Converted some examples to use patterns/engines

Change-Id: If7154019f1cb5e723069ff35f6301fce048323b5
This commit is contained in:
Anastasia Karpinska 2013-09-04 11:27:39 +04:00 committed by Ivan A. Melnikov
parent 6ee4d32fc2
commit a638c9864f
6 changed files with 146 additions and 119 deletions

View File

@ -0,0 +1,55 @@
import logging
import os
import sys
logging.basicConfig(level=logging.ERROR)
my_dir_path = os.path.dirname(os.path.abspath(__file__))
sys.path.insert(0, os.path.join(os.path.join(my_dir_path, os.pardir),
os.pardir))
from taskflow import blocks
from taskflow.engines.action_engine import engine as eng
from taskflow import task
# This examples shows how LinearFlow and ParallelFlow can be used
# together to execute calculations in parallel and then use the
# result for the next task. Adder task is used for all calculations
# and arguments' bindings are used to set correct parameters to the task.
class Provider(task.Task):
def __init__(self, name, *args):
super(Provider, self).__init__(name)
self._provide = args
def execute(self):
return self._provide
class Adder(task.Task):
def __init__(self, name):
super(Adder, self).__init__(name)
def execute(self, x, y):
return x + y
flow = blocks.LinearFlow().add(
# x1 = 2, y1 = 3, x2 = 5, x3 = 8
blocks.Task(Provider("provide-adder", 2, 3, 5, 8),
save_as=('x1', 'y1', 'x2', 'y2')),
blocks.ParallelFlow().add(
# z1 = x1+y1 = 5
blocks.Task(Adder("add"), save_as='z1', rebind_args=['x1', 'y1']),
# z2 = x2+y2 = 13
blocks.Task(Adder("add"), save_as='z2', rebind_args=['x2', 'y2'])),
# r = z1+z2 = 18
blocks.Task(Adder("add"), save_as='r', rebind_args=['z1', 'z2']))
engine = eng.MultiThreadedActionEngine(flow)
engine.run()
print engine.storage.fetch_all()

View File

@ -8,93 +8,58 @@ my_dir_path = os.path.dirname(os.path.abspath(__file__))
sys.path.insert(0, os.path.join(os.path.join(my_dir_path, os.pardir),
os.pardir))
from taskflow.patterns import linear_flow as lf
from taskflow import blocks
from taskflow.engines.action_engine import engine as eng
from taskflow import task
def flow_notify(state, details):
print("'%s' entered state: %s" % (details['flow'], state))
# In this example LinearFlow is used to group four tasks to
# calculate value. Added task is used twice. In the first case
# it uses default parameters ('x' and 'y') and in the second one
# arguments are binding with 'z' and 'd' keys from engine storage.
# Multiplier task uses binding too, but explicitly shows that 'z'
# parameter is binded with 'a' key from engine storage.
def task_notify(state, details):
print("'%s' entered state: %s" % (details['runner'], state))
# This class is used to populate requirements that further tasks need to run
# by populating those tasks from a dictionary and returning said dictionary
# when the flow runs (so that further tasks can use those values). You can
# think of this a needed bootstrapping of a flow in a way.
class Provider(task.Task):
def __init__(self, name, **kwargs):
super(Provider, self).__init__(name)
self.provides.update(kwargs.keys())
self._provide = kwargs
def execute(self, context):
def __init__(self, name, *args):
super(Provider, self).__init__(name)
self._provide = args
def execute(self):
return self._provide
class Adder(task.Task):
def __init__(self, name, x_name, y_name, provides_name):
super(Adder, self).__init__(name)
self.requires.update([x_name, y_name])
self.provides.update([provides_name])
self._provides_name = provides_name
def execute(self, context, **kwargs):
return {
self._provides_name: sum(kwargs.values()),
}
def __init__(self, name):
super(Adder, self).__init__(name)
def execute(self, x, y):
return x + y
class Multiplier(task.Task):
def __init__(self, name, z_name, by_how_much):
def __init__(self, name, multiplier):
super(Multiplier, self).__init__(name)
self.requires.update([z_name])
self._by_how_much = by_how_much
self._z_name = z_name
self._multiplier = multiplier
def execute(self, context, **kwargs):
return kwargs.pop(self._z_name) * self._by_how_much
def execute(self, z):
return z * self._multiplier
flow = lf.Flow("calc-them")
flow.add(Provider("provide-adder", x=2, y=3, d=5))
flow = blocks.LinearFlow().add(
# x = 2, y = 3, d = 5
blocks.Task(Provider("provide-adder", 2, 3, 5), save_as=('x', 'y', 'd')),
# z = x+y = 5
blocks.Task(Adder("add"), save_as='z'),
# a = z+d = 10
blocks.Task(Adder("add"), save_as='a', rebind_args=['z', 'd']),
# r = a*3 = 30
blocks.Task(Multiplier("multi", 3), save_as='r', rebind_args={'z': 'a'}))
# Add x + y to produce z (5)
flow.add(Adder('add', 'x', 'y', 'z'))
engine = eng.SingleThreadedActionEngine(flow)
engine.run()
# Add z + d to produce a (5 + 5)
flow.add(Adder('add', 'z', 'd', 'a'))
# Multiple a by 3 (30)
multi_uuid = flow.add(Multiplier('multi', 'a', 3))
# Get notified of the state changes the flow is going through.
flow.notifier.register('*', flow_notify)
# Get notified of the state changes the flows tasks/runners are going through.
flow.task_notifier.register('*', task_notify)
# Context is typically passed in openstack, it is not needed here.
print '-' * 7
print 'Running'
print '-' * 7
context = {}
flow.run(context)
# This will have the last results and the task that produced that result,
# but we don't care about the task that produced it and just want the result
# itself.
print '-' * 11
print 'All results'
print '-' * 11
for (uuid, v) in flow.results.items():
print '%s => %s' % (uuid, v)
multi_results = flow.results[multi_uuid]
print '-' * 15
print "Multiply result"
print '-' * 15
print(multi_results)
assert multi_results == 30, "Example is broken"
print engine.storage.fetch_all()

