def citations_nb_counts():
"""Get number of citations for the record `recid`."""
recid = request.view_args.get("recid")
if recid is None:
return
from intbitset import intbitset
from invenio.legacy.bibrank.citation_searcher import get_cited_by, get_cited_by_count
if CFG_BIBRANK_SHOW_CITATION_LINKS:
if CFG_INSPIRE_SITE:
from invenio.legacy.search_engine import search_unit
citers_recids = intbitset(get_cited_by(recid))
citeable_recids = search_unit(p="citeable", f="collection")
return len(citers_recids & citeable_recids)
else:
return get_cited_by_count(recid)
return 0
def citations_nb_counts():
"""Get number of citations for the record `recid`."""
recid = request.view_args.get('recid')
if recid is None:
return
from intbitset import intbitset
from invenio.legacy.bibrank.citation_searcher import (get_cited_by,
get_cited_by_count)
if CFG_BIBRANK_SHOW_CITATION_LINKS:
if CFG_INSPIRE_SITE:
from invenio.legacy.search_engine import search_unit
citers_recids = intbitset(get_cited_by(recid))
citeable_recids = search_unit(p='citeable', f='collection')
return len(citers_recids & citeable_recids)
else:
return get_cited_by_count(recid)
return 0
def deleted_recids_cache(cache={}):
if 'deleted_records' not in cache:
cache['deleted_records'] = search_unit(p='DELETED', f='980', m='a')
return cache['deleted_records']
def deleted_recids_cache(cache={}):
if "deleted_records" not in cache:
cache["deleted_records"] = search_unit(p="DELETED", f="collection", m="a")
return cache["deleted_records"]
def deleted_recids_cache(cache={}):
if 'deleted_records' not in cache:
cache['deleted_records'] = search_unit(p='DELETED', f='980', m='a')
return cache['deleted_records']
def ref_analyzer(citation_informations, dicts,
updated_recids, tags, do_catchup=True):
"""Analyze the citation informations and calculate the citation weight
and cited by list dictionary.
"""
citations_weight = dicts['cites_weight']
citations = dicts['cites']
references = dicts['refs']
selfcites = dicts['selfcites']
selfrefs = dicts['selfrefs']
authorcites = dicts['authorcites']
def step(msg_prefix, recid, done, total):
if done % 30 == 0:
task_sleep_now_if_required()
if done % 1000 == 0:
mesg = "%s done %s of %s" % (msg_prefix, done, total)
write_message(mesg)
task_update_progress(mesg)
write_message("Processing: %s" % recid, verbose=9)
def add_to_dicts(citer, cited):
# Make sure we don't add ourselves
# Workaround till we know why we are adding ourselves.
if citer == cited:
return
if cited not in citations_weight:
citations_weight[cited] = 0
# Citations and citations weight
if citer not in citations.setdefault(cited, []):
citations[cited].append(citer)
citations_weight[cited] += 1
# References
if cited not in references.setdefault(citer, []):
references[citer].append(cited)
# dict of recid -> institute_give_publ_id
records_info, references_info = citation_informations
t1 = os.times()[4]
write_message("Phase 0: temporarily remove changed records from " \
"citation dictionaries; they will be filled later")
if do_catchup:
for somerecid in updated_recids:
try:
del citations[somerecid]
except KeyError:
pass
for somerecid in updated_recids:
try:
del references[somerecid]
except KeyError:
pass
# Try to find references based on 999C5r
# e.g 8 -> ([astro-ph/9889],[hep-ph/768])
# meaning: rec 8 contains these in bibliography
write_message("Phase 1: Report numbers references")
done = 0
for thisrecid, refnumbers in iteritems(references_info['report-numbers']):
step("Report numbers references", thisrecid, done,
len(references_info['report-numbers']))
done += 1
for refnumber in (r for r in refnumbers if r):
field = 'reportnumber'
refnumber = standardize_report_number(refnumber)
# Search for "hep-th/5644654 or such" in existing records
recids = get_recids_matching_query(p=refnumber, f=field)
write_message("These match searching %s in %s: %s" % \
(refnumber, field, list(recids)), verbose=9)
if not recids:
insert_into_missing(thisrecid, refnumber)
else:
remove_from_missing(refnumber)
if len(recids) > 1:
