def execute_workflow_step(workflow,
task_id,
job_data,
cwl_args=None,
executor=None):
"""
Constructs and executes single step workflow based on the "workflow"
and "task_id". "cwl_args" can be used to update default parameters
used for loading and runtime contexts. Exports json file with the
execution results.
"""
cwl_args = {} if cwl_args is None else cwl_args
executor = SingleJobExecutor() if executor is None else executor
step_tmp_folder, step_cache_folder, step_outputs_folder, step_report = get_temp_folders(
task_id=task_id, job_data=job_data)
default_cwl_args = get_default_cwl_args(cwl_args)
default_cwl_args.update({ # add execution specific parameters
"tmp_outdir_prefix": step_cache_folder + "/",
"tmpdir_prefix": step_cache_folder + "/",
"cidfile_dir": step_tmp_folder,
"cidfile_prefix": task_id,
"basedir": os.getcwd(
), # job should already have abs path for inputs, so this is useless
"outdir": step_outputs_folder
})
workflow_step_path = os.path.join(step_tmp_folder,
task_id + "_step_workflow.cwl")
fast_cwl_step_load( # will save new worlflow to "workflow_step_path"
workflow=workflow,
target_id=task_id,
cwl_args=default_cwl_args,
location=workflow_step_path)
_stderr = sys.stderr # to trick the logger
sys.stderr = sys.__stderr__
step_outputs, step_status = executor(
slow_cwl_load(workflow=workflow_step_path, cwl_args=default_cwl_args),
job_data, RuntimeContext(default_cwl_args))
sys.stderr = _stderr
if step_status != "success":
raise ValueError
# To remove "http://commonwl.org/cwltool#generation": 0 (copied from cwltool)
visit_class(step_outputs, ("File", ), MutationManager().unset_generation)
dump_json(step_outputs, step_report)
return step_outputs, step_report
def execute(self, context):
post_status(context)
self.cwlwf, it_is_workflow = load_cwl(
self.dag.default_args["cwl_workflow"], self.dag.default_args)
self.cwl_step = [
step for step in self.cwlwf.steps
if self.task_id == step.id.split("#")[-1]
][0] if it_is_workflow else self.cwlwf
_logger.info('{0}: Running!'.format(self.task_id))
upstream_task_ids = [t.task_id for t in self.upstream_list] + \
([self.reader_task_id] if self.reader_task_id else [])
_logger.debug('{0}: Collecting outputs from: \n{1}'.format(
self.task_id, json.dumps(upstream_task_ids, indent=4)))
upstream_data = self.xcom_pull(context=context,
task_ids=upstream_task_ids)
_logger.info('{0}: Upstream data: \n {1}'.format(
self.task_id, json.dumps(upstream_data, indent=4)))
promises = {}
for data in upstream_data: # upstream_data is an array with { promises and outdir }
promises = merge(promises, data["promises"])
if "outdir" in data:
self.outdir = data["outdir"]
_d_args = self.dag.default_args
if not self.outdir:
self.outdir = _d_args['tmp_folder']
_logger.debug('{0}: Step inputs: {1}'.format(
self.task_id, json.dumps(self.cwl_step.tool["inputs"], indent=4)))
_logger.debug('{0}: Step outputs: {1}'.format(
self.task_id, json.dumps(self.cwl_step.tool["outputs"], indent=4)))
jobobj = {}
for inp in self.cwl_step.tool["inputs"]:
jobobj_id = shortname(inp["id"]).split("/")[-1]
source_ids = []
promises_outputs = []
try:
source_field = inp["source"] if it_is_workflow else inp.get(
"id")
source_ids = [shortname(s)
for s in source_field] if isinstance(
source_field,
list) else [shortname(source_field)]
promises_outputs = [
promises[source_id] for source_id in source_ids
if source_id in promises
]
except:
_logger.warning(
"{0}: Couldn't find source field in step input: {1}".
