from glob import glob
import json
import logging
-from optparse import OptionParser
+import netrc
+from optparse import OptionParser, OptionGroup
import os
from pprint import pprint, pformat
import shlex
from StringIO import StringIO
-import time
+import stat
+from subprocess import Popen, PIPE
import sys
+import time
import types
import urllib
import urllib2
import urlparse
from htsworkflow.util import api
-from htsworkflow.pipelines.sequences import \
- create_sequence_table, \
- scan_for_sequences
-from htsworkflow.pipelines import qseq2fastq
-from htsworkflow.pipelines import srf2fastq
+from htsworkflow.util.rdfhelp import \
+ dafTermOntology, \
+ fromTypedNode, \
+ get_model, \
+ get_serializer, \
+ load_into_model, \
+ submissionOntology
+from htsworkflow.submission.daf import \
+ DAFMapper, \
+ MetadataLookupException, \
+ get_submission_uri
+from htsworkflow.submission.condorfastq import CondorFastqExtract
+
+logger = logging.getLogger('ucsc_gather')
def main(cmdline=None):
parser = make_parser()
else:
logging.basicConfig(level = logging.WARNING )
- apidata = {'apiid': opts.apiid, 'apikey': opts.apikey }
+ apidata = api.make_auth_from_opts(opts, parser)
+
+ model = get_model(opts.load_model)
+ mapper = DAFMapper(opts.name, opts.daf, model)
+ submission_uri = get_submission_uri(opts.name)
+
+ if opts.library_url is not None:
+ mapper.library_url = opts.library_url
+
+ if opts.load_rdf is not None:
+ load_into_model(model, 'turtle', opts.load_rdf, submission_uri)
- if opts.host is None or opts.apiid is None or opts.apikey is None:
- parser.error("Please specify host url, apiid, apikey")
+ if opts.make_ddf and opts.daf is None:
+ parser.error("Please specify your daf when making ddf files")
if len(args) == 0:
parser.error("I need at least one library submission-dir input file")
if opts.make_tree_from is not None:
make_tree_from(opts.make_tree_from, library_result_map)
- if opts.daf is not None:
+ if opts.link_daf:
link_daf(opts.daf, library_result_map)
if opts.fastq:
- build_fastqs(opts.host,
- apidata,
- opts.sequence,
- library_result_map,
- force=opts.force)
+ extractor = CondorFastqExtract(opts.host, apidata, opts.sequence,
+ force=opts.force)
+ extractor.build_fastqs(library_result_map)
- if opts.ini:
- make_submission_ini(opts.host, apidata, library_result_map)
+ if opts.scan_submission:
+ scan_submission_dirs(mapper, library_result_map)
- if opts.makeddf:
- make_all_ddfs(library_result_map, opts.daf, force=opts.force)
+ if opts.make_ddf:
+ make_all_ddfs(mapper, library_result_map, opts.daf, force=opts.force)
+ if opts.print_rdf:
+ writer = get_serializer()
+ print writer.serialize_model_to_string(model)
+
def make_parser():
- # Load defaults from the config files
- config = SafeConfigParser()
- config.read([os.path.expanduser('~/.htsworkflow.ini'), '/etc/htsworkflow.ini'])
-
- sequence_archive = None
- apiid = None
- apikey = None
- apihost = None
- SECTION = 'sequence_archive'
- if config.has_section(SECTION):
- sequence_archive = config.get(SECTION, 'sequence_archive',sequence_archive)
- sequence_archive = os.path.expanduser(sequence_archive)
- apiid = config.get(SECTION, 'apiid', apiid)
- apikey = config.get(SECTION, 'apikey', apikey)
- apihost = config.get(SECTION, 'host', apihost)
-
parser = OptionParser()
+ model = OptionGroup(parser, 'model')
+ model.add_option('--name', help="Set submission name")
+ model.add_option('--load-model', default=None,
+ help="Load model database")
+ model.add_option('--load-rdf', default=None,
+ help="load rdf statements into model")
+ model.add_option('--print-rdf', action="store_true", default=False,
+ help="print ending model state")
+ parser.add_option_group(model)
# commands
- parser.add_option('--make-tree-from',
+ commands = OptionGroup(parser, 'commands')
+ commands.add_option('--make-tree-from',
help="create directories & link data files",
default=None)
- parser.add_option('--fastq', help="generate scripts for making fastq files",
- default=False, action="store_true")
-
- parser.add_option('--ini', help="generate submission ini file", default=False,
- action="store_true")
-
- parser.add_option('--makeddf', help='make the ddfs', default=False,
