import requests

import json

import sys

import os

import pandas

import time



username = ""

password = ""



old_column_name = ["id", "Openness", "Openness_e", "Adventurousness", "Adventurousness_e", "Artistic interests", "Artistic interests_e", "Emotionality", "Emotionality_e", "Imagination", "Imagination_e", "Intellect", "Intellect_e", "Authority-challenging", "Authority-challenging_e", "Conscientiousness", "Conscientiousness_e", "Achievement striving", "Achievement striving_e", "Cautiousness", "Cautiousness_e", "Dutifulness", "Dutifulness_e", "Orderliness", "Orderliness_e", "Self-discipline", "Self-discipline_e", "Self-efficacy", "Self-efficacy_e", "Extraversion", "Extraversion_e", "Activity level", "Activity level_e", "Assertiveness", "Assertiveness_e", "Cheerfulness", "Cheerfulness_e", "Excitement-seeking", "Excitement-seeking_e", "Outgoing", "Outgoing_e", "Gregariousness", "Gregariousness_e", "Agreeableness", "Agreeableness_e", "Altruism", "Altruism_e", "Cooperation", "Cooperation_e", "Modesty", "Modesty_e", "Uncompromising", "Uncompromising_e", "Sympathy", "Sympathy_e", "Trust", "Trust_e", "Emotional range", "Emotional range_e", "Fiery", "Fiery_e", "Prone to worry", "Prone to worry_e", "Melancholy", "Melancholy_e", "Immoderation", "Immoderation_e", "Self-consciousness", "Self-consciousness_e", "Susceptible to stress", "Susceptible to stress_e", "Challenge", "Challenge_e", "Closeness", "Closeness_e", "Curiosity", "Curiosity_e", "Excitement", "Excitement_e", "Harmony", "Harmony_e", "Ideal", "Ideal_e", "Liberty", "Liberty_e", "Love", "Love_e", "Practicality", "Practicality_e", "Self-expression", "Self-expression_e", "Stability", "Stability_e", "Structure", "Structure_e", "Conservation", "Conservation_e", "Openness to change", "Openness to change_e", "Hedonism", "Hedonism_e", "Self-enhancement", "Self-enhancement_e", "Self-transcendence", "Self-transcendence_e"]



column_name = ['id']

for column in old_column_name[1:]:

    column_name.append(column)

    column_name.append('R_' + column)

 

# def pre_aggregate(dir_name, file_list):

#     """Aggregate the data for each person and output a txt file for each person to directory 'Aggregated_files'



#     Args:

#     =====

#         dir_name:  The name of the directory that containing all the data to be aggregated 

#         file_list: The list of all the files in the target directory

        

#     Returns:

#     ========

#         output_file_list: The list of filenames in the output directory 'Aggregated_files'

#     """

#     name_list = []

#     i = 0

#     out_dir_name ='Aggregated_files'

#     if not os.path.exists(out_dir_name):

#         os.mkdir(out_dir_name)

#     for file_name in file_list:

#         name_list = []

#         with open(dir_name + '/' +file_name, "rb") as f:

#             for line in f:

#                 # get rid of the first line

#                 if i == 0:

#                     i += 1

#                     continue

#                 try:

#                     [cid, ordernum, fname, lname, content] = line.strip().split('\t')

#                 except:

#                     [cid, ordernum, fname, lname] = line.strip().split('\t')

#                     print line

#                 if '?' in lname:

#                     lname = lname.strip('?')

#                 if (fname, lname) not in name_list:

#                     fout = open(out_dir_name + '/'+ '_'.join((fname, lname))+'.txt', "w")

#                     fout.write(fname+'\t'+lname+'\n')

#                     fout.write(cid+'\t'+content+'\n\n')

#                     name_list.append((fname, lname))

#                 else:

#                     fout = open(out_dir_name + '/'+ '_'.join((fname, lname))+'.txt', "a")

#                     fout.write(cid+'\t'+content+'\n'+'\n')

#                 fout.close()

#         output_file_list = ['_'.join((x[0], x[1]))+'.txt' for x in name_list]

#         return output_file_list





# def read_data(dir_name, file_list):

#     """Read all the data from a specific directory

    

