total = mm.size()
while mm.tell() < total:
yield next(i), mm.readline()
mm.close()
# map :: (k1,v1) -> [ (k2, v2) ]
def mapfn(k1, v):
for w in v.bytes.split():
yield w, 1
# reduce :: (k2, [v2]) -> [ (k3, v3) ]
def reducefn(k2, v):
return sum(v)
# Server
s = Server()
s.connect()
s.mapfn = mapfn
s.reducefn = reducefn
s.datafn = datafn
start = time.time()
s.start()
stop = time.time()
print stop-start
#print s.results()
print sorted(s.results().iteritems(), key=lambda x: x[1], reverse=True)[1:25]
########NEW FILE########
yield next(i), mm.readline()
mm.close()
# map :: (k1,v1) -> [ (k2, v2) ]
def mapfn(k1, v):
for w in v.bytes.split():
yield w, 1
# reduce :: (k2, [v2]) -> [ (k3, v3) ]
def reducefn(k2, v):
return sum(v)
# Server
s = Server()
s.connect()
s.mapfn = mapfn
s.reducefn = reducefn
s.datafn = datafn
start = time.time()
s.start()
stop = time.time()
print stop - start
# print s.results()
print sorted(s.results().iteritems(), key=lambda x: x[1], reverse=True)[1:25]
from kaylee import Server
# Example
# -----------------------------------------------
f = open('mobydick.txt')
data = dict(enumerate(f.readlines()))
def mapfn(k, v):
for w in v.split():
yield w, 1
def reducefn(k, v):
return sum(v)
# Server
s = Server()
s.connect()
s.mapfn = mapfn
s.reducefn = reducefn
s.data = data
start = time.time()
s.start()
stop = time.time()
print stop-start
#print s.results()
