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
225
226
227
228
229
230
231
|
from dataclasses import asdict, dataclass
from typing import List, Optional
@dataclass
class GrobidAddress:
addr_line: Optional[str] = None
post_code: Optional[str] = None
settlement: Optional[str] = None
country: Optional[str] = None
country_code: Optional[str] = None # XXX
@dataclass
class GrobidAffiliation:
institution: Optional[str] = None
department: Optional[str] = None
laboratory: Optional[str] = None
address: Optional[GrobidAddress] = None
@dataclass
class GrobidAuthor:
full_name: Optional[str]
given_name: Optional[str] = None
middle: Optional[str] = None # XXX
surname: Optional[str] = None
suffix: Optional[str] = None # XXX
email: Optional[str] = None # XXX
affiliation: Optional[GrobidAffiliation] = None
def to_csl_dict(self) -> dict:
d = dict(
given=self.given_name,
family=self.surname,
suffix=self.suffix,
)
return _simplify_dict(d)
def _csl_date(s: Optional[str]) -> Optional[list]:
if not s:
return None
# YYYY
if len(s) >= 4 and s[0:4].isdigit():
year = int(s[0:4])
else:
return None
# YYYY-MM
if len(s) >= 7 and s[4] == '-' and s[5:7].isdigit():
month = int(s[5:7])
else:
return [[year]]
# YYYY-MM-DD
if len(s) == 10 and s[7] == '-' and s[8:10].isdigit():
day = int(s[8:10])
return [[year, month, day]]
else:
return [[year, month]]
def test_csl_date() -> None:
assert _csl_date("1998") == [[1998]]
assert _csl_date("1998-03") == [[1998, 3]]
assert _csl_date("1998-03-12") == [[1998, 3, 12]]
assert _csl_date("1998-blah") == [[1998]]
assert _csl_date("asdf") is None
@dataclass
class GrobidCitation:
authors: List[GrobidAuthor]
index: Optional[int] = None
id: Optional[str] = None
date: Optional[str] = None
issue: Optional[str] = None
journal: Optional[str] = None # XXX: venue? other?
publisher: Optional[str] = None
title: Optional[str] = None
url: Optional[str] = None
volume: Optional[str] = None
pages: Optional[str] = None
first_page: Optional[str] = None # XXX
last_page: Optional[str] = None # XXX
unstructured: Optional[str] = None
arxiv_id: Optional[str] = None
doi: Optional[str] = None
pmid: Optional[str] = None
pmcid: Optional[str] = None
oa_url: Optional[str] = None
note: Optional[str] = None
def to_dict(self) -> dict:
return _simplify_dict(asdict(self))
def to_csl_dict(self, default_type: str = "article-journal") -> dict:
"""
Transforms in to Citation Style Language (CSL) JSON schema
"""
csl = dict(
type=default_type,
author=[a.to_csl_dict() for a in self.authors],
issued=_csl_date(self.date),
publisher=self.publisher,
title=self.title,
page=self.pages,
URL=self.url,
DOI=self.doi,
PMID=self.pmid,
PMCID=self.pmcid,
note=self.note,
# fields with '-' in the key name
**{
"container-title": self.journal,
"page-first": self.first_page,
})
# numeric fields
if self.issue and self.issue.isdigit():
csl['issue'] = int(self.issue)
if self.volume and self.volume.isdigit():
csl['volume'] = int(self.volume)
return _simplify_dict(csl)
@dataclass
class GrobidJournal:
name: Optional[str] = None
abbrev: Optional[str] = None
publisher: Optional[str] = None
volume: Optional[str] = None
issue: Optional[str] = None
issn: Optional[str] = None
eissn: Optional[str] = None
@dataclass
class GrobidHeader:
authors: List[GrobidAuthor]
title: Optional[str] = None
date: Optional[str] = None
doi: Optional[str] = None
journal: Optional[GrobidJournal] = None
@dataclass
class GrobidDocument:
grobid_version: str
grobid_timestamp: str
header: GrobidHeader
pdf_md5: Optional[str] = None
language_code: Optional[str] = None
citations: Optional[List[GrobidCitation]] = None
abstract: Optional[str] = None
body: Optional[str] = None
acknowledgement: Optional[str] = None
annex: Optional[str] = None
def to_dict(self) -> dict:
"""
Returns a dict version of this object which has no 'None' fields
(recursively), and is appropriate for serializing to JSON with
json.dumps().
If you did want all the fields, you could use dataclasses.asdict()
directly on thing object.
"""
return _simplify_dict(asdict(self))
def to_legacy_dict(self) -> dict:
"""
Returns a dict in the old "grobid2json" format.
"""
d = self.to_dict()
# all header fields at top-level
d.update(d.pop('header', {}))
# files not in the old schema
d.pop('pdf_md5', None)
for c in d.get('citations', []):
c.pop('note', None)
# author changes
for a in d['authors']:
a['name'] = a.pop('full_name')
addr = a.get('affiliation', {}).get('address')
if addr and addr.get('post_code'):
addr['postCode'] = addr.pop('post_code')
for c in d['citations'] or []:
for a in c['authors']:
a['name'] = a.pop('full_name')
return d
def remove_encumbered(self) -> None:
"""
This helper function removes fields from this object which might raise
copyright concerns.
"""
self.abstract = None
self.body = None
self.acknowledgement = None
self.annex = None
def _simplify_dict(d: dict) -> dict:
"""
Recursively remove empty dict values from a dict and all sub-lists and
sub-dicts.
TODO: should this return Optional[dict]?
"""
if d in [None, {}, '']:
return {}
for k in list(d.keys()):
if isinstance(d[k], dict):
d[k] = _simplify_dict(d[k])
elif isinstance(d[k], list):
for i in range(len(d[k])):
if isinstance(d[k][i], dict):
d[k][i] = _simplify_dict(d[k][i])
if d[k] in [None, {}, '']:
d.pop(k)
return d
|