__author__ = "Johannes Köster"
__contributors__ = ["Per Unneberg"]
__copyright__ = "Copyright 2015, Johannes Köster"
__email__ = "koester@jimmy.harvard.edu"
__license__ = "MIT"
import os
import json
import re
import inspect
import textwrap
from itertools import chain
from collections import Mapping
import multiprocessing
import string
import shlex
import sys
from snakemake.io import regex, Namedlist, Wildcards, _load_configfile
from snakemake.logging import logger
from snakemake.exceptions import WorkflowError
import snakemake
[docs]def validate(data, schema):
"""Validate data with JSON schema at given path.
Args:
data (object): data to validate. Can be a config dict or a pandas data frame.
schema (str): Path to JSON schema used for validation. The schema can also be
in YAML format. If validating a pandas data frame, the schema has to
describe a row record (i.e., a dict with column names as keys pointing
to row values). See http://json-schema.org. The path is interpreted
relative to the Snakefile when this function is called.
"""
try:
import jsonschema
except ImportError:
raise WorkflowError("The Python 3 package jsonschema must be installed "
"in order to use the validate directive.")
if not os.path.isabs(schema):
frame = inspect.currentframe().f_back
# if workflow object is not available this has not been started from a workflow
if "workflow" in frame.f_globals:
workflow = frame.f_globals["workflow"]
schema = os.path.join(workflow.current_basedir, schema)
schema = _load_configfile(schema, filetype="Schema")
if not isinstance(data, dict):
try:
import pandas as pd
if isinstance(data, pd.DataFrame):
for i, record in enumerate(data.to_dict("records")):
record = {k: v for k, v in record.items() if not pd.isnull(v)}
try:
jsonschema.validate(record, schema)
except jsonschema.exceptions.ValidationError as e:
raise WorkflowError(
"Error validating row {} of data frame.".format(i),
e)
return
except ImportError:
pass
raise WorkflowError("Unsupported data type for validation.")
else:
try:
jsonschema.validate(data, schema)
except jsonschema.exceptions.ValidationError as e:
raise WorkflowError("Error validating config file.", e)
[docs]def simplify_path(path):
"""Return a simplified version of the given path."""
relpath = os.path.relpath(path)
if relpath.startswith("../../"):
return path
else:
return relpath
[docs]def linecount(filename):
"""Return the number of lines of given file.
Args:
filename (str): the path to the file
"""
with open(filename) as f:
return sum(1 for l in f)
[docs]def listfiles(pattern, restriction=None, omit_value=None):
"""Yield a tuple of existing filepaths for the given pattern.
Wildcard values are yielded as the second tuple item.
Args:
pattern (str): a filepattern. Wildcards are specified in snakemake syntax, e.g. "{id}.txt"
restriction (dict): restrict to wildcard values given in this dictionary
omit_value (str): wildcard value to omit
Yields:
tuple: The next file matching the pattern, and the corresponding wildcards object
"""
pattern = os.path.normpath(pattern)
first_wildcard = re.search("{[^{]", pattern)
if first_wildcard:
dirname = os.path.dirname(pattern[:first_wildcard.start()])
if not dirname:
dirname = "."
else:
dirname = os.path.dirname(pattern)
pattern = re.compile(regex(pattern))
for dirpath, dirnames, filenames in os.walk(dirname):
for f in chain(filenames, dirnames):
if dirpath != ".":
f = os.path.normpath(os.path.join(dirpath, f))
match = re.match(pattern, f)
if match:
wildcards = Namedlist(fromdict=match.groupdict())
if restriction is not None:
invalid = any(omit_value not in v and v != wildcards[k]
for k, v in restriction.items())
if not invalid:
yield f, wildcards
else:
yield f, wildcards
[docs]def makedirs(dirnames):
"""Recursively create the given directory or directories without
reporting errors if they are present.
"""
if isinstance(dirnames, str):
dirnames = [dirnames]
for dirname in dirnames:
os.makedirs(dirname, exist_ok=True)
[docs]def report(text, path,
stylesheet=os.path.join(os.path.dirname(__file__), "report.css"),
defaultenc="utf8",
template=None,
metadata=None, **files):
"""Create an HTML report using python docutils.
