def getOptionsAndScripts(elem):
jobOptionsEls = Evaluate('options_atn', elem)
prescriptEls = Evaluate('prescript', elem)
jobOptions = [getText(e.childNodes) for e in jobOptionsEls]
prescripts = [getText(e.childNodes) for e in prescriptEls]
return (jobOptions,prescripts)
def handleDQ2Tag(self, tag):
siteTag = tag.getElementsByTagName('dq2site')
dq2site = getText(siteTag[0].childNodes)
dsTag = tag.getElementsByTagName('dq2dataset')
dq2dsName = getText(dsTag[0].childNodes)
typeTag = tag.getElementsByTagName('datatype')
type = getText(typeTag[0].childNodes)
fileTags = tag.getElementsByTagName('dq2file')
dq2files = []
for t in fileTags:dq2files.append(getText(t.childNodes))
return (dq2dsName, dq2files, type)
class TryIt:
def __init__(self, job, dict):
self.job = job
self.dict = dict
self.error = False
self.messages = []
def runFn(self, fn, tag):
try:
fn(self.job.getElementsByTagName(tag), self.dict)
except:
self.error = True
m = 'Error building descriptor in method %s\nTraceback:%s' % (fn.__name__, exc2string2())
self.messages.append(m)
def runXPathFn(self, fn, xpath):
try:
fn(Evaluate(xpath, self.job), self.dict)
except:
self.error = True
m = 'Error building descriptor in method %s\nTraceback:%s' % (fn.__name__, exc2string2())
self.messages.append(m)
def handleSimpleTextTag(self, (tagName, dictKey) ):
nodes = self.job.getElementsByTagName(tagName)
if nodes: self.dict[dictKey] = string.strip(getText(nodes[0].childNodes))
def handleAthenaDataSets_OldStyle(self, datasets, dict):
"""
Old dataset declaration: <dataset>/blah/blah</dataset>
Old style means dc2 data.
"""
if not datasets: return
logger.debug('Processing ' + str(len(datasets)) + ' old style <dataset> tags')
datasetInstanceList = []
datasetType = 'dc2'
datasetRepeat = 1
for dsElementNode in datasets:
logicalDataSetName = string.strip(getText(dsElementNode.childNodes))
physDataSetName = self.getPhysicalDataset(logicalDataSetName)
dsFactory = DataSetFactory(logger)
datasetInstance = dsFactory.create(datasetType,physDataSetName,physDataSetName, 'standard') # physical = logical
datasetInstanceList.append(datasetInstance)
if not dict.has_key('datasets'):
dict['datasets'] = {}
dict['datasets'].setdefault(datasetType,[]).extend(datasetInstanceList)
def handleAthenaGroups(self, groups, dict):
# KLUDGE: Athena added to job group to make it the same as the
# name of job group definition version
logger.debug('groups '+str(groups))
if groups:
groupName = string.strip(getText(groups[0].childNodes))
if not groupName.startswith('Athena'): groupName = 'Athena'+groupName
dict['jobGroup'] = groupName
def handleFurtherAthenaOptions(self, furtherOptions, dict):
# there can be only one joboptions tag.
# these tags allow chaining of athena jobs. Would have been
# better had jobOptions been a list from the begining.
logger.debug("furtherOptions" +str(furtherOptions))
dict['furtherJobOptions'] = []
if furtherOptions:
tags = furtherOptions[0].getElementsByTagName("jobOptionItem")
dict['furtherJobOptions'] = [getText(jot.childNodes)
for jot in tags]
def handleAthenaDataSet(self, node, dict):
datasetType = getElementNodes(node.childNodes)[0].tagName
datasetName = getText(node.getElementsByTagName('datasetName')[0].childNodes)
datasetRepeat = self.getDatasetRepeat(node)
physDataSetName = self.getPhysicalDataset(datasetName)
dsFactory = DataSetFactory(logger)
dataSetInstance = dsFactory.create(datasetType,physDataSetName,physDataSetName,'standard') # physical = logical
[dict.setdefault(datasetType,[]).append(dataSetInstance) for i in range(datasetRepeat)]
logger.info('handleAthenaDataset no of data sets %d' % (len(dict)))
def doit(self, node, tagName):
print node.nodeName, tagName
try:
nodes = node.getElementsByTagName(tagName)
if not nodes: return
node = nodes[0]
self.dict[tagName] = getText(node.childNodes)
except:
m = """Could not retrieve text for tagname %s, ignoring
Traceback:\n %s""" % (tagName, exc2string2())
logger.error(m)
def getDatasetRepeat(self,datasetTag):
datasetRepeat = 1 # default
repeatTags = datasetTag.getElementsByTagName('datasetRepeat')
if not repeatTags: return datasetRepeat
datasetRepeat = getText(repeatTags[0].childNodes)
try:
datasetRepeat = int(datasetRepeat)
except:
msg = 'The tag <datasetRepeat> must take a non-zero positive integer as content. Instead, found: ' + str(datasetRepeat) + '\n'
msg += 'This Athena job will now be rejected.\n'
logger.error(msg)
raise RTTInputError(msg)
return datasetRepeat
def handleAthenaOptions(self, options, dict):
logger.debug('options '+str(options))
dict['jobOptions'] = []
if options:
dict['jobOptions'] = [string.strip(
self.fixAthenaJobOptions(getText(options[0].childNodes)))]
def handleAthenaPreconfigOptions(self, preconfigs, dict):
dict['preconfigJobOptions'] = []
if preconfigs:
dict['preconfigJobOptions'] = [string.strip(
self.fixAthenaJobOptions(getText(preconfigs[0].childNodes)))]
def handleBatchWallTime(self, walltime, dict):
if not walltime: return
dict['suggestedBatchWallTime'] = getText(walltime[0].childNodes)
def handleQueues(self, queues, dict):
logger.debug('queues ...'+str(queues))
if queues:
dict['suggestedQueue'] =string.strip(getText(queues[0].childNodes))
