from pyhanlp import *

import utils


def open_file(tokens):
    sentences = []
    with open(tokens, "r", encoding="utf-8") as f:
        for sentence in f.readlines():
            sentence = sentence.strip()
            sentences.append(sentence)
    return sentences


def tokenizier(sentences):
    use_cases = []
    for sentence in sentences:
        condition = [[], "", ""]
        else_flag = False
        # s: result of segment
        s = HanLP.segment(sentence)
        control = 999
        for i in range(len(s)):
            if "或" in str(s[i].word) and str(s[i].nature) == 'c':
                if not condition[0].__contains__(s[i - 1].word):
                    condition[0].append(s[i - 1].word)
                if not condition[0].__contains__(s[i + 1].word):
                    condition[0].append(s[i + 1].word)
                else_flag = True
            elif str(s[i].nature) == 'p':
                condition[1] = s[i].word
                control = i
            elif str(s[i].nature) == 'c':
                control = i
                condition[1] = s[i].word
            elif str(s[i].nature) == 'd':
                control = i
                condition[1] = s[i].word
            elif str(s[i].nature) == 'v' and condition[1] != "":
                condition[1] += s[i].word
            elif str(s[i].nature) != 'c' and i > control:
                condition[2] += s[i].word
            else:
                condition[0].append(s[i].word)
        # 去重
        condition[0] = list(set(condition[0]))
        if not else_flag:
            string = "".join(condition[0])
            condition[0].clear()
            condition[0] = [string]
        use_cases.append(condition)
    return use_cases


def generateConditions(useCases):
    allConditions = []

    for case in useCases:
        if len(case[0]) > 1:
            for i in case[0]:
                string = ""
                string += i
                string += case[1] + case[2]
                allConditions.append(string)
        else:
            string = ""
            string = "".join(case[0]) + "".join(case[1:])
            allConditions.append(string)
    with open("allConditions.txt", "w", encoding="utf-8") as f:
        for line in allConditions:
            f.write(line)
            f.write('\n')


def generateUseCases(filename):
    cases = []
    with open(filename, "r") as f:
        for line in f.readlines():
            cases.append(line.replace("\n", ""))

    for case in cases:
        single_case = utils.case_template
        if "不为" in case:
            start = case.index("不")
            except_condition = case[start+2:]
            attribute_name = case[:start]
            if except_condition == '空':
                with open("./generated/" + case + ".txt", 'w', encoding='utf-8') as file_write:
                    single_case[attribute_name] = "test"
                    print(single_case.__str__())
                    file_write.write(single_case.__str__())


if __name__ == '__main__':
    sentences = open_file("main_info.txt")
    useCases = tokenizier(sentences)
    generateConditions(useCases)
    generateUseCases("allConditions.txt")