def fill():
alldicts = {}
from invenio.legacy.bibrank.tag_based_indexer import fromDB
serialized_weights = cache.get('citations_weights')
if serialized_weights:
weights = deserialize_via_marshal(serialized_weights)
else:
weights = fromDB('citation')
alldicts['citations_weights'] = weights
# for cited:M->N queries, it is interesting to cache also
# some preprocessed citationdict:
alldicts['citations_keys'] = intbitset(weights.keys())
# Citation counts
alldicts['citations_counts'] = [t for t in iteritems(weights)]
alldicts['citations_counts'].sort(key=itemgetter(1), reverse=True)
# Self-cites
serialized_weights = cache.get('selfcites_weights')
if serialized_weights:
selfcites = deserialize_via_marshal(serialized_weights)
else:
selfcites = fromDB('selfcites')
selfcites_weights = {}
for recid, counts in alldicts['citations_counts']:
selfcites_weights[recid] = counts - selfcites.get(recid, 0)
alldicts['selfcites_weights'] = selfcites_weights
alldicts['selfcites_counts'] = [(recid, selfcites_weights.get(recid, cites)) for recid, cites in alldicts['citations_counts']]
alldicts['selfcites_counts'].sort(key=itemgetter(1), reverse=True)
return alldicts
def fill():
alldicts = {}
from invenio.legacy.bibrank.tag_based_indexer import fromDB
redis = get_redis()
serialized_weights = redis.get('citations_weights')
if serialized_weights:
weights = deserialize_via_marshal(serialized_weights)
else:
weights = fromDB('citation')
alldicts['citations_weights'] = weights
# for cited:M->N queries, it is interesting to cache also
# some preprocessed citationdict:
alldicts['citations_keys'] = intbitset(weights.keys())
# Citation counts
alldicts['citations_counts'] = [t for t in iteritems(weights)]
alldicts['citations_counts'].sort(key=itemgetter(1), reverse=True)
# Self-cites
serialized_weights = redis.get('selfcites_weights')
if serialized_weights:
selfcites = deserialize_via_marshal(serialized_weights)
else:
selfcites = fromDB('selfcites')
selfcites_weights = {}
for recid, counts in alldicts['citations_counts']:
selfcites_weights[recid] = counts - selfcites.get(recid, 0)
alldicts['selfcites_weights'] = selfcites_weights
alldicts['selfcites_counts'] = [(recid, selfcites_weights.get(recid, cites)) for recid, cites in alldicts['citations_counts']]
alldicts['selfcites_counts'].sort(key=itemgetter(1), reverse=True)
return alldicts
def process_updates(rank_method_code):
"""
This is what gets executed first when the task is started.
It handles the --rebuild option. If that option is not specified
we fall back to the process_one()
"""
write_message("Running rank method: %s" % rank_method_code, verbose=0)
selfcites_config = read_configuration(rank_method_code)
config = {
'algorithm':
selfcites_config.get(rank_method_code, "algorithm"),
'friends_threshold':
selfcites_config.get(rank_method_code, "friends_threshold")
}
quick = task_get_option("quick") != "no"
if not quick:
return rebuild_tables(rank_method_code, config)
tags = get_authors_tags()
recids, end_date = fetch_concerned_records(rank_method_code,
task_get_option("id"))
citations_fun = get_citations_fun(config['algorithm'])
weights = fromDB(rank_method_code)
write_message("recids %s" % str(recids))
total = len(recids)
for count, recid in enumerate(recids):
task_sleep_now_if_required(can_stop_too=True)
msg = "Extracting for %s (%d/%d)" % (recid, count + 1, total)
task_update_progress(msg)
write_message(msg)
process_one(recid, tags, citations_fun, weights)
intoDB(weights, end_date, rank_method_code)
store_weights_cache(weights)
write_message("Complete")
return True
def process_updates(rank_method_code):
"""
This is what gets executed first when the task is started.
It handles the --rebuild option. If that option is not specified
we fall back to the process_one()
"""
write_message("Running rank method: %s" % rank_method_code, verbose=0)
selfcites_config = read_configuration(rank_method_code)
config = {
'algorithm': selfcites_config.get(rank_method_code, "algorithm"),
'friends_threshold': selfcites_config.get(rank_method_code, "friends_threshold")
}
quick = task_get_option("quick") != "no"
if not quick:
return rebuild_tables(rank_method_code, config)
tags = get_authors_tags()
recids, end_date = fetch_concerned_records(rank_method_code,
task_get_option("id"))
citations_fun = get_citations_fun(config['algorithm'])
weights = fromDB(rank_method_code)
write_message("recids %s" % str(recids))
total = len(recids)
for count, recid in enumerate(recids):
task_sleep_now_if_required(can_stop_too=True)
msg = "Extracting for %s (%d/%d)" % (recid, count + 1, total)
task_update_progress(msg)
write_message(msg)
process_one(recid, tags, citations_fun, weights)
intoDB(weights, end_date, rank_method_code)
store_weights_cache(weights)
write_message("Complete")
return True