format(self.task_id, json.dumps(inp, indent=4)))
_logger.info(
'{0}: For input {1} with source_ids: {2} found upstream outputs: \n{3}'
.format(self.task_id, jobobj_id, source_ids, promises_outputs))
if len(promises_outputs) > 1:
if inp.get("linkMerge", "merge_nested") == "merge_flattened":
jobobj[jobobj_id] = flatten(promises_outputs)
else:
jobobj[jobobj_id] = promises_outputs
# Should also check if [None], because in this case we need to take default value
elif len(promises_outputs) == 1 and (promises_outputs[0]
is not None):
jobobj[jobobj_id] = promises_outputs[0]
elif "valueFrom" in inp:
jobobj[jobobj_id] = None
elif "default" in inp:
d = copy.copy(inp["default"])
jobobj[jobobj_id] = d
else:
continue
_logger.debug('{0}: Collected job object: \n {1}'.format(
self.task_id, json.dumps(jobobj, indent=4)))
def _post_scatter_eval(shortio, cwl_step):
_value_from = {
shortname(i["id"]).split("/")[-1]: i["valueFrom"]
for i in cwl_step.tool["inputs"] if "valueFrom" in i
}
_logger.debug('{0}: Step inputs with valueFrom: \n{1}'.format(
self.task_id, json.dumps(_value_from, indent=4)))
def value_from_func(k, v):
if k in _value_from:
return expression.do_eval(_value_from[k],
shortio,
self.cwlwf.tool.get(
"requirements", []),
None,
None, {},
context=v)
else:
return v
return {k: value_from_func(k, v) for k, v in shortio.items()}
job = _post_scatter_eval(jobobj, self.cwl_step)
_logger.info('{0}: Final job data: \n {1}'.format(
self.task_id, json.dumps(job, indent=4)))
_d_args['outdir'] = tempfile.mkdtemp(
prefix=os.path.join(self.outdir, "step_tmp"))
_d_args['tmpdir_prefix'] = os.path.join(_d_args['outdir'], 'cwl_tmp_')
_d_args['tmp_outdir_prefix'] = os.path.join(_d_args['outdir'],
'cwl_outdir_')
_d_args["record_container_id"] = True
_d_args["cidfile_dir"] = _d_args['outdir']
_d_args["cidfile_prefix"] = self.task_id
_logger.debug('{0}: Runtime context: \n {1}'.format(self, _d_args))
executor = SingleJobExecutor()
runtimeContext = RuntimeContext(_d_args)
runtimeContext.make_fs_access = getdefault(
runtimeContext.make_fs_access, StdFsAccess)
for inp in self.cwl_step.tool["inputs"]:
if inp.get("not_connected"):
del job[shortname(inp["id"].split("/")[-1])]
_stderr = sys.stderr
sys.stderr = sys.__stderr__
(output, status) = executor(
self.cwl_step.embedded_tool if it_is_workflow else self.cwl_step,
job,
runtimeContext,
logger=_logger)
sys.stderr = _stderr
if not output and status == "permanentFail":
raise ValueError
_logger.debug('{0}: Embedded tool outputs: \n {1}'.format(
self.task_id, json.dumps(output, indent=4)))
promises = {}
for out in self.cwl_step.tool["outputs"]:
out_id = shortname(out["id"])
jobout_id = out_id.split("/")[-1]
try:
promises[out_id] = output[jobout_id]
except:
continue
# Unsetting the Generation from final output object
visit_class(promises, ("File", ), MutationManager().unset_generation)
data = {"promises": promises, "outdir": self.outdir}
_logger.info('{0}: Output: \n {1}'.format(self.task_id,
json.dumps(data, indent=4)))
return data
def executor(self, tool, job_order, **kwargs):
final_output = []
final_status = []
def output_callback(out, processStatus):
final_status.append(processStatus)
final_output.append(out)
if "basedir" not in kwargs:
raise WorkflowException("Must provide 'basedir' in kwargs")
output_dirs = set()
if kwargs.get("outdir"):
finaloutdir = os.path.abspath(kwargs.get("outdir"))
else:
finaloutdir = None
if kwargs.get("tmp_outdir_prefix"):
kwargs["outdir"] = tempfile.mkdtemp(
prefix=kwargs["tmp_outdir_prefix"])
else:
kwargs["outdir"] = tempfile.mkdtemp()
output_dirs.add(kwargs["outdir"])
kwargs["mutation_manager"] = MutationManager()
jobReqs = None
if "cwl:requirements" in job_order:
jobReqs = job_order["cwl:requirements"]
elif ("cwl:defaults" in tool.metadata
and "cwl:requirements" in tool.metadata["cwl:defaults"]):
jobReqs = tool.metadata["cwl:defaults"]["cwl:requirements"]
if jobReqs:
for req in jobReqs:
tool.requirements.append(req)
if kwargs.get("default_container"):
tool.requirements.insert(
0, {
"class": "DockerRequirement",
"dockerPull": kwargs["default_container"]
})
jobs = tool.job(job_order, output_callback, **kwargs)
try:
for runnable in jobs:
if runnable:
builder = kwargs.get("builder", None)
if builder is not None:
runnable.builder = builder
if runnable.outdir:
output_dirs.add(runnable.outdir)
runnable.run(**kwargs)
else:
time.sleep(1)
except WorkflowException as e:
raise e
except Exception as e:
log.error("Got exception")
raise WorkflowException(str(e))
# wait for all processes to finish
self.wait()
if final_output and final_output[0] and finaloutdir:
final_output[0] = relocateOutputs(final_output[0], finaloutdir,
output_dirs,
kwargs.get("move_outputs"),
kwargs["make_fs_access"](""))
if kwargs.get("rm_tmpdir"):
cleanIntermediate(output_dirs)
if final_output and final_status:
return (final_output[0], final_status[0])
else:
return (None, "permanentFail")
def main(args=None, stdout=sys.stdout):
"""Main method for toil-cwl-runner."""
