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
|
import sys
import json
import itertools
import fatcat_client
from .common import EntityImporter, clean
def or_none(s):
if s is None:
return None
if len(s) == 0:
return None
return s
def truthy(s):
if s is None:
return None
s = s.lower()
if s in ('true', 't', 'yes', 'y', '1'):
return True
elif s in ('false', 'f', 'no', 'n', '0'):
return False
else:
return None
class JournalMetadataImporter(EntityImporter):
"""
Imports journal metadata ("containers") by ISSN, currently from a custom
munged JSON format (see ../extra/journal_metadata/).
See guide for details on the many 'extra' fields used here.
"""
def __init__(self, api, **kwargs):
eg_desc = kwargs.get('editgroup_description',
"Automated import of container-level metadata, by ISSN. Metadata from Internet Archive munging.")
eg_extra = kwargs.get('editgroup_extra', dict())
eg_extra['agent'] = eg_extra.get('agent', 'fatcat_tools.JournalMetadataImporter')
super().__init__(api,
editgroup_description=eg_desc,
editgroup_extra=eg_extra)
def want(self, raw_record):
if raw_record.get('issnl') and raw_record.get('name'):
return True
return False
def parse_record(self, row):
"""
row is a python dict (parsed from JSON).
returns a ContainerEntity (or None if invalid or couldn't parse)
"""
if not row.get('name'):
# Name is required (by schema)
return None
extra = dict()
for key in ('issne', 'issnp', 'languages', 'country', 'urls', 'abbrev',
'coden', 'aliases', 'original_name', 'first_year', 'last_year',
'platform', 'default_license', 'road', 'mimetypes',
'sherpa_romeo', 'kbart'):
if row.get(key):
extra[key] = row[key]
# TODO: not including for now: norwegian, dois/crossref, ia
extra_doaj = dict()
if row.get('doaj'):
if row['doaj'].get('as_of'):
extra_doaj['as_of'] = row['doaj']['as_of']
if row['doaj'].get('works'):
extra_doaj['works'] = row['doaj']['works']
if extra_doaj:
extra['doaj'] = extra_doaj
extra_ia = dict()
# TODO: would like an ia.longtail_ia flag
if row.get('sim'):
# NB: None case of the .get() here is blech, but othrwise
# extra['ia'].get('sim') would be false-y, breaking 'any_ia_sim' later on
extra_ia['sim'] = {
'year_spans': row['sim'].get('year_spans'),
}
if extra_ia:
extra['ia'] = extra_ia
ce = fatcat_client.ContainerEntity(
issnl=row['issnl'],
container_type=None, # TODO
name=clean(row.get('name')),
publisher=clean(row.get('publisher')),
wikidata_qid=None, # TODO
extra=extra)
return ce
def try_update(self, ce):
existing = None
try:
existing = self.api.lookup_container(issnl=ce.issnl)
except fatcat_client.rest.ApiException as err:
if err.status != 404:
raise err
# doesn't exist, need to update
return True
# eventually we'll want to support "updates", but for now just skip if
# entity already exists
if existing:
self.counts['exists'] += 1
return False
return True
def insert_batch(self, batch):
self.api.create_container_batch(batch,
autoaccept=True,
description=self.editgroup_description,
extra=json.dumps(self.editgroup_extra))
|