#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
Convert InDesign tagged text to ConTeXt
"""
import sys, os
import re

quote = u'$&_%'

rePatterns = {
	# paragraph styles
	ur'^<pstyle:Ü 1\.>((\d\.)*\s+)?(.+)$' : ur'\\chapter{\3}\n',
	ur'^<pstyle:Ü 1\.1>((\d\.)*\s+)?(.+)$' : ur'\\section{\3}\n',
	ur'^<pstyle:Ü 1\.1\.1>((\d\.)*\s+)?(.+)$' : ur'\\subsection{\3}\n',
	ur'^<pstyle:Ü 1\.1\.1\.1>((\d\.)*\s+)?(.+)$' : ur'\\subsubsection{\3}\n',
	# character styles
	ur'<ct:Bold>(.+?)<ct:>' : ur'{\\bf \1}',
	#ur'<cf:Arial>(.*?)<cf:Times New Roman>' : ur'\\otherfont{\1}',
	
	u'<.*?>' : u'', # delete all other tags

	# lines that start with dotted numbers = section titles
	ur'^\d+\s+(.+)$' : ur'\\chapter{\1}\n',
	ur'^\d+\.\d+\.?\s+(.+)$' : ur'\\section{\1}\n',
	ur'^\d+\.\d+\.\d+\.?\s+(.+)$' : ur'\\subsection{\1}\n',
	ur'^\d+\.\d+\.\d+\.\d+\.?\s+(.+)\$' : ur'\\subsubsection{\1}\n',
	
	ur'^(\s*)[–\-·•]\s+' : ur'\1\\item\t', # itemization (lines starting with bullet etc.)
	ur'^(\s*)(\d+)\.?\)\s+' : ur'\1\\item[\2]\t', # itemization (numerical)
	ur'([Zusovz])\.([Baguo])\.' : ur'\1.\\,\2.', # u.a., s.o., o.g., z.B.
	ur'[„"“](.*?)[“”"]' : ur'\\quotation{\1}', # German quotation
	ur'[\'’,](.*?)[\'’‘]' : ur'\\quote{\1}', # German single quotation
	#ur'"(.*?)"' : ur'\\quotation{\1}', # quotation?
	ur' (\.\?\!:;)' : ur'\1', # spaces in front of punctuation
	ur'{\\em\s+}' : ur'', # empty emphasizing
	ur' (%|°)' : ur'\\,\1', # spaces in front of measure units
	u' - ' : u' – ', # en dash
	ur'(\d{4})\s*(\-|–)\s*(\d{4})' : ur'\1–\3', # year numbers
	
	u' +' : u' ', # multiple spaces
	u'^\s+$' : u'\n', # make empty lines really empty

#	ur'' : ur'',
	
}

reres = {}
status = {
	'item' : False
}

# collect parameters
if len(sys.argv) > 1:
	sourcename = sys.argv[1]
	if len(sys.argv) > 2:
		targetname = sys.argv[2]
	else:
		targetname = sourcename.replace('.txt', '.tex')
else:
	print "file name?"
	sys.exit()

# compile regular expressions
for k in rePatterns:
	p = re.compile(k)
	reres[p] = rePatterns[k]

source = open(sourcename, 'rU')
target = open(targetname, 'w')

# convert lines
for line in source.readlines():
	line = unicode(line, 'utf-16be') # "unicode" encoded InDesign tagged text is UTF-16 big-endian encoded!
	for p in reres:
		line = p.sub(reres[p], line)
	for c in quote:
		line = line.replace(c, u'\\'+c)
	if '\\item ' in line and not status['item']:
		target.write('\\startitemize[]\n')
		status['item'] = True
	if status['item'] and not '\\item ' in line:
		target.write('\\stopitemize\n')
		status['item'] = False
	target.write(line.encode('utf-8')) # write UTF-8

source.close()
target.close()

print "%s completed" % targetname