def do_upgrade():
logger = Logger("Rabbit m_name upgrade script")
warnings.filterwarnings('ignore')
run_sql(
"alter table aidPERSONIDPAPERS add `m_name` VARCHAR(255) not null after name"
)
run_sql("alter table aidPERSONIDPAPERS add INDEX `m_name-b` (`m_name`)")
present_bibrefs = set(
run_sql("select bibref_table, bibref_value from aidPERSONIDPAPERS"))
total_updates = len(present_bibrefs)
records_for_rabbit = set()
for i, bibref in enumerate(present_bibrefs):
logger.update_status(
float(i) / total_updates,
'%s out of %s (%s)' % (str(i), str(total_updates), str(bibref)))
try:
name = get_name_by_bibref(bibref)
except AssertionError, error:
if "A bibref must have exactly one name" in error.message:
records_for_rabbit.add(bibref[1])
else:
raise error
else:
m_name = create_matchable_name(name)
run_sql(
"update aidPERSONIDPAPERS set name=%s, m_name=%s where bibref_table=%s "
"and bibref_value=%s ", (name, m_name, bibref[0], bibref[1]))
def show_papers(personid, external_id=None, orcid=None, inspire=None):
search = 'select * from aidPERSONIDPAPERS where personid='
search += str(personid) + ' and flag>-2'
result = run_sql(search)
hep_records = ''
for personid, table, bibref, bibrec, author, match, flag, cul, date \
in result:
#for personid, table, bibref, bibrec, author in result:
#rec = AmendableRecord(get_bibrecord(bibrec))
position = -1
author_name = get_name_by_bibref((table, bibref))
for key, value in AmendableRecord(get_bibrecord(bibrec)).\
iterfields(['{0}__%'.format(table, )]):
if (key[0] == '700__a' or key[0] == '100__a') and \
value == author_name:
position = key[1]
if position >= 0:
for key, value in AmendableRecord(get_bibrecord(bibrec)).\
iterfields(['{0}__%'.format(table, )]):
if key[1] == position and key[0] in \
('{0}__a'.format(table), '{0}__i'.
format(table), '{0}__j'.format(table),
'{0}__k'.format(table), '{0}__m'.format(table), ):
if value.replace('ORCID:', '') == external_id and \
value.replace('ORCID:', '') != orcid and \
value != inspire:
hep_records += " " + " ".join([str(bibrec),
author, value, '\n'])
if hep_records:
return hep_records
return None
def show_papers(personid, external_id=None, orcid=None, inspire=None):
search = 'select * from aidPERSONIDPAPERS where personid='
search += str(personid) + ' and flag>-2'
result = run_sql(search)
hep_records = ''
for personid, table, bibref, bibrec, author, match, flag, cul, date \
in result:
#for personid, table, bibref, bibrec, author in result:
#rec = AmendableRecord(get_bibrecord(bibrec))
position = -1
author_name = get_name_by_bibref((table, bibref))
for key, value in AmendableRecord(get_bibrecord(bibrec)).\
iterfields(['{0}__%'.format(table, )]):
if (key[0] == '700__a' or key[0] == '100__a') and \
value == author_name:
position = key[1]
if position >= 0:
for key, value in AmendableRecord(get_bibrecord(bibrec)).\
iterfields(['{0}__%'.format(table, )]):
if key[1] == position and key[0] in \
('{0}__a'.format(table), '{0}__i'.
format(table), '{0}__j'.format(table),
'{0}__k'.format(table), '{0}__m'.format(table), ):
if value.replace('ORCID:', '') == external_id and \
value.replace('ORCID:', '') != orcid and \
value != inspire:
hep_records += " " + " ".join([str(bibrec),
author, value, '\n'])
if hep_records:
return hep_records
return None
def export_to_dot(cs, fname, graph_info, extra_edge=None):
from invenio.bibauthorid_dbinterface import get_name_by_bibref
fptr = open(fname, "w")
fptr.write("graph wedgy {\n")
fptr.write(" overlap=prism\n")
for idx, bib in enumerate(graph_info):
fptr.write(' %d [color=black label="%s"];\n' %
(idx, get_name_by_bibref(idx)))
if extra_edge:
v1, v2, (prob, cert) = extra_edge
fptr.write(' %d -- %d [color=green label="p: %.2f, c: %.2f"];\n' %
(v1, v2, prob, cert))
for clus in cs.clusters:
fptr.write(" %s [color=blue];\n" %
" -- ".join(str(x) for x in clus.bibs))
fptr.write("".join(" %d -- %d [color=red]\n" % (b1, b2)
for b1 in clus.bibs for h in clus.hate
for b2 in h.bibs))
fptr.write("}")
def process_authors(pids):
# get the author - paper associations for authors
# with id: min_pid <= id < max_pid (excluding rejected papers)
associations = set(
run_sql("""select personid, bibref_table, bibref_value, bibrec
from aidPERSONIDPAPERS
where flag > -2
and personid in ( %s ) """ %
' , '.join("'" + str(p) + "'" for p in pids)))
# get mapping of: (bibref_table, bibref_value) -> name
bibref_to_name = dict()
