def fix_duplicate_orcids(p):
from collections import defaultdict
from papers.name import most_similar_author
counts = defaultdict(int)
for a in p.authors:
if a.orcid:
counts[a.orcid] += 1
author_name_pairs = p.author_name_pairs()
best_indices = {}
for orcid, count in list(counts.items()):
if count >= 2:
try:
n = Name.objects.get(researcher__orcid=orcid)
best_author = most_similar_author(n.pair, author_name_pairs)
best_indices[best_author] = orcid
except Name.DoesNotExist:
logger.exception('DUPLICATE ORCID WITH NO RESEARCHER_ID')
logger.info(best_indices)
for idx, a in enumerate(p.authors):
if a.orcid and counts[a.orcid] >= 2:
if best_indices.get(idx) != a.orcid:
# delete
p.authors_list[idx]['orcid'] = None
p.authors_list[idx]['researcher_id'] = None
if best_indices:
p.save()
p.update_index()
def get_form_initial_data(self):
data = super(HALProtocol, self).get_form_initial_data()
data['first_name'] = self.user.first_name
data['last_name'] = self.user.last_name
# Abstract
if self.paper.abstract:
data['abstract'] = kill_html(self.paper.abstract)
else:
self.paper.consolidate_metadata(wait=False)
# Topic
topic_text = ''
if 'abstract' in data:
topic_text = data['abstract']
else:
topic_text = self.paper.title
data['topic'] = self.predict_topic(topic_text)
if data['topic'] == 'OTHER':
del data['topic']
# Depositing author
most_similar_idx = None
first, last = (self.user.first_name, self.user.last_name)
if first and last:
most_similar_idx = most_similar_author((first,last),
self.paper.author_name_pairs())
data['depositing_author'] = most_similar_idx
return data
def get_form_initial_data(self, **kwargs):
data = super(HALProtocol, self).get_form_initial_data(**kwargs)
data['first_name'] = self.user.first_name
data['last_name'] = self.user.last_name
# Abstract
if self.paper.abstract:
data['abstract'] = kill_html(self.paper.abstract)
else:
self.paper.consolidate_metadata(wait=False)
# Topic
topic_text = ''
if 'abstract' in data:
topic_text = data['abstract']
else:
topic_text = self.paper.title
data['topic'] = self.predict_topic(topic_text)
if data['topic'] == 'OTHER':
del data['topic']
# Depositing author
most_similar_idx = None
first, last = (self.user.first_name, self.user.last_name)
if first and last:
most_similar_idx = most_similar_author(
(first, last), self.paper.author_name_pairs())
data['depositing_author'] = most_similar_idx
return data
def fix_duplicate_orcids(p):
from collections import defaultdict
from papers.name import most_similar_author
counts = defaultdict(int)
for a in p.authors:
if a.orcid:
counts[a.orcid] += 1
author_name_pairs = p.author_name_pairs()
best_indices = {}
for orcid, count in list(counts.items()):
if count >= 2:
try:
n = Name.objects.get(researcher__orcid=orcid)
best_author = most_similar_author(n.pair, author_name_pairs)
best_indices[best_author] = orcid
except Name.DoesNotExist:
logger.exception('DUPLICATE ORCID WITH NO RESEARCHER_ID')
logger.info(best_indices)
for idx, a in enumerate(p.authors):
if a.orcid and counts[a.orcid] >= 2:
if best_indices.get(idx) != a.orcid:
# delete
p.authors_list[idx]['orcid'] = None
p.authors_list[idx]['researcher_id'] = None
if best_indices:
p.save()
p.update_index()
def affiliate_author_with_orcid(ref_name, orcid, authors, initial_orcids=None):
"""
Given a reference name and an ORCiD for a researcher, find out which
author in the list is the most likely to be that author. This function
is run on author lists of papers listed in the ORCiD record so we expect
that one of the authors should be the same person as the ORCiD holder.
This just finds the most similar name and returns the appropriate orcids
list (None everywhere except for the most similar name where it is the ORCiD).
"""
max_sim_idx = most_similar_author(ref_name, authors)
orcids = [None] * len(authors)
if initial_orcids and len(initial_orcids) == len(authors):
orcids = initial_orcids
if max_sim_idx is not None:
orcids[max_sim_idx] = orcid
return orcids