def __getitem__(self, x):
"""Retrieve the job object from the registry: registry[x].
If 'x' is a job id (int) then a single job object is returned or IndexError.
If 'x' is a name (string) then a unique same name is returned, otherwise [].
If 'x' is a job object then it is returned if it belongs to the registry, otherwise None.
If 'x' is not of any of the types above, raise TypeError.
or by name. If retrieved by name then the job must be unique, otherwise the RegistryKeyError is raised.
If the input is incorrect, RegistryAccessError is raised.
"""
if isinstance(x, int):
try:
return addProxy(self.objects[x])
except IndexError:
raise RegistryIndexError('list index out of range')
if isinstance(x, str):
ids = []
for i in self.objects.keys():
j = self.objects[i]
if j.name == x:
ids.append(j.id)
if len(ids) > 1:
raise RegistryKeyError('object "%s" not unique' % x)
if len(ids) == 0:
raise RegistryKeyError('object "%s" not found' % x)
return addProxy(self.objects[ids[0]])
raise RegistryAccessError('Expected int or string (job name).')
def __getitem__(self, x):
"""Retrieve the job object from the registry: registry[x].
If 'x' is a job id (int) then a single job object is returned or IndexError.
If 'x' is a name (string) then a unique same name is returned, otherwise [].
If 'x' is a job object then it is returned if it belongs to the registry, otherwise None.
If 'x' is not of any of the types above, raise TypeError.
or by name. If retrieved by name then the job must be unique, otherwise the RegistryKeyError is raised.
If the input is incorrect, RegistryAccessError is raised.
"""
if isinstance(x, int):
try:
return addProxy(self.objects[x])
except IndexError:
raise RegistryIndexError('list index out of range')
if isinstance(x, str):
ids = []
for i in self.objects.keys():
j = self.objects[i]
if j.name == x:
ids.append(j.id)
if len(ids) > 1:
raise RegistryKeyError('object "%s" not unique' % x)
if len(ids) == 0:
raise RegistryKeyError('object "%s" not found' % x)
return addProxy(self.objects[ids[0]])
raise RegistryAccessError('Expected int or string (job name).')
def createNewJob(self):
"""Create any jobs required for this unit"""
j = GPI.Job()
j.backend = self._getParent().backend.clone()
# copy form ourselves or the parent transform depending on what's
# specified
fields = [
'application', 'splitter', 'inputfiles', 'inputdata',
'inputsandbox', 'outputfiles', 'postprocessors'
]
for f in fields:
if (f == "postprocessors"
and len(getattr(self, f).process_objects) > 0):
j.postprocessors = copy.deepcopy(addProxy(self).postprocessors)
elif (f != "postprocessors" and getattr(self, f)):
setattr(j, f, copy.deepcopy(getattr(self, f)))
elif (f == "postprocessors"
and len(getattr(self._getParent(), f).process_objects) > 0):
j.postprocessors = copy.deepcopy(
addProxy(self._getParent()).postprocessors)
elif (f != "postprocessors" and getattr(self._getParent(), f)):
setattr(j, f, copy.deepcopy(getattr(self._getParent(), f)))
return j
def select(self, minid=None, maxid=None, **attrs):
from Ganga.GPIDev.Lib.Job.Job import Job
if isType(minid, Job):
if minid.master is not None:
minid = minid.master.id
else:
minid = minid.id
if maxid is None:
maxid = minid
if isType(maxid, Job):
if maxid.master is not None:
maxid = maxid.master.id
else:
maxid = maxid.id
logger = getLogger()
this_repr = repr.Repr()
from Ganga.GPIDev.Base.Proxy import addProxy
attrs_str = ""
## Loop through all possible input combinations to constructa string representation of the attrs from possible inputs
## Required to flatten the additional arguments into a flat string in attrs_str
for a in attrs:
if isclass(attrs[a]):
this_attr = addProxy(attrs[a]())
else:
from Ganga.GPIDev.Base.Objects import GangaObject
if isType(attrs[a], GangaObject):
this_attr = addProxy(attrs[a])
else:
if type(attrs[a]) is str:
from Ganga.GPIDev.Base.Proxy import getRuntimeGPIObject