store_citation_warning('multiple-matches', refnumber)
msg = "Whoops: record '%d' report number value '%s' " \
"matches many records; taking only the first one. %s" % \
(thisrecid, refnumber, repr(recids))
write_message(msg, stream=sys.stderr)
for recid in list(recids)[:1]: # take only the first one
add_to_dicts(thisrecid, recid)
mesg = "done fully"
write_message(mesg)
task_update_progress(mesg)
t2 = os.times()[4]
# Try to find references based on 999C5s
# e.g. Phys.Rev.Lett. 53 (1986) 2285
write_message("Phase 2: Journal references")
done = 0
for thisrecid, refs in iteritems(references_info['journals']):
step("Journal references", thisrecid, done,
len(references_info['journals']))
done += 1
for reference in (r for r in refs if r):
p = reference
field = 'journal'
# check reference value to see whether it is well formed:
if not re_CFG_JOURNAL_PUBINFO_STANDARD_FORM_REGEXP_CHECK.match(p):
store_citation_warning('not-well-formed', p)
msg = "Whoops, record '%d' reference value '%s' " \
"is not well formed; skipping it." % (thisrecid, p)
write_message(msg, stream=sys.stderr)
continue # skip this ill-formed value
recids = search_unit(p, field) - INTBITSET_OF_DELETED_RECORDS
write_message("These match searching %s in %s: %s" \
% (reference, field, list(recids)), verbose=9)
if not recids:
insert_into_missing(thisrecid, p)
else:
remove_from_missing(p)
if len(recids) > 1:
store_citation_warning('multiple-matches', p)
msg = "Whoops: record '%d' reference value '%s' " \
"matches many records; taking only the first one. %s" % \
(thisrecid, p, repr(recids))
write_message(msg, stream=sys.stderr)
for recid in list(recids)[:1]: # take only the first one
add_to_dicts(thisrecid, recid)
mesg = "done fully"
write_message(mesg)
task_update_progress(mesg)
t3 = os.times()[4]
# Try to find references based on 999C5a
# e.g. 10.1007/BF03170733
write_message("Phase 3: DOI references")
done = 0
for thisrecid, refs in iteritems(references_info['doi']):
step("DOI references", thisrecid, done, len(references_info['doi']))
done += 1
for reference in (r for r in refs if r):
p = reference
field = 'doi'
recids = get_recids_matching_query(p, field)
write_message("These match searching %s in %s: %s" \
% (reference, field, list(recids)), verbose=9)
if not recids:
insert_into_missing(thisrecid, p)
else:
remove_from_missing(p)
if len(recids) > 1:
store_citation_warning('multiple-matches', p)
msg = "Whoops: record '%d' DOI value '%s' " \
"matches many records; taking only the first one. %s" % \
(thisrecid, p, repr(recids))
write_message(msg, stream=sys.stderr)
for recid in list(recids)[:1]: # take only the first one
add_to_dicts(thisrecid, recid)
mesg = "done fully"
write_message(mesg)
task_update_progress(mesg)
t4 = os.times()[4]
# Search for stuff like CERN-TH-4859/87 in list of refs
write_message("Phase 4: report numbers catchup")
done = 0
for thisrecid, reportcodes in iteritems(records_info['report-numbers']):
step("Report numbers catchup", thisrecid, done,
len(records_info['report-numbers']))
done += 1
for reportcode in (r for r in reportcodes if r):
if reportcode.startswith('arXiv'):
std_reportcode = standardize_report_number(reportcode)
report_pattern = r'^%s( *\[[a-zA-Z.-]*\])?' % \
re.escape(std_reportcode)
recids = get_recids_matching_query(report_pattern,
tags['refs_report_number'],
'r')
else:
recids = get_recids_matching_query(reportcode,
tags['refs_report_number'],
'e')
for recid in recids:
add_to_dicts(recid, thisrecid)
mesg = "done fully"
write_message(mesg)
task_update_progress(mesg)
# Find this record's pubinfo in other records' bibliography
write_message("Phase 5: journals catchup")
done = 0
t5 = os.times()[4]
for thisrecid, rec_journals in iteritems(records_info['journals']):
step("Journals catchup", thisrecid, done,
len(records_info['journals']))
done += 1
for journal in rec_journals:
journal = journal.replace("\"", "")
# Search the publication string like
# Phys. Lett., B 482 (2000) 417 in 999C5s
recids = search_unit(p=journal, f=tags['refs_journal'], m='a') \
