def __init__(self, infile, outfile, reference=None, *args, **kargs):
""".. rubric:: constructor
:param str infile: input BAM file
:param str outfile: output CRAM filename
:param str reference: reference file in :term:`FASTA` format
"""
super(BAM2CRAM, self).__init__(infile, outfile, *args, **kargs)
self._default_method = "samtools"
self.reference = reference
if self.reference is None:
logger.debug("No reference provided. Infering from input file")
# try to find the local file replacing .sam by .fa
reference = infile.replace(".cram", ".fa")
if os.path.exists(reference):
logger.debug(
"Reference found from inference ({})".format(reference))
else:
logger.debug("No reference found.")
msg = "Please enter the reference corresponding "
msg += "to the input BAM file:"
reference = input(msg)
if os.path.exists(reference) is False:
raise IOError("Reference required")
else:
logger.debug("Reference exist ({}).".format(reference))
self.reference = reference
self.threads = cpu_count()
def __init__(self, infile, outfile, reference=None, *args, **kargs):
""".. rubric:: constructor
:param str infile: input SAM file
:param str outfile: output filename
:param str reference: reference file in :term:`FASTA` format
command used::
samtools view -SCh
.. note:: the API related to the third argument may change in the future.
"""
super(SAM2CRAM, self).__init__(infile, outfile, *args, **kargs)
self.reference = reference
if self.reference is None:
logger.debug("No reference provided. Infering from input file")
# try to find the local file replacing .sam by .fa
reference = infile.replace(".sam", ".fa")
if os.path.exists(reference):
logger.debug("Reference found from inference ({})".format(reference))
else:
logger.debug("No reference found.")
msg = "Please enter the reference corresponding "
msg += "to the input SAM file:"
reference = raw_input(msg)
if os.path.exists(reference) is False:
raise IOError("Reference required")
else:
logger.debug("Reference exist ({}).".format(reference))
self.reference = reference
self.threads = cpu_count()
def __init__(self, infile, outfile):
""".. rubric:: constructor
:param str infile: The path of the input file.
:param str outfile: The path of The output file
"""
if os.path.exists(infile) is False:
msg = "Incorrect input file: %s" % infile
_log.error(msg)
raise ValueError(msg)
self.infile = infile
self.outfile = outfile
self.threads = cpu_count()
self._execute_mode = "subprocess" # set to shell to call shell() method
def __init__(self, infile, outfile, reference=None, *args, **kargs):
""".. rubric:: constructor
:param str infile: input FASTQ file
:param str outfile: output filename
:param str reference: reference file in :term:`FASTA` format
command used::
samtools view -@ <thread> -Sh -T <reference> in.cram > out.sam
.. note:: the API related to the third argument may change in the future.
"""
super(CRAM2FASTQ, self).__init__(infile, outfile, *args, **kargs)
self.threads = cpu_count()
def __init__(self, infile, outfile):
""".. rubric:: constructor
:param str infile: The path of the input file.
:param str outfile: The path of The output file
"""
# do not check the existence of the input file because it could be just a prefix
# if os.path.exists(infile) is False:
# msg = "Incorrect input file: %s" % infile
# _log.error(msg)
# raise ValueError(msg)
if not outfile:
outfile = generate_outfile_name(infile, self.output_ext[0])
self.infile = infile
self.outfile = outfile
self.threads = cpu_count()
self._execute_mode = "shell" #"subprocess" # set to shell to call shell() method
self.logger = logger
class ConvBase(metaclass=ConvMeta):
"""Base class for all converters.
To build a new converter, create a new class which inherits from
:class:`ConvBase` and implement method that performs the conversion.
The name of the converter method must start with ``_method_``.
For instance: ::
class Fastq2Fasta(ConvBase):
def _method_python(self, *args, **kwargs):
# include your code here. You can use the infile and outfile
# attributes.
self.infile
self.outfile
"""
# specify the extensions of the input file, can be a sequence (must be
# overridden in subclasses)
input_ext = None
# specify the extensions of the output file, can be a sequence (must be
# overridden in subclasses)
output_ext = None
# list available methods
available_methods = []
# default method should be provided
_default_method = None
_library_to_install = None
_is_compressor = False
# threads to be used by default if argument is required in a method
threads = cpu_count()
def __init__(self, infile, outfile):
""".. rubric:: constructor
:param str infile: the path of the input file.
