def id_lookup(paper_id, idtype):
"""Take an ID of type PMID, PMCID, or DOI and lookup the other IDs.
If the DOI is not found in Pubmed, try to obtain the DOI by doing a
reverse-lookup of the DOI in CrossRef using article metadata.
Parameters
----------
paper_id : str
ID of the article.
idtype : str
Type of the ID: 'pmid', 'pmcid', or 'doi
Returns
-------
ids : dict
A dictionary with the following keys: pmid, pmcid and doi.
"""
if idtype not in ('pmid', 'pmcid', 'doi'):
raise ValueError("Invalid idtype %s; must be 'pmid', 'pmcid', "
"or 'doi'." % idtype)
ids = {'doi': None, 'pmid': None, 'pmcid': None}
pmc_id_results = pmc_client.id_lookup(paper_id, idtype)
# Start with the results of the PMC lookup and then override with the
# provided ID
ids['pmid'] = pmc_id_results.get('pmid')
ids['pmcid'] = pmc_id_results.get('pmcid')
ids['doi'] = pmc_id_results.get('doi')
ids[idtype] = paper_id
# If we gave a DOI, then our work is done after looking for PMID and PMCID
if idtype == 'doi':
return ids
# If we gave a PMID or PMCID, we need to check to see if we got a DOI.
# If we got a DOI back, we're done.
elif ids.get('doi'):
return ids
# If we get here, then we've given PMID or PMCID and don't have a DOI yet.
# If we gave a PMCID and have neither a PMID nor a DOI, then we'll run
# into problems later on when we try to the reverse lookup using CrossRef.
# So we bail here and return what we have (PMCID only) with a warning.
if ids.get('pmcid') and ids.get('doi') is None and ids.get('pmid') is None:
logger.warning('%s: PMCID without PMID or DOI' % ids.get('pmcid'))
return ids
# To clarify the state of things at this point:
assert ids.get('pmid') is not None
assert ids.get('doi') is None
# As a last result, we try to get the DOI from CrossRef (which internally
# tries to get the DOI from Pubmed in the process of collecting the
# necessary metadata for the lookup):
ids['doi'] = crossref_client.doi_query(ids['pmid'])
# It may still be None, but at this point there's nothing we can do...
return ids
def id_lookup(paper_id, idtype):
"""Take an ID of type PMID, PMCID, or DOI and lookup the other IDs.
If the DOI is not found in Pubmed, try to obtain the DOI by doing a
reverse-lookup of the DOI in CrossRef using article metadata.
Parameters
----------
paper_id : string
ID of the article.
idtype : 'pmid', 'pmcid', or 'doi
Type of the ID.
Returns
-------
ids : dict
A dictionary with the following keys: pmid, pmcid and doi.
"""
if idtype not in ('pmid', 'pmcid', 'doi'):
raise ValueError("Invalid idtype %s; must be 'pmid', 'pmcid', "
"or 'doi'." % idtype)
ids = {'doi': None, 'pmid': None, 'pmcid': None}
pmc_id_results = pmc_client.id_lookup(paper_id, idtype)
# Start with the results of the PMC lookup and then override with the
# provided ID
ids['pmid'] = pmc_id_results.get('pmid')
ids['pmcid'] = pmc_id_results.get('pmcid')
ids['doi'] = pmc_id_results.get('doi')
ids[idtype] = paper_id
# If we gave a DOI, then our work is done after looking for PMID and PMCID
if idtype == 'doi':
return ids
# If we gave a PMID or PMCID, we need to check to see if we got a DOI.
# If we got a DOI back, we're done.
elif ids.get('doi'):
return ids
# If we get here, then we've given PMID or PMCID and don't have a DOI yet.
# If we gave a PMCID and have neither a PMID nor a DOI, then we'll run
# into problems later on when we try to the reverse lookup using CrossRef.
# So we bail here and return what we have (PMCID only) with a warning.
if ids.get('pmcid') and ids.get('doi') is None and ids.get('pmid') is None:
logger.warning('%s: PMCID without PMID or DOI' % ids.get('pmcid'))
return ids
# To clarify the state of things at this point:
assert ids.get('pmid') is not None
assert ids.get('doi') is None
# As a last result, we try to get the DOI from CrossRef (which internally
# tries to get the DOI from Pubmed in the process of collecting the
# necessary metadata for the lookup):
ids['doi'] = crossref_client.doi_query(ids['pmid'])
# It may still be None, but at this point there's nothing we can do...
return ids
def get_missing_pmids(self, tr_data):
"Try to get missing pmids using the pmc client."
num_missing = 0
num_found = 0
logger.debug("Getting missing pmids.")
# TODO: This is very slow...should find a way to speed it up.
for tr_entry in tr_data:
if tr_entry['pmid'] is None:
num_missing += 1
ret = id_lookup(tr_entry['pmcid'])
if 'pmid' in ret.keys():
tr_entry['pmid'] = ret['pmid']
num_found += 1
''' # The web api does not support this much access, sadly.