+ commands.add_option('--fastq', default=False, action="store_true",
+ help="generate scripts for making fastq files")
+ commands.add_option('--scan-submission', default=False, action="store_true",
+ help="Import metadata for submission into our model")
+ commands.add_option('--link-daf', default=False, action="store_true",
+ help="link daf into submission directories")
+ commands.add_option('--make-ddf', help='make the ddfs', default=False,
action="store_true")
+ parser.add_option_group(commands)
- parser.add_option('--daf', default=None, help='specify daf name')
parser.add_option('--force', default=False, action="store_true",
help="Force regenerating fastqs")
-
- # configuration options
- parser.add_option('--apiid', default=apiid, help="Specify API ID")
- parser.add_option('--apikey', default=apikey, help="Specify API KEY")
- parser.add_option('--host', default=apihost,
- help="specify HTSWorkflow host",)
- parser.add_option('--sequence', default=sequence_archive,
- help="sequence repository")
-
+ parser.add_option('--daf', default=None, help='specify daf name')
+ parser.add_option('--library-url', default=None,
+ help="specify an alternate source for library information")
# debugging
parser.add_option('--verbose', default=False, action="store_true",
help='verbose logging')
parser.add_option('--debug', default=False, action="store_true",
help='debug logging')
+ api.add_auth_options(parser)
+
return parser
-
def make_tree_from(source_path, library_result_map):
"""Create a tree using data files from source path.
"""
if not os.path.exists(lib_path):
logging.info("Making dir {0}".format(lib_path))
os.mkdir(lib_path)
- source_lib_dir = os.path.join(source_path, lib_path)
+ source_lib_dir = os.path.abspath(os.path.join(source_path, lib_path))
if os.path.exists(source_lib_dir):
pass
for filename in os.listdir(source_lib_dir):
logging.info(
'LINK {0} to {1}'.format(source_pathname, target_pathname))
-def build_fastqs(host, apidata, sequences_path, library_result_map,
- force=False ):
- """
- Generate condor scripts to build any needed fastq files
-
- Args:
- host (str): root of the htsworkflow api server
- apidata (dict): id & key to post to the server
- sequences_path (str): root of the directory tree to scan for files
- library_result_map (list): [(library_id, destination directory), ...]
- """
- qseq_condor_header = """
-Universe=vanilla
-executable=%(exe)s
-error=log/qseq2fastq.err.$(process).log
-output=log/qseq2fastq.out.$(process).log
-log=log/qseq2fastq.log
-
-""" % {'exe': sys.executable }
- qseq_condor_entries = []
- srf_condor_header = """
-Universe=vanilla
-executable=%(exe)s
-output=log/srf_pair_fastq.out.$(process).log
-error=log/srf_pair_fastq.err.$(process).log
-log=log/srf_pair_fastq.log
-environment="PYTHONPATH=/home/diane/lib/python2.6/site-packages:/home/diane/proj/solexa/gaworkflow PATH=/woldlab/rattus/lvol0/mus/home/diane/bin:/usr/bin:/bin"
-
-""" % {'exe': sys.executable }
- srf_condor_entries = []
- lib_db = find_archive_sequence_files(host,
- apidata,
- sequences_path,
- library_result_map)
-
- needed_targets = find_missing_targets(library_result_map, lib_db, force)
-
- for target_pathname, available_sources in needed_targets.items():
- logging.debug(' target : %s' % (target_pathname,))
- logging.debug(' candidate sources: %s' % (available_sources,))
- if available_sources.has_key('qseq'):
- source = available_sources['qseq']
- qseq_condor_entries.append(
- condor_qseq_to_fastq(source.path,
- target_pathname,
- source.flowcell,
- force=force)
- )
- elif available_sources.has_key('srf'):
- source = available_sources['srf']
- mid = getattr(source, 'mid_point', None)
- srf_condor_entries.append(
- condor_srf_to_fastq(source.path,
- target_pathname,
- source.paired,
- source.flowcell,
- mid,
- force=force)
- )
- else:
- print " need file", target_pathname
-
- if len(srf_condor_entries) > 0:
- make_submit_script('srf.fastq.condor',
- srf_condor_header,
- srf_condor_entries)
-
- if len(qseq_condor_entries) > 0:
- make_submit_script('qseq.fastq.condor',
- qseq_condor_header,
- qseq_condor_entries)
-
-
-def find_missing_targets(library_result_map, lib_db, force=False):
- """
- Check if the sequence file exists.
