class TestRules(unittest.TestCase):
"""Test rules"""
def setUp(self):
self.workflow = Workflow("foo")
name = self.workflow.add_rule(name="bar")
self.workflow._rules["bar"].set_output(*(), **{"foo":"bar"})
self.workflow._rules["bar"].set_params(*(), **{"cmd":"foo", "options":["foo", "bar"]})
def test_create_rule(self):
self.assertListEqual(["bar"], [x.name for x in self.workflow.rules])
create_rule_from_existing(name="foo", template="bar", workflow=self.workflow)
self.assertListEqual(["bar", "foo"], [x.name for x in self.workflow.rules])
@raises(NoRulesException)
def test_create_rule_empty_workflow(self):
create_rule_from_existing(name="foo", template="bar", workflow=Workflow("foo"))
@raises(AssertionError)
def test_create_rule_wrong_workflow(self):
create_rule_from_existing(name="foo", template="bar", workflow=None)
@raises(UnknownRuleException)
def test_create_rule_wrong_template(self):
create_rule_from_existing(name="bar", template="foo", workflow=self.workflow)
def test_create_rule_add_output(self):
create_rule_from_existing(name="foo", template="bar", workflow=self.workflow, **{'output':((), {'bar':'foo'})})
self.assertDictEqual({'bar':'foo'}, dict(self.workflow.get_rule("foo").output))
def test_create_rule_add_params(self):
create_rule_from_existing(name="foo", template="bar", workflow=self.workflow, **{'params':((), {'cmd':'bar', 'options':['foo']})})
self.assertDictEqual({'cmd':'bar', 'options':['foo']}, dict(self.workflow.get_rule("foo").params))
def bash_completion(snakefile="Snakefile"):
if not len(sys.argv) >= 2:
print("Calculate bash completion for snakemake. This tool shall not be invoked by hand.")
sys.exit(1)
prefix = sys.argv[2]
if prefix.startswith("-"):
opts = [
action.option_strings[0] for action in get_argument_parser()._actions
if action.option_strings and action.option_strings[0].startswith(prefix)]
print(*opts, sep="\n")
else:
files = glob.glob("{}*".format(prefix))
if files:
print(*files, sep="\n")
elif os.path.exists(snakefile):
workflow = Workflow(snakefile=snakefile, snakemakepath=get_snakemake_path())
workflow.include(snakefile)
workflow_files = sorted(set(
file for file in workflow.concrete_files
if file.startswith(prefix)))
if workflow_files:
print(*workflow_files, sep="\n")
rules = [
rule.name for rule in workflow.rules
if rule.name.startswith(prefix)]
if rules:
print(*rules, sep="\n")
sys.exit(0)
def load_rule_args(snakefile, rule_name, default_wildcards=None, change_dir=False):
"""
Returns a rule object for some default arguments.
Example usage:
```
try:
snakemake
except NameError:
snakefile_path = os.getcwd() + "/Snakefile"
snakemake = load_rule_args(
snakefile = snakefile_path,
rule_name = 'create_prediction_target',
default_wildcards={
'ds_dir': 'full_data_samplefilter'
}
)
```
"""
# save current working dir for later
cwd = os.getcwd()
try:
if default_wildcards == None:
default_wildcards = dict()
# change to snakefile directory
os.chdir(os.path.dirname(snakefile))
# load workflow
workflow = Workflow(snakefile=snakefile)
workflow.include(snakefile)
# get rule
rule = workflow.get_rule(rule_name)
smk_input = dict(rule.expand_input(default_wildcards)[0])
smk_output = dict(rule.expand_output(default_wildcards)[0])
smk_params = dict(rule.expand_params(
default_wildcards, rule.input, rule.output, AttrDict(rule.resources)))
# setup rule arguments
retval = SnakemakeRuleArgs(
input=smk_input,
params=smk_params,
output=smk_output,
wildcards=default_wildcards
)
return retval
finally:
if not change_dir:
# change back to previous working directory
os.chdir(cwd)
def snake(args):
chunk_path = 'mtsv-chunk'
fm_build_path = 'mtsv-build'
print(chunk_path)
print(fm_build_path)
print(args)
print(partition(args))
# for i in args:
# print(i)
workflow = Workflow("__file__", overwrite_workdir=args.working_dir)
if args.cluster_cfg is not None:
workflow.cluster_cfg = args.cluster_cfg
snakemake.workflow.rules = Rules()
snakemake.workflow.config = dict()
def get_workflow(self):
""" make sure there is a workflow object
TODO:
* allow multiple workflows?
