tips_xml_pat.search( c ):
				return 1
		return 0

	def set_sfmt_mismatch( self ):
		self.has_sfmt_mismatch = 1

	def set_named_fmt_mismatch( self ):
		self.has_named_fmt_mismatch = 1

	def set_fmt_missing_sd( self ):
		self.has_fmt_missing_sd = 1

	def set_context_error( self ):
		self.has_context_error = 1

	def set_named_fmt_mismatch( self ):
		self.has_named_fmt_mismatch = 1

	def set_xml_error( self ):
		self.has_xml_error = 1

	def set_lastchar_error( self ):
		self.has_lastchar_error = 1

def read_msgs( fname ):
	empty_pat   = re.compile( r'^ \s* $',      re.VERBOSE )
	comment_pat = re.compile( r'\#',           re.VERBOSE )
	msgid_pat   = re.compile( r'msgid \s+ "',  re.VERBOSE )
	msgstr_pat  = re.compile( r'msgstr \s+ "', re.VERBOSE )
	str_pat     = re.compile( r'"',            re.VERBOSE )
	old_pat     = re.compile( r'\#~ \s+ ',     re.VERBOSE )

	msgnr = 1
	f = open( fname )
	lines = f.readlines()
	
	# parse it like a statemachine
	NONE   = 0			# Nothing detected, yet
	CMNT   = 1			# Inside comment part
	MSGID  = 2			# Inside msgid part
	MSGSTR = 3			# Inside msgstr part
	STR    = 4			# A continuation string
	OLD    = 5			# An old pattern with #~

	state = NONE
	msg = None
	msgs = []

	for ix in range( len(lines) ):	# Use line numbers for messages
		line = lines[ix]
		lineno = ix + 1

		m = empty_pat.match( line )
		if m:
			continue	# Empty lines are not interesting

		# What's the next state?
		if  old_pat.match( line ):
			next_state = OLD
		elif comment_pat.match( line ):
			next_state = CMNT
		elif msgid_pat.match( line ):
			next_state = MSGID
		elif msgstr_pat.match( line ):
			next_state = MSGSTR
		elif str_pat.match( line ):
			next_state = STR
		else:
			next_state = NONE

		#print "%(state)d->%(next_state)d\t%(line)s" % vars()
		if state == NONE:
			# expect msgid or comment or old stuff
			if next_state == CMNT:
				state = CMNT
				msgnr += 1
				msg = Msgid( msgnr, lineno ) # Start with an empty new item
				msgs.append( msg )
				msg.add_cmnt( line )

			elif next_state == MSGID:
				state = MSGID
				msgnr += 1
				msg = Msgid( msgnr, lineno ) # Start with an empty new item
				msgs.append( msg )
				msg.add_msgid( line )

			elif next_state == MSGSTR:
				print 'WARNING: Wild msgstr at %(fname)s:%(lineno)d' % vars()
				state = MSGSTR
				msgnr += 1
				msg = Msgid( msgnr, lineno ) # Start with an empty new item
				msgs.append( msg )
				msg.add_msgstr( line )

			elif next_state == STR:
				print 'WARNING: Wild string at %(fname)s:%(lineno)d' % vars()

			elif next_state == OLD:
				pass	# Just skip

		elif state == CMNT:
			if next_state == CMNT:
				if msg:
					msg.add_cmnt( line )
				else:
					# Note. We may need to do something about these comments
					# Skip for now
					pass

			elif next_state == MSGID:
				state = MSGID
				if not msg:
					msgnr += 1
					msg = Msgid( msgnr, lineno ) # Start with an empty new item
					msgs.append( msg )
				msg.add_msgid( line )

			elif next_state == MSGSTR:
				print 'WARNING: Wild msgstr at %(fname)s:%(lineno)d' % vars()
				state = MSGSTR
				msgnr += 1
				msg = Msgid( msgnr, lineno ) # Start with an empty new item
				msgs.append( msg )
				msg.add_msgstr( line )