- This requires computing what the sequence name is and checking
- to see if it can be found in the sequence location.
-
- Adds seq.paired flag to sequences listed in lib_db[*]['lanes']
- """
- fastq_paired_template = '%(lib_id)s_%(flowcell)s_c%(cycle)s_l%(lane)s_r%(read)s.fastq'
- fastq_single_template = '%(lib_id)s_%(flowcell)s_c%(cycle)s_l%(lane)s.fastq'
- # find what targets we're missing
- needed_targets = {}
- for lib_id, result_dir in library_result_map:
- lib = lib_db[lib_id]
- lane_dict = make_lane_dict(lib_db, lib_id)
-
- for lane_key, sequences in lib['lanes'].items():
- for seq in sequences:
- seq.paired = lane_dict[seq.flowcell]['paired_end']
- lane_status = lane_dict[seq.flowcell]['status']
-
- if seq.paired and seq.read is None:
- seq.read = 1
- filename_attributes = {
- 'flowcell': seq.flowcell,
- 'lib_id': lib_id,
- 'lane': seq.lane,
- 'read': seq.read,
- 'cycle': seq.cycle
- }
- # skip bad runs
- if lane_status == 'Failed':
- continue
- if seq.flowcell == '30DY0AAXX':
- # 30DY0 only ran for 151 bases instead of 152
- # it is actually 76 1st read, 75 2nd read
- seq.mid_point = 76
-
- # end filters
- if seq.paired:
- target_name = fastq_paired_template % filename_attributes
- else:
- target_name = fastq_single_template % filename_attributes
-
- target_pathname = os.path.join(result_dir, target_name)
- if force or not os.path.exists(target_pathname):
- t = needed_targets.setdefault(target_pathname, {})
- t[seq.filetype] = seq
-
- return needed_targets
-
def link_daf(daf_path, library_result_map):
if not os.path.exists(daf_path):
base_daf = os.path.basename(daf_path)
for lib_id, result_dir in library_result_map:
+ if not os.path.exists(result_dir):
+ raise RuntimeError("Couldn't find target directory %s" %(result_dir,))
submission_daf = os.path.join(result_dir, base_daf)
if not os.path.exists(submission_daf):
+ if not os.path.exists(daf_path):
+ raise RuntimeError("Couldn't find daf: %s" %(daf_path,))
os.link(daf_path, submission_daf)
-def make_submission_ini(host, apidata, library_result_map, paired=True):
- #attributes = get_filename_attribute_map(paired)
- view_map = NameToViewMap(host, apidata)
-
- candidate_fastq_src = {}
-
+def scan_submission_dirs(view_map, library_result_map):
+ """Look through our submission directories and collect needed information
+ """
for lib_id, result_dir in library_result_map:
- order_by = ['order_by=files', 'view', 'replicate', 'cell',
- 'readType', 'mapAlgorithm', 'insertLength' ]
- inifile = ['[config]']
- inifile += [" ".join(order_by)]
- inifile += ['']
- line_counter = 1
- result_ini = os.path.join(result_dir, result_dir+'.ini')
-
- # write other lines
- submission_files = os.listdir(result_dir)
- fastqs = {}
- fastq_attributes = {}
- for f in submission_files:
- attributes = view_map.find_attributes(f, lib_id)
- if attributes is None:
- raise ValueError("Unrecognized file: %s" % (f,))
-
- ext = attributes["extension"]
- if attributes['view'] is None:
- continue
- elif attributes.get("type", None) == 'fastq':
- fastqs.setdefault(ext, set()).add(f)
- fastq_attributes[ext] = attributes
- else:
- inifile.extend(
- make_submission_section(line_counter,
- [f],
- attributes
- )
- )
- inifile += ['']
- line_counter += 1
- # add in fastqs on a single line.