#     Args:

#     =====

#         dir_name:  The name of the directory that containing all the aggregated data

#         file_list: The list of all the files in the target directory



#     Returns:

#     ========

#         text: A list of tuple which contains the text, id, and type of each file

#     """

#     text = []

#     for file_name in file_list:

#         if file_name[-3:] == "txt":

#             content = process_txt_file(dir_name + "/" + file_name)

#             tmp = file_name[:-4]

#             text.append((content, tmp.split('_')[0], tmp.split('_')[1]))

# #        elif file_name[-3:] == "csv":

# #            data = process_csv_file(dir_name + "/" + file_name)

# #            if data is not None:

# #                text.append((data, file_name[:-4], "csv"))

#     return text



# def process_txt_file(file_name):

#     """Clean the txt file

    

#     Args:

#     =====

#         file_name: The name of the file which you want to process



#     Returns:

#     ========

#         content: pure text of the file

#     """

#     i = 0

#     content = ""

#     with open(file_name, "rb") as f:

#         for line in f:

#             # get rid of the first line

#             if i == 0:

#                 i += 1

#                 continue

#             content += " ".join(line.strip().split(" ")[1:])

#         f.close()

#     return content



# def process_csv_file(file_name):

#     """Clean the csv file and get rid of useless csv files

    

#     Args:

#     =====

#         file_name: The name of the file which you want to process



#     Returns:

#     ========

#         content: pure text of the file

#     """

#     file = pandas.read_csv(file_name)

#     if "tweet" not in file.columns:

#         return None

#     else:

#         content = " ".join(file["tweet"].tolist())

#         return content



def process_txt_file_2(dir_name, file_name):

    i = 0

    text = []

    with open(dir_name + "/" + file_name, "rb") as f:

        for line in f:

            # get rid of the first line

            if i == 0:

                i += 1

                continue

            if line == '\r\n' or line == '\n':

                continue

            if len(text) > 0 and text[-1][0] == line.split('\t')[0]:

                text[-1][1] += line.split('\t')[1]

                continue

            text.append(line.split('\t'))

        f.close()

    return text



def parse_json(tree, id):

    """Get all the personalities and their corresponding sampling error from



    Args:

    =====

        tree: a nested dictionary which contains all the information that we get from IBM watson

        id:   the id of the current user

        type: file type



    Returns:

    ========

        personalities: a list of all the personalities

    """

    personalities = [id]

    if "tree" not in tree:

        personalities.extend(["0"] * (len(column_name) - 1))

    else:

        # for name in tree["tree"]["children"]:

        #     for personality in name["children"][0]["children"]:

        #         personalities.extend([str(personality["percentage"]), str(personality["sampling_error"])])

        #         if name["name"] == "Big 5":

        #             for sub_personality in personality["children"]:

        #                 personalities.extend([str(sub_personality["percentage"]), str(sub_personality["sampling_error"])])



        for name in tree["tree"]["children"]:

            for personality in name["children"][0]["children"]:

                personalities.extend([str(personality["percentage"]), str(personality["raw_score"]), str(personality["sampling_error"]), str(personality["raw_sampling_error"])])

                if name["name"] == "Big 5":

                    for sub_personality in personality["children"]:

                        personalities.extend([str(sub_personality["percentage"]), str(sub_personality["raw_score"]), str(sub_personality["sampling_error"]), str(sub_personality["raw_sampling_error"])])

    return personalities



def analyze_text(text):

    """ Use IBM watson to analyze the text

    """

    total_personalities = map(lambda info: parse_json(

    json.loads(requests.post("https://gateway.watsonplatform.net/personality-insights/api/v2/profile", auth=(username, password), headers = {"content-type": "text/plain"}, params = {"include_raw": True}, data = info[1]).text), info[0]), text)

    return total_personalities



def save_to_txt_file(output_filename, header, data):

    """Save result to file

    

    Args:

    =====

        header: schema

        data: the data you want to save

    """

    with open(output_filename, "wb") as f:

        f.write("\t".join(header) + "\n")

        final_data = "\n".join(map(lambda x: "\t".join(x), data))

        f.write(final_data)

        f.close()



def main():

    start = time.time()

    dir_name = sys.argv[1]

    file_list = os.listdir(dir_name)

    if not os.path.isdir('output/'):

        os.mkdir("output/")

    for file_name in file_list:

        data = process_txt_file_2(dir_name, file_name)

#        for tag in data:

#            print tag

        columns = analyze_text(data)

        save_to_txt_file('output/' + file_name, column_name, columns)

    end = time.time()

    print end - start



if __name__ == '__main__':

    main()