this_attr = getRuntimeGPIObject(attrs[a], True)
else:
this_attr = attrs[a]
full_str = str(this_attr)
split_str = full_str.split('\n')
for line in split_str:
line = line.strip()
flat_str = ''.join(split_str)
attrs_str += ", %s=\"%s\"" % (a, flat_str)
logger.debug("Attrs_Str: %s" % attrs_str)
logger.debug(
"Constructing slice: %s" % ("%s.select(minid='%s', maxid='%s'%s)" %
(self.name, this_repr.repr(minid),
this_repr.repr(maxid), attrs_str)))
this_slice = self.__class__("%s.select(minid='%s', maxid='%s'%s)" %
(self.name, this_repr.repr(minid),
this_repr.repr(maxid), attrs_str))
def append(id, obj):
this_slice.objects[id] = obj
self.do_select(append, minid, maxid, **attrs)
return this_slice
def setParameter(self, **args):
"""Use: setParameter(processName="HWW") to set the processName in all applications to "HWW"
Warns if applications are not affected because they lack the parameter"""
for name, parm in args.iteritems():
for tf in [t for t in self.transforms if t.application]:
if name in tf.application.getNodeData():
addProxy(tf.application).__setattr__(name, parm)
else:
logger.warning("Transform %i was not affected!", tf.name)
def setParameter(self, **args):
"""Use: setParameter(processName="HWW") to set the processName in all applications to "HWW"
Warns if applications are not affected because they lack the parameter"""
for name, parm in args.iteritems():
for tf in [t for t in self.transforms if t.application]:
if name in tf.application.getNodeData():
addProxy(tf.application).__setattr__(name, parm)
else:
logger.warning("Transform %i was not affected!", tf.name)
def select(self, minid=None, maxid=None, **attrs):
import repr
from Ganga.GPIDev.Lib.Job.Job import Job
if isType(minid, Job):
if minid.master:
minid = minid.master.id
else:
minid = minid.id
if maxid is None:
maxid = minid
if isType(maxid, Job):
if maxid.master:
maxid = maxid.master.id
else:
maxid = maxid.id
logger = getLogger()
this_repr = repr.Repr()
from Ganga.GPIDev.Base.Proxy import addProxy
attrs_str = ""
## Loop through all possible input combinations to constructa string representation of the attrs from possible inputs
## Required to flatten the additional arguments into a flat string in attrs_str
for a in attrs:
from inspect import isclass
if isclass(attrs[a]):
this_attr = addProxy(attrs[a]())
else:
from Ganga.GPIDev.Base.Objects import GangaObject
if isType(attrs[a], GangaObject):
this_attr = addProxy(attrs[a])
else:
if type(attrs[a]) is str:
from Ganga.GPIDev.Base.Proxy import getRuntimeGPIObject
this_attr = getRuntimeGPIObject(attrs[a], True)
else:
this_attr = attrs[a]
full_str = str(this_attr)
split_str = full_str.split('\n')
for line in split_str:
line = line.strip()
flat_str = ''.join(split_str)
attrs_str += ", %s=\"%s\"" % (a, flat_str)
logger.debug("Attrs_Str: %s" % attrs_str)
logger.debug("Constructing slice: %s" % ("%s.select(minid='%s', maxid='%s'%s)" % (self.name, this_repr.repr(minid), this_repr.repr(maxid), attrs_str)))
this_slice = self.__class__("%s.select(minid='%s', maxid='%s'%s)" % (self.name, this_repr.repr(minid), this_repr.repr(maxid), attrs_str))
def append(id, obj):
this_slice.objects[id] = obj
self.do_select(append, minid, maxid, **attrs)
return this_slice
def createNewJob(self):
"""Create any jobs required for this unit"""
import copy
j = makeRegisteredJob()
j.backend = self._getParent().backend.clone()
j.application = self._getParent().application.clone()
if self.inputdata:
j.inputdata = self.inputdata.clone()
j.inputfiles = copy.deepcopy(self._getParent().inputfiles)
trf = self._getParent()
task = trf._getParent()
j.inputsandbox = self._getParent().inputsandbox
j.outputfiles = copy.deepcopy(self._getParent().outputfiles)
if len(self._getParent().postprocessors.process_objects) > 0:
j.postprocessors = copy.deepcopy(
addProxy(self._getParent()).postprocessors)
if trf.splitter:
j.splitter = trf.splitter.clone()