-
- for extension, fastq_files in fastqs.items():
- inifile.extend(
- make_submission_section(line_counter,
- fastq_files,
- fastq_attributes[extension])
- )
- inifile += ['']
- line_counter += 1
-
- f = open(result_ini,'w')
- f.write(os.linesep.join(inifile))
-
+ logging.info("Importing %s from %s" % (lib_id, result_dir))
+ try:
+ view_map.import_submission_dir(result_dir, lib_id)
+ except MetadataLookupException, e:
+ logging.error("Skipping %s: %s" % (lib_id, str(e)))
-def make_lane_dict(lib_db, lib_id):
- """
- Convert the lane_set in a lib_db to a dictionary
- indexed by flowcell ID
- """
- result = []
- for lane in lib_db[lib_id]['lane_set']:
- result.append((lane['flowcell'], lane))
- return dict(result)
-
-
-def make_all_ddfs(library_result_map, daf_name, make_condor=True, force=False):
+def make_all_ddfs(view_map, library_result_map, daf_name, make_condor=True, force=False):
dag_fragment = []
for lib_id, result_dir in library_result_map:
- ininame = result_dir+'.ini'
- inipathname = os.path.join(result_dir, ininame)
- if os.path.exists(inipathname):
- dag_fragment.extend(
- make_ddf(ininame, daf_name, True, make_condor, result_dir)
- )
+ submissionNode = view_map.get_submission_node(result_dir)
+ dag_fragment.extend(
+ make_ddf(view_map, submissionNode, daf_name, make_condor, result_dir)
+ )
if make_condor and len(dag_fragment) > 0:
dag_filename = 'submission.dagman'
f.close()
-def make_ddf(ininame, daf_name, guess_ddf=False, make_condor=False, outdir=None):
+def make_ddf(view_map, submissionNode, daf_name, make_condor=False, outdir=None):
"""
Make ddf files, and bonus condor file
"""
dag_fragments = []
- curdir = os.getcwd()
+
+ name = fromTypedNode(view_map.model.get_target(submissionNode, submissionOntology['name']))
+ if name is None:
+ logging.error("Need name for %s" % (str(submissionNode)))
+ return []
+
+ ddf_name = name + '.ddf'
if outdir is not None:
- os.chdir(outdir)
- output = sys.stdout
- ddf_name = None
- if guess_ddf:
- ddf_name = make_ddf_name(ininame)
- print ddf_name
- output = open(ddf_name,'w')
+ outfile = os.path.join(outdir, ddf_name)
+ output = open(outfile,'w')
+ else:
+ output = sys.stdout
+
+ # filename goes first
+ variables = ['filename']
+ variables.extend(view_map.get_daf_variables())
+ output.write('\t'.join(variables))
+ output.write(os.linesep)
+
+ nameTerm = dafTermOntology['name']
- file_list = read_ddf_ini(ininame, output)
+ submission_views = view_map.model.get_targets(submissionNode, submissionOntology['has_view'])
+ file_list = []
+ for viewNode in submission_views:
+ record = []
+ for variable_name in variables:
+ varNode = dafTermOntology[variable_name]
+ values = list(view_map.model.get_targets(viewNode, varNode))
+
+ if variable_name == 'view':
+ nameNode = view_map.model.get_target(values[0], nameTerm)
+ values = [fromTypedNode(nameNode)]
+ else:
+ values = [ fromTypedNode(v) for v in values ]
+ if variable_name == 'filename':
+ file_list.extend(values)
+
+ if len(values) == 0:
+ attribute = "#None#"
+ elif len(values) == 1:
+ attribute = values[0]
+ else:
+ attribute = ",".join(values)
+ record.append(attribute)
+ output.write('\t'.join(record))
+ output.write(os.linesep)
+
+ logging.info(
+ "Examined {0}, found files: {1}".format(
+ str(submissionNode), ", ".join(file_list)))
file_list.append(daf_name)
- if ddf_name is not None:
- file_list.append(ddf_name)
-
+ file_list.append(ddf_name)
+
if make_condor:
- archive_condor = make_condor_archive_script(ininame, file_list)
- upload_condor = make_condor_upload_script(ininame)
+ print name, file_list
+ archive_condor = make_condor_archive_script(name, file_list, outdir)
