def test_do_links(self):
src_tmp_dir = tempfile.mkdtemp()
dst_tmp_dir = os.path.join(src_tmp_dir, 'dst' )
safe_makedir(dst_tmp_dir)
src_file_path = os.path.join(src_tmp_dir, 'file1.txt')
dst_file_path = os.path.join(dst_tmp_dir, 'file1.txt')
open(src_file_path, 'w').close()
do_hardlink([src_file_path], dst_tmp_dir)
assert(filecmp.cmp(src_file_path, dst_file_path))
os.remove(dst_file_path)
do_symlink([src_file_path], dst_tmp_dir)
assert(filecmp.cmp(src_file_path, dst_file_path))
def test_do_links(self):
src_tmp_dir = tempfile.mkdtemp()
dst_tmp_dir = os.path.join(src_tmp_dir, 'dst' )
safe_makedir(dst_tmp_dir)
src_file_path = os.path.join(src_tmp_dir, 'file1.txt')
dst_file_path = os.path.join(dst_tmp_dir, 'file1.txt')
open(src_file_path, 'w').close()
do_hardlink([src_file_path], dst_tmp_dir)
assert(filecmp.cmp(src_file_path, dst_file_path))
os.remove(dst_file_path)
do_symlink([src_file_path], dst_tmp_dir)
assert(filecmp.cmp(src_file_path, dst_file_path))
def test_do_links(self):
src_tmp_dir = tempfile.mkdtemp()
dst_tmp_dir = os.path.join(src_tmp_dir, "dst")
safe_makedir(dst_tmp_dir)
src_file_path = os.path.join(src_tmp_dir, "file1.txt")
dst_file_path = os.path.join(dst_tmp_dir, "file1.txt")
open(src_file_path, "w").close()
do_hardlink([src_file_path], dst_tmp_dir)
assert filecmp.cmp(src_file_path, dst_file_path)
os.remove(dst_file_path)
do_symlink([src_file_path], dst_tmp_dir)
assert filecmp.cmp(src_file_path, dst_file_path)
def setup_analysis_directory_structure(fc_dir, projects_to_analyze,
restrict_to_projects=None, restrict_to_samples=None,
create_files=True,
fallback_libprep=None,
quiet=False,
config=None, config_file_path=None):
"""
Copy and sort files from their CASAVA-demultiplexed flowcell structure
into their respective project/sample/libPrep/FCIDs. This collects samples
split across multiple flowcells.
:param str fc_dir: The directory created by CASAVA for this flowcell.
:param dict config: The parsed configuration file.
:param set projects_to_analyze: A dict (of Project objects, or empty)
:param bool create_files: Alter the filesystem (as opposed to just parsing flowcells) (default True)
:param str fallback_libprep: If libprep cannot be determined, use this value if supplied (default None)
:param list restrict_to_projects: Specific projects within the flowcell to process exclusively
:param list restrict_to_samples: Specific samples within the flowcell to process exclusively
:returns: A list of NGIProject objects that need to be run through the analysis pipeline
:rtype: list
:raises KeyError: If a required configuration key is not available.
