def sample_remote_indexes():
print()
print("SimIndexCollection with remote backend indexes")
processes = []
for i in range(2):
port = 9000 + i
process = Process(target=sim_server.start_sim_index_server,
kwargs={'port': port, 'logRequests': False})
process.daemon = True
processes.append(process)
for process in processes:
process.start()
print("Waiting for servers to start")
time.sleep(1)
remote_index_coll = SimIndexCollection()
for i in range(2):
port = 9000 + i
remote_index_coll.add_shards(
RemoteSimIndex("http://localhost:{}/RPC2".format(port)))
remote_index_coll.set_query_scorer('tfidf')
remote_index_coll.index_urls('http://www.stanford.edu/',
'http://www.berkeley.edu',
'http://www.ucla.edu',
'http://www.mit.edu')
pprint(remote_index_coll.query('stanford university'))
for process in processes:
process.terminate()
def sample_sim_index_collection():
# SimIndexCollection
print()
print("SimIndexCollection: build a collection, index some urls, and query it")
indexes = (MemorySimIndex(), MemorySimIndex())
index_coll = SimIndexCollection()
index_coll.add_shards(*indexes)
index_coll.set_query_scorer('tfidf')
index_coll.index_urls('http://www.stanford.edu/',
'http://www.berkeley.edu',
'http://www.ucla.edu',
'http://www.mit.edu')
pprint(index_coll.query('stanford university'))
def sample_sim_index_collection():
# SimIndexCollection
print()
print(
"SimIndexCollection: build a collection, index some urls, and query it"
)
indexes = (MemorySimIndex(), MemorySimIndex())
index_coll = SimIndexCollection()
index_coll.add_shards(*indexes)
index_coll.set_query_scorer('tfidf')
index_coll.index_urls('http://www.stanford.edu/',
'http://www.berkeley.edu', 'http://www.ucla.edu',
'http://www.mit.edu')
pprint(index_coll.query('stanford university'))
def sample_remote_indexes():
print()
print("SimIndexCollection with remote backend indexes")
processes = []
for i in range(2):
port = 9000 + i
process = Process(target=sim_server.start_sim_index_server,
kwargs={
'port': port,
'logRequests': False
})
process.daemon = True
processes.append(process)
for process in processes:
process.start()
print("Waiting for servers to start")
time.sleep(1)
remote_index_coll = SimIndexCollection()
for i in range(2):
port = 9000 + i
remote_index_coll.add_shards(
RemoteSimIndex("http://localhost:{}/RPC2".format(port)))
remote_index_coll.set_query_scorer('tfidf')
remote_index_coll.index_urls('http://www.stanford.edu/',
'http://www.berkeley.edu',
'http://www.ucla.edu', 'http://www.mit.edu')
pprint(remote_index_coll.query('stanford university'))
for process in processes:
process.terminate()
