config/config.py

import re
import config.long_responses as long
from config.valid_text import is_valid_input

def message_probability(
        user_message, recognised_words,
        single_response=False, required_words=None
):
    if required_words is None:
        required_words = []
    message_certainty = 0
    has_required_words = True

    # Counts how many words are present in each predefined message
    for word in user_message:
        if word in recognised_words:
            message_certainty += 1

    # Calculates the percent of recognised words in a user message
    percentage = float(message_certainty) / float(len(recognised_words))

    # Checks that the required words are in the string
    for word in required_words:
        if word not in user_message:
            has_required_words = False
            break

    # Must either have the required words, or be a single response
    if has_required_words or single_response:
        return int(percentage * 100)
    else:
        return 0


def check_all_messages(message):
    highest_prob_list = {}

    # Simplifies response creation / adds it to the dict
    def response(bot_response, list_of_words, single_response=False, required_words=[]):
        nonlocal highest_prob_list
        highest_prob_list[bot_response] = message_probability(message, list_of_words, single_response, required_words)

    # Responses -------------------------------------------------------------------------------------------------------

    response("بله", ["بله", "بلی", "بلخ", "یله", "پله", "gfi",
                     "آره", "آری", "آرخ",  "yes", "ya",
                     "bale", "ari", "are", "fgi", "hvi",
                     "Hvd", "Hvd", "fgd",  "hsj", "isj",
                     "nhvn", "هست", "است",  "دارد",
                     "یس",], single_response=True)
    response("خیر", [
                "نه", "منفی", "odv", "kodv", "خیر", "no",
                "ni", "نخیر", "نیست", "نبود", "نمیشود",
                "نمی شود", "نمی\u200cشود", "نیست",
                "ندیدم", "دخ", "nist", "nabod", "na",
                "nah", "noch", "nuch", "nooch", "manfi",
            ], single_response=True)


    # Longer responses

    response(long.R_L01, ["اختلالات", "حرکتی", "Movement", "disorders","حرکت"],
             required_words=["حرکت"])
    response(long.R_L02, ["تعمیم یافته", "صورت", "گردن", "اندامهای", "رقصاک", "فوقانی", "تحتانی", "تنه"],
             required_words=["رقصاک"])
    response(long.R_L03, ["چشمی", "عصبی", "حرکت"
    ], required_words=["عصبی"])
    response(long.R_L04, ["بیماری‌", "عصبی", "متا", "بولیکی", "نورومتابولیکی"
    ], required_words=["عصبی"])
    response(long.R_L05, ["روان‌پزشکی", "روانپریشی", "روان", "پزشکی", "روانپزشکی"
    ], required_words=["روانپزشکی"])
    response(long.R_L06, ["هیپرهموسیستئینمی", "hyperhomocysteinemia","ژنتیکی"
    ], required_words=["ژنتیکی"])
    response(long.R_L07, ["درگیری", "طناب", "نخاعی", "spinal", "cord", "نخاعی"
    ], required_words=["نخاعی"])
    response(long.R_L08, ["کوبالامین", "داخل", "سلولی", "intracellular", "cobalamin","سلولی"
    ], required_words=["سلولی"])

    response(long.R_L09, ["لیپاز", "اسید", "لیزوزومال", "LAL-D", "کمبود"
    ], required_words=["کمبود"])
    response(long.R_L10, ["هیدرولیپوآمید", "دهیدروژناز", "کمبود"
    ], required_words=["کمبود"])
    response(long.R_L11, ["فسفاتمی","هایپوفسفاتمی", "هیپوفسفاتمی", "hypophosphatemia", "HP", "ژنتیکی"
    ], required_words=["ژنتیکی"])
    response(long.R_L12, ["آنسفالوپاتی", "هیپرآمونمیک", "غیر", "کبدی", "جراحی", "چاقی", "Nonhepatic", "hyperammonemic", "encephalopathy", "bariatric", "surgery", "NHE-BS"
    ], required_words=["کبدی"])
    response(long.R_L13, ["لرزش", "میتو", "کندری", "میتوکندری", "ژنتیکی"
    ], required_words=["ژنتیکی"])
    response(long.R_L13, ["اندام", "لرزش", "ترمور", "Tremor", "TRM", "عصبی"
    ], required_words=["عصبی"])


    #
    #
    # "اختلالات حرکتی": "L01",
    # " رقصاک تعمیم یافته  در صورت ، گردن ، اندامهای فوقانی/تحتانی ، تنه ": "L02",
    # "اختلالات حرکات چشم و علائم عصبی": "L03",
    # "بیماریهای عصبی": "L04",
    # "هوموستئینمی شدید": "L06",
    # "تظاهرات روانی": "L05",
    # "درگیری نخاع": "L07",
    # "متابولیسم کبدی داخل سلولی": "L08",
    # "کمبود لیپاز اسید لیزوزومی": "L09",
    # "کمبود دی هیدرو لیپوامید دی هیدروژناز": "L10",
    #
    #
    # "فسفاتمی": "L11",
    # "عوارض" ,"آنسفالوپاتی", "غیرکبدی" , "جراحی", "باریاتریک": "L12",
    # "لرزش اندام فوقانی": "L13",
    # "بیماریهای میتوکندری": "L14",
    #
    #
    #
    # "کمبود", "هیدرولیپوآمید", "دهیدروژناز"
    # "هایپوفسفاتمی", "هیپوفسفاتمی", "hypophosphatemia", "HP"
    # "آنسفالوپاتی", "هیپرآمونمیک", "غیر", "کبدی", "جراحی", "چاقی", "Nonhepatic", "hyperammonemic", "encephalopathy", "bariatric", "surgery", "NHE-BS
    # "لرزش", "ترمور", "Tremor", "لرزش", "TRM"


    best_match = max(highest_prob_list, key=highest_prob_list.get)
    # print(highest_prob_list)
    # print(f'Best match = {best_match} | Score: {highest_prob_list[best_match]}')

    return long.unknown() if highest_prob_list[best_match] < 1 else best_match


# Used to get the response
def get_response(user_input):
    split_message = re.split(r'\s+|[,;?!.-]\s*', user_input.lower())
    response = check_all_messages(split_message)
    return response


#
# # Testing the response system
# while True:
#     print('Bot: ' + get_response(input('You: ')))