thread_list = []
for tr_entry in tr_data:
if tr_entry['pmid'] is None:
th = Thread(target=lookup_pmid, args=[tr_entry])
thread_list.append(th)
N = min(10, len(thread_list))
logger.debug("Starting %d threading pool." % N)
active_threads = []
for _ in range(N):
th = thread_list.pop()
th.start()
active_threads.append(th)
while len(thread_list):
for th in active_threads[:]:
if not th.is_alive():
th.join()
active_threads.remove(th)
if len(thread_list):
new_th = thread_list.pop()
new_th.start()
active_threads.append(th)
sleep(0.1)
for th in active_threads:
th.join()
'''
logger.debug("Found %d/%d new pmids." % (num_found, num_missing))
return
def get_text_content_for_pmids(pmids):
"""Get text content for articles given a list of their pmids
Parameters
----------
pmids : list of str
Returns
-------
text_content : list of str
"""
pmc_pmids = set(pmc_client.filter_pmids(pmids, source_type='fulltext'))
pmc_ids = []
for pmid in pmc_pmids:
pmc_id = pmc_client.id_lookup(pmid, idtype='pmid')['pmcid']
if pmc_id:
pmc_ids.append(pmc_id)
else:
pmc_pmids.discard(pmid)
pmc_xmls = []
failed = set()
for pmc_id in pmc_ids:
if pmc_id is not None:
pmc_xmls.append(pmc_client.get_xml(pmc_id))
else:
failed.add(pmid)
time.sleep(0.5)
remaining_pmids = set(pmids) - pmc_pmids | failed
abstracts = []
for pmid in remaining_pmids:
abstract = pubmed_client.get_abstract(pmid)
abstracts.append(abstract)
time.sleep(0.5)
return [
text_content for source in (pmc_xmls, abstracts)
for text_content in source if text_content is not None
]
def get_text_content_for_pmids(pmids):
"""Get text content for articles given a list of their pmids
Parameters
----------
pmids : list of str
Returns
-------
text_content : list of str
"""
pmc_pmids = set(pmc_client.filter_pmids(pmids, source_type='fulltext'))
pmc_ids = []
for pmid in pmc_pmids:
pmc_id = pmc_client.id_lookup(pmid, idtype='pmid')['pmcid']
if pmc_id:
pmc_ids.append(pmc_id)
else:
pmc_pmids.discard(pmid)
pmc_xmls = []
failed = set()
for pmc_id in pmc_ids:
if pmc_id is not None:
pmc_xmls.append(pmc_client.get_xml(pmc_id))
else:
failed.append(pmid)
time.sleep(0.5)
remaining_pmids = set(pmids) - pmc_pmids | failed
abstracts = []
for pmid in remaining_pmids:
abstract = pubmed_client.get_abstract(pmid)
abstracts.append(abstract)
time.sleep(0.5)
return [text_content for source in (pmc_xmls, abstracts)
for text_content in source if text_content is not None]
def test_id_lookup_pmcid_no_prefix_idtype():
ids = pmc_client.id_lookup('4322985', idtype='pmcid')
assert ids['doi'] == example_ids['doi']
assert ids['pmid'] == example_ids['pmid']
assert ids['pmcid'] == example_ids['pmcid']
assert unicode_strs(ids)
def test_id_lookup_pmcid_idtype():
ids = pmc_client.id_lookup('PMC4322985', idtype='pmcid')
assert (ids['doi'] == example_ids['doi'])
assert (ids['pmid'] == example_ids['pmid'])
assert (ids['pmcid'] == example_ids['pmcid'])
assert unicode_strs(ids)
def test_invalid_idtype():
ids = pmc_client.id_lookup('DOI10.18632/oncotarget.2555', idtype='foo')
def test_id_lookup_doi_prefix_no_idtype():
ids = pmc_client.id_lookup('DOI10.18632/oncotarget.2555')
assert ids['doi'] == example_ids['doi']
assert ids['pmid'] == example_ids['pmid']
assert ids['pmcid'] == example_ids['pmcid']
assert unicode_strs(ids)
def test_id_lookup_pmcid_no_prefix_idtype():
ids = pmc_client.id_lookup('4322985', idtype='pmcid')
assert ids['doi'] == example_ids['doi']
assert ids['pmid'] == example_ids['pmid']
assert ids['pmcid'] == example_ids['pmcid']
assert unicode_strs(ids)
def test_id_lookup_doi_prefix_no_idtype():
ids = pmc_client.id_lookup('DOI10.18632/oncotarget.2555')
assert ids['doi'] == example_ids['doi']
assert ids['pmid'] == example_ids['pmid']
assert ids['pmcid'] == example_ids['pmcid']
assert unicode_strs(ids)
def test_invalid_idtype():
ids = pmc_client.id_lookup('DOI10.18632/oncotarget.2555', idtype='foo')
def test_id_lookup_pmid_with_prefix_no_idtype():
ids = pmc_client.id_lookup('PMID25361007')
assert ids['doi'] == example_ids['doi']
assert ids['pmid'] == example_ids['pmid']
assert ids['pmcid'] == example_ids['pmcid']
assert unicode_strs(ids)
def test_id_lookup_pmid_with_prefix_no_idtype():
ids = pmc_client.id_lookup('PMID25361007')
assert ids['doi'] == example_ids['doi']
assert ids['pmid'] == example_ids['pmid']
assert ids['pmcid'] == example_ids['pmcid']
assert unicode_strs(ids)
def test_id_lookup_pmid_no_prefix_no_idtype():
ids = pmc_client.id_lookup("25361007")
assert ids["doi"] == example_ids["doi"]
assert ids["pmid"] == example_ids["pmid"]
assert ids["pmcid"] == example_ids["pmcid"]
assert unicode_strs(ids)
def test_invalid_idtype():
ids = pmc_client.id_lookup("DOI10.18632/oncotarget.2555", idtype="foo")
def test_id_lookup_doi_prefix_no_idtype():
ids = pmc_client.id_lookup("DOI10.18632/oncotarget.2555")
assert ids["doi"] == example_ids["doi"]
assert ids["pmid"] == example_ids["pmid"]
assert ids["pmcid"] == example_ids["pmcid"]
assert unicode_strs(ids)
def test_id_lookup_pmcid_no_prefix_idtype():
ids = pmc_client.id_lookup("4322985", idtype="pmcid")
assert ids["doi"] == example_ids["doi"]
assert ids["pmid"] == example_ids["pmid"]
assert ids["pmcid"] == example_ids["pmcid"]
assert unicode_strs(ids)