* what king of options to allow?
* allow options every time or just first?
"""
if self.workflow is None:
# create a new workflow object with some basic defaults
# create a blank file just so snakemake has something to hang on to
# (this file cannot be read from on some Windows systems...)
self.tempfiles['root'] = tempfile.NamedTemporaryFile('w')
self.workflow = Workflow(snakefile=self.tempfiles['root'].name)
return self.workflow
def create_workflow(snakefile):
workflow = Workflow(snakefile=snakefile, use_conda=True)
try:
workflow.include(snakefile,
overwrite_first_rule=True,
print_compilation=False)
workflow.check()
except (Exception, BaseException) as ex:
print_exception(ex, workflow.linemaps)
success = False
return workflow, success
def load_rule_args(snakefile, rule_name, default_wildcards=None, change_dir=False, create_dir=True, root=None):
"""
Returns a rule object for some default arguments.
Example usage:
```
snakefile_path = os.getcwd() + "/Snakefile"
try:
snakemake
except NameError:
snakemake = load_rule_args(
snakefile = snakefile_path,
rule_name = 'create_prediction_target',
default_wildcards={
'ds_dir': 'full_data_samplefilter'
},
# root = "./" # path relative to snakefile
)
```
"""
# save current working dir for later
cwd = os.getcwd()
if root is None:
root = os.path.dirname(snakefile)
else:
if not os.path.isabs(root):
root = os.path.join(os.path.dirname(snakefile), root)
log.info("root dir: %s", root)
try:
if default_wildcards == None:
default_wildcards = dict()
# change to root directory
os.chdir(root)
# load workflow
workflow = Workflow(snakefile=snakefile)
workflow.include(snakefile)
# get rule
rule = workflow.get_rule(rule_name)
smk_resources = AttrDict(rule.resources)
smk_input = dict(rule.expand_input(default_wildcards)[0])
smk_output = dict(rule.expand_output(default_wildcards)[0])
smk_params = dict(rule.expand_params(
default_wildcards,
rule.input,
rule.output,
smk_resources
))
# Make paths in snakemake inputs and outputs absolute
smk_input = map_custom_wd(workflow, smk_input, root)
smk_output = map_custom_wd(workflow, smk_output, root)
if create_dir:
mk_dirs(smk_output)
# setup rule arguments
retval = SnakemakeRuleArgs(
resources=smk_resources,
input=smk_input,
params=smk_params,
output=smk_output,
wildcards=default_wildcards
)
return retval
finally:
if not change_dir:
# change back to previous working directory
os.chdir(cwd)
logger.info("rule '{}' up to date".format(dest))
parser = argparse.ArgumentParser("Copy/sync rules to a given directory")
parser.add_argument('Snakefile', help="Snakefile to import")
parser.add_argument('-n', '--dry-run', action="store_true", help="Dry run")
parser.add_argument('-d',
'--outdir',
action="store",
default=os.curdir,
help="Snakefile to import")
args = parser.parse_args()
snakefile = os.path.abspath(args.Snakefile)
workflow = Workflow(snakefile=snakefile)
try:
workflow.include(snakefile,
overwrite_first_rule=True,
print_compilation=False)
workflow.check()
except (Exception, BaseException) as ex:
print_exception(ex, workflow.linemaps)
success = False
# Map the rules included from snakemake_rules
DEST = args.outdir
rules = {
x: os.path.join(DEST, os.path.relpath(x, SNAKEMAKE_RULES_PATH))
for x in workflow.included if x.startswith(SNAKEMAKE_RULES_PATH)
def snakemake(snakefile,
listrules=False,
cores=1,
resources=None,
workdir=None,
targets=None,
dryrun=False,
touch=False,
forcetargets=False,
forceall=False,
forcerun=None,
prioritytargets=None,
stats=None,
printreason=False,
printshellcmds=False,
printdag=False,
printrulegraph=False,
nocolor=False,
quiet=False,
keepgoing=False,
cluster=None,
immediate_submit=False,
standalone=False,
ignore_ambiguity=False,
snakemakepath=None,
lock=True,
unlock=False,
cleanup_metadata=None,
force_incomplete=False,
ignore_incomplete=False,
list_version_changes=False,
list_code_changes=False,
list_input_changes=False,
list_params_changes=False,
summary=False,
output_wait=3,
print_compilation=False,
debug=False,
notemp=False,
nodeps=False,
jobscript=None,
timestamp=False):
"""
Run snakemake on a given snakefile.