			elif next_state == STR:
				print 'WARNING: Wild string at %(fname)s:%(lineno)d' % vars()

			elif next_state == OLD:
				msg = None
				pass	# Just skip

		elif state == MSGID:
			if next_state == CMNT:
				# Hmmm. A comment here?
				print 'WARNING: Unexpted comment at %(fname)s:%(lineno)d' % vars()

			elif next_state == MSGID:
				raise Exception( 'Unexpected msgid at %(fname)s:%(lineno)d' % vars() )

			elif next_state == MSGSTR:
				state = MSGSTR
				msg.add_msgstr( line )

			elif next_state == STR:
				msg.add_msgid( line )

			elif next_state == OLD:
				msg = None
				pass	# Just skip

		elif state == MSGSTR:
			if next_state == CMNT:
				# A comment probably starts a new item
				state = CMNT
				msgnr += 1
				msg = Msgid( msgnr, lineno )
				msgs.append( msg )
				msg.add_cmnt( line )

			elif next_state == MSGID:
				state = MSGID
				msgnr += 1
				msg = Msgid( msgnr, lineno )
				msgs.append( msg )
				msg.add_msgid( line )

			elif next_state == MSGSTR:
				raise Exception( 'Unexpected msgstr at %(fname)s:%(lineno)d' % vars() )

			elif next_state == STR:
				msg.add_msgstr( line )

			elif next_state == OLD:
				msg = None
				pass	# Just skip

		else:
			raise Exception( 'Unexpected state in po parsing (state = %d)' % state )

	# Strip items with just comments. (Can this happen?)
	msgs1 = []
	for m in msgs:
		if not m.msgid() and not m.msgstr():
			#print "INFO: No msgid or msgstr at %s:%s" % ( fname, m.lineno )
			pass
		else:
			msgs1.append( m )
	msgs = msgs1
	return msgs

def analyze_msgs( fname, msgs, nr_templates = None, nth = 0 ):
	nr_fuzzy = 0
	nr_untranslated = 0
	nr_sfmt_mismatches = 0
	nr_named_fmt_mismatches = 0
	nr_fmt_missing_sd = 0
	nr_context_errors = 0
	nr_xml_errors = 0
	nr_lastchar_errors = 0

	# A pattern to find %() without s or d
	# Here is a command to use for testing
	# print re.compile('% \( \w+ \) \d* (\D|$)', re.VERBOSE).findall( '%(event_name)s: %(place)s%(endnotes)s. ' )
	find_named_fmt_pat2 = re.compile('% \( \w+ \) \d* (\D|$)', re.VERBOSE)

	# A pattern to find all %()
	find_named_fmt_pat = re.compile('% \( \w+ \) \d* \D', re.VERBOSE)

	# Special XML characters
	# It is not allowed to have a quote, an ampersand or an angle bracket
	xml_chars_pat = re.compile( r'(?<=\W) > | " | & (?!(quot|nbsp|gt|amp);)', re.VERBOSE )

	for msg in msgs:
		msgid = msg.msgid()
		msgstr = msg.msgstr()
		#print
		#print "msgid: %(msgid)s" % vars()
		#print "msgstr: %(msgstr)s" % vars()

		if not msgstr:
			nr_untranslated += 1
			continue

		if msg.is_fuzzy:
			nr_fuzzy += 1
			# Skip fuzzies or not?
			# continue

		cnt1 = msgid.count('%s')
		cnt2 = msgstr.count('%s')
		if cnt1 != cnt2:
			nr_sfmt_mismatches += 1
			msg.set_sfmt_mismatch()

		# Same number of named formats?
		fmts1 = find_named_fmt_pat.findall( msgid )
		fmts2 = find_named_fmt_pat.findall( msgstr )
		if len( fmts1 ) != len( fmts2 ):
			if not msg.has_sfmt_mismatch:
				nr_sfmt_mismatches += 1
				msg.set_sfmt_mismatch()