# change the first event for GaussSplitter
from GangaLHCb.Lib.Splitters.GaussSplitter import GaussSplitter
if isType(trf.splitter, GaussSplitter):
events_per_unit = j.splitter.eventsPerJob * \
j.splitter.numberOfJobs
j.splitter.firstEventNumber += self.getID() * events_per_unit
else:
j.splitter = SplitByFiles()
return j
def split(self, job):
subjobs = []
# sort out multiple arg splitting
if (self.attribute != '' or len(self.values) > 0) and len(self.multi_attrs) > 0:
raise ApplicationConfigurationError("Setting both 'attribute'/'values' and 'multi_attrs' is unsupported")
if self.attribute != '':
attrlist = [self.attribute]
values = []
for v in self.values:
values.append([v])
else:
# check we have enough values in the dictionary
numjobs = -1
attrlist = []
for attr in self.multi_attrs:
if numjobs == -1:
numjobs = len(self.multi_attrs[attr])
else:
if len(self.multi_attrs[attr]) != numjobs:
raise ApplicationConfigurationError(
"Number of values for '%s' doesn't equal others '%d'" % (attr, numjobs))
attrlist.append(attr)
# now get everything organised
values = []
for i in range(0, numjobs):
valtmp = []
for attr in attrlist:
valtmp.append(self.multi_attrs[attr][i])
values.append(valtmp)
# check we have enough values to cover the attributes
for vallist in values:
if len(attrlist) != len(vallist):
raise ApplicationConfigurationError(
"Number of attributes to split over doesn't equal number of values in list '%s'" % vallist)
# now perform the split
for vallist in values:
# for each list of values, set the attributes
j = addProxy(self.createSubjob(job))
for i in range(0, len(attrlist)):
attrs = attrlist[i].split('.')
obj = j
for attr in attrs[:-1]:
obj = getattr(obj, attr)
attr = attrs[-1]
setattr(obj, attr, vallist[i])
logger.debug('set %s = %s to subjob.' %
(attrlist[i], getattr(obj, attr)))
subjobs.append(stripProxy(j))
return subjobs
def browseBK(gui=True):
"""Return an LHCbDataset from the GUI LHCb Bookkeeping.
Utility function to launch the new LHCb bookkeeping from inside Ganga.
The function returns an LHCbDataset object.
After browsing and selecting the desired datafiles, click on the
\"Save as ...\" button. The Browser will quit and save the seleted files
as an LHCbDataset object
Usage:
# retrieve an LHCbDataset object with the selected files and store
# them in the variable l
l = browseBK()
# retrieve an LHCbDataset object with the selected files and store
# them in the jobs inputdata field, ready for submission
j.inputdata=browseBK()
"""
import Ganga.Utility.logging
from Ganga.GPIDev.Base.Proxy import addProxy
logger = Ganga.Utility.logging.getLogger()
try:
from GangaLHCb.Lib.DIRAC.Bookkeeping import Bookkeeping
from Ganga.GPI import LHCbDataset
except ImportError:
logger.warning('Could not start Bookkeeping Browser')
return None
bkk = Bookkeeping()
return addProxy(bkk.browse(gui))
def createNewJob(self):
"""Create any jobs required for this unit"""
import copy
j = GPI.Job()
j.backend = self._getParent().backend.clone()
j.application = self._getParent().application.clone()
if self.inputdata:
j.inputdata = self.inputdata.clone()
j.inputfiles = copy.deepcopy(self._getParent().inputfiles)
trf = self._getParent()
task = trf._getParent()
j.inputsandbox = self._getParent().inputsandbox
j.outputfiles = copy.deepcopy(self._getParent().outputfiles)
if len(self._getParent().postprocessors.process_objects) > 0:
j.postprocessors = copy.deepcopy(
addProxy(self._getParent()).postprocessors)
if trf.splitter:
j.splitter = trf.splitter.clone()
# change the first event for GaussSplitter
from GangaLHCb.Lib.Splitters.GaussSplitter import GaussSplitter
if isType(trf.splitter, GaussSplitter):
events_per_unit = j.splitter.eventsPerJob * \
j.splitter.numberOfJobs
j.splitter.firstEventNumber = self.getID() * events_per_unit
else:
j.splitter = SplitByFiles()
return j
def getDataset(self):
'''Gets the dataset from the bookkeeping for current path, etc.'''