Note: at the moment, this function is not thread-safe!
Arguments
snakefile -- the snakefile.
list -- list rules.
jobs -- maximum number of parallel jobs (default: 1).
directory -- working directory (default: current directory).
rule -- execute this rule (default: first rule in snakefile).
dryrun -- print the rules that would be executed,
but do not execute them.
forcethis -- force the selected rule to be executed
forceall -- force all rules to be executed
time_measurements -- measure the running times of all rules
lock -- lock the working directory
"""
init_logger(nocolor=nocolor, stdout=dryrun, debug=debug, timestamp=timestamp)
if not os.path.exists(snakefile):
logger.error("Error: Snakefile \"{}\" not present.".format(snakefile))
return False
if workdir:
olddir = os.getcwd()
workflow = Workflow(
snakefile=snakefile, snakemakepath=snakemakepath,
jobscript=jobscript)
if standalone:
try:
# set the process group
os.setpgrp()
except:
# ignore: if it does not work we can still work without it
pass
success = True
try:
workflow.include(snakefile, workdir=workdir,
overwrite_first_rule=True, print_compilation=print_compilation)
workflow.check()
if not print_compilation:
if listrules:
workflow.list_rules()
else:
if not printdag and not printrulegraph:
# handle subworkflows
subsnakemake = partial(
snakemake,
cores=cores,
resources=resources,
dryrun=dryrun,
touch=touch,
printreason=printreason,
printshellcmds=printshellcmds,
nocolor=nocolor,
quiet=quiet,
keepgoing=keepgoing,
cluster=cluster,
immediate_submit=immediate_submit,
standalone=standalone,
ignore_ambiguity=ignore_ambiguity,
snakemakepath=snakemakepath,
lock=lock,
unlock=unlock,
cleanup_metadata=cleanup_metadata,
force_incomplete=force_incomplete,
ignore_incomplete=ignore_incomplete,
output_wait=output_wait,
debug=debug,
notemp=notemp,
nodeps=nodeps,
jobscript=jobscript,
timestamp=timestamp)
for subworkflow in workflow.subworkflows:
logger.warning("Executing subworkflow {}.".format(subworkflow.name))
if not subsnakemake(subworkflow.snakefile, workdir=subworkflow.workdir, targets=subworkflow.targets):
success = False
if workflow.subworkflows:
logger.warning("Executing main workflow.")
if success:
success = workflow.execute(
targets=targets, dryrun=dryrun, touch=touch,
cores=cores, forcetargets=forcetargets,
forceall=forceall, forcerun=forcerun,
prioritytargets=prioritytargets, quiet=quiet,
keepgoing=keepgoing, printshellcmds=printshellcmds,
printreason=printreason, printrulegraph=printrulegraph,
printdag=printdag, cluster=cluster,
immediate_submit=immediate_submit,
ignore_ambiguity=ignore_ambiguity,
workdir=workdir, stats=stats,
force_incomplete=force_incomplete,
ignore_incomplete=ignore_incomplete,
list_version_changes=list_version_changes,
list_code_changes=list_code_changes,
list_input_changes=list_input_changes,
list_params_changes=list_params_changes,
summary=summary,
output_wait=output_wait,
nolock=not lock,
unlock=unlock,
resources=resources,
notemp=notemp,
nodeps=nodeps,
cleanup_metadata=cleanup_metadata
)
except (Exception, BaseException) as ex:
print_exception(ex, workflow.linemaps)
success = False
if workdir:
os.chdir(olddir)
if workflow.persistence:
workflow.persistence.unlock()
return success
def setUp(self):
self.workflow = Workflow("foo")
name = self.workflow.add_rule(name="bar")
self.workflow._rules["bar"].set_output(*(), **{"foo":"bar"})
self.workflow._rules["bar"].set_params(*(), **{"cmd":"foo", "options":["foo", "bar"]})
def _load_snakefile(self, file_path):
"""Load the Snakefile"""
logger.error("loading snakefile {}".format(file_path))
workflow = Workflow(snakefile=file_path)
workflow.include(file_path)
return workflow.rules
def snakemake(snakefile,
listrules=False,
list_target_rules=False,
cores=1,
nodes=1,
local_cores=1,
resources=dict(),
config=dict(),
configfile=None,
config_args=None,
workdir=None,
targets=None,
dryrun=False,
touch=False,
forcetargets=False,
forceall=False,
forcerun=[],
prioritytargets=[],
stats=None,
printreason=False,
printshellcmds=False,
printdag=False,
printrulegraph=False,
printd3dag=False,
nocolor=False,
quiet=False,
keepgoing=False,
cluster=None,
cluster_config=None,
cluster_sync=None,
drmaa=None,
jobname="snakejob.{rulename}.{jobid}.sh",
immediate_submit=False,
standalone=False,
ignore_ambiguity=False,
snakemakepath=None,
lock=True,
unlock=False,
cleanup_metadata=None,
force_incomplete=False,
ignore_incomplete=False,
list_version_changes=False,
list_code_changes=False,
list_input_changes=False,
list_params_changes=False,
list_resources=False,
summary=False,
detailed_summary=False,
latency_wait=3,
benchmark_repeats=1,
wait_for_files=None,
print_compilation=False,
debug=False,
notemp=False,
nodeps=False,
keep_target_files=False,
allowed_rules=None,
jobscript=None,
timestamp=False,
greediness=None,
no_hooks=False,
overwrite_shellcmd=None,
updated_files=None,
log_handler=None,
keep_logger=False,
verbose=False):