- INTBITSET_OF_DELETED_RECORDS
write_message("These records match %s in %s: %s" \
% (journal, tags['refs_journal'], list(recids)), verbose=9)
for recid in recids:
add_to_dicts(recid, thisrecid)
mesg = "done fully"
write_message(mesg)
task_update_progress(mesg)
write_message("Phase 6: DOI catchup")
done = 0
t6 = os.times()[4]
for thisrecid, dois in iteritems(records_info['doi']):
step("DOI catchup", thisrecid, done, len(records_info['doi']))
done += 1
for doi in dois:
# Search the publication string like
# Phys. Lett., B 482 (2000) 417 in 999C5a
recids = search_unit(p=doi, f=tags['refs_doi'], m='a') \
- INTBITSET_OF_DELETED_RECORDS
write_message("These records match %s in %s: %s" \
% (doi, tags['refs_doi'], list(recids)), verbose=9)
for recid in recids:
add_to_dicts(recid, thisrecid)
mesg = "done fully"
write_message(mesg)
task_update_progress(mesg)
write_message("Phase 7: remove empty lists from dicts")
# Remove empty lists in citation and reference
keys = citations.keys()
for k in keys:
if not citations[k]:
del citations[k]
keys = references.keys()
for k in keys:
if not references[k]:
del references[k]
if task_get_task_param('verbose') >= 3:
# Print only X first to prevent flood
write_message("citation_list (x is cited by y):")
write_message(dict(islice(iteritems(citations), 10)))
write_message("size: %s" % len(citations))
write_message("reference_list (x cites y):")
write_message(dict(islice(iteritems(references), 10)))
write_message("size: %s" % len(references))
write_message("selfcitedbydic (x is cited by y and one of the " \
"authors of x same as y's):")
write_message(dict(islice(iteritems(selfcites), 10)))
write_message("size: %s" % len(selfcites))
write_message("selfdic (x cites y and one of the authors of x " \
"same as y's):")
write_message(dict(islice(iteritems(selfrefs), 10)))
write_message("size: %s" % len(selfrefs))
write_message("authorcitdic (author is cited in recs):")
write_message(dict(islice(iteritems(authorcites), 10)))
write_message("size: %s" % len(authorcites))
t7 = os.times()[4]
write_message("Execution time for analyzing the citation information " \
"generating the dictionary:")
write_message("... checking ref report numbers: %.2f sec" % (t2-t1))
write_message("... checking ref journals: %.2f sec" % (t3-t2))
write_message("... checking ref DOI: %.2f sec" % (t4-t3))
write_message("... checking rec report numbers: %.2f sec" % (t5-t4))
write_message("... checking rec journals: %.2f sec" % (t6-t5))
write_message("... checking rec DOI: %.2f sec" % (t7-t6))
write_message("... total time of ref_analyze: %.2f sec" % (t7-t1))
return citations_weight, citations, references, selfcites, \
selfrefs, authorcites
deserialize_via_marshal
from invenio.modules.indexer.tokenizers.BibIndexJournalTokenizer import \
CFG_JOURNAL_PUBINFO_STANDARD_FORM, \
CFG_JOURNAL_PUBINFO_STANDARD_FORM_REGEXP_CHECK
from invenio.legacy.search_engine import search_pattern, search_unit
from invenio.legacy.bibrecord import get_fieldvalues
from invenio.modules.formatter.utils import parse_tag
from invenio.modules.knowledge.api import get_kb_mappings
from invenio.legacy.bibsched.bibtask import write_message, task_get_option, \
task_update_progress, task_sleep_now_if_required, \
task_get_task_param
from invenio.ext.logging import register_exception
from invenio.legacy.bibindex.engine import get_field_tags
INTBITSET_OF_DELETED_RECORDS = search_unit(p='DELETED', f='980', m='a')
re_CFG_JOURNAL_PUBINFO_STANDARD_FORM_REGEXP_CHECK = re.compile(CFG_JOURNAL_PUBINFO_STANDARD_FORM_REGEXP_CHECK)
def get_recids_matching_query(p, f, m='e'):
"""Return set of recIDs matching query for pattern p in field f."""
return search_pattern(p=p, f=f, m=m) - INTBITSET_OF_DELETED_RECORDS
def get_citation_weight(rank_method_code, config, chunk_size=20000):
"""return a dictionary which is used by bibrank daemon for generating
the index of sorted research results by citation information
"""
begin_time = time.time()
quick = task_get_option("quick") != "no"