if not self.path:
return None
if not self.type in ['Path', 'RunsByDate', 'Run', 'Production']:
raise GangaException('Type="%s" is not valid.' % self.type)
if not self.type is 'RunsByDate':
if self.startDate:
msg = 'startDate not supported for type="%s".' % self.type
raise GangaException(msg)
if self.endDate:
msg = 'endDate not supported for type="%s".' % self.type
raise GangaException(msg)
if self.selection:
msg = 'selection not supported for type="%s".' % self.type
raise GangaException(msg)
cmd = "getDataset('%s','%s','%s','%s','%s','%s')" % (self.path, self.dqflag, self.type, self.startDate, self.endDate, self.selection)
from Ganga.GPIDev.Lib.GangaList.GangaList import GangaList
knownLists = [tuple, list, GangaList]
if isType(self.dqflag, knownLists):
cmd = "getDataset('%s',%s,'%s','%s','%s','%s')" % (self.path, self.dqflag, self.type, self.startDate,
self.endDate, self.selection)
result = get_result(cmd, 'BK query error.', 'BK query error.')
logger.debug("Finished Running Command")
files = []
value = result['Value']
if 'LFNs' in value:
files = value['LFNs']
if not type(files) is list: # i.e. a dict of LFN:Metadata
# if 'LFNs' in files: # i.e. a dict of LFN:Metadata
files = files.keys()
logger.debug("Creating DiracFile objects")
## Doesn't work not clear why
from GangaDirac.Lib.Files.DiracFile import DiracFile
#new_files = []
#def _createDiracLFN(this_file):
# return DiracFile(lfn = this_file)
#GangaObject.__createNewList(new_files, files, _createDiracLFN)
logger.debug("Creating new list")
new_files = [DiracFile(lfn=f) for f in files]
#new_files = [DiracFile(lfn=_file) for _file in files]
#for f in files:
# new_files.append(DiracFile(lfn=f))
#ds.extend([DiracFile(lfn = f)])
logger.info("Constructing LHCbDataset")
from GangaLHCb.Lib.LHCbDataset import LHCbDataset
logger.debug("Imported LHCbDataset")
ds = LHCbDataset(files=new_files, fromRef=True)
logger.debug("Returning Dataset")
return addProxy(ds)
def browseBK(gui=True):
"""Return an LHCbDataset from the GUI LHCb Bookkeeping.
Utility function to launch the new LHCb bookkeeping from inside Ganga.
The function returns an LHCbDataset object.
After browsing and selecting the desired datafiles, click on the
\"Save as ...\" button. The Browser will quit and save the seleted files
as an LHCbDataset object
Usage:
# retrieve an LHCbDataset object with the selected files and store
# them in the variable l
l = browseBK()
# retrieve an LHCbDataset object with the selected files and store
# them in the jobs inputdata field, ready for submission
j.inputdata=browseBK()
"""
import Ganga.Utility.logging
from Ganga.GPIDev.Base.Proxy import addProxy
logger = Ganga.Utility.logging.getLogger()
try:
from GangaLHCb.Lib.Backends.Bookkeeping import Bookkeeping
from Ganga.GPI import LHCbDataset
except ImportError:
logger.warning("Could not start Bookkeeping Browser")
return None
bkk = Bookkeeping()
return addProxy(bkk.browse(gui))
def _addToInterface(interface, name, _object):
if isType(_object, GangaObject):
setattr(interface, name, addProxy(_object))
elif isclass(_object) and issubclass(_object, GangaObject):
setattr(interface, name, getProxyClass(_object))
else:
setattr(interface, name, _object)
def _addToInterface(interface, name, _object):
if isType(_object, GangaObject):
setattr(interface, name, addProxy(_object))
elif isclass(_object) and issubclass(_object, GangaObject):
setattr(interface, name, getProxyClass(_object))
else:
setattr(interface, name, _object)
def _wrap(obj):
if isType(obj, GangaObject):
return addProxy(obj)
if isType(obj, RegistrySlice):
return obj._proxyClass(obj)
if isType(obj, list):
return map(addProxy, obj)
return obj
def setUp(self):
# make a list of lists containing GangaObjects
self.filelist = []
for _ in range(10):
self.filelist.append([self._makeRandomTFile() for _ in range(3)])
# make an empty GangaList
self.gangalist = addProxy(makeGangaList([]))
def testContains(self):