"""Run snakemake on a given snakefile.
This function provides access to the whole snakemake functionality. It is not thread-safe.
Args:
snakefile (str): the path to the snakefile
listrules (bool): list rules (default False)
list_target_rules (bool): list target rules (default False)
cores (int): the number of provided cores (ignored when using cluster support) (default 1)
nodes (int): the number of provided cluster nodes (ignored without cluster support) (default 1)
local_cores (int): the number of provided local cores if in cluster mode (ignored without cluster support) (default 1)
resources (dict): provided resources, a dictionary assigning integers to resource names, e.g. {gpu=1, io=5} (default {})
config (dict): override values for workflow config
workdir (str): path to working directory (default None)
targets (list): list of targets, e.g. rule or file names (default None)
dryrun (bool): only dry-run the workflow (default False)
touch (bool): only touch all output files if present (default False)
forcetargets (bool): force given targets to be re-created (default False)
forceall (bool): force all output files to be re-created (default False)
forcerun (list): list of files and rules that shall be re-created/re-executed (default [])
prioritytargets (list): list of targets that shall be run with maximum priority (default [])
stats (str): path to file that shall contain stats about the workflow execution (default None)
printreason (bool): print the reason for the execution of each job (default false)
printshellcmds (bool): print the shell command of each job (default False)
printdag (bool): print the dag in the graphviz dot language (default False)
printrulegraph (bool): print the graph of rules in the graphviz dot language (default False)
printd3dag (bool): print a D3.js compatible JSON representation of the DAG (default False)
nocolor (bool): do not print colored output (default False)
quiet (bool): do not print any default job information (default False)
keepgoing (bool): keep goind upon errors (default False)
cluster (str): submission command of a cluster or batch system to use, e.g. qsub (default None)
cluster_config (str): configuration file for cluster options (default None)
cluster_sync (str): blocking cluster submission command (like SGE 'qsub -sync y') (default None)
drmaa (str): if not None use DRMAA for cluster support, str specifies native args passed to the cluster when submitting a job
jobname (str): naming scheme for cluster job scripts (default "snakejob.{rulename}.{jobid}.sh")
immediate_submit (bool): immediately submit all cluster jobs, regardless of dependencies (default False)
standalone (bool): kill all processes very rudely in case of failure (do not use this if you use this API) (default False)
ignore_ambiguity (bool): ignore ambiguous rules and always take the first possible one (default False)
snakemakepath (str): path to the snakemake executable (default None)
lock (bool): lock the working directory when executing the workflow (default True)
unlock (bool): just unlock the working directory (default False)
cleanup_metadata (bool): just cleanup metadata of output files (default False)
force_incomplete (bool): force the re-creation of incomplete files (default False)
ignore_incomplete (bool): ignore incomplete files (default False)
list_version_changes (bool): list output files with changed rule version (default False)
list_code_changes (bool): list output files with changed rule code (default False)
list_input_changes (bool): list output files with changed input files (default False)
list_params_changes (bool): list output files with changed params (default False)
summary (bool): list summary of all output files and their status (default False)
latency_wait (int): how many seconds to wait for an output file to appear after the execution of a job, e.g. to handle filesystem latency (default 3)
benchmark_repeats (int): number of repeated runs of a job if declared for benchmarking (default 1)
wait_for_files (list): wait for given files to be present before executing the workflow
list_resources (bool): list resources used in the workflow (default False)
summary (bool): list summary of all output files and their status (default False). If no option is specified a basic summary will be ouput. If 'detailed' is added as an option e.g --summary detailed, extra info about the input and shell commands will be included
detailed_summary (bool): list summary of all input and output files and their status (default False)
print_compilation (bool): print the compilation of the snakefile (default False)
debug (bool): allow to use the debugger within rules
notemp (bool): ignore temp file flags, e.g. do not delete output files marked as temp after use (default False)
nodeps (bool): ignore dependencies (default False)
keep_target_files (bool): Do not adjust the paths of given target files relative to the working directory.
