def getSMRTCellInfoFromFullPath(fullPath):
cellPath = os.path.dirname(os.path.dirname(fullPath))
primaryFolder = os.path.basename(os.path.dirname(fullPath))
context = SecondaryJobService.getCellContext(os.path.basename(fullPath))
limsCode = LIMSMapper.limsCodeFromCellPath(cellPath)
return {'SMRTCellPath' : cellPath,
'PrimaryFolder' : primaryFolder,
'Context' : context,
'LIMSCode' : limsCode}
def _getSMRTCellInfo(self, jobID):
inputs = self._getJobInputs(jobID)
inputData = []
for i in inputs:
inputDict = {'SMRTCellPath' : i.get('collectionPathUri'),
'PrimaryFolder' : i.get('primaryResultsFolder'),
'Context' : 'none',
'LIMSCode' : LIMSMapper.limsCodeFromCellPath(i.get('collectionPathUri'))
}
inputData.append(inputDict)
return inputData
def createSecondaryJobs():
secondaryJobObjects = {}
# Do it by condition
conditions = n.unique(csv["Name"])
newJobDefs = {}
for cond in conditions:
condRows = csv[csv["Name"] == cond]
# If this is a new secondary job, populate the necessary database tables
if csvType == "newJob":
uniqueJobs = n.unique(
zip(
condRows["SecondaryServerName"],
condRows["SecondaryProtocol"],
condRows["SecondaryReference"],
)
)
for job in uniqueJobs:
msg = "Creating SecondaryJob for job info: %s" % str(job)
MU.logMsg(classString, msg, "info")
# First make the job, but don't save it
serverName = job[0]
protocol = job[1]
reference = job[2]
secondaryServer = SecondaryAnalysisServer.objects.get(serverName=serverName)
sjs = SecondaryJobServiceFactory.create(secondaryServer)
# protocolEntry = sjs.getModelProtocolInfo(job[1])
# referenceEntry = sjs.getModelReferenceInfo(job[2])
#
# jobDef = {'protocol' : simplejson.dumps(protocolEntry),
# 'reference' : simplejson.dumps(referenceEntry),
# 'server' : secondaryServer}
jobDef = {"protocol": protocol, "reference": reference, "server": secondaryServer}
# Now add the cells
jobRows = condRows[
(condRows["SecondaryServerName"] == serverName)
& (condRows["SecondaryProtocol"] == protocol)
& (condRows["SecondaryReference"] == reference)
]
jobCells = n.unique(zip(jobRows["SMRTCellPath"], jobRows["PrimaryFolder"]))
smrtCells = []
for c in jobCells:
path, primaryFolder = tuple(c)
msg = "Creating or accessing SMRTCell for data path: %s" % os.path.join(path, primaryFolder)
MU.logMsg(classString, msg, "info")
if os.path.exists(path):
# This is a data path
limsCode = LIMSMapper.limsCodeFromCellPath(path)
cell = SMRTCell.objects.get_or_create(
path=path, primaryFolder=primaryFolder, limsCode=limsCode
)
smrtCells.append(cell[0])
else:
# this is a LIMS Code
dataPath = LIMSMapper.cellPathFromLimsCode(path)
cell = SMRTCell.objects.get_or_create(
path=dataPath, primaryFolder=primaryFolder, limsCode=path
)
smrtCells.append(cell[0])
# Add the SMRT Cells
jobDef["cells"] = smrtCells
hasJob = False
for pk, jd in newJobDefs.iteritems():
if jobDef == jd:
hasJob = True
jobObj = SecondaryJob.objects.get(id=pk)
if not hasJob:
cells = jobDef.pop("cells")
jobObj = SecondaryJob(**jobDef)
jobObj.save()
jobObj.cells.add(*cells)
jobDef["cells"] = cells
newJobDefs[jobObj.id] = jobDef
msg = "Successfully created and saved SecondaryJob: %s" % str(model_to_dict(jobObj))
MU.logMsg(classString, msg, "info")
# Link secondary job to condition
if not secondaryJobObjects.has_key(cond):
secondaryJobObjects[cond] = [jobObj]
else:
secondaryJobObjects[cond].append(jobObj)
else:
# Job already exists
for job, serverName in zip(condRows["SecondaryJobID"], condRows["SecondaryServerName"]):
server = SecondaryAnalysisServer.objects.get(serverName=serverName)
newJob, created = SecondaryJob.objects.get_or_create(jobID=job, server=server)