+ upload_condor = make_condor_upload_script(name, outdir)
dag_fragments.extend(
- make_dag_fragment(ininame, archive_condor, upload_condor)
+ make_dag_fragment(name, archive_condor, upload_condor)
)
- os.chdir(curdir)
-
return dag_fragments
-def read_ddf_ini(filename, output=sys.stdout):
- """
- Read a ini file and dump out a tab delmited text file
- """
- file_list = []
- config = SafeConfigParser()
- config.read(filename)
-
- order_by = shlex.split(config.get("config", "order_by"))
-
- output.write("\t".join(order_by))
- output.write(os.linesep)
- sections = config.sections()
- sections.sort()
- for section in sections:
- if section == "config":
- # skip the config block
- continue
- values = []
- for key in order_by:
- v = config.get(section, key)
- values.append(v)
- if key == 'files':
- file_list.extend(parse_filelist(v))
-
- output.write("\t".join(values))
- output.write(os.linesep)
- return file_list
-
-
def read_library_result_map(filename):
"""
Read a file that maps library id to result directory.
return results
-def make_condor_archive_script(ininame, files):
+def make_condor_archive_script(name, files, outdir=None):
script = """Universe = vanilla
Executable = /bin/tar
queue
"""
+ if outdir is None:
+ outdir = os.getcwd()
for f in files:
- if not os.path.exists(f):
+ pathname = os.path.join(outdir, f)
+ if not os.path.exists(pathname):
raise RuntimeError("Missing %s" % (f,))
- context = {'archivename': make_submission_name(ininame),
+ context = {'archivename': make_submission_name(name),
'filelist': " ".join(files),
'initialdir': os.getcwd(),
'user': os.getlogin()}
- condor_script = make_condor_name(ininame, 'archive')
+ condor_script = os.path.join(outdir, make_condor_name(name, 'archive'))
condor_stream = open(condor_script,'w')
condor_stream.write(script % context)
condor_stream.close()
return condor_script
-def make_condor_upload_script(ininame):
+def make_condor_upload_script(name, outdir=None):
script = """Universe = vanilla
Executable = /usr/bin/lftp
-arguments = -c put ../%(archivename)s -o ftp://detrout@encodeftp.cse.ucsc.edu/%(archivename)s
+arguments = -c put ../%(archivename)s -o ftp://%(ftpuser)s:%(ftppassword)s@%(ftphost)s/%(archivename)s
Error = upload.err.$(Process).log
Output = upload.out.$(Process).log
queue
"""
- context = {'archivename': make_submission_name(ininame),
- 'initialdir': os.getcwd(),
- 'user': os.getlogin()}
-
- condor_script = make_condor_name(ininame, 'upload')
+ if outdir is None:
+ outdir = os.getcwd()
+
+ auth = netrc.netrc(os.path.expanduser("~diane/.netrc"))
+
+ encodeftp = 'encodeftp.cse.ucsc.edu'
+ ftpuser = auth.hosts[encodeftp][0]
+ ftppassword = auth.hosts[encodeftp][2]
+ context = {'archivename': make_submission_name(name),
+ 'initialdir': outdir,
+ 'user': os.getlogin(),
+ 'ftpuser': ftpuser,
+ 'ftppassword': ftppassword,
+ 'ftphost': encodeftp}
+
+ condor_script = os.path.join(outdir, make_condor_name(name, 'upload'))
condor_stream = open(condor_script,'w')
condor_stream.write(script % context)
condor_stream.close()
+ os.chmod(condor_script, stat.S_IREAD|stat.S_IWRITE)
+
return condor_script
return contents
-def condor_srf_to_fastq(srf_file, target_pathname, paired, flowcell=None,
- mid=None, force=False):
- py = srf2fastq.__file__
- args = [ py, srf_file, ]
- if paired:
- args.extend(['--left', target_pathname])
- # this is ugly. I did it because I was pregenerating the target
- # names before I tried to figure out what sources could generate
- # those targets, and everything up to this point had been
- # one-to-one. So I couldn't figure out how to pair the
- # target names.
