Dynamic-Memory-Networks-for-QA-System/utils.py at master · cuteboydot/Dynamic-Memory-Networks-for-QA-System · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import numpy as np
import os
import re
import datetime
import pickle
import random

def tokenize(sent):
    return [x.strip() for x in re.split('(\W+)?', sent) if x.strip()]

def data_loader(file_train, file_test):
    print("data_loader start!")

    data_list_train = []
    data_list_test = []
    data_idx_list_train = []
    data_idx_list_test = []
    max_story_len = 0
    max_words_len = 0

    rdic = []
    rdic.append("<nil>")

    # read file for train
    with open(file_train) as f:
        lines_train = f.readlines()

    for line_train in lines_train:
        line = line_train.lower()

        idx, line = line.split(" ", 1)
        idx = int(idx)

        #words = tokenize(line)
        #idx = int(words[0])
        #words = words[1:]

        if idx == 1:
            stories = []

        if "\t" in line:
            que, ans, sup = line.split('\t')
            que = tokenize(que)
            if que[-1] == "?":
                que = que[:-1]

            #que = words[:-3]
            #ans = words[-2]
            #sup = words[-1]
            data_list_train.append([stories, que, ans])
            stories = stories.copy()

            max_story_len = max(max_story_len, len(stories))
            max_words_len = max(max_words_len, len(que))

            for q in que:
                if q not in rdic:
                    rdic.append(q)

            if ans not in rdic:
                rdic.append(ans)
        else:
            #sen = words[:-1]
            sen = tokenize(line)
            if sen[-1] == ".":
                sen = sen[:-1]
            stories.append(sen)

            max_words_len = max(max_words_len, len(sen))

            for s in sen:
                if s not in rdic:
                    rdic.append(s)

    # read file for test
    with open(file_test) as f:
        lines_test = f.readlines()

    for line_test in lines_test:
        line = line_test.lower()

        idx, line = line.split(" ", 1)
        idx = int(idx)

        if idx == 1:
            stories = []

        if "\t" in line:
            que, ans, sup = line.split('\t')
            que = tokenize(que)
            if que[-1] == "?":
                que = que[:-1]

            data_list_test.append([stories, que, ans])
            stories = stories.copy()

            max_story_len = max(max_story_len, len(stories))
            max_words_len = max(max_words_len, len(que))

            for q in que:
                if q not in rdic:
                    rdic.append(q)

            if ans not in rdic:
                rdic.append(ans)
        else:
            sen = tokenize(line)
            if sen[-1] == ".":
                sen = sen[:-1]
            stories.append(sen)

            max_words_len = max(max_words_len, len(sen))

            for s in sen:
                if s not in rdic:
                    rdic.append(s)

    # make dic
    dic = {w:i for i,w in enumerate(rdic)}
    max_len = [max_story_len, max_words_len]

    # make dic[idx] list for train
    for data in data_list_train:
        stories = data[0]
        question = data[1]
        answer = data[2]

        stories_idx = []
        for story in stories:
            story_idx = [dic[w] for w in story]
            stories_idx.append(story_idx)

        question_idx = [dic[q] for q in question]
        answer_idx = dic[answer]

        data_idx_list_train.append([stories_idx, question_idx, answer_idx])

    # make dic[idx] list for test
    for data in data_list_test:
        stories = data[0]
        question = data[1]
        answer = data[2]

        stories_idx = []
        for story in stories:
            story_idx = [dic[w] for w in story]
            stories_idx.append(story_idx)

        question_idx = [dic[q] for q in question]

        answer_idx = dic[answer]

        data_idx_list_test.append([stories_idx, question_idx, answer_idx])

    print("data_loader end!")
    print()

    return data_list_train, data_list_test, data_idx_list_train, data_idx_list_test, rdic, dic, max_len