# Add other job info in here if job was newly created...
if True: # created:
sjs = SecondaryJobServiceFactory.create(server)
jobID = newJob.jobID
jobInfo = sjs.getModelJobInfo(jobID)
# Add protocol and reference info
# protocol = SecondaryJobService.getSingleItem(jobInfo.get('protocol', {'name' : 'unknown'}))
# reference = SecondaryJobService.getSingleItem(jobInfo.get('reference', {'name' : 'unknown'}))
# newJob.protocol = simplejson.dumps(protocol)
# newJob.reference = simplejson.dumps(reference)
newJob.protocol = jobInfo.get("protocol", "unknown")
newJob.reference = jobInfo.get("reference", "unknown")
newJob.save()
# Get the SMRT Cells
smrtCells = jobInfo.get("inputs", [])
smrtCellObjs = []
for c in smrtCells:
cell = SMRTCell.objects.get_or_create(
path=c["SMRTCellPath"], primaryFolder=c["PrimaryFolder"], limsCode=c["LIMSCode"]
)
smrtCellObjs.append(cell[0])
# Now add the SMRT Cells to the new job
[newJob.cells.add(x) for x in smrtCellObjs]
# Link secondary job to condition
if not secondaryJobObjects.has_key(cond):
secondaryJobObjects[cond] = [newJob]
else:
secondaryJobObjects[cond].append(newJob)
return secondaryJobObjects
def validateDefinition(self):
csv = self.csv
allHeaders = self.getValidHeaderValues()
minHeaders = self.getValidHeaderValues(True)
csvHeaders = filter(lambda x: not x.startswith('p_'), csv.dtype.names)
if csv is None:
msg = '[%s] is not a valid file name or stream' % str(self.csv)
return (False, msg)
# Check to see that valid headers were supplied
jobType = self.getDefinitionType()
if not jobType:
msg = 'Invalid headers supplied in input csv: %s' % str(csvHeaders)
return (False, msg)
# Check if csvFN can be parsed
if not csv.shape:
msg = 'Invalid CSV file name or stream: file does not exist or stream can not be opened'
return (False, msg)
# Check for unpopulated default columns
minColnames = minHeaders.get(jobType)
wrngclmns = filter(lambda x: n.dtype(x[1]) == n.dtype(bool) and x[0] in minColnames, csv.dtype.descr)
if wrngclmns:
msg = 'Incorrectly formatted CSV file:\n Column(s) [%s] have not been populated' % ', '.join([c[0] for c in wrngclmns])
return (False, msg)
# Check if the file contains the correct default column names
if filter(lambda x: x not in csvHeaders, minColnames):
msg = 'Incorrectly formatted CSV file:\n Missing default column names from %s' % str(minColnames)
return (False, msg)
# Check to ensure SecondaryServerName maps to valid server
try:
secondaryServers = [SecondaryAnalysisServer.objects.get(serverName=x) for x in csv['SecondaryServerName']]
secondaryServers = dict((x.serverName, x) for x in secondaryServers)
serverNames = n.unique(secondaryServers.keys())
dataHandlerDict = dict([(s, SecondaryJobServiceFactory.create(secondaryServers.get(s))) for s in serverNames])
except ObjectDoesNotExist:
msg = 'Invalid SecondaryServerName. Valid values are: %s' % (', '.join([x.serverName for x in SecondaryAnalysisServer.objects.all()]))
return (False, msg)
# Check for correct naming of conditions
if filter(lambda x: re.findall(r'[^A-Za-z0-9_\.\-]', str(x)), csv['Name']):
msg = 'Incorrectly formatted CSV file:\n Condition names can only contain: alphanumeric characters, dashes (-), underscores (_) and dots (.)'
return (False, msg)
# Check if the non-default columns have a p_ prefix
allColnames = allHeaders.get(jobType)
extras = filter(lambda x: x not in allColnames, csv.dtype.names)
if extras and not filter(lambda x: x.startswith('p_'), extras):
msg = 'Incorrectly formatted CSV file:\n Extra parameters need to be named using a "p_" prefix'
return (False, msg)
# Check new job specific settings
if jobType == 'newJob':
serverPropDict = dict([(s, dataHandlerDict[s].makePropertyDict(('ReferenceNames', 'ProtocolNames'))) for s in serverNames])
# Check if protocols and references provided exists in secondary server's protocol list
for s,p,r in zip(csv['SecondaryServerName'], csv['SecondaryProtocol'], csv['SecondaryReference']):
serverProtocols = serverPropDict[s].get('ProtocolNames', [])
serverReferences = serverPropDict[s].get('ReferenceNames', [])
referenceWhitelist = ['LIMSTemplates', 'LIMSTemplate']
if len(filter(lambda x: x==p, serverProtocols)) != 1:
msg = 'SecondaryProtocol does not map to a single name on server: %s' % p
return (False, msg)
if len(filter(lambda x: x==r, serverReferences)) != 1 and r not in referenceWhitelist:
msg = 'SecondaryReference does not map to a single name on server: %s' % r
return (False, msg)
# Check whether primary folder names are contained within the given run codes and pls.h5/bas.h5 files exist
for s,d,p in zip(csv['SecondaryServerName'], csv['SMRTCellPath'], csv['PrimaryFolder']):
if os.path.exists(d):
# A data path was given, not a LIMSCode
scv = SMRTCellDataPathValidator(d, p)
valid, msg = scv.isValid()
if not valid:
return (False, msg)
# This is for internal PacBio martin server validation
elif valid and re.findall('^martin', s.lower()):
limsPath = LIMSMapper.limsCodeFromCellPath(d)
if not limsPath:
return (False, 'Martin jobs must be given cells with a valid LIMS Code: supplied [%s]' % d)
else:
# Path provided might be a LIMSCode - check with the LIMSHandler to see if it validates
limsPath = LIMSMapper.limsCodeFromCellPath(d)
if not limsPath:
return (False, 'SMRTCellPath [%s] is not a valid path, nor a valid LIMS Code' % d)
else:
scv = SMRTCellDataPathValidator(limsPath, p)
valid, msg = scv.isValid()
if not valid:
return (False, msg)
