Here is an example of a script as a local native Python task: This example shows that we can still take advantage of pre-processing and variable substitution. This will work on the manual and comments section, these will also be pre-processed.
There are several ways of doing this. The following are examples.
definition file:
The default ECF_MICRO is %, this may interfere with your Python scripts. In this case you either redefine it, in the task definition or directly in the Python script.
task python
edit ECF_EXTN .py. # search for python.py
edit ECF_MICRO '$'
edit ECF_JOB_CMD 'python $ECF_JOB$ > $ECF_JOBOUT$ 2>&1'
label info "none"
meter step -1 100 100
event my_event |
Notice that the ECF_JOB_CMD calls Python. This allows us to change the Python version, within the viewer. Alternatively it can be omitted, providing we add "#!/usr/bin/env python" as the first line of our Python script.
Header: head.py
#////////1/////////2/////////3/////////4/////////5/////////6/////////7/////////8
# Name :
# Author : Avi
# Revision : $Revision: #10 $
#
# Copyright 2009- ECMWF.
# This software is licensed under the terms of the Apache Licence version 2.0
# which can be obtained at http://www.apache.org/licenses/LICENSE-2.0.
# In applying this licence, ECMWF does not waive the privileges and immunities
# granted to it by virtue of its status as an intergovernmental organisation
# nor does it submit to any jurisdiction.
#////////1/////////2/////////3/////////4/////////5/////////6/////////7/////////8
import datetime,time
import os, signal, platform
import ecflow
print("cwd: " + os.getcwd())
print("PYTHONPATH=====================================================")
try:
print(os.environ['PYTHONPATH'].split(os.pathsep))
except KeyError:
print("Could not get PYTHONPATH")
print("LD_LIBRARY_PATH=====================================================")
try:
print(os.environ['LD_LIBRARY_PATH'].split(os.pathsep))
except KeyError:
print("Could not get LD_LIBRARY_PATH")
class Client(object):
"""Encapsulate communication with the ecflow server. This will automatically call
the child command init()/complete(), for job start/finish. It will also
handle exceptions and signals, by calling the abort child command.
*ONLY* one instance of this class, should be used. Otherwise zombies will be created.
"""
def __init__(self,init_add_vars = False):
print ("Creating Client")
self.ci = ecflow.Client()
self.ci.set_host_port("%ECF_HOST%","%ECF_PORT%")
self.ci.set_child_pid(os.getpid())
self.ci.set_child_path("%ECF_NAME%")
self.ci.set_child_password("%ECF_PASS%")
self.ci.set_child_try_no(%ECF_TRYNO%)
if init_add_vars:
self.ci.set_child_init_add_vars({"name1":"1", "name2":"2"})
self.ci.set_child_init_add_vars([ecflow.Variable("name3","3"), ecflow.Variable("name4","4")])
self.ci.set_child_complete_del_vars(["name","name1","name2","name3","name4"])
print(" Only wait 20 seconds, if the server cannot be contacted (note default is 24 hours) before failing")
self.ci.set_child_timeout(20)
self.ci.set_zombie_child_timeout(10)
# Abort the task for the following signals
signal.signal(signal.SIGINT, self.signal_handler)
signal.signal(signal.SIGHUP, self.signal_handler)
signal.signal(signal.SIGQUIT, self.signal_handler)
signal.signal(signal.SIGILL, self.signal_handler)
signal.signal(signal.SIGTRAP, self.signal_handler)
signal.signal(signal.SIGIOT, self.signal_handler)
signal.signal(signal.SIGBUS, self.signal_handler)
signal.signal(signal.SIGFPE, self.signal_handler)
signal.signal(signal.SIGUSR1, self.signal_handler)
signal.signal(signal.SIGUSR2, self.signal_handler)
signal.signal(signal.SIGPIPE, self.signal_handler)
signal.signal(signal.SIGTERM, self.signal_handler)
signal.signal(signal.SIGXCPU, self.signal_handler)
if platform.system() != "Darwin":
signal.signal(signal.SIGPWR, self.signal_handler)
def at_time(self):
%ecfmicro #
return datetime.datetime.fromtimestamp(time.time()).strftime('%H:%M:%S')
#ecfmicro %
def signal_handler(self,signum, frame):
print(' Aborting: Signal handler called with signal ', signum)
self.ci.child_abort("Signal handler called with signal " + str(signum));
def __enter__(self):
print('Calling init at: ' + self.at_time())
self.ci.child_init()
return self.ci
def __exit__(self,ex_type,value,tb):
print (" Client:__exit__: ex_type:" + str(ex_type) + " value:" + str(value) + "\n traceback:" + str(tb))
if ex_type is not None:
print('Calling abort ' + self.at_time())
self.ci.child_abort("Aborted with exception type " + str(ex_type) + ":" + str(value))
sys.exit(0)
print('Calling complete at: ' + self.at_time())
self.ci.child_complete()
sys.exit(0)
|
task wrapper:
$include <head.py>
# This will also handle call to sys.exit(), i.e. Client.__exit__ will still be called.
with Client() as ci:
# *******************************************************************************
# This is where the main work is done.
# *******************************************************************************
for i in range(1,100):
ci.child_meter('step',i)
ci.child_label('info', "value_" + str(i))
time.sleep(1)
ci.child_event('my_event')
print "Finished event,meter and label child commands"
$manual
This is the manual section. Instead of calling python from the ECF_JOB_CMD we could alternatively place,
#!/bin/env/python
on the first line of this file.
$end
$comment
Note: We do not need a include a tail.py, the head.py does it all.
$end |