"""
LOG.info("Setting up analysis for demultiplexed data in source folder \"{}\"".format(fc_dir))
if not restrict_to_projects: restrict_to_projects = []
if not restrict_to_samples: restrict_to_samples = []
config["quiet"] = quiet # Hack because I enter here from a script sometimes
pattern="(.+(?:{}|{}))\/.+".format(config["analysis"]["sthlm_root"], config["analysis"]["upps_root"])
matches=re.match(pattern, fc_dir)
if matches:
flowcell_root=matches.group(1)
else:
LOG.error("cannot guess which project the flowcell {} belongs to".format(fc_dir))
raise RuntimeError
analysis_top_dir = os.path.abspath(os.path.join(flowcell_root,config["analysis"]["top_dir"]))
try:
safe_makedir(analysis_top_dir)
except OSError as e:
LOG.error('Error: Analysis top directory {} does not exist and could not '
'be created.'.format(analysis_top_dir))
fc_dir = fc_dir if os.path.isabs(fc_dir) else os.path.join(analysis_top_dir, fc_dir)
if not os.path.exists(fc_dir):
LOG.error("Error: Flowcell directory {} does not exist".format(fc_dir))
return []
# Map the directory structure for this flowcell
try:
fc_dir_structure = parse_flowcell(fc_dir)
except (OSError, ValueError) as e:
LOG.error("Error when processing flowcell dir \"{}\": {}".format(fc_dir, e))
return []
fc_full_id = fc_dir_structure['fc_full_id']
if not fc_dir_structure.get('projects'):
LOG.warn("No projects found in specified flowcell directory \"{}\"".format(fc_dir))
# Iterate over the projects in the flowcell directory
for project in fc_dir_structure.get('projects', []):
project_name = project['project_name']
project_original_name = project['project_original_name']
samplesheet_path = fc_dir_structure.get("samplesheet_path")
try:
# Maps e.g. "Y.Mom_14_01" to "P123"
project_id = get_project_id_from_name(project_name)
except (CharonError, RuntimeError, ValueError) as e:
LOG.warn('Could not retrieve project id from Charon (record missing?). '
'Using project name ("{}") as project id '
'(error: {})'.format(project_name, e))
project_id = project_name
# If specific projects are specified, skip those that do not match
if restrict_to_projects and project_name not in restrict_to_projects and \
project_id not in restrict_to_projects:
LOG.debug("Skipping project {} (not in restrict_to_projects)".format(project_name))
continue
LOG.info("Setting up project {}".format(project.get("project_name")))
# Create a project directory if it doesn't already exist, including
# intervening "DATA" directory
project_dir = os.path.join(analysis_top_dir, "DATA", project_id)
project_sl_dir = os.path.join(analysis_top_dir, "DATA", project_name)
project_analysis_dir = os.path.join(analysis_top_dir, "ANALYSIS", project_id)
project_analysis_sl_dir = os.path.join(analysis_top_dir, "ANALYSIS", project_name)
if create_files:
safe_makedir(project_dir, 0o2770)
safe_makedir(project_analysis_dir, 0o2770)
if not project_dir == project_sl_dir and \
not os.path.exists(project_sl_dir):
os.symlink(project_dir, project_sl_dir)
if not project_analysis_dir == project_analysis_sl_dir and \
not os.path.exists(project_analysis_sl_dir):
os.symlink(project_analysis_dir, project_analysis_sl_dir)
try:
project_obj = projects_to_analyze[project_dir]
except KeyError:
project_obj = NGIProject(name=project_name, dirname=project_id,
project_id=project_id,
base_path=analysis_top_dir)
projects_to_analyze[project_dir] = project_obj
# Iterate over the samples in the project
for sample in project.get('samples', []):
sample_name = sample['sample_name']
# If specific samples are specified, skip those that do not match
if restrict_to_samples and sample_name not in restrict_to_samples:
LOG.debug("Skipping sample {}: not in specified samples "
"{}".format(sample_name, ", ".join(restrict_to_samples)))
continue
LOG.info("Setting up sample {}".format(sample_name))
# Create a directory for the sample if it doesn't already exist
sample_dir = os.path.join(project_dir, sample_name)
if create_files: safe_makedir(sample_dir, 0o2770)
# This will only create a new sample object if it doesn't already exist in the project
sample_obj = project_obj.add_sample(name=sample_name, dirname=sample_name)
# Get the Library Prep ID for each file
pattern = re.compile(".*\.(fastq|fq)(\.gz|\.gzip|\.bz2)?$")
fastq_files = filter(pattern.match, sample.get('files', []))