:param str outfile: the path of The output file
"""
if not outfile:
outfile = generate_outfile_name(infile, self.output_ext[0])
self.infile = infile
self.outfile = outfile
# execute mode can be shell or subprocess.
self._execute_mode = "shell"
# The logger to be set to INFO, DEBUG, WARNING, ERROR, CRITICAL
self.logger = logger
def __call__(self, *args, method_name=None, **kwargs):
"""
:param str method: the method to be found in :attr:`available_methods`
:param *args: positional arguments
:param *kwargs: keyword arguments
"""
# If method provided, use it
if "method" in kwargs:
method_name = kwargs["method"]
del kwargs["method"]
# If not, but there is one argument, presumably this is
# the method
method_name = method_name or self.default
# If not, we need to check the name
# "dummy" is a method used to evaluate the cost of the
# execute() method for the benchmark
if method_name not in self.available_methods + ['dummy']:
msg = "Methods available are {}".format(self.available_methods)
_log.error(msg)
raise ValueError(msg)
_log.info("{}> Executing {} method ".format(self.name, method_name))
# reference to the method requested
method_reference = getattr(self, "_method_{}".format(method_name))
# call the method itself
t1 = time.time()
method_reference(*args, **kwargs)
t2 = time.time()
_log.info("Took {} seconds ".format(t2 - t1))
#FIXME property not use
@property
def name(self):
"""
The name of the class
"""
return type(self).__name__
def _method_dummy(self, *args, **kwargs):
# The execute commands has a large initialisation cost (about a second)
# This commands does not and can be used to evaluate that cost
self.execute("")
def shell(self, cmd):
from bioconvert.core.shell import shell
_log.info("CMD: {}".format(cmd))
shell(cmd)
def execute(self, cmd, ignore_errors=False, verbose=False, shell=False):
if shell is True or self._execute_mode == "shell":
self.shell(cmd)
return
_log.info("CMD: {}".format(cmd))
self._execute(cmd, ignore_errors, verbose)
def _execute(self, cmd, ignore_errors=False, verbose=False):
"""
Execute a command in a sub-shell
:param str cmd: the command to execute
:param ignore_errors: If True the result is returned whatever the
return value of the sub-shell.
Otherwise a Runtime error is raised when the sub-shell
return a non zero value
:param verbose: If true displays errors on standard error
:return: the result of the command
:rtype: a :class:`StringIO` instance
"""
try:
process_ = Popen(cmd,
shell=True,
stdout=PIPE,
stderr=PIPE,
stdin=None)
except Exception as err:
msg = "Failed to execute Command: '{}'. error: '{}'".format(cmd, err)
raise RuntimeError(msg)
inputs = [process_.stdout, process_.stderr]
output = StringIO()
errors = StringIO()
while process_.poll() is None:
# select has 3 parameters, 3 lists, the sockets, the fileobject to watch
# in reading, writing, the errors
# in addition a timeout option (the call is blocking while a fileObject
# is not ready to be processed)
# by return we get 3 lists with the fileObject to be processed
# in reading, writing, errors.
readable, writable, exceptional = select.select(inputs, [], [], 1)
while readable and inputs:
for flow in readable:
data = flow.read()
if not data:
# the flow ready in reading which has no data
# is a closed flow
# thus we must stop to watch it
inputs.remove(flow)
if flow is process_.stdout:
output.write(data.decode("utf-8"))
elif flow is process_.stderr:
errors.write(data.decode("utf-8"))
print(process_.stderr)
readable, writable, exceptional = select.select(inputs, [], [], 1)
errors = errors.getvalue().strip()
if verbose:
if errors:
print(errors, file=sys.stderr)
if process_.returncode != 0:
if not ignore_errors:
raise RuntimeError(errors)
else:
return output
def boxplot_benchmark(self, N=5, rerun=True, include_dummy=False,
to_exclude=[], to_include=[], rot_xticks=90,
boxplot_args={}):
"""Simple wrapper to call :class:`Benchmark` and plot the results
see :class:`~bioconvert.core.benchmark.Benchmark` for details.
