def __init__(self):
def cache_filler():
alldicts = {}
try:
res = run_sql(
"SELECT object_name,object_value FROM rnkCITATIONDATA")
except OperationalError:
# database problems, return empty cache
return {}
for row in res:
object_name = row[0]
object_value = row[1]
try:
object_value_dict = deserialize_via_marshal(object_value)
except:
object_value_dict = {}
alldicts[object_name] = object_value_dict
if object_name == 'citationdict':
# for cited:M->N queries, it is interesting to cache also
# some preprocessed citationdict:
alldicts['citationdict_keys'] = object_value_dict.keys()
alldicts['citationdict_keys_intbitset'] = intbitset(
object_value_dict.keys())
return alldicts
def timestamp_verifier():
res = run_sql(
"""SELECT DATE_FORMAT(last_updated, '%Y-%m-%d %H:%i:%s')
FROM rnkMETHOD WHERE name='citation'""")
if res:
return res[0][0]
else:
return '0000-00-00 00:00:00'
DataCacher.__init__(self, cache_filler, timestamp_verifier)
def __init__(self):
def cache_filler():
alldicts = {}
try:
res = run_sql("SELECT object_name,object_value FROM rnkCITATIONDATA")
except OperationalError:
# database problems, return empty cache
return {}
for row in res:
object_name = row[0]
object_value = row[1]
try:
object_value_dict = deserialize_via_marshal(object_value)
except:
object_value_dict = {}
alldicts[object_name] = object_value_dict
if object_name == 'citationdict':
# for cited:M->N queries, it is interesting to cache also
# some preprocessed citationdict:
alldicts['citationdict_keys'] = object_value_dict.keys()
alldicts['citationdict_keys_intbitset'] = intbitset(object_value_dict.keys())
return alldicts
def timestamp_verifier():
res = run_sql("""SELECT DATE_FORMAT(last_updated, '%Y-%m-%d %H:%i:%s')
FROM rnkMETHOD WHERE name='citation'""")
if res:
return res[0][0]
else:
return '0000-00-00 00:00:00'
DataCacher.__init__(self, cache_filler, timestamp_verifier)
def __init__(self):
def cache_filler():
alldicts = {}
try:
res = run_sql("""SELECT object_name,object_value FROM rnkCITATIONDATA""")
except OperationalError:
# database problems, return empty cache
return {}
for row in res:
object_name = row[0]
object_value = row[1]
try:
object_value_dict = marshal.loads(decompress(object_value))
except:
object_value_dict = {}
alldicts[object_name] = object_value_dict
if object_name == "citationdict":
# for cited:M->N queries, it is interesting to cache also
# some preprocessed citationdict:
alldicts["citationdict_keys"] = object_value_dict.keys()
alldicts["citationdict_keys_intbitset"] = intbitset(object_value_dict.keys())
return alldicts
def timestamp_verifier():
return get_table_update_time("rnkCITATIONDATA")
DataCacher.__init__(self, cache_filler, timestamp_verifier)
def __init__(self):
@gcfix
def fill():
alldicts = {}
from invenio.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 weights.iteritems()]
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 cache_filler():
self.cache = None # misfire from pylint: disable=W0201
# this is really defined in DataCacher
return fill()
from invenio.bibrank_tag_based_indexer import get_lastupdated
def timestamp_verifier():
citation_lastupdate = get_lastupdated('citation')
if citation_lastupdate:
return citation_lastupdate.strftime("%Y-%m-%d %H:%M:%S")
else:
return "0000-00-00 00:00:00"
DataCacher.__init__(self, cache_filler, timestamp_verifier)
def __init__(self):
def fill():
alldicts = {}
from invenio.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 weights.iteritems()]
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 cache_filler():
self.cache = None # misfire from pylint: disable=W0201
# this is really defined in DataCacher
return fill()
from invenio.bibrank_tag_based_indexer import get_lastupdated
def timestamp_verifier():
citation_lastupdate = get_lastupdated('citation')
if citation_lastupdate:
return citation_lastupdate.strftime("%Y-%m-%d %H:%M:%S")
else:
return "0000-00-00 00:00:00"
DataCacher.__init__(self, cache_filler, timestamp_verifier)
def __init__(self):
def initial_fill():
alldicts = {}
from invenio.bibrank_tag_based_indexer import fromDB
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 weights.iteritems()]
alldicts['citations_counts'].sort(key=itemgetter(1), reverse=True)
# Self-cites
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 incremental_fill():
self.cache = None
return initial_fill()
def cache_filler():
if self.cache:
cache = incremental_fill()
else:
cache = initial_fill()
return cache
from invenio.bibrank_tag_based_indexer import get_lastupdated
def timestamp_verifier():
citation_lastupdate = get_lastupdated('citation')
if citation_lastupdate:
return citation_lastupdate.strftime("%Y-%m-%d %H:%M:%S")
else:
return "0000-00-00 00:00:00"
DataCacher.__init__(self, cache_filler, timestamp_verifier)