"""Tests __contains__"""
plist = [addProxy(x) for x in self.plain1]
self.assertEqual(plist, self.proxied1)
for p in plist:
self.assertTrue(isProxy(p))
self.assertIn(p, self.proxied1, 'Proxied list should contain each proxied object')
def setUp(self):
# make a list of lists containing GangaObjects
self.filelist = []
for _ in range(10):
self.filelist.append([self._makeRandomTFile() for _ in range(3)])
# make an empty GangaList
self.gangalist = addProxy(makeGangaList([]))
def testContains(self):
"""Tests __contains__"""
plist = [addProxy(x) for x in self.plain1]
assert plist == self.proxied1
for p in plist:
assert isProxy(p)
assert p in self.proxied1, 'Proxied list should contain each proxied object'
def testContains(self):
"""Tests __contains__"""
plist = [addProxy(x) for x in self.plain1]
self.assertEqual(plist, self.proxied1)
for p in plist:
self.assertTrue(isProxy(p))
self.assertIn(p, self.proxied1, 'Proxied list should contain each proxied object')
def testContains(self):
"""Tests __contains__"""
plist = [addProxy(x) for x in self.plain1]
assert plist == self.proxied1
for p in plist:
assert isProxy(p)
assert p in self.proxied1, 'Proxied list should contain each proxied object'
def setUp(self):
super(TestNestedLists, self).setUp()
# make a list of lists containing GangaObjects
self.filelist = []
self.gangalist = None
for _ in range(10):
self.filelist.append([self._makeRandomTFile() for _ in range(3)])
# make an empty GangaList
self.gangalist = addProxy(makeGangaList([]))
def testRemove(self):
t = addProxy(TFile(name='bar'))
self.proxied1.insert(7, t)
list_len = len(self.proxied1)
self.proxied1.remove(t)
assert len(self.proxied1) == list_len - 1
assert t not in self.proxied1
assert t._impl not in self.proxied1._impl
def testRemove(self):
t = addProxy(TFile(name='bar'))
self.proxied1.insert(7, t)
list_len = len(self.proxied1)
self.proxied1.remove(t)
assert len(self.proxied1) == list_len - 1
assert t not in self.proxied1
assert t._impl not in self.proxied1._impl
def split(self, job):
subjobs = []
for arg in self.args:
j = addProxy(self.createSubjob(job))
# Add new arguments to subjob
j.application.args = arg
logger.debug('Arguments for split job is: ' + str(arg))
subjobs.append(stripProxy(j))
return subjobs
def getDiracFiles():
import os
from GangaDirac.Lib.Files.DiracFile import DiracFile
from Ganga.GPIDev.Lib.GangaList.GangaList import GangaList
filename = DiracFile.diracLFNBase().replace('/', '-') + '.lfns'
logger.info('Creating list, this can take a while if you have a large number of SE files, please wait...')
execute('dirac-dms-user-lfns &> /dev/null', shell=True, timeout=None)
g = GangaList()
with open(filename[1:], 'r') as lfnlist:
lfnlist.seek(0)
g.extend((DiracFile(lfn='%s' % lfn.strip()) for lfn in lfnlist.readlines()))
return addProxy(g)
def testAppend(self):
t = addProxy(TFile(name='foo'))
self.plain1.append(t)
assert self.plain1[-1] == t
assert self.plain1.pop() == t
self.proxied1.append(t)
assert self.proxied1[-1] == t
assert self.proxied1[-1] is t, 'Identity Test'
assert isProxy(self.proxied1[-1]), 'Make sure we get back a proxy'
assert self.proxied1.pop() == t
def testAppend(self):
t = addProxy(TFile(name='foo'))
self.plain1.append(t)
assert self.plain1[-1] == t
assert self.plain1.pop() == t
self.proxied1.append(t)
assert self.proxied1[-1] == t
assert self.proxied1[-1] is t, 'Identity Test'
assert isProxy(self.proxied1[-1]), 'Make sure we get back a proxy'
assert self.proxied1.pop() == t
def __call__(self, this_id):
""" Retrieve an object by id.
"""
if isinstance(this_id, str):
if this_id.isdigit():
this_id = int(this_id)
else:
matches = [o for o in self.objects if fnmatch.fnmatch(o._getRegistry()._getName(o), this_id)]
if len(matches) > 1:
logger.error('Multiple Matches: Wildcards are allowed for ease of matching, however')
logger.error(' to keep a uniform response only one item may be matched.')