# For each fastq file, create the libprep and seqrun objects
# and add the fastq file to the seqprep object
# Note again that these objects only get created if they don't yet exist;
# if they do exist, the existing object is returned
for fq_file in fastq_files:
# Try to parse from SampleSheet
try:
if not samplesheet_path: raise ValueError()
lane_num = re.match(r'[\w-]+_L\d{2}(\d)_\w+', fq_file).groups()[0]
libprep_name = determine_library_prep_from_samplesheet(samplesheet_path,
project_original_name,
sample_name,
lane_num)
except (IndexError, ValueError) as e:
LOG.debug('Unable to determine library prep from sample sheet file '
'("{}"); try to determine from Charon'.format(e))
try:
# Requires Charon access
libprep_name = determine_library_prep_from_fcid(project_id, sample_name, fc_full_id)
LOG.debug('Found libprep name "{}" in Charon'.format(libprep_name))
except ValueError:
charon_session = CharonSession()
libpreps = charon_session.sample_get_libpreps(project_id, sample_name).get('libpreps')
if len(libpreps) == 1:
libprep_name = libpreps[0].get('libprepid')
LOG.warn('Project "{}" / sample "{}" / seqrun "{}" / fastq "{}" '
'has no libprep information in Charon, but only one '
'library prep is present in Charon ("{}"). Using '
'this as the library prep.'.format(project_name,
sample_name,
fc_full_id,
fq_file,
libprep_name))
elif fallback_libprep:
libprep_name = fallback_libprep
LOG.warn('Project "{}" / sample "{}" / seqrun "{}" / fastq "{}" '
'has no libprep information in Charon, but a fallback '
'libprep value of "{}" was supplied -- using this '
'value.'.format(project_name,
sample_name,
fc_full_id,
fq_file,
libprep_name,
fallback_libprep))
else:
error_text = ('Project "{}" / sample "{}" / seqrun "{}" / fastq "{}" '
'has no libprep information in Charon. Skipping '
'analysis.'.format(project_name, sample_name,
fc_full_id, fq_file))
LOG.error(error_text)
if not config.get('quiet'):
mail_analysis(project_name=project_name,
sample_name=sample_name,
level="ERROR",
info_text=error_text)
continue
libprep_object = sample_obj.add_libprep(name=libprep_name,
dirname=libprep_name)
libprep_dir = os.path.join(sample_dir, libprep_name)
if create_files: safe_makedir(libprep_dir, 0o2770)
seqrun_object = libprep_object.add_seqrun(name=fc_full_id,
dirname=fc_full_id)
seqrun_dir = os.path.join(libprep_dir, fc_full_id)
if create_files: safe_makedir(seqrun_dir, 0o2770)
seqrun_object.add_fastq_files(fq_file)
if fastq_files and create_files:
src_sample_dir = os.path.join(fc_dir_structure['fc_dir'],
project['data_dir'],
project['project_dir'],
sample['sample_dir'])
for libprep_obj in sample_obj:
for seqrun_obj in libprep_obj:
src_fastq_files = [os.path.join(src_sample_dir, fastq_file) for
fastq_file in seqrun_obj.fastq_files]
seqrun_dst_dir = os.path.join(project_obj.base_path, project_obj.dirname,
sample_obj.dirname, libprep_obj.dirname,
seqrun_obj.dirname)
LOG.info("Symlinking fastq files from {} to {}...".format(src_sample_dir, seqrun_dir))
try:
do_symlink(src_fastq_files, seqrun_dir)
except OSError:
error_text = ('Could not symlink files for project/sample'
'libprep/seqrun {}/{}/{}/{}'.format(project_obj,
sample_obj,
libprep_obj,
seqrun_obj))
LOG.error(error_text)
if not config.get('quiet'):
mail_analysis(project_name=project_name,
sample_name=sample_name,
level="ERROR",
info_text=error_text)
continue
return projects_to_analyze
def setup_analysis_directory_structure(fc_dir,
projects_to_analyze,
restrict_to_projects=None,
restrict_to_samples=None,
create_files=True,
fallback_libprep=None,
quiet=False,
config=None,
config_file_path=None):
"""
Copy and sort files from their CASAVA-demultiplexed flowcell structure
into their respective project/sample/libPrep/FCIDs. This collects samples
split across multiple flowcells.
:param str fc_dir: The directory created by CASAVA for this flowcell.
:param dict config: The parsed configuration file.
:param set projects_to_analyze: A dict (of Project objects, or empty)
:param bool create_files: Alter the filesystem (as opposed to just parsing flowcells) (default True)
:param str fallback_libprep: If libprep cannot be determined, use this value if supplied (default None)
:param list restrict_to_projects: Specific projects within the flowcell to process exclusively
:param list restrict_to_samples: Specific samples within the flowcell to process exclusively
:returns: A list of NGIProject objects that need to be run through the analysis pipeline
:rtype: list
:raises KeyError: If a required configuration key is not available.