This is deprecated in favor of the --report flag.
Attention: This function needs Python docutils to be installed for the
python installation you use with Snakemake.
All keywords not listed below are intepreted as paths to files that shall
be embedded into the document. They keywords will be available as link
targets in the text. E.g. append a file as keyword arg via F1=input[0]
and put a download link in the text like this:
.. code:: python
report('''
==============
Report for ...
==============
Some text. A link to an embedded file: F1_.
Further text.
''', outputpath, F1=input[0])
Instead of specifying each file as a keyword arg, you can also expand
the input of your rule if it is completely named, e.g.:
report('''
Some text...
''', outputpath, **input)
Args:
text (str): The "restructured text" as it is expected by python docutils.
path (str): The path to the desired output file
stylesheet (str): An optional path to a css file that defines the style of the document. This defaults to <your snakemake install>/report.css. Use the default to get a hint how to create your own.
defaultenc (str): The encoding that is reported to the browser for embedded text files, defaults to utf8.
template (str): An optional path to a docutils HTML template.
metadata (str): E.g. an optional author name or email address.
"""
try:
import snakemake.report
except ImportError:
raise WorkflowError(
"Python 3 package docutils needs to be installed to use the report function.")
snakemake.report.report(text, path,
stylesheet=stylesheet,
defaultenc=defaultenc,
template=template,
metadata=metadata, **files)
[docs]def R(code):
"""Execute R code.
This is deprecated in favor of the ``script`` directive.
This function executes the R code given as a string.
The function requires rpy2 to be installed.
Args:
code (str): R code to be executed
"""
try:
import rpy2.robjects as robjects
except ImportError:
raise ValueError(
"Python 3 package rpy2 needs to be installed to use the R function.")
robjects.r(format(textwrap.dedent(code), stepout=2))
class Unformattable:
def __init__(self, errormsg="This cannot be used for formatting"):
self.errormsg = errormsg
def __str__(self):
raise ValueError(self.errormsg)
[docs]def read_job_properties(jobscript,
prefix="# properties",
pattern=re.compile("# properties = (.*)")):
"""Read the job properties defined in a snakemake jobscript.
This function is a helper for writing custom wrappers for the
snakemake --cluster functionality. Applying this function to a
jobscript will return a dict containing information about the job.
"""
with open(jobscript) as jobscript:
for m in map(pattern.match, jobscript):
if m:
return json.loads(m.group(1))
[docs]def min_version(version):
"""Require minimum snakemake version, raise workflow error if not met."""
import pkg_resources
if pkg_resources.parse_version(
snakemake.__version__) < pkg_resources.parse_version(version):
raise WorkflowError(
"Expecting Snakemake version {} or higher.".format(version))
[docs]def update_config(config, overwrite_config):
"""Recursively update dictionary config with overwrite_config.
See
http://stackoverflow.com/questions/3232943/update-value-of-a-nested-dictionary-of-varying-depth
for details.
Args:
config (dict): dictionary to update
overwrite_config (dict): dictionary whose items will overwrite those in config
"""
def _update(d, u):
for (key, value) in u.items():
if (isinstance(value, Mapping)):
d[key] = _update(d.get(key, {}), value)
else:
d[key] = value
return d
_update(config, overwrite_config)
[docs]def available_cpu_count():
"""
Return the number of available virtual or physical CPUs on this system.
The number of available CPUs can be smaller than the total number of CPUs
when the cpuset(7) mechanism is in use, as is the case on some cluster
systems.
Adapted from http://stackoverflow.com/a/1006301/715090
"""
try:
with open('/proc/self/status') as f:
status = f.read()
m = re.search(r'(?m)^Cpus_allowed:\s*(.*)$', status)
if m:
res = bin(int(m.group(1).replace(',', ''), 16)).count('1')
if res > 0:
return min(res, multiprocessing.cpu_count())
except IOError:
pass
return multiprocessing.cpu_count()