bibrefs = set([(table, ref) for _, table, ref, _ in associations])
counter = 0
for bibref in bibrefs:
counter += 1
bibref_to_name[bibref] = get_name_by_bibref(bibref)
write_message('Got name: %s for bibref: %s.' %
(bibref_to_name[bibref], str(bibref)))
# get mapping of: author -> affiliations
pid_to_affiliations = dict()
counterr = 0
for pid in pids:
counterr += 1
pid_to_affiliations[pid] = get_affiliations_for_author(pid)
write_message('Got affiliations: %s for author: %s.' %
(pid_to_affiliations[pid], str(pid)))
# get the affiliated records for this group of authors: (rec, name, affiliation, ... + some extra info)
affiliated_records = list()
for pid, table, ref, rec in associations:
# we don't want to keep records which have no affiliation/s
affiliations, year = pid_to_affiliations[pid]
if not affiliations:
continue
affiliated_records += [
rec, bibref_to_name[(table, ref)],
serialize(affiliations), table, ref, year
]
# flush data to db
if affiliated_records:
write_message('Populating table with %s records.' %
len(affiliated_records))
populate_table('bibEDITAFFILIATIONS_tmp', [
'bibrec', 'name', 'affiliations', 'bibref_table', 'bibref_value',
'year'
],
affiliated_records,
empty_table_first=False)
return len(affiliated_records)
def get_message_body(self):
"""Return the body of the message to be reported by the exception"""
msg = ['Found wrong signature claimed to profile ']
try:
cname = get_canonical_name_of_author(self.pid)[0]
except IndexError:
cname = self.pid
msg.append("%s/author/profile/%s" % (CFG_SITE_URL, cname))
sig_name = get_name_by_bibref(self.signature[0:2])
p_sigs = [(x, get_name_by_bibref(x[0:2]))
for x in self.present_signatures]
p_sig_strings = ",".join('%s (%s on record %s)' % (x[0], x[1], x[0][2])
for x in p_sigs)
msg.append(
"want to move %s (%s on record %s) to this profile but [%s] are already present and claimed"
% (self.signature, sig_name, self.signature[2], p_sig_strings))
msg.append("%s/record/%s" % (CFG_SITE_URL, self.signature[2]))
# msg.append(self.message)
return '\n'.join(msg)
def process_authors(pids):
# get the author - paper associations for authors
# with id: min_pid <= id < max_pid (excluding rejected papers)
associations = set(run_sql("""select personid, bibref_table, bibref_value, bibrec
from aidPERSONIDPAPERS
where flag > -2
and personid in ( %s ) """ % ' , '.join("'"+str(p)+"'" for p in pids)))
# get mapping of: (bibref_table, bibref_value) -> name
bibref_to_name = dict()
bibrefs = set([(table, ref) for _, table, ref, _ in associations])
counter = 0
for bibref in bibrefs:
counter += 1
bibref_to_name[bibref] = get_name_by_bibref(bibref)
write_message('Got name: %s for bibref: %s.' % (bibref_to_name[bibref], str(bibref)))
# get mapping of: author -> affiliations
pid_to_affiliations = dict()
counterr = 0
for pid in pids:
counterr += 1
pid_to_affiliations[pid] = get_affiliations_for_author(pid)
write_message('Got affiliations: %s for author: %s.' % (pid_to_affiliations[pid], str(pid)))
# get the affiliated records for this group of authors: (rec, name, affiliation, ... + some extra info)
affiliated_records = list()
for pid, table, ref, rec in associations:
# we don't want to keep records which have no affiliation/s
affiliations, year = pid_to_affiliations[pid]
if not affiliations:
continue
affiliated_records += [rec, bibref_to_name[(table, ref)], serialize(affiliations), table, ref, year]
# flush data to db
if affiliated_records:
write_message('Populating table with %s records.' % len(affiliated_records))
populate_table('bibEDITAFFILIATIONS_tmp', ['bibrec','name','affiliations','bibref_table','bibref_value', 'year'], affiliated_records, empty_table_first=False)
return len(affiliated_records)
def export_to_dot(cs, fname, graph_info, extra_edge=None):
from invenio.bibauthorid_dbinterface import get_name_by_bibref
fptr = open(fname, "w")
fptr.write("graph wedgy {\n")
fptr.write(" overlap=prism\n")
for idx, bib in enumerate(graph_info):
fptr.write(' %d [color=black label="%s"];\n' % (idx, get_name_by_bibref(idx)))
if extra_edge:
v1, v2, (prob, cert) = extra_edge
fptr.write(' %d -- %d [color=green label="p: %.2f, c: %.2f"];\n' % (v1, v2, prob, cert))
for clus in cs.clusters:
fptr.write(" %s [color=blue];\n" % " -- ".join(str(x) for x in clus.bibs))
fptr.write("".join(" %d -- %d [color=red]\n" % (b1, b2)
for b1 in clus.bibs for h in clus.hate for b2 in h.bibs))
fptr.write("}")
