#!/usr/bin/env python
import logging
-logging.basicConfig(level=logging.INFO, format="run-command: %(message)s")
+
+logger = logging.getLogger('run-command')
+log_handler = logging.StreamHandler()
+log_handler.setFormatter(logging.Formatter("run-command: %(message)s"))
+logger.addHandler(log_handler)
+logger.setLevel(logging.INFO)
import arvados
import re
items = ["$(dir %s/%s/)" % (prefix, l) for l in os.listdir(fn)]
elif stat.S_ISREG(mode):
with open(fn) as f:
- items = [line for line in f]
+ items = [line.rstrip("\r\n") for line in f]
return items
else:
return None
stdinfile = None
rcode = 1
+def recursive_foreach(params, fvars):
+ var = fvars[0]
+ fvars = fvars[1:]
+ items = get_items(params, params[var])
+ logger.info("parallelizing on %s with items %s" % (var, items))
+ if items is not None:
+ for i in items:
+ params = copy.copy(params)
+ params[var] = i
+ if len(fvars) > 0:
+ recursive_foreach(params, fvars)
+ else:
+ arvados.api().job_tasks().create(body={
+ 'job_uuid': arvados.current_job()['uuid'],
+ 'created_by_job_task_uuid': arvados.current_task()['uuid'],
+ 'sequence': 1,
+ 'parameters': params
+ }
+ ).execute()
+ else:
+ logger.error("parameter %s with value %s in task.foreach yielded no items" % (var, params[var]))
+ sys.exit(1)
+
try:
if "task.foreach" in jobp:
if arvados.current_task()['sequence'] == 0:
- var = jobp["task.foreach"]
- items = get_items(jobp, jobp[var])
- logging.info("parallelizing on %s with items %s" % (var, items))
- if items is not None:
- for i in items:
- params = copy.copy(jobp)
- params[var] = i
- arvados.api().job_tasks().create(body={
- 'job_uuid': arvados.current_job()['uuid'],
- 'created_by_job_task_uuid': arvados.current_task()['uuid'],
- 'sequence': 1,
- 'parameters': params
- }
- ).execute()
- if "task.vwd" in jobp:
- # Base vwd collection will be merged with output fragments from
- # the other tasks by crunch.
- arvados.current_task().set_output(subst.do_substitution(jobp, jobp["task.vwd"]))
- else:
- arvados.current_task().set_output(None)
- sys.exit(0)
- else:
+ # This is the first task to start the other tasks and exit
+ fvars = jobp["task.foreach"]
+ if isinstance(fvars, basestring):
+ fvars = [fvars]
+ if not isinstance(fvars, list) or len(fvars) == 0:
+ logger.error("value of task.foreach must be a string or non-empty list")
sys.exit(1)
+ recursive_foreach(jobp, jobp["task.foreach"])
+ if "task.vwd" in jobp:
+ # Set output of the first task to the base vwd collection so it
+ # will be merged with output fragments from the other tasks by
+ # crunch.
+ arvados.current_task().set_output(subst.do_substitution(jobp, jobp["task.vwd"]))
+ else:
+ arvados.current_task().set_output(None)
+ sys.exit(0)
else:
+ # This is the only task so taskp/jobp are the same
taskp = jobp
if "task.vwd" in taskp:
stdoutname = subst.do_substitution(taskp, taskp["task.stdout"])
stdoutfile = open(stdoutname, "wb")
- logging.info("{}{}{}".format(' '.join(cmd), (" < " + stdinname) if stdinname is not None else "", (" > " + stdoutname) if stdoutname is not None else ""))
-
+ logger.info("{}{}{}".format(' '.join(cmd), (" < " + stdinname) if stdinname is not None else "", (" > " + stdoutname) if stdoutname is not None else ""))
+except subst.SubstitutionError as e:
+ logger.error(str(e))
+ logger.error("task parameters were:")
+ logger.error(pprint.pformat(taskp))
+ sys.exit(1)
except Exception as e:
- logging.exception("caught exception")
- logging.error("task parameters was:")
- logging.error(pprint.pformat(taskp))
+ logger.exception("caught exception")
+ logger.error("task parameters were:")
+ logger.error(pprint.pformat(taskp))
sys.exit(1)
try:
rcode = sp.wait()
if sig.sig is not None:
- logging.critical("terminating on signal %s" % sig.sig)
+ logger.critical("terminating on signal %s" % sig.sig)
sys.exit(2)
else:
- logging.info("completed with exit code %i (%s)" % (rcode, "success" if rcode == 0 else "failed"))
+ logger.info("completed with exit code %i (%s)" % (rcode, "success" if rcode == 0 else "failed"))
except Exception as e:
- logging.exception("caught exception")
+ logger.exception("caught exception")
# restore default signal handlers.
signal.signal(signal.SIGINT, signal.SIG_DFL)
for l in links:
os.unlink(l)
-logging.info("the following output files will be saved to keep:")
+logger.info("the following output files will be saved to keep:")
subprocess.call(["find", ".", "-type", "f", "-printf", "run-command: %12.12s %h/%f\\n"], stdout=sys.stderr)
-logging.info("start writing output to keep")
+logger.info("start writing output to keep")
if "task.vwd" in taskp:
if "task.foreach" in jobp:
# Just a single task, so do merge with the original collection
outcollection = vwd.checkin(subst.do_substitution(taskp, taskp["task.vwd"]), outdir, merge=True).manifest_text()
else:
- outcollection = robust_put.upload(outdir)
+ outcollection = robust_put.upload(outdir, logger)
api.job_tasks().update(uuid=arvados.current_task()['uuid'],
body={