cwllogger.removeHandler(defaultStreamHandler)
config = Config()
config.cwl = True
parser = argparse.ArgumentParser()
addOptions(parser, config)
parser.add_argument("cwltool", type=str)
parser.add_argument("cwljob", nargs=argparse.REMAINDER)
# Will override the "jobStore" positional argument, enables
# user to select jobStore or get a default from logic one below.
parser.add_argument("--jobStore", type=str)
parser.add_argument("--not-strict", action="store_true")
parser.add_argument("--quiet", dest="logLevel", action="store_const",
const="ERROR")
parser.add_argument("--basedir", type=str)
parser.add_argument("--outdir", type=str, default=os.getcwd())
parser.add_argument("--version", action='version', version=baseVersion)
dockergroup = parser.add_mutually_exclusive_group()
dockergroup.add_argument(
"--user-space-docker-cmd",
help="(Linux/OS X only) Specify a user space docker command (like "
"udocker or dx-docker) that will be used to call 'pull' and 'run'")
dockergroup.add_argument(
"--singularity", action="store_true", default=False,
help="[experimental] Use Singularity runtime for running containers. "
"Requires Singularity v2.3.2+ and Linux with kernel version v3.18+ or "
"with overlayfs support backported.")
dockergroup.add_argument(
"--no-container", action="store_true", help="Do not execute jobs in a "
"Docker container, even when `DockerRequirement` "
"is specified under `hints`.")
parser.add_argument(
"--preserve-environment", type=str, nargs='+',
help="Preserve specified environment variables when running"
" CommandLineTools", metavar=("VAR1 VAR2"), default=("PATH",),
dest="preserve_environment")
parser.add_argument(
"--destBucket", type=str,
help="Specify a cloud bucket endpoint for output files.")
parser.add_argument(
"--beta-dependency-resolvers-configuration", default=None)
parser.add_argument("--beta-dependencies-directory", default=None)
parser.add_argument(
"--beta-use-biocontainers", default=None, action="store_true")
parser.add_argument(
"--beta-conda-dependencies", default=None, action="store_true")
parser.add_argument("--tmpdir-prefix", type=Text,
help="Path prefix for temporary directories",
default="tmp")
parser.add_argument("--tmp-outdir-prefix", type=Text,
help="Path prefix for intermediate output directories",
default="tmp")
parser.add_argument(
"--force-docker-pull", action="store_true", default=False,
dest="force_docker_pull",
help="Pull latest docker image even if it is locally present")
parser.add_argument(
"--no-match-user", action="store_true", default=False,
help="Disable passing the current uid to `docker run --user`")
# mkdtemp actually creates the directory, but
# toil requires that the directory not exist,
# so make it and delete it and allow
# toil to create it again (!)
workdir = tempfile.mkdtemp()
os.rmdir(workdir)
if args is None:
args = sys.argv[1:]
# we use workdir as jobStore:
options = parser.parse_args([workdir] + args)
# if tmpdir_prefix is not the default value, set workDir too
if options.tmpdir_prefix != 'tmp':
options.workDir = options.tmpdir_prefix
if options.provisioner and not options.jobStore:
raise NoSuchJobStoreException(
'Please specify a jobstore with the --jobStore option when specifying a provisioner.')
