Here is the one possible answer:
#!/usr/bin/env python2.7
import os
import ecflow
defs = ecflow.Defs()
suite = defs.add_suite("data_aquisition")
suite.add_repeat( ecflow.RepeatDay(1) )
suite.add_variable("ECF_HOME", os.getenv("HOME") + "/course")
suite.add_variable("ECF_INCLUDE", os.getenv("HOME") + "/course")
suite.add_variable("ECF_FILES", os.getenv("HOME") + "/course/data")
suite.add_variable("SLEEP","2")
for city in ( "exeter", "toulouse", "offenbach", "washington", "tokyo", "melbourne", "montreal" ) :
fcity = suite.add_family(city)
fcity.add_task("archive")
for obs_type in ( "observations", "fields", "images" ):
type_fam = fcity.add_family(obs_type)
if city in ("exeter", "toulouse", "offenbach"): type_fam.add_time("00:00 23:00 01:00")
if city in ("washington") : type_fam.add_time("00:00 23:00 03:00")
if city in ("tokyo") : type_fam.add_time("12:00")
if city in ("melbourne") : type_fam.add_day( "monday" )
if city in ("montreal") : type_fam.add_date(1, 0, 0)
type_fam.add_task("get")
type_fam.add_task("process").add_trigger("get eq complete")
type_fam.add_task("store").add_trigger("get eq complete")
It is also possible to automatically generate the ecf script using the python api:
defs.generate_scripts()
This enables testing of the suite definition, with out worrying about the ecf script‘s.
| Note |
|---|
| When there are no event‘s, meter‘s or label‘s in the suite definition, the content of the generated scripts are identical. Hence this functionality should only be used as debug aid for the definition |