import nltk

class Word:

    def __init__(self, posi, token, stem, pos):
        self.__token = token
        self.__stem = stem
        self.__pos = pos
        self.__posi = posi

    def get_token(self):
        return self.__token

    def get_stem(self):
        return self.__stem

    def get_pos(self):
        return self.__pos

    def get_posi(self):
        return self.__posi


def annotate_corpus_with_nltk_pipeline(input_filename, language):
    '''Liest einen Text ein und speichert diesen annotiert in einem Dictionary mit den Wörtern als Instanzen.'''
    with open(input_filename, 'r', encoding='utf8') as f:
        read_text = f.read()

    dic_text = {}
    list_text = nltk.sent_tokenize(read_text, language=language)
    for line in list_text:
        word_list_obj = []
        word_list = nltk.word_tokenize(line, language=language)
        pos_tag_tuples_list = nltk.pos_tag(word_list, tagset=None, lang=language[0:3])
        for posi,mtuple in enumerate(pos_tag_tuples_list):
            if mtuple[1] == 'CD' or mtuple[1].startswith('N'):
                wn_pos_tag = 'n'
            elif mtuple[1] == 'MD' or mtuple[1].startswith('V'):
                wn_pos_tag = 'v'
            elif mtuple[1].startswith('J'):
                wn_pos_tag = 'a'
            elif mtuple[1].startswith('R'):
                wn_pos_tag = 'r'
            else:
                wn_pos_tag = 'n'
            stem = nltk.stem.WordNetLemmatizer.lemmatize(self=None, word=mtuple[0], pos=wn_pos_tag)
            word = Word(posi, mtuple[0], stem, mtuple[1])
            word_list_obj.append(word)
        dic_text[list_text.index(line)] = word_list_obj
    return dic_text