# Check to make sure that protocol is split/merge-able.
if 'ExtractBy' in csvHeaders:
for s, p, e in zip(csv['SecondaryServerName'], csv['SecondaryProtocol'], csv['ExtractBy']):
sjs = dataHandlerDict.get(s)
if e and not sjs.protocolIsSplittable(p):
msg = 'SecondaryProtocol [%s] is not be extractable by [%s]. Does the protocol generate a cmp.h5?' % (p, e)
return (False, msg)
elif jobType == 'existingJob':
# Check to make sure all job IDs are integers
try:
[int(x) for x in csv['SecondaryJobID']]
except ValueError:
msg = 'Invalid SecondaryJobID provided. All IDs must be integers: %s' % csv['SecondaryJobID']
return (False, msg)
# Check to make sure that jobs actually exist on specified servers
for s,j in zip(csv['SecondaryServerName'], csv['SecondaryJobID']):
sdh = dataHandlerDict.get(s)
try:
sdh.singleJobExists(j)
except SecondaryJobServiceError: # yes, this is bad... i know
msg = 'Single Job ID [%s] does not exist on server [%s]' % (j,s)
return (False, msg)
# Check to make sure that protocol is split/merge-able.
if 'ExtractBy' in csvHeaders:
for s, j, e in zip(csv['SecondaryServerName'], csv['SecondaryJobID'], csv['ExtractBy']):
sjs = dataHandlerDict.get(s)
jobInfo = sjs.getModelJobInfo(j)
#protocol = SecondaryJobService.getSingleItem(jobInfo.get('protocol')).get('name')
protocol = jobInfo.get('protocol')
if protocol and protocol != 'unknown' and e and not sjs.protocolIsSplittable(protocol):
msg = 'SecondaryProtocol [%s] is not be extractable by [%s]. Does the protocol generate a cmp.h5?' % (protocol, e)
return (False, msg)
# COMMON TO NEW AND EXISTING JOBS
# Check for uniqueness of column values within conditions
for cond in n.unique(csv['Name']):
condRows = csv[csv['Name'] == cond]
notUnique = ['SecondaryJobID', 'SecondaryServerName', 'SMRTCellPath', 'PrimaryFolder', 'SecondaryProtocol']
if filter(lambda x: len(n.unique(condRows[x])) != 1, [k for k in condRows.dtype.names if k not in notUnique and not k.startswith('p_')]):
msg = 'For condition name=%s some of the attributes are NOT unique' % cond
return (False, msg)
# Check to make sure that merged existing jobs all have the same reference
if 'SecondaryJobID' in condRows.dtype.names:
references = []
for s, j in zip(condRows['SecondaryServerName'], condRows['SecondaryJobID']):
sjs = dataHandlerDict.get(s)
jobInfo = sjs.getModelJobInfo(j)
#reference = SecondaryJobService.getSingleItem(jobInfo.get('reference'))
#refName = reference.get('name')
refName = jobInfo.get('reference')
if not refName in references:
references.append(refName)
if len(references) > 1:
msg = 'Cannot merge two jobs with different reference sequences! Condition: %s' % condRows['Name']
return (False, msg)
# Check to make sure that split/merge was specified correctly
if 'ExtractBy' in csvHeaders:
for e in csv['ExtractBy']:
bOps = ['>', '<', '==', '!=', '!==', '>=', '<=']
combs = ['&', '|']
opsUsed = filter(lambda x: x in e, bOps)
if not opsUsed:
msg = 'Illegal syntax for ExtractBy field, please use binary operator statement (e.g. Readlength > 1000)'
return (False, msg)
combsUsed = filter(lambda x: x in e, combs)
if combsUsed:
e = [x.strip() for x in re.split(r'[&\|]+', e)]
else:
e = [e]
validExtract = ExperimentDefinitionValidator.getValidExtractByValues()
if not all([any([v in x for v in validExtract]) for x in e]):
msg = 'Invalid ExtractBy [%s]. Please select valid ExtractBy option: %s' % (e, ', '.join(validExtract))
return (False, msg)
return (True, 'CSV file passed validation')