		# Do we have the same named formats?
		fmts1.sort()
		fmts2.sort()
		if fmts1 != fmts2:
			nr_named_fmt_mismatches += 1
			msg.set_named_fmt_mismatch()

		# Any formats missing format letter?
		fmts = find_named_fmt_pat2.findall( msgstr )
		for f in fmts:
			if not f in ('s', 'd'):
				nr_fmt_missing_sd += 1
				msg.set_fmt_missing_sd()
				break

		# Runaway context. In the translated part we only to see
		# the translation of the word after the |
		if msgid.count('|') > 0 and msgstr.count('|') > 0 and msgid != msgstr:
			nr_context_errors += 1
			msg.set_context_error()

		# XML errors
		# Only look at messages in the tips.xml
		if msg.is_tips_xml():
			if xml_chars_pat.search( msgstr ):
				nr_xml_errors += 1
				msg.set_xml_error()

		# Last character of msgid? White space? Period?
		if not msg.is_fuzzy and (msgid[-1:].isspace() != msgstr[-1:].isspace() or (msgid[-1:] == '.') != (msgstr[-1:] == '.')):
			nr_lastchar_errors += 1
			msg.set_lastchar_error()

	nr_msgs = len(msgs)
	if nth > 0:
		print
		print "====================================="
	print "%-20s%s"     % ( "File:",              fname )
	print "%-20s%d"     % ( "Template total:",    nr_templates )
	print "%-20s%d"     % ( "PO total:",          nr_msgs )
	print "%-20s%d"     % ( "Fuzzy:",             nr_fuzzy )
	print "%-20s%d"     % ( "Untranslated:",      nr_untranslated )
	print "%-20s%d"     % ( "%s mismatches:",     nr_sfmt_mismatches )
	print "%-20s%d"     % ( "%() name mismatches:",   nr_named_fmt_mismatches )
	print "%-20s%d"     % ( "%() missing s/d:",   nr_fmt_missing_sd )
	print "%-20s%d"     % ( "Runaway context:",   nr_context_errors )
	print "%-20s%d"     % ( "XML special chars:", nr_xml_errors )
	print "%-20s%d"     % ( "Last character:",    nr_lastchar_errors )

	po_coverage = (1.0 - (float(nr_untranslated) / float(nr_msgs))) * 100
	print "%-20s%5.2f%%" % ( "PO Coverage:",       po_coverage )

	template_coverage = po_coverage * float(nr_msgs) / float(nr_templates)
	print "%-20s%5.2f%%" % ( "Template Coverage:", template_coverage )

	if nr_sfmt_mismatches:
		print
		print "-------- %s mismatches --------------"
		for m in msgs:
			if m.has_sfmt_mismatch:
				m.diag()

	if nr_named_fmt_mismatches:
		print
		print "-------- %() name mismatches --------------"
		for m in msgs:
			if m.has_named_fmt_mismatch:
				m.diag()

	if nr_fmt_missing_sd:
		print
		print "-------- %() without 's' or 'd' mismatches --------------"
		for m in msgs:
			if m.has_fmt_missing_sd:
				m.diag()

	if nr_context_errors:
		print
		print "-------- Runaway context in translation ---------"
		for m in msgs:
			if m.has_context_error:
				m.diag()

	if nr_xml_errors:
		print
		print "-------- unescaped XML special characters ---------"
		for m in msgs:
			if m.has_xml_error:
				m.diag()

	if nr_lastchar_errors:
		print
		print "-------- last character not identical ---------"
		for m in msgs:
			if m.has_lastchar_error:
				m.diag()

def main():
	try:
		pot_msgs = read_msgs( 'gramps.pot' )
		nr_templates = len( pot_msgs )
		#analyze_msgs( 'gramps.pot', pot_msgs )
		nth = 0
		for fname in sys.argv[1:]:
			msgs = read_msgs( fname )
			analyze_msgs( fname, msgs, nr_templates, nth )
			nth += 1

	except Exception, e:
		print e

if __name__ == "__main__":
	main()