Arabic text to numeric representation

ara2bin.py

# -*- coding: utf-8 -*-

from lang_trans.arabic import buckwalter
import re
from odf import text, teletype 
from odf.opendocument import load

# consonants
solar = r"[tvd*rzs$SDTZln]"
lunar = r"['bjHxEgqfkmh]"
hamza = r"[<>]"

def harf2bin(line):

	line = line.strip()
	
	# pre-processing
	# e.g. اللَّهِ  => ألْلَاْهِيْ

	# al
	line = re.sub(r"Al" + solar, "1", line)
	line = re.sub(r"Alo?" + lunar, "10", line)

	# tanween
	line = re.sub(r"[NFK]", "10", line)

	# long vowels
	line = re.sub(r"[uia][wyYA]", "10", line)

	# ashba3
	line = re.sub(r"[uia]$", "10", line)
 
	# short vowels
	line = re.sub(r"[uia]", "1", line)

	# sukun
	line = re.sub(r"o", "0", line)

	# shadda
	line = re.sub(r"~", "10", line)

	# remove consonants
	line = re.sub(r"(" + lunar + "|" + solar + "|" + hamza + ")", "", line)
 
	# remove long vowels used as consonants
	line = re.sub(r"[wyYA]1", "", line)
 
	return line


# load text from line by parameter or fetch from the cloud
textdoc = load("Sample Text.odt")
paras = textdoc.getElementsByType(text.P)

for p in paras:
	t = teletype.extractText(p)
	b = buckwalter.transliterate(t)
	print(b)
	print(harf2bin(b))