"""
LOG.info(
"Setting up analysis for demultiplexed data in source folder \"{}\"".
format(fc_dir))
if not restrict_to_projects: restrict_to_projects = []
if not restrict_to_samples: restrict_to_samples = []
config[
"quiet"] = quiet # Hack because I enter here from a script sometimes
#Checks flowcell path to establish which group owns it
pattern = ".+({}|{})\/.+".format(config["analysis"]["sthlm_root"],
config["analysis"]["upps_root"])
matches = re.match(pattern, fc_dir)
if matches:
flowcell_uppnexid = matches.group(1)
else:
LOG.error(
"cannot guess which project (sthlm/uppsala) the flowcell {} belongs to"
.format(fc_dir))
raise RuntimeError
analysis_top_dir = os.path.abspath(
os.path.join(config["analysis"]["base_root"], flowcell_uppnexid,
config["analysis"]["top_dir"]))
try:
safe_makedir(analysis_top_dir)
except OSError as e:
LOG.error(
'Error: Analysis top directory {} does not exist and could not '
'be created.'.format(analysis_top_dir))
fc_dir = fc_dir if os.path.isabs(fc_dir) else os.path.join(
analysis_top_dir, fc_dir)
if not os.path.exists(fc_dir):
LOG.error("Error: Flowcell directory {} does not exist".format(fc_dir))
return []
# Map the directory structure for this flowcell
try:
fc_dir_structure = parse_flowcell(fc_dir)
except (OSError, ValueError) as e:
LOG.error("Error when processing flowcell dir \"{}\": {}".format(
fc_dir, e))
return []
fc_full_id = fc_dir_structure['fc_full_id']
if not fc_dir_structure.get('projects'):
LOG.warning(
"No projects found in specified flowcell directory \"{}\"".format(
fc_dir))
# Iterate over the projects in the flowcell directory
for project in fc_dir_structure.get('projects', []):
project_name = project['project_name']
project_original_name = project['project_original_name']
samplesheet_path = fc_dir_structure.get("samplesheet_path")
# parse the samplesheet and get the expected sample numbers assigned by bcl2fastq
samplesheet_sample_numbers = get_sample_numbers_from_samplesheet(
samplesheet_path) if samplesheet_path else None
try:
# Maps e.g. "Y.Mom_14_01" to "P123"
project_id = get_project_id_from_name(project_name)
except (CharonError, RuntimeError, ValueError) as e:
LOG.warning(
'Could not retrieve project id from Charon (record missing?). '
'Using project name ("{}") as project id '
'(error: {})'.format(project_name, e))
project_id = project_name
# If specific projects are specified, skip those that do not match
if restrict_to_projects and project_name not in restrict_to_projects and \
project_id not in restrict_to_projects:
LOG.debug(
"Skipping project {} (not in restrict_to_projects)".format(
project_name))
continue
LOG.info("Setting up project {}".format(project.get("project_name")))
# Create a project directory if it doesn't already exist, including
# intervening "DATA" directory
project_dir = os.path.join(analysis_top_dir, "DATA", project_id)
project_sl_dir = os.path.join(analysis_top_dir, "DATA", project_name)
project_analysis_dir = os.path.join(analysis_top_dir, "ANALYSIS",
project_id)
project_analysis_sl_dir = os.path.join(analysis_top_dir, "ANALYSIS",
project_name)
if create_files:
safe_makedir(project_dir, 0o2770)
safe_makedir(project_analysis_dir, 0o2770)
if not project_dir == project_sl_dir and \
not os.path.exists(project_sl_dir):
os.symlink(project_dir, project_sl_dir)
if not project_analysis_dir == project_analysis_sl_dir and \
not os.path.exists(project_analysis_sl_dir):
os.symlink(project_analysis_dir, project_analysis_sl_dir)
try:
project_obj = projects_to_analyze[project_dir]
except KeyError:
project_obj = NGIProject(name=project_name,
dirname=project_id,
project_id=project_id,
base_path=analysis_top_dir)
projects_to_analyze[project_dir] = project_obj
# Iterate over the samples in the project
for sample in project.get('samples', []):
sample_name = sample['sample_name']
# If specific samples are specified, skip those that do not match
if restrict_to_samples and sample_name not in restrict_to_samples:
LOG.debug("Skipping sample {}: not in specified samples "