allowed_rules (set): Restrict allowed rules to the given set. If None or empty, all rules are used.
jobscript (str): path to a custom shell script template for cluster jobs (default None)
timestamp (bool): print time stamps in front of any output (default False)
greediness (float): set the greediness of scheduling. This value between 0 and 1 determines how careful jobs are selected for execution. The default value (0.5 if prioritytargets are used, 1.0 else) provides the best speed and still acceptable scheduling quality.
overwrite_shellcmd (str): a shell command that shall be executed instead of those given in the workflow. This is for debugging purposes only.
updated_files(list): a list that will be filled with the files that are updated or created during the workflow execution
verbose(bool): show additional debug output (default False)
log_handler (function): redirect snakemake output to this custom log handler, a function that takes a log message dictionary (see below) as its only argument (default None). The log message dictionary for the log handler has to following entries:
:level:
the log level ("info", "error", "debug", "progress", "job_info")
:level="info", "error" or "debug":
:msg:
the log message
:level="progress":
:done:
number of already executed jobs
:total:
number of total jobs
:level="job_info":
:input:
list of input files of a job
:output:
list of output files of a job
:log:
path to log file of a job
:local:
whether a job is executed locally (i.e. ignoring cluster)
:msg:
the job message
:reason:
the job reason
:priority:
the job priority
:threads:
the threads of the job
Returns:
bool: True if workflow execution was successful.
"""
if updated_files is None:
updated_files = list()
if cluster or cluster_sync or drmaa:
cores = sys.maxsize
else:
nodes = sys.maxsize
if cluster_config:
cluster_config = load_configfile(cluster_config)
else:
cluster_config = dict()
if not keep_logger:
setup_logger(handler=log_handler,
quiet=quiet,
printreason=printreason,
printshellcmds=printshellcmds,
nocolor=nocolor,
stdout=dryrun,
debug=verbose,
timestamp=timestamp)
if greediness is None:
greediness = 0.5 if prioritytargets else 1.0
else:
if not (0 <= greediness <= 1.0):
logger.error("Error: greediness must be a float between 0 and 1.")
return False
if not os.path.exists(snakefile):
logger.error("Error: Snakefile \"{}\" not present.".format(snakefile))
return False
snakefile = os.path.abspath(snakefile)
cluster_mode = (cluster is not None) + (cluster_sync is not
None) + (drmaa is not None)
if cluster_mode > 1:
logger.error("Error: cluster and drmaa args are mutually exclusive")
return False
if debug and (cores > 1 or cluster_mode):
logger.error(
"Error: debug mode cannot be used with more than one core or cluster execution.")
