harmonize_metadata.py
8.96 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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
import argparse
import json
import os
METADATA_ORDER = ['Identifier', 'Language', 'Licence', 'PublicationDate', 'DocumentTitle', 'ArticleTitle',
'DocumentType', 'Source', 'Domain', 'No_of_sentences', 'No_of_words', 'No_of_punctuation',
'No_of_tokens', 'Author', 'SourceType', 'Keywords', 'Url', 'Style', 'Subdomain', 'Issn_isbn_eisbn',
'ScientificField', 'EnTitle', 'EnKeywords', 'EnAbstract', 'PublishingCompany', 'IssueYear',
'IssueVolume', 'IssueNumber', 'PageRange', 'Reviewer', 'Translator', 'OriginalType']
def main():
args = parse_arguments()
if not os.path.isdir(args.input):
print('Error: Input must be a root corpora directory!')
metadata_summary = harmonize_metadata(args.input, args.summary, args.output)
sentences = 0
tokens = 0
for meta in metadata_summary:
sentences += meta['No_of_sentences']
tokens += meta['No_of_tokens']
def parse_arguments():
parser = argparse.ArgumentParser(description='Harmonize metadata and write metadata summary to json file.')
required_arguments = parser.add_argument_group('required arguments')
required_arguments.add_argument('-i', '--input', help='corpora root directory', required=True)
required_arguments.add_argument('-o', '--output', help='output directory', required=True)
required_arguments.add_argument('-s', '--summary', help='path to json metadata summary file', required=True)
return parser.parse_args()
def harmonize_metadata(root_directory, summary_json_path, harmonized_outpath):
metadata_summary = []
for root, dirs, files in os.walk(root_directory):
for filename in files:
if filename.endswith('.conllup') or filename.endswith('.conllu'):
src = os.path.join(root, filename)
metadata = get_all_metadata(src)
year = get_year(src)
year_path = os.path.join(harmonized_outpath, year)
os.makedirs(year_path, exist_ok=True)
dst = os.path.join(year_path, filename)
copy_and_harmonize(src, dst, metadata)
metadata_summary.append(metadata)
with open(summary_json_path, 'w') as f:
json.dump(metadata_summary, f, indent=4)
return metadata_summary
def get_all_metadata(filepath):
metadata = {'No_of_sentences': 0,
'No_of_words': 0,
'No_of_punctuation': 0,
'No_of_tokens': 0,
'SourceType': 'Publishing House',
'Style': 'scientific',
'Licence': 'CC0'}
with open(filepath, 'r') as conllup_file:
for line in conllup_file:
line = line.strip()
if is_segment(line):
metadata['No_of_tokens'] += 1
pos = line.split('\t')[3]
if pos == 'PUNCT':
metadata['No_of_punctuation'] += 1
else:
metadata['No_of_words'] += 1
elif is_metadata(line):
name, value = get_metadata(line)
if name == 'sent_id':
metadata['No_of_sentences'] += 1
elif name == 'newdoc id' and value:
metadata['Identifier'] = value
elif name == 'language' and value:
metadata['Language'] = value
elif name == 'license' and value:
metadata['Licence'] = value
elif name == 'date' and value:
if len(value) == 4:
metadata['PublicationDate'] = f'{value}-01-01'
elif len(value) == 7:
metadata['PublicationDate'] = f'{value}-01'
elif len(value) == 10:
metadata['PublicationDate'] = value
else:
print(f'Something wrong with the publication date: {value}.')
elif name == 'title' and value:
metadata['DocumentTitle'] = value
elif name == 'content_type' and value:
if value == 'abstract':
metadata['DocumentType'] = 'abstract'
elif value == 'full_text':
metadata['DocumentType'] = 'paper'
else:
print(f'Unknown content_type value: {value}.')
elif name == 'journal' and value:
metadata['Source'] = value
elif name == 'domain' and value:
metadata['Domain'] = value
elif name == 'authors' and value:
metadata['Author'] = value
elif name == 'url' and value:
metadata['Url'] = value
elif name == 'keywords' and value:
metadata['Keywords'] = value
elif name == 'disciplines' and value:
metadata['Subdomain'] = value
elif name == 'fields' and value:
metadata['ScientificField'] = value
elif name == 'entitle' and value:
metadata['EnTitle'] = value
elif name == 'enkeywords' and value:
metadata['EnKeywords'] = value
elif name == 'enabstract' and value:
metadata['EnAbstract'] = value
elif name == 'publishing_company' and value:
metadata['PublishingCompany'] = value
elif name == 'issue_year' and value:
metadata['IssueYear'] = value
elif name == 'issue_volume' and value:
metadata['IssueVolume'] = value
elif name == 'issue_number' and value:
metadata['IssueNumber'] = value
elif name == 'page_range' and value:
metadata['PageRange'] = value
elif name == 'reviewers' and value:
metadata['Reviewer'] = value
elif name == 'translators' and value:
metadata['Translator'] = value
elif name == 'type' and value:
metadata['OriginalType'] = value
elif name == 'source':
pass
elif name == 'text':
pass
elif name == 'newpar id':
pass
elif name == 'global.columns':
pass
else:
print(f'Unknown metadata name: {name}.')
return metadata
def is_segment(line):
if line and line[0].isdigit():
return True
return False
def is_metadata(line):
if line.startswith('#'):
return True
return False
def get_metadata(line):
name_value_pair = line.split('=', 1)
name = name_value_pair[0].lstrip('#').strip()
value = name_value_pair[1].strip()
return name, value
def is_full_text(filepath):
with open(filepath, 'r') as conllup_file:
for line in conllup_file:
line = line.strip()
if is_segment(line):
continue
elif is_metadata(line):
name, value = get_metadata(line)
if name == 'content_type':
if value.strip() == 'full_text':
return True
else:
return False
return False
def get_year(filepath):
with open(filepath, 'r') as conllup_file:
for line in conllup_file:
line = line.strip()
if is_segment(line):
continue
elif is_metadata(line):
name, value = get_metadata(line)
if name == 'date':
return value.split('-')[0]
return 0
def copy_and_harmonize(src, dst, metadata):
harmonized_lines = []
with open(src, 'r') as conllup_file:
for line in conllup_file:
line = line.strip()
if is_metadata(line):
name, value = get_metadata(line)
if name == 'newdoc id':
harmonized_lines.append(line)
for meta_name in METADATA_ORDER:
if meta_name in metadata:
harmonized_lines.append(f'# {meta_name} = {metadata[meta_name]}')
elif meta_name not in ['Issn_isbn_eisbn', 'Translator', 'EnTitle', 'EnAbstract', 'IssueVolume',
'ArticleTitle', 'Reviewer', 'EnKeywords', 'PageRange', 'Keywords',
'IssueNumber']:
print(f'Missing {meta_name} in {src}.')
elif name in ['sent_id', 'global.columns', 'newpar id', 'text']:
harmonized_lines.append(line)
else:
pass
else:
harmonized_lines.append(line)
with open(dst, 'w') as dst_file:
dst_file.write('\n'.join(harmonized_lines))
if __name__ == '__main__':
main()