use_container = not options.no_container
if options.logLevel:
cwllogger.setLevel(options.logLevel)
outdir = os.path.abspath(options.outdir)
tmp_outdir_prefix = os.path.abspath(options.tmp_outdir_prefix)
tmpdir_prefix = os.path.abspath(options.tmpdir_prefix)
fileindex = {}
existing = {}
conf_file = getattr(options,
"beta_dependency_resolvers_configuration", None)
use_conda_dependencies = getattr(options, "beta_conda_dependencies", None)
job_script_provider = None
if conf_file or use_conda_dependencies:
dependencies_configuration = DependenciesConfiguration(options)
job_script_provider = dependencies_configuration
options.default_container = None
runtime_context = cwltool.context.RuntimeContext(vars(options))
runtime_context.find_default_container = functools.partial(
find_default_container, options)
runtime_context.workdir = workdir
runtime_context.move_outputs = "leave"
runtime_context.rm_tmpdir = False
loading_context = cwltool.context.LoadingContext(vars(options))
with Toil(options) as toil:
if options.restart:
outobj = toil.restart()
else:
loading_context.hints = [{
"class": "ResourceRequirement",
"coresMin": toil.config.defaultCores,
"ramMin": toil.config.defaultMemory / (2**20),
"outdirMin": toil.config.defaultDisk / (2**20),
"tmpdirMin": 0
}]
loading_context.construct_tool_object = toil_make_tool
loading_context.resolver = cwltool.resolver.tool_resolver
loading_context.strict = not options.not_strict
options.workflow = options.cwltool
options.job_order = options.cwljob
uri, tool_file_uri = cwltool.load_tool.resolve_tool_uri(
options.cwltool, loading_context.resolver,
loading_context.fetcher_constructor)
options.tool_help = None
options.debug = options.logLevel == "DEBUG"
job_order_object, options.basedir, jobloader = \
cwltool.main.load_job_order(
options, sys.stdin, loading_context.fetcher_constructor,
loading_context.overrides_list, tool_file_uri)
document_loader, workflowobj, uri = \
cwltool.load_tool.fetch_document(
uri, loading_context.resolver,
loading_context.fetcher_constructor)
document_loader, avsc_names, processobj, metadata, uri = \
cwltool.load_tool.validate_document(
document_loader, workflowobj, uri,
loading_context.enable_dev, loading_context.strict, False,
loading_context.fetcher_constructor, False,
loading_context.overrides_list,
do_validate=loading_context.do_validate)
loading_context.overrides_list.extend(
metadata.get("cwltool:overrides", []))
try:
tool = cwltool.load_tool.make_tool(
document_loader, avsc_names, metadata, uri,
loading_context)
except cwltool.process.UnsupportedRequirement as err:
logging.error(err)
return 33
runtime_context.secret_store = SecretStore()
initialized_job_order = cwltool.main.init_job_order(
job_order_object, options, tool, jobloader, sys.stdout,
secret_store=runtime_context.secret_store)
fs_access = cwltool.stdfsaccess.StdFsAccess(options.basedir)
fill_in_defaults(
tool.tool["inputs"], initialized_job_order, fs_access)
def path_to_loc(obj):
if "location" not in obj and "path" in obj:
obj["location"] = obj["path"]
del obj["path"]
def import_files(tool):
visit_class(tool, ("File", "Directory"), path_to_loc)
visit_class(tool, ("File", ), functools.partial(
add_sizes, fs_access))
normalizeFilesDirs(tool)
adjustDirObjs(tool, functools.partial(
get_listing, fs_access, recursive=True))
adjustFileObjs(tool, functools.partial(
uploadFile, toil.importFile, fileindex, existing,
skip_broken=True))
tool.visit(import_files)
for inp in tool.tool["inputs"]:
def set_secondary(fileobj):
if isinstance(fileobj, Mapping) \
and fileobj.get("class") == "File":
if "secondaryFiles" not in fileobj:
fileobj["secondaryFiles"] = [
{"location": cwltool.builder.substitute(
fileobj["location"], sf), "class": "File"}
for sf in inp["secondaryFiles"]]
if isinstance(fileobj, MutableSequence):
for entry in fileobj:
set_secondary(entry)
if shortname(inp["id"]) in initialized_job_order \
and inp.get("secondaryFiles"):
set_secondary(initialized_job_order[shortname(inp["id"])])
import_files(initialized_job_order)
visitSteps(tool, import_files)
try:
runtime_context.use_container = use_container
runtime_context.tmpdir = os.path.realpath(tmpdir_prefix)
runtime_context.tmp_outdir_prefix = os.path.realpath(
tmp_outdir_prefix)
runtime_context.job_script_provider = job_script_provider
runtime_context.force_docker_pull = options.force_docker_pull
runtime_context.no_match_user = options.no_match_user
(wf1, _) = makeJob(tool, {}, None, runtime_context)
except cwltool.process.UnsupportedRequirement as err:
logging.error(err)
return 33
wf1.cwljob = initialized_job_order
if wf1 is CWLJob: # Clean up temporary directories only created with CWLJobs.
