- Notifications
You must be signed in to change notification settings - Fork 783
Expand file tree
/
Copy pathdata_utils.py
More file actions
Latest commit
executable file
·146 lines (117 loc) · 5.47 KB
/
Copy pathdata_utils.py
File metadata and controls
executable file
·146 lines (117 loc) · 5.47 KB
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
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Utilities for downloading data from WMT, tokenizing, vocabularies."""
from __future__ importabsolute_import
from __future__ importdivision
from __future__ importprint_function
importos
importre
fromsix.movesimporturllib
fromtensorflow.python.platformimportgfile
# Special vocabulary symbols - we always put them at the start.
_PAD=b"_PAD"
_GO=b"_GO"
_EOS=b"_EOS"
_UNK=b"_UNK"
_START_VOCAB= [_PAD, _GO, _EOS, _UNK]
PAD_ID=0
GO_ID=1
EOS_ID=2
UNK_ID=3
# Regular expressions used to tokenize.
_WORD_SPLIT=re.compile(b"([.,!?\"':;)(])")
_DIGIT_RE=re.compile(br"\d")
defbasic_tokenizer(sentence):
"""Very basic tokenizer: split the sentence into a list of tokens."""
words= []
forspace_separated_fragmentinsentence.strip().split():
words.extend(re.split(_WORD_SPLIT, space_separated_fragment))
return [wforwinwordsifw]
defcreate_vocabulary(vocabulary_path, data_path, max_vocabulary_size,
tokenizer=None, normalize_digits=True):
ifnotgfile.Exists(vocabulary_path):
print("Creating vocabulary %s from %s"% (vocabulary_path, data_path))
vocab= {}
withgfile.GFile(data_path, mode="rb") asf:
counter=0
forlineinf:
counter+=1
ifcounter%100000==0:
print(" processing line %d"%counter)
tokens=tokenizer(line) iftokenizerelsebasic_tokenizer(line)
forwintokens:
word=re.sub(_DIGIT_RE, b"0", w) ifnormalize_digitselsew
ifwordinvocab:
vocab[word] +=1
else:
vocab[word] =1
vocab_list=_START_VOCAB+sorted(vocab, key=vocab.get, reverse=True)
print('>> Full Vocabulary Size :',len(vocab_list))
iflen(vocab_list) >max_vocabulary_size:
vocab_list=vocab_list[:max_vocabulary_size]
withgfile.GFile(vocabulary_path, mode="wb") asvocab_file:
forwinvocab_list:
vocab_file.write(w+b"\n")
definitialize_vocabulary(vocabulary_path):
ifgfile.Exists(vocabulary_path):
rev_vocab= []
withgfile.GFile(vocabulary_path, mode="rb") asf:
rev_vocab.extend(f.readlines())
rev_vocab= [line.strip() forlineinrev_vocab]
vocab=dict([(x, y) for (y, x) inenumerate(rev_vocab)])
returnvocab, rev_vocab
else:
raiseValueError("Vocabulary file %s not found.", vocabulary_path)
defsentence_to_token_ids(sentence, vocabulary, tokenizer=None, normalize_digits=True):
iftokenizer:
words=tokenizer(sentence)
else:
words=basic_tokenizer(sentence)
ifnotnormalize_digits:
return [vocabulary.get(w, UNK_ID) forwinwords]
# Normalize digits by 0 before looking words up in the vocabulary.
return [vocabulary.get(re.sub(_DIGIT_RE, b"0", w), UNK_ID) forwinwords]
defdata_to_token_ids(data_path, target_path, vocabulary_path,
tokenizer=None, normalize_digits=True):
ifnotgfile.Exists(target_path):
print("Tokenizing data in %s"%data_path)
vocab, _=initialize_vocabulary(vocabulary_path)
withgfile.GFile(data_path, mode="rb") asdata_file:
withgfile.GFile(target_path, mode="w") astokens_file:
counter=0
forlineindata_file:
counter+=1
ifcounter%100000==0:
print(" tokenizing line %d"%counter)
token_ids=sentence_to_token_ids(line, vocab, tokenizer,
normalize_digits)
tokens_file.write(" ".join([str(tok) fortokintoken_ids]) +"\n")
defprepare_custom_data(working_directory, train_enc, train_dec, test_enc, test_dec, enc_vocabulary_size, dec_vocabulary_size, tokenizer=None):
# Create vocabularies of the appropriate sizes.
enc_vocab_path=os.path.join(working_directory, "vocab%d.enc"%enc_vocabulary_size)
dec_vocab_path=os.path.join(working_directory, "vocab%d.dec"%dec_vocabulary_size)
create_vocabulary(enc_vocab_path, train_enc, enc_vocabulary_size, tokenizer)
create_vocabulary(dec_vocab_path, train_dec, dec_vocabulary_size, tokenizer)
# Create token ids for the training data.
enc_train_ids_path=train_enc+ (".ids%d"%enc_vocabulary_size)
dec_train_ids_path=train_dec+ (".ids%d"%dec_vocabulary_size)
data_to_token_ids(train_enc, enc_train_ids_path, enc_vocab_path, tokenizer)
data_to_token_ids(train_dec, dec_train_ids_path, dec_vocab_path, tokenizer)
# Create token ids for the development data.
enc_dev_ids_path=test_enc+ (".ids%d"%enc_vocabulary_size)
dec_dev_ids_path=test_dec+ (".ids%d"%dec_vocabulary_size)
data_to_token_ids(test_enc, enc_dev_ids_path, enc_vocab_path, tokenizer)
data_to_token_ids(test_dec, dec_dev_ids_path, dec_vocab_path, tokenizer)
return (enc_train_ids_path, dec_train_ids_path, enc_dev_ids_path, dec_dev_ids_path, enc_vocab_path, dec_vocab_path)