"""
self._benchmark = Benchmark(self, N=N, to_exclude=to_exclude,
to_include=to_include)
self._benchmark.include_dummy = include_dummy
data = self._benchmark.plot(rerun=rerun, rot_xticks=rot_xticks,
boxplot_args=boxplot_args)
return data
def _get_default_method(self):
if self._default_method is None:
return self.available_methods[0]
elif self._default_method not in self.available_methods:
return self.available_methods[0]
else:
return self._default_method
default = property(_get_default_method)
def install_tool(self, executable):
"""Install the given tool, using the script:
bioconvert/install_script/install_executable.sh
if the executable is not already present
:param executable: executable to install
:return: nothing
"""
# imported but not unused (when we don't have bioconvert_path)
# import bioconvert
from bioconvert import bioconvert_data
if shutil.which(executable) is None:
logger.info("Installing tool : " + executable)
# Assigned but never used, says flake8
# bioconvert_path = bioconvert.__path__[0]
script = bioconvert_data(
'install_' + executable + '.sh', where="../misc")
subprocess.call(['sh', script])
@classmethod
def add_argument_to_parser(cls, sub_parser):
sub_parser.description = cls.get_description()
for arg in itertools.chain(cls.get_common_arguments_for_converter(), cls.get_additional_arguments()):
arg.add_to_sub_parser(sub_parser)
@classmethod
def get_description(cls):
return "Allow to convert file in '%s' to '%s' format." % ConvMeta.split_converter_to_format(cls.__name__)
@classmethod
def get_additional_arguments(cls):
return []
@staticmethod
def get_common_arguments():
yield ConvArg(
names="input_file",
nargs="?",
default=None,
type=ConvArg.file,
help="The path to the file to convert.",
)
yield ConvArg(
names="output_file",
nargs="?",
default=None,
type=ConvArg.file,
output_argument=True,
help="The path where the result will be stored.",
)
yield ConvArg(
names=["-f", "--force", ],
action="store_true",
help="if outfile exists, it is overwritten with this option",
)
yield ConvArg(
names=["-v", "--verbosity", ],
default=bioconvert.logger.level,
help="Set the outpout verbosity.",
choices=["DEBUG", "INFO", "WARNING", "ERROR", "CRITICAL"],
)
yield ConvArg(
names=["--raise-exception", ],
action="store_true",
help="Let exception ending the execution be raised and displayed",
)
yield ConvArg(
names=["-m", "--batch", ],
default=False,
action="store_true",
help="Allow conversion of a set of files using wildcards. You "
"must use quotes to escape the wildcards. For instance: "
"--batch 'test*fastq' ")
yield ConvArg(
names=["-b", "--benchmark", ],
default=False,
action="store_true",
help="Running all available methods",
)
yield ConvArg(
names=["-N", "--benchmark-N", ],
default=5,
type=int,
help="Number of trials for each methods",
)
yield ConvArg(
names=["-a", "--allow-indirect-conversion", ],
default=False,
action="store_true",
help="Allow to chain converter when direct conversion is absent",
)
@classmethod
def get_common_arguments_for_converter(cls):
for a in ConvBase.get_common_arguments():
yield a
try:
# Some converter does not have any method and work in __call__, so preventing to crash by searching for them
yield ConvArg(
names=["-c", "--method", ],
nargs="?",
default=cls._get_default_method(cls),
help="The method to use to do the conversion.",
choices=cls.available_methods,
)
except Exception as e:
_log.warning("converter '{}' does not seems to have methods: {}".format(cls.__name__, e))
pass
yield ConvArg(
names=["-s", "--show-methods", ],
default=False,
action="store_true",
help="A converter may have several methods",
)