wf1.addFollowOnFn(cleanTempDirs, wf1)
outobj = toil.start(wf1)
outobj = resolve_indirect(outobj)
# Stage files. Specify destination bucket if specified in CLI
# options. If destination bucket not passed in,
# options.destBucket's value will be None.
toilStageFiles(
toil,
outobj,
outdir,
fileindex,
existing,
export=True,
destBucket=options.destBucket)
if not options.destBucket:
visit_class(outobj, ("File",), functools.partial(
compute_checksums, cwltool.stdfsaccess.StdFsAccess("")))
visit_class(outobj, ("File", ), MutationManager().unset_generation)
stdout.write(json.dumps(outobj, indent=4))
return 0
def executor(self, tool, job_order, runtimeContext, **kwargs):
"""Executor method."""
final_output = []
final_status = []
def output_callback(out, processStatus):
final_status.append(processStatus)
final_output.append(out)
if not runtimeContext.basedir:
raise WorkflowException('`runtimeContext` should contain a '
'`basedir`')
output_dirs = set()
if runtimeContext.outdir:
finaloutdir = os.path.abspath(runtimeContext.outdir)
else:
finaloutdir = None
if runtimeContext.tmp_outdir_prefix:
runtimeContext.outdir = tempfile.mkdtemp(
prefix=runtimeContext.tmp_outdir_prefix)
else:
runtimeContext.outdir = tempfile.mkdtemp()
output_dirs.add(runtimeContext.outdir)
runtimeContext.mutation_manager = MutationManager()
jobReqs = None
if "cwl:requirements" in job_order:
jobReqs = job_order["cwl:requirements"]
elif ("cwl:defaults" in tool.metadata
and "cwl:requirements" in tool.metadata["cwl:defaults"]):
jobReqs = tool.metadata["cwl:defaults"]["cwl:requirements"]
if jobReqs:
for req in jobReqs:
tool.requirements.append(req)
if not runtimeContext.default_container:
runtimeContext.default_container = 'frolvlad/alpine-bash'
runtimeContext.docker_outdir = os.path.join(runtimeContext.working_dir,
"cwl/docker_outdir")
runtimeContext.docker_tmpdir = os.path.join(runtimeContext.working_dir,
"cwl/docker_tmpdir")
runtimeContext.docker_stagedir = os.path.join(
runtimeContext.working_dir, "cwl/docker_stagedir")
jobs = tool.job(job_order, output_callback, runtimeContext)
try:
for runnable in jobs:
if runnable:
if runtimeContext.builder:
runnable.builder = runtimeContext.builder
if runnable.outdir:
output_dirs.add(runnable.outdir)
runnable.run(runtimeContext)
else:
# log.error(
# "Workflow cannot make any more progress"
# )
# break
time.sleep(1)
except WorkflowException as e:
traceback.print_exc()
raise e
except Exception as e:
traceback.print_exc()
raise WorkflowException(str(e))
# wait for all processes to finish
self.wait()
if final_output and final_output[0] and finaloutdir:
final_output[0] = relocateOutputs(
final_output[0], finaloutdir, output_dirs,
runtimeContext.move_outputs, runtimeContext.make_fs_access(""))
if runtimeContext.rm_tmpdir:
cleanIntermediate(output_dirs)
if final_output and final_status:
return str(final_output[0]), str(final_status[0])
else:
return None, "permanentFail"
def execute_workflow_step(workflow,
task_id,
job_data,
cwl_args=None,
executor=None):
"""
Constructs and executes single step workflow based on the "workflow"
and "task_id". "cwl_args" can be used to update default parameters
used for loading and runtime contexts. Exports json file with the
execution results. If the step was evaluated as the one that need to
be skipped, the output "skipped" will set to True and the step_report
file will include "nulls". This function doesn't remove any temporary
data in both success and failure scenarios.