return False
overwrite_config = dict()
if configfile:
overwrite_config.update(load_configfile(configfile))
if config:
overwrite_config.update(config)
if workdir:
olddir = os.getcwd()
if not os.path.exists(workdir):
logger.info(
"Creating specified working directory {}.".format(workdir))
os.makedirs(workdir)
workdir = os.path.abspath(workdir)
os.chdir(workdir)
workflow = Workflow(snakefile=snakefile,
snakemakepath=snakemakepath,
jobscript=jobscript,
overwrite_shellcmd=overwrite_shellcmd,
overwrite_config=overwrite_config,
overwrite_workdir=workdir,
overwrite_configfile=configfile,
config_args=config_args,
debug=debug)
if standalone:
try:
# set the process group
os.setpgrp()
except:
# ignore: if it does not work we can still work without it
pass
success = True
try:
workflow.include(snakefile,
overwrite_first_rule=True,
print_compilation=print_compilation)
workflow.check()
if not print_compilation:
if listrules:
workflow.list_rules()
elif list_target_rules:
workflow.list_rules(only_targets=True)
elif list_resources:
workflow.list_resources()
else:
# if not printdag and not printrulegraph:
# handle subworkflows
subsnakemake = partial(snakemake,
cores=cores,
nodes=nodes,
local_cores=local_cores,
resources=resources,
dryrun=dryrun,
touch=touch,
printreason=printreason,
printshellcmds=printshellcmds,
nocolor=nocolor,
quiet=quiet,
keepgoing=keepgoing,
cluster=cluster,
cluster_config=cluster_config,
cluster_sync=cluster_sync,
drmaa=drmaa,
jobname=jobname,
immediate_submit=immediate_submit,
standalone=standalone,
ignore_ambiguity=ignore_ambiguity,
snakemakepath=snakemakepath,
lock=lock,
unlock=unlock,
cleanup_metadata=cleanup_metadata,
force_incomplete=force_incomplete,
ignore_incomplete=ignore_incomplete,
latency_wait=latency_wait,
benchmark_repeats=benchmark_repeats,
verbose=verbose,
notemp=notemp,
nodeps=nodeps,
jobscript=jobscript,
timestamp=timestamp,
greediness=greediness,
no_hooks=no_hooks,
overwrite_shellcmd=overwrite_shellcmd,
config=config,
config_args=config_args,
keep_logger=True)
success = workflow.execute(
targets=targets,
dryrun=dryrun,
touch=touch,
cores=cores,
nodes=nodes,
local_cores=local_cores,
forcetargets=forcetargets,
forceall=forceall,
forcerun=forcerun,
prioritytargets=prioritytargets,
quiet=quiet,
keepgoing=keepgoing,
printshellcmds=printshellcmds,
printreason=printreason,
printrulegraph=printrulegraph,
printdag=printdag,
cluster=cluster,
cluster_config=cluster_config,
cluster_sync=cluster_sync,
jobname=jobname,
drmaa=drmaa,
printd3dag=printd3dag,
immediate_submit=immediate_submit,
ignore_ambiguity=ignore_ambiguity,
stats=stats,
force_incomplete=force_incomplete,
ignore_incomplete=ignore_incomplete,
list_version_changes=list_version_changes,
list_code_changes=list_code_changes,
list_input_changes=list_input_changes,
list_params_changes=list_params_changes,
summary=summary,
latency_wait=latency_wait,
benchmark_repeats=benchmark_repeats,
wait_for_files=wait_for_files,
detailed_summary=detailed_summary,
nolock=not lock,
unlock=unlock,
resources=resources,
notemp=notemp,
nodeps=nodeps,
keep_target_files=keep_target_files,
cleanup_metadata=cleanup_metadata,
subsnakemake=subsnakemake,
updated_files=updated_files,
allowed_rules=allowed_rules,
greediness=greediness,
no_hooks=no_hooks)
except BrokenPipeError:
# ignore this exception and stop. It occurs if snakemake output is piped into less and less quits before reading the whole output.
# in such a case, snakemake shall stop scheduling and quit with error 1
success = False
except (Exception, BaseException) as ex:
print_exception(ex, workflow.linemaps)
success = False
if workdir:
os.chdir(olddir)
if workflow.persistence:
workflow.persistence.unlock()
if not keep_logger:
logger.cleanup()
return success
def _create_snakemake_dag(snakefile: str,
configfiles: Optional[List[str]] = None,
**kwargs: Any) -> DAG:
"""Create ``snakemake.dag.DAG`` instance.
The code of this function comes from the Snakemake codebase and is adapted
to fullfil REANA purposes of getting the needed metadata.
:param snakefile: Path to Snakefile.
:type snakefile: string
:param configfiles: List of config files paths.
:type configfiles: List
:param kwargs: Snakemake args.