"{}".format(sample_name,
", ".join(restrict_to_samples)))
continue
LOG.info("Setting up sample {}".format(sample_name))
# Create a directory for the sample if it doesn't already exist
sample_dir = os.path.join(project_dir, sample_name)
if create_files: safe_makedir(sample_dir, 0o2770)
# This will only create a new sample object if it doesn't already exist in the project
sample_obj = project_obj.add_sample(name=sample_name,
dirname=sample_name)
# Get the Library Prep ID for each file
pattern = re.compile(".*\.(fastq|fq)(\.gz|\.gzip|\.bz2)?$")
fastq_files = list(filter(pattern.match, sample.get('files', [])))
# For each fastq file, create the libprep and seqrun objects
# and add the fastq file to the seqprep object
# Note again that these objects only get created if they don't yet exist;
# if they do exist, the existing object is returned
for fq_file in fastq_files:
# Try to use assignment from SampleSheet
samplesheet_sample = match_fastq_sample_number_to_samplesheet(
fq_file, samplesheet_sample_numbers, project_id)
if samplesheet_sample is not None and \
samplesheet_sample[6] is not None:
libprep_name = samplesheet_sample[6]
else:
LOG.debug(
'Unable to determine library prep from sample sheet file; try to determine from Charon'
)
try:
# Requires Charon access
libprep_name = determine_library_prep_from_fcid(
project_id, sample_name, fc_full_id)
LOG.debug('Found libprep name "{}" in Charon'.format(
libprep_name))
except ValueError:
charon_session = CharonSession()
libpreps = charon_session.sample_get_libpreps(
project_id, sample_name).get('libpreps')
if len(libpreps) == 1:
libprep_name = libpreps[0].get('libprepid')
LOG.warning(
'Project "{}" / sample "{}" / seqrun "{}" / fastq "{}" '
'has no libprep information in Charon, but only one '
'library prep is present in Charon ("{}"). Using '
'this as the library prep.'.format(
project_name, sample_name, fc_full_id,
fq_file, libprep_name))
elif fallback_libprep:
libprep_name = fallback_libprep
LOG.warning(
'Project "{}" / sample "{}" / seqrun "{}" / fastq "{}" '
'has no libprep information in Charon, but a fallback '
'libprep value of "{}" was supplied -- using this '
'value.'.format(project_name, sample_name,
fc_full_id, fq_file,
libprep_name))
else:
error_text = (
'Project "{}" / sample "{}" / seqrun "{}" / fastq "{}" '
'has no libprep information in Charon. Skipping '
'analysis.'.format(project_name, sample_name,
fc_full_id, fq_file))
LOG.error(error_text)
if not config.get('quiet'):
mail_analysis(project_name=project_name,
sample_name=sample_name,
level="ERROR",
info_text=error_text)
continue
libprep_object = sample_obj.add_libprep(name=libprep_name,
dirname=libprep_name)
libprep_dir = os.path.join(sample_dir, libprep_name)
if create_files: safe_makedir(libprep_dir, 0o2770)
seqrun_object = libprep_object.add_seqrun(name=fc_full_id,
dirname=fc_full_id)
seqrun_dir = os.path.join(libprep_dir, fc_full_id)
if create_files: safe_makedir(seqrun_dir, 0o2770)
seqrun_object.add_fastq_files(fq_file)
if fastq_files and create_files:
src_sample_dir = os.path.join(fc_dir_structure['fc_dir'],
project['data_dir'],
project['project_dir'],
sample['sample_dir'])
for libprep_obj in sample_obj:
for seqrun_obj in libprep_obj:
src_fastq_files = [
os.path.join(src_sample_dir, fastq_file)
for fastq_file in seqrun_obj.fastq_files
]
seqrun_dst_dir = os.path.join(project_obj.base_path,
"DATA",
project_obj.dirname,
sample_obj.dirname,
libprep_obj.dirname,
seqrun_obj.dirname)
LOG.info(
"Symlinking fastq files from {} to {}...".format(
src_sample_dir, seqrun_dst_dir))
try:
do_symlink(src_fastq_files, seqrun_dst_dir)
except OSError:
error_text = (
'Could not symlink files for project/sample'
'libprep/seqrun {}/{}/{}/{}'.format(
project_obj, sample_obj, libprep_obj,
seqrun_obj))
LOG.error(error_text)
if not config.get('quiet'):
mail_analysis(project_name=project_name,
sample_name=sample_name,
level="ERROR",
info_text=error_text)
continue
return projects_to_analyze