"""
cwl_args = {} if cwl_args is None else cwl_args
executor = SingleJobExecutor() if executor is None else executor
step_tmp_folder, step_cache_folder, step_outputs_folder, step_report = get_temp_folders(
task_id=task_id, job_data=job_data)
default_cwl_args = get_default_cwl_args(cwl_args)
default_cwl_args.update({ # add execution specific parameters
"tmp_outdir_prefix": step_cache_folder + "/",
"tmpdir_prefix": step_cache_folder + "/",
"cidfile_dir": step_tmp_folder,
"cidfile_prefix": task_id,
"basedir": os.getcwd(
), # job should already have abs path for inputs, so this is useless
"outdir": step_outputs_folder
})
workflow_step_path = os.path.join(step_tmp_folder,
task_id + "_step_workflow.cwl")
fast_cwl_step_load( # will save new worlflow to "workflow_step_path"
workflow=workflow,
target_id=task_id,
cwl_args=default_cwl_args,
location=workflow_step_path)
workflow_data = slow_cwl_load(workflow=workflow_step_path,
cwl_args=default_cwl_args)
skipped = True
step_outputs = {
output_id: None
for output_id, _ in get_items(workflow_data.tool["outputs"])
}
if need_to_run(workflow_data, job_data, task_id):
skipped = False
_stderr = sys.stderr # to trick the logger
sys.stderr = sys.__stderr__
step_outputs, step_status = executor(workflow_data, job_data,
RuntimeContext(default_cwl_args))
sys.stderr = _stderr
if step_status != "success":
raise ValueError("Failed to run workflow step")
# To remove "http://commonwl.org/cwltool#generation": 0 (copied from cwltool)
visit_class(step_outputs, ("File", ),
MutationManager().unset_generation)
dump_json(step_outputs, step_report)
return step_outputs, step_report, skipped
def cwlmain(
self,
argsl=None, # type: List[str]
args=None, # type: argparse.Namespace
job_order_object=None, # type: MutableMapping[Text, Any]
stdin=sys.stdin, # type: IO[Any]
stdout=None, # type: Union[TextIO, codecs.StreamWriter]
stderr=sys.stderr, # type: IO[Any]
versionfunc=versionstring, # type: Callable[[], Text]
logger_handler=None, #
custom_schema_callback=None, # type: Callable[[], None]
executor=None, # type: Callable[..., Tuple[Dict[Text, Any], Text]]
loadingContext=None, # type: LoadingContext
runtimeContext=None # type: RuntimeContext
): # type: (...) -> int
if not stdout:
stdout = codecs.getwriter('utf-8')(sys.stdout)
_logger.removeHandler(defaultStreamHandler)
if logger_handler:
stderr_handler = logger_handler
else:
stderr_handler = logging.StreamHandler(stderr)
_logger.addHandler(stderr_handler)
try:
if args is None:
args = arg_parser().parse_args(argsl)
if args.workflow and "--outdir" not in argsl:
outputPath = args.workflow.split('/')[-1].split('.')[0]
setattr(
args, "outdir",
os.getcwd() + "/" + outputPath + "/" +
datetime.datetime.now().strftime('%Y-%m-%d-%H%M'))
if runtimeContext is None:
runtimeContext = RuntimeContext(vars(args))
else:
runtimeContext = runtimeContext.copy()
rdflib_logger = logging.getLogger("rdflib.term")
rdflib_logger.addHandler(stderr_handler)
rdflib_logger.setLevel(logging.ERROR)
if args.quiet:
_logger.setLevel(logging.WARN)
if runtimeContext.debug:
_logger.setLevel(logging.DEBUG)
rdflib_logger.setLevel(logging.DEBUG)
if args.timestamps:
formatter = logging.Formatter("[%(asctime)s] %(message)s",
"%Y-%m-%d %H:%M:%S")
stderr_handler.setFormatter(formatter)
# version
if args.version:
return versionfunc(), 0
else:
_logger.info(versionfunc())
if args.print_supported_versions:
return "\n".join(supportedCWLversions(args.enable_dev)), 0
if not args.workflow:
if os.path.isfile("CWLFile"):
setattr(args, "workflow", "CWLFile")
else:
_logger.error("")
_logger.error(
"CWL document required, no input file was provided")
arg_parser().print_help()
return "CWL document required, no input file was provided", 1
if args.relax_path_checks:
command_line_tool.ACCEPTLIST_RE = command_line_tool.ACCEPTLIST_EN_RELAXED_RE
if args.ga4gh_tool_registries:
ga4gh_tool_registries[:] = args.ga4gh_tool_registries
if not args.enable_ga4gh_tool_registry:
del ga4gh_tool_registries[:]
