#!/usr/bin/python
# -*- coding=utf-8 -*-
# ------------------------------------------------------------------------
# Name:        lemmatizer
# Purpose:     Arabic Lemmatizer
#
# Author:      Taha Zerrouki (taha.zerrouki[at]gmail.com)
#
# Created:     26-08-2020
# Copyright:   (c) Taha Zerrouki 2011
# Licence:     GPL
# ------------------------------------------------------------------------
"""
Syntaxic Analysis
"""

# ~ from operator import xor
# ~ import functools
# ~ import operator
import pprint
import re
import pyarabic.araby as araby


class Templater:
    """
    Arabic Templater
    """

    def __init__(self, cache_path=False):
        """ """
        # create analexer
        pass

    def extract_wazns(self, lemmas_roots_list):
        """
        Extract a list of wazns
        """
        wazns = []
        for lemmas, roots, word_type in lemmas_roots_list:
            if word_type == "stopword":
                continue
            lemmas = list(set(lemmas))
            roots = list(set(roots))
            for lemma in lemmas:

                for root in roots:
                    # uniq
                    wazns.append(self.extract_wazn(lemma, root))
        wazns = list(set(wazns))
        return wazns

    @staticmethod
    def is_invalid_wazn(wazn):
        """
        Test if the wazn is valid
        test invalid letters in a wazn, before make it a template
        here F, Ain, Lam, are coded 0,1,2
        """
        # حروف غير اشتقاقية
        if bool(re.search(r"[بثجحخذرزسشصضظعفقكلهء]", wazn)):
            return True
        # if the numbers are missplaced
        if bool(re.search(r"[^0][دط]", wazn)):
            return True
        # if the numbers are missplaced
        if bool(re.search(r"[12].*0|2.*1", wazn)):
            return True
        return False

    def extract_wazn(self, lemma, root):
        """
        extract a wazn from a given lemma and root
        """
        if not root or len(root) < 3:
            return ""
        else:
            wazn = lemma[::-1]
            if len(root) == 4:
                wazn = wazn.replace(root[2], "2", 2)
            else:
                #  doubled root
                if root[2] == root[1]:
                    wazn = wazn.replace(root[2], "1", 1)
                else:
                    wazn = wazn.replace(root[2], "2", 1)
            wazn = wazn.replace(root[1], "1", 1)
            wazn = wazn.replace(root[0], "0", 1)
            wazn = wazn[::-1]
            if self.is_invalid_wazn(wazn):
                return ""
            wazn = wazn.replace("2", "ل")
            wazn = wazn.replace("1", "ع")
            wazn = wazn.replace("0", "ف")
            return wazn


def mainly():
    """
    main test
    """
    # #test syn
    text = "إلى البيت"
    result = []
    lemmer = Lemmatizer()
    result = lemmer.analyze_text(text)
    # the result contains objects
    pprint.pprint(result)


if __name__ == "__main__":
    mainly()