- # With this at least the command will run correctly.
- # however if we rename the default targets, this'll break
- # also I think it'll generate it twice.
- args.extend(['--right',
- target_pathname.replace('_r1.fastq', '_r2.fastq')])
- else:
- args.extend(['--single', target_pathname ])
- if flowcell is not None:
- args.extend(['--flowcell', flowcell])
-
- if mid is not None:
- args.extend(['-m', str(mid)])
-
- if force:
- args.extend(['--force'])
-
- script = """
-arguments="%s"
-queue
-""" % (" ".join(args),)
-
- return script
-
-
-def condor_qseq_to_fastq(qseq_file, target_pathname, flowcell=None, force=False):
- py = qseq2fastq.__file__
- args = [py, '-i', qseq_file, '-o', target_pathname ]
- if flowcell is not None:
- args.extend(['-f', flowcell])
- script = """
-arguments="%s"
-queue
-""" % (" ".join(args))
-
- return script
-
-def find_archive_sequence_files(host, apidata, sequences_path,
- library_result_map):
- """
- Find all the archive sequence files possibly associated with our results.
-
- """
- logging.debug("Searching for sequence files in: %s" %(sequences_path,))
-
- lib_db = {}
- seq_dirs = set()
- #seq_dirs = set(os.path.join(sequences_path, 'srfs'))
- candidate_lanes = {}
- for lib_id, result_dir in library_result_map:
- lib_info = get_library_info(host, apidata, lib_id)
- lib_info['lanes'] = {}
- lib_db[lib_id] = lib_info
-
- for lane in lib_info['lane_set']:
- lane_key = (lane['flowcell'], lane['lane_number'])
- candidate_lanes[lane_key] = lib_id
- seq_dirs.add(os.path.join(sequences_path,
- 'flowcells',
- lane['flowcell']))
- logging.debug("Seq_dirs = %s" %(unicode(seq_dirs)))
- candidate_seq_list = scan_for_sequences(seq_dirs)
-
- # at this point we have too many sequences as scan_for_sequences
- # returns all the sequences in a flowcell directory
- # so lets filter out the extras
-
- for seq in candidate_seq_list:
- lane_key = (seq.flowcell, seq.lane)
- lib_id = candidate_lanes.get(lane_key, None)
- if lib_id is not None:
- lib_info = lib_db[lib_id]
- lib_info['lanes'].setdefault(lane_key, set()).add(seq)
-
- return lib_db
-
-
-class NameToViewMap(object):
- """Determine view attributes for a given submission file name
- """
- def __init__(self, root_url, apidata):
- self.root_url = root_url
- self.apidata = apidata
-
- self.lib_cache = {}
- self.lib_paired = {}
- # ma is "map algorithm"
- ma = 'TH1014'
-
- self.patterns = [
- ('*.bai', None),
- ('*.splices.bam', 'Splices'),
- ('*.bam', self._guess_bam_view),
- ('junctions.bed', 'Junctions'),
- ('*.jnct', 'Junctions'),
- ('*.plus.bigwig', 'PlusSignal'),
- ('*.minus.bigwig', 'MinusSignal'),
- ('*.bigwig', 'Signal'),
- ('*.tar.bz2', None),
- ('*.condor', None),
- ('*.daf', None),
- ('*.ddf', None),
- ('*.?ufflinks-0.9.0?genes.expr', 'GeneDeNovo'),
- ('*.?ufflinks-0.9.0?transcripts.expr', 'TranscriptDeNovo'),
- ('*.?ufflinks-0.9.0?transcripts.gtf', 'GeneModel'),
- ('*.GENCODE-v3c?genes.expr', 'GeneGencV3c'),
- ('*.GENCODE-v3c?transcripts.expr', 'TranscriptGencV3c'),