if custom_schema_callback:
custom_schema_callback()
elif args.enable_ext:
res = pkg_resources.resource_stream(__name__, 'extensions.yml')
use_custom_schema("v1.0", "http://commonwl.org/cwltool",
res.read())
res.close()
else:
use_standard_schema("v1.0")
if loadingContext is None:
loadingContext = LoadingContext(vars(args))
else:
loadingContext = loadingContext.copy()
loadingContext.disable_js_validation = \
args.disable_js_validation or (not args.do_validate)
loadingContext.construct_tool_object = getdefault(
loadingContext.construct_tool_object,
workflow.default_make_tool)
loadingContext.resolver = getdefault(loadingContext.resolver,
tool_resolver)
try:
uri, tool_file_uri = resolve_tool_uri(
args.workflow,
resolver=loadingContext.resolver,
fetcher_constructor=loadingContext.fetcher_constructor)
except:
return "Can't find file " + args.workflow, 0
try_again_msg = "" if args.debug else ", try again with --debug for more information"
try:
job_order_object, input_basedir, jobloader = load_job_order(
args, stdin, loadingContext.fetcher_constructor,
loadingContext.overrides_list, tool_file_uri)
if args.overrides:
loadingContext.overrides_list.extend(
load_overrides(
file_uri(os.path.abspath(args.overrides)),
tool_file_uri))
document_loader, workflowobj, uri = fetch_document(
uri,
resolver=loadingContext.resolver,
fetcher_constructor=loadingContext.fetcher_constructor)
if args.print_deps:
# printdeps(workflowobj, document_loader, stdout, args.relative_deps, uri)
result = returndeps(workflowobj, document_loader, stdout,
args.relative_deps, uri)
return result, 0
document_loader, avsc_names, processobj, metadata, uri \
= validate_document(document_loader, workflowobj, uri,
enable_dev=loadingContext.enable_dev,
strict=loadingContext.strict,
preprocess_only=(args.print_pre or args.pack),
fetcher_constructor=loadingContext.fetcher_constructor,
skip_schemas=args.skip_schemas,
overrides=loadingContext.overrides_list,
do_validate=loadingContext.do_validate)
if args.print_pre:
# stdout.write(json_dumps(processobj, indent=4))
return json_dumps(processobj, indent=4), 0
loadingContext.overrides_list.extend(
metadata.get("cwltool:overrides", []))
tool = make_tool(document_loader, avsc_names, metadata, uri,
loadingContext)
if args.make_template:
yaml.safe_dump(generate_input_template(tool),
sys.stdout,
default_flow_style=False,
indent=4,
block_seq_indent=2)
return yaml.safe_dump(generate_input_template(tool),
indent=4), 0
if args.validate:
_logger.info("Tool definition is valid")
return "Tool definition is valid", 0
if args.pack:
stdout.write(
print_pack(document_loader, processobj, uri, metadata))
return print_pack(document_loader, processobj, uri,
metadata), 0
if args.print_rdf:
stdout.write(
printrdf(tool, document_loader.ctx,
args.rdf_serializer))
return printrdf(tool, document_loader.ctx,
args.rdf_serializer), 0
if args.print_dot:
printdot(tool, document_loader.ctx, stdout)
return "args.print_dot still not solved", 0
except (validate.ValidationException) as exc:
_logger.error(u"Tool definition failed validation:\n%s",
exc,
exc_info=args.debug)
infor = "Tool definition failed validation:\n%s" + exc + args.debug
return infor, 1
except (RuntimeError, WorkflowException) as exc:
_logger.error(u"Tool definition failed initialization:\n%s",
exc,
exc_info=args.debug)
infor = "Tool definition failed initialization:\n%s" + exc + args.debug
return infor, 1
except Exception as exc:
_logger.error(
u"I'm sorry, I couldn't load this CWL file%s.\nThe error was: %s",
try_again_msg,
exc if not args.debug else "",
exc_info=args.debug)
return "I'm sorry, I couldn't load this CWL file", 1
if isinstance(tool, int):
return tool, 0
# If on MacOS platform, TMPDIR must be set to be under one of the
# shared volumes in Docker for Mac
# More info: https://dockstore.org/docs/faq
if sys.platform == "darwin":
default_mac_path = "/private/tmp/docker_tmp"
if runtimeContext.tmp_outdir_prefix == DEFAULT_TMP_PREFIX:
runtimeContext.tmp_outdir_prefix = default_mac_path
for dirprefix in ("tmpdir_prefix", "tmp_outdir_prefix",
"cachedir"):
if getattr(runtimeContext, dirprefix) and getattr(
runtimeContext, dirprefix) != DEFAULT_TMP_PREFIX:
sl = "/" if getattr(runtimeContext, dirprefix).endswith("/") or dirprefix == "cachedir" \
else ""
setattr(
runtimeContext, dirprefix,
os.path.abspath(getattr(runtimeContext, dirprefix)) +
sl)
if not os.path.exists(
os.path.dirname(getattr(runtimeContext,
dirprefix))):
try:
os.makedirs(
os.path.dirname(
getattr(runtimeContext, dirprefix)))
except Exception as e:
_logger.error("Failed to create directory: %s", e)
infor = "Failed to create directory: %s" + e + ""
return infor, 1
if args.cachedir:
if args.move_outputs == "move":
runtimeContext.move_outputs = "copy"
runtimeContext.tmp_outdir_prefix = args.cachedir
runtimeContext.secret_store = getdefault(
runtimeContext.secret_store, SecretStore())
try:
initialized_job_order_object = init_job_order(
job_order_object,
args,
tool,
jobloader,
stdout,
print_input_deps=args.print_input_deps,
relative_deps=args.relative_deps,
input_basedir=input_basedir,
secret_store=runtimeContext.secret_store)
except SystemExit as err:
return err.code
if not executor:
if args.parallel:
executor = MultithreadedJobExecutor()
else:
executor = SingleJobExecutor()
assert executor is not None
if isinstance(initialized_job_order_object, int):
return initialized_job_order_object
try:
runtimeContext.basedir = input_basedir
del args.workflow
del args.job_order
conf_file = getattr(args,
"beta_dependency_resolvers_configuration",
None) # Text
use_conda_dependencies = getattr(args,
"beta_conda_dependencies",
None) # Text
job_script_provider = None # type: Optional[DependenciesConfiguration]
if conf_file or use_conda_dependencies:
runtimeContext.job_script_provider = DependenciesConfiguration(
args)
runtimeContext.find_default_container = \
functools.partial(find_default_container, args)
runtimeContext.make_fs_access = getdefault(
runtimeContext.make_fs_access, StdFsAccess)
(out, status) = executor(tool,
initialized_job_order_object,
runtimeContext,
logger=_logger)
# This is the workflow output, it needs to be written
if out is not None:
def loc_to_path(obj):
for field in ("path", "nameext", "nameroot",
"dirname"):
if field in obj:
del obj[field]
if obj["location"].startswith("file://"):
obj["path"] = uri_file_path(obj["location"])
visit_class(out, ("File", "Directory"), loc_to_path)
# Unsetting the Generation fron final output object
visit_class(out, ("File", ),
MutationManager().unset_generation)
if isinstance(out, string_types):
stdout.write(out)
else:
stdout.write(
json_dumps(
out,
indent=4, # type: ignore
ensure_ascii=False))
stdout.write("\n")
if hasattr(stdout, "flush"):
stdout.flush() # type: ignore
if status != "success":
_logger.warning(u"Final process status is %s", status)
infor = "Final process status is %s" + status + ""
return infor, 1
_logger.info(u"Final process status is %s", status)
return out, status
except (validate.ValidationException) as exc:
_logger.error(u"Input object failed validation:\n%s",
exc,
exc_info=args.debug)
infor = "Input object failed validation:\n%s" + exc + args.debug
return infor, 1
except UnsupportedRequirement as exc:
_logger.error(
u"Workflow or tool uses unsupported feature:\n%s",
exc,
exc_info=args.debug)
infor = "Workflow or tool uses unsupported feature:\n%s" + exc + args.debug
return infor, 3
except WorkflowException as exc:
_logger.error(u"Workflow error%s:\n%s",
try_again_msg,
strip_dup_lineno(six.text_type(exc)),
exc_info=args.debug)
infor = "Workflow error%s:\n%s" + try_again_msg + strip_dup_lineno(
six.text_type(exc)) + args.debug
return infor, 1
except Exception as exc:
_logger.error(u"Unhandled error%s:\n %s",
try_again_msg,
exc,
exc_info=args.debug)
infor = "Unhandled error%s:\n %s" + try_again_msg + exc + args.debug
return infor, 1
finally:
_logger.removeHandler(stderr_handler)
_logger.addHandler(defaultStreamHandler)