:type kwargs: Any
"""
overwrite_config = dict()
if configfiles is None:
configfiles = []
for f in configfiles:
# get values to override. Later configfiles override earlier ones.
overwrite_config.update(load_configfile(f))
# convert provided paths to absolute paths
configfiles = list(map(os.path.abspath, configfiles))
workflow = Workflow(
snakefile=snakefile,
overwrite_configfiles=configfiles,
overwrite_config=overwrite_config,
)
workflow.include(snakefile=snakefile, overwrite_first_rule=True)
workflow.check()
# code copied and adapted from `snakemake.workflow.Workflow.execute()`
# in order to build the DAG and calculate the job dependencies.
# https://github.com/snakemake/snakemake/blob/75a544ba528b30b43b861abc0ad464db4d6ae16f/snakemake/workflow.py#L525
def rules(items):
return map(
workflow._rules.__getitem__,
filter(workflow.is_rule, items),
)
if kwargs.get("keep_target_files"):
def files(items):
return filterfalse(workflow.is_rule, items)
else:
def files(items):
relpath = (lambda f: f if os.path.isabs(f) or f.startswith(
"root://") else os.path.relpath(f))
return map(relpath, filterfalse(workflow.is_rule, items))
if not kwargs.get("targets"):
targets = ([workflow.first_rule]
if workflow.first_rule is not None else list())
prioritytargets = kwargs.get("prioritytargets", [])
forcerun = kwargs.get("forcerun", [])
until = kwargs.get("until", [])
omit_from = kwargs.get("omit_from", [])
priorityrules = set(rules(prioritytargets))
priorityfiles = set(files(prioritytargets))
forcerules = set(rules(forcerun))
forcefiles = set(files(forcerun))
untilrules = set(rules(until))
untilfiles = set(files(until))
omitrules = set(rules(omit_from))
omitfiles = set(files(omit_from))
targetrules = set(
chain(
rules(targets),
filterfalse(Rule.has_wildcards, priorityrules),
filterfalse(Rule.has_wildcards, forcerules),
filterfalse(Rule.has_wildcards, untilrules),
))
targetfiles = set(
chain(files(targets), priorityfiles, forcefiles, untilfiles))
dag = DAG(
workflow,
workflow.rules,
targetrules=targetrules,
targetfiles=targetfiles,
omitfiles=omitfiles,
omitrules=omitrules,
)
workflow.persistence = Persistence(dag=dag)
dag.init()
dag.update_checkpoint_dependencies()
dag.check_dynamic()
return dag
#!/usr/bin/env python3
import sys, os
import unittest
from snakemake.workflow import Workflow
""" Can I unit test a Snakefile?
Of course I can!
Or, at least, I can test any functions defined at the top level.
Importing the functions from the Snakefile requres parsing it with the
Snakemake internals, and setting a couple of environment things.
"""
os.environ['TOOLBOX'] = 'dummy'
sf = os.path.join(os.path.dirname(__file__), '..', 'Snakefile.qc')
wf = Workflow(sf, overwrite_config=dict(runid='170221_K00166_0183_AHHT3HBBXX'))
wf.include(sf)
# I can now import top-level functions like so:
split_fq_name = wf.globals['split_fq_name']
class T(unittest.TestCase):
def test_split_fq_name(self):
"""Test the function which claims to work as follows...
Break out components from the name of a a FASTQ file.
eg. 10749/10749DMpool03/170221_K00166_0183_AHHT3HBBXX_8_10749DM0001L01_1.fastq.gz
eg. 170221_K00166_0183_AHHT3HBBXX_1_unassigned_1.fastq.gz
"""
# eg. 1
self.assertEqual(
split_fq_name(
n = 0
with open(input.fastq) as f:
for _ in f:
n += 1
with open(output.counts, 'w') as f:
print(n / 4, file=f)
"""
from snakemake.workflow import Workflow, Rules
import snakemake.workflow
from snakemake import shell
from snakemake.logging import setup_logger
setup_logger()
workflow = Workflow(__file__)
snakemake.workflow.rules = Rules()
snakemake.workflow.config = dict()
### Output from snakemake --print-compilation follows (reformatted)
workflow.include("pipeline.conf")
shell.prefix("set -euo pipefail;")
@workflow.rule(name='all', lineno=6, snakefile='.../Snakefile')
@workflow.input("reads.counts")
@workflow.norun()
@workflow.run
def __all(input, output, params, wildcards, threads, resources, log, version):