- ('*.GENCODE-v4?genes.expr', 'GeneGencV4'),
- ('*.GENCODE-v4?transcripts.expr', 'TranscriptGencV4'),
- ('*.GENCODE-v4?transcript.expr', 'TranscriptGencV4'),
- ('*_1.75mers.fastq', 'FastqRd1'),
- ('*_2.75mers.fastq', 'FastqRd2'),
- ('*_r1.fastq', 'FastqRd1'),
- ('*_r2.fastq', 'FastqRd2'),
- ('*.fastq', 'Fastq'),
- ('*.gtf', 'GeneModel'),
- ('*.ini', None),
- ('*.log', None),
- ('paired-end-distribution*', 'InsLength'),
- ('*.stats.txt', 'InsLength'),
- ('*.srf', None),
- ('*.wig', None),
- ('*.zip', None),
- ]
-
- self.views = {
- None: {"MapAlgorithm": "NA"},
- "Paired": {"MapAlgorithm": ma},
- "Aligns": {"MapAlgorithm": ma},
- "Single": {"MapAlgorithm": ma},
- "Splices": {"MapAlgorithm": ma},
- "Junctions": {"MapAlgorithm": ma},
- "PlusSignal": {"MapAlgorithm": ma},
- "MinusSignal": {"MapAlgorithm": ma},
- "Signal": {"MapAlgorithm": ma},
- "GeneModel": {"MapAlgorithm": ma},
- "GeneDeNovo": {"MapAlgorithm": ma},
- "TranscriptDeNovo": {"MapAlgorithm": ma},
- "GeneGencV3c": {"MapAlgorithm": ma},
- "TranscriptGencV3c": {"MapAlgorithm": ma},
- "GeneGencV4": {"MapAlgorithm": ma},
- "TranscriptGencV4": {"MapAlgorithm": ma},
- "FastqRd1": {"MapAlgorithm": "NA", "type": "fastq"},
- "FastqRd2": {"MapAlgorithm": "NA", "type": "fastq"},
- "Fastq": {"MapAlgorithm": "NA", "type": "fastq" },
- "GeneModel": {"MapAlgorithm": ma},
- "InsLength": {"MapAlgorithm": ma},
- }
- # view name is one of the attributes
- for v in self.views.keys():
- self.views[v]['view'] = v
-
- def find_attributes(self, pathname, lib_id):
- """Looking for the best extension
- The 'best' is the longest match
-
- :Args:
- filename (str): the filename whose extention we are about to examine
- """
- path, filename = os.path.splitext(pathname)
- if not self.lib_cache.has_key(lib_id):
- self.lib_cache[lib_id] = get_library_info(self.root_url,
- self.apidata, lib_id)
-
- lib_info = self.lib_cache[lib_id]
- if lib_info['cell_line'].lower() == 'unknown':
- logging.warn("Library %s missing cell_line" % (lib_id,))
- attributes = {
- 'cell': lib_info['cell_line'],
- 'replicate': lib_info['replicate'],
- }
- is_paired = self._is_paired(lib_id, lib_info)
-
- if is_paired:
- attributes.update(self.get_paired_attributes(lib_info))
- else:
- attributes.update(self.get_single_attributes(lib_info))
-
- for pattern, view in self.patterns:
- if fnmatch.fnmatch(pathname, pattern):
- if callable(view):
- view = view(is_paired=is_paired)
-
- attributes.update(self.views[view])
- attributes["extension"] = pattern
- return attributes
-
-
- def _guess_bam_view(self, is_paired=True):
- """Guess a view name based on library attributes
- """
- if is_paired:
- return "Paired"
- else:
- return "Aligns"
-
-
- def _is_paired(self, lib_id, lib_info):
- """Determine if a library is paired end"""
- if len(lib_info["lane_set"]) == 0:
- return False
-
- if not self.lib_paired.has_key(lib_id):
- is_paired = 0
- isnot_paired = 0
- failed = 0
- # check to see if all the flowcells are the same.