View File

@ -1,7 +1,5 @@
Calling jim.
Context = {'joe_number': 444, 'jim_number': 555}
Calling joe.
Context = {'joe_number': 444, 'jim_number': 555}
Calling jim 555.
Calling joe 444.
Calling 444 and apologizing.
Calling 555 and apologizing.
Flow failed: IOError('Suzzie not home right now.',)

View File

@ -8,44 +8,48 @@ my_dir_path = os.path.dirname(os.path.abspath(__file__))
sys.path.insert(0, os.path.join(os.path.join(my_dir_path, os.pardir),
os.pardir))
from taskflow import decorators
from taskflow.patterns import linear_flow as lf
from taskflow import blocks
from taskflow.engines.action_engine import engine as eng
from taskflow import task
def undo_call(context, result, cause):
print("Calling %s and apologizing." % result)
class CallJim(task.Task):
def execute(self, jim_number, *args, **kwargs):
print("Calling jim %s." % jim_number)
def revert(self, jim_number, *args, **kwargs):
print("Calling %s and apologizing." % jim_number)
@decorators.task(revert=undo_call)
def call_jim(context):
print("Calling jim.")
print("Context = %s" % (context))
return context['jim_number']
class CallJoe(task.Task):
def execute(self, joe_number, *args, **kwargs):
print("Calling joe %s." % joe_number)
def revert(self, joe_number, *args, **kwargs):
print("Calling %s and apologizing." % joe_number)
@decorators.task(revert=undo_call)
def call_joe(context):
print("Calling joe.")
print("Context = %s" % (context))
return context['joe_number']
class CallSuzzie(task.Task):
def execute(self, suzzie_number, *args, **kwargs):
raise IOError("Suzzie not home right now.")
def revert(self, suzzie_number, *args, **kwargs):
# TODO(imelnikov): this method should not be requred
pass
@decorators.task
def call_suzzie(context):
raise IOError("Suzzie not home right now.")
flow = blocks.LinearFlow().add(blocks.Task(CallJim),
blocks.Task(CallJoe),
blocks.Task(CallSuzzie))
engine = eng.SingleThreadedActionEngine(flow)
flow = lf.Flow("call-them")
flow.add(call_jim)
flow.add(call_joe)
flow.add(call_suzzie)
context = {
engine.storage.inject({
"joe_number": 444,
"jim_number": 555,
}
"suzzie_number": 666
})
try:
flow.run(context)
engine.run()
except Exception as e:
print "Flow failed: %r" % e

View File

@ -1,4 +1,2 @@
Calling jim.
Context = {'joe_number': 444, 'jim_number': 555}
Calling joe.
Context = {'joe_number': 444, 'jim_number': 555}
Calling jim 555.
Calling joe 444.

View File

@ -8,28 +8,35 @@ my_dir_path = os.path.dirname(os.path.abspath(__file__))
sys.path.insert(0, os.path.join(os.path.join(my_dir_path, os.pardir),
os.pardir))
from taskflow import decorators
from taskflow.patterns import linear_flow as lf
from taskflow import blocks
from taskflow.engines.action_engine import engine as eng
from taskflow import task
@decorators.task
def call_jim(context):
print("Calling jim.")
print("Context = %s" % (context))
class CallJim(task.Task):
def __init__(self):
super(CallJim, self).__init__()
def execute(self, jim_number, *args, **kwargs):
print("Calling jim %s." % jim_number)
@decorators.task
def call_joe(context):
print("Calling joe.")
print("Context = %s" % (context))
class CallJoe(task.Task):
def __init__(self):
super(CallJoe, self).__init__()
flow = lf.Flow("call-them")
flow.add(call_jim)
flow.add(call_joe)
def execute(self, joe_number, *args, **kwargs):
print("Calling joe %s." % joe_number)
context = {
flow = blocks.LinearFlow().add(blocks.Task(CallJim),
blocks.Task(CallJoe))
engine = eng.SingleThreadedActionEngine(flow)
engine.storage.inject({
"joe_number": 444,
"jim_number": 555,
}
flow.run(context)
})
engine.run()