- # otherwise we might need to do something complicated
- for flowcell in lib_info["lane_set"]:
- # yes there's also a status code, but this comparison
- # is easier to read
- if flowcell["status"].lower() == "failed":
- # ignore failed flowcell
- failed += 1
- pass
- elif flowcell["paired_end"]:
- is_paired += 1
- else:
- isnot_paired += 1
-
- logging.debug("Library %s: %d paired, %d single, %d failed" % \
- (lib_info["library_id"], is_paired, isnot_paired, failed))
-
- if is_paired > isnot_paired:
- self.lib_paired[lib_id] = True
- elif is_paired < isnot_paired:
- self.lib_paired[lib_id] = False
- else:
- raise RuntimeError("Equal number of paired & unpaired lanes."\
- "Can't guess library paired status")
-
- return self.lib_paired[lib_id]
-
- def get_paired_attributes(self, lib_info):
- if lib_info['insert_size'] is None:
- errmsg = "Library %s is missing insert_size, assuming 200"
- logging.warn(errmsg % (lib_info["library_id"],))
- insert_size = 200
- else:
- insert_size = lib_info['insert_size']
- return {'insertLength': insert_size,
- 'readType': '2x75'}
-
- def get_single_attributes(self, lib_info):
- return {'insertLength':'ilNA',
- 'readType': '1x75D'
- }
-
def make_submission_section(line_counter, files, attributes):
"""
Create a section in the submission ini file
return ".".join(elements)
-def make_submit_script(target, header, body_list):
- """
- write out a text file
-
- this was intended for condor submit scripts
-
- Args:
- target (str or stream):
- if target is a string, we will open and close the file
- if target is a stream, the caller is responsible.
-
- header (str);
- header to write at the beginning of the file
- body_list (list of strs):
- a list of blocks to add to the file.
- """
- if type(target) in types.StringTypes:
- f = open(target,"w")
- else:
- f = target
- f.write(header)
- for entry in body_list:
- f.write(entry)
- if type(target) in types.StringTypes:
- f.close()
-
def parse_filelist(file_string):
return file_string.split(",")
if not os.path.exists(f):
raise RuntimeError("%s does not exist" % (f,))
+def make_md5sum(filename):
+ """Quickly find the md5sum of a file
+ """
+ md5_cache = os.path.join(filename+".md5")
+ print md5_cache
+ if os.path.exists(md5_cache):
+ logging.debug("Found md5sum in {0}".format(md5_cache))
+ stream = open(md5_cache,'r')
+ lines = stream.readlines()
+ md5sum = parse_md5sum_line(lines, filename)
+ else:
+ md5sum = make_md5sum_unix(filename, md5_cache)
+ return md5sum
+
+def make_md5sum_unix(filename, md5_cache):
+ cmd = ["md5sum", filename]
+ logging.debug("Running {0}".format(" ".join(cmd)))
+ p = Popen(cmd, stdout=PIPE)
+ stdin, stdout = p.communicate()
+ retcode = p.wait()
+ logging.debug("Finished {0} retcode {1}".format(" ".join(cmd), retcode))
+ if retcode != 0:
+ logging.error("Trouble with md5sum for {0}".format(filename))
+ return None
+ lines = stdin.split(os.linesep)
+ md5sum = parse_md5sum_line(lines, filename)
+ if md5sum is not None:
+ logging.debug("Caching sum in {0}".format(md5_cache))
+ stream = open(md5_cache, "w")
+ stream.write(stdin)
+ stream.close()
+ return md5sum
+
+def parse_md5sum_line(lines, filename):
+ md5sum, md5sum_filename = lines[0].split()
+ if md5sum_filename != filename:
+ errmsg = "MD5sum and I disagre about filename. {0} != {1}"
+ logging.error(errmsg.format(filename, md5sum_filename))
+ return None
+ return md5sum
if __name__ == "__main__":
main()
projects / htsworkflow.git / blobdiff