logger.error(' If you wanted a slice, please use the select method')
raise RegistryKeyError("Multiple matches for id='%s':%s" % (this_id, str(map(lambda x: x._getRegistry()._getName(x), matches))))
if len(matches) < 1:
return
return addProxy(matches[0])
try:
return addProxy(self.objects[this_id])
except KeyError as err:
logger.debug('Object id=%d not found' % this_id)
logger.debug("%s" % err)
raise RegistryKeyError('Object id=%d not found' % this_id)
def split(self, job):
from Ganga.GPIDev.Lib.Job import Job
subjobs = []
for run in self.Files.keys():
j = addProxy(self.createSubjob(job))
# j.splitter = None
# j.merger = None
jp = stripProxy(j)
jp._splitter_data = self.optionsString(run)
subjobs.append(jp)
print "Submitting jobs for %d runs" % (len(subjobs))
return subjobs
def split(self, job):
subjobs = []
for arg in self.args:
j = addProxy(self.createSubjob(job,['application']))
# Add new arguments to subjob
app = copy.deepcopy(job.application)
app.args = arg
j.application = app
logger.debug('Arguments for split job is: ' + str(arg))
subjobs.append(stripProxy(j))
return subjobs
def getDiracFiles():
import os
from GangaDirac.Lib.Files.DiracFile import DiracFile
from Ganga.GPIDev.Lib.GangaList.GangaList import GangaList
filename = DiracFile.diracLFNBase().replace('/', '-') + '.lfns'
logger.info(
'Creating list, this can take a while if you have a large number of SE files, please wait...'
)
execute('dirac-dms-user-lfns &> /dev/null', shell=True, timeout=None)
g = GangaList()
with open(filename[1:], 'r') as lfnlist:
lfnlist.seek(0)
g.extend(
(DiracFile(lfn='%s' % lfn.strip()) for lfn in lfnlist.readlines()))
return addProxy(g)
def __call__(self, this_id):
""" Retrieve an object by id.
"""
if isinstance(this_id, str):
if this_id.isdigit():
this_id = int(this_id)
else:
matches = [
o for o in self.objects
if fnmatch.fnmatch(o._getRegistry()._getName(o), this_id)
]
if len(matches) > 1:
logger.error(
'Multiple Matches: Wildcards are allowed for ease of matching, however'
)
logger.error(
' to keep a uniform response only one item may be matched.'
)
logger.error(
' If you wanted a slice, please use the select method'
)
raise RegistryKeyError(
"Multiple matches for id='%s':%s" %
(this_id,
str(
map(lambda x: x._getRegistry()._getName(x),
matches))))
if len(matches) < 1:
return
return addProxy(matches[0])
try:
return addProxy(self.objects[this_id])
except KeyError as err:
logger.debug('Object id=%d not found' % this_id)
logger.deubg("%s" % str(err))
raise RegistryKeyError('Object id=%d not found' % this_id)
def createNewJob(self):
"""Create any jobs required for this unit"""
j = GPI.Job()
j.backend = self._getParent().backend.clone()
# copy form ourselves or the parent transform depending on what's
# specified
fields = ['application', 'splitter', 'inputfiles',
'inputdata', 'inputsandbox', 'outputfiles', 'postprocessors']
for f in fields:
if (f == "postprocessors" and len(getattr(self, f).process_objects) > 0):
j.postprocessors = copy.deepcopy(addProxy(self).postprocessors)
elif (f != "postprocessors" and getattr(self, f)):
setattr(j, f, copy.deepcopy(getattr(self, f)))
elif (f == "postprocessors" and len(getattr(self._getParent(), f).process_objects) > 0):
j.postprocessors = copy.deepcopy(
addProxy(self._getParent()).postprocessors)
elif (f != "postprocessors" and getattr(self._getParent(), f)):
setattr(j, f, copy.deepcopy(getattr(self._getParent(), f)))
return j
def split(self, job):
import os
subjobs = []
subsets = splitCSVFile(job.application.csvfile, self.nbevents)
# Less files than number of jobs wanted => easy
logger.info('Creating %d subjobs ...', len(allLines))
# Base for the naming of each subjob's CSV file
tmpname = os.path.basename(incsvfile)
if len(tmpname.split('.')) > 1:
patterncsv = '.'.join(
tmpname.split('.')[0:-1]) + "_sub%d." + tmpname.split('.')[-1]
else:
patterncsv = tmpname + "_sub%d"
# Base for the naming of each subjob's output file
tmpname = os.path.basename(job.application.outputfile)
if len(tmpname.split('.')) > 1:
patternout = '.'.join(
tmpname.split('.')[0:-1]) + "_sub%d." + tmpname.split('.')[-1]
else:
patternout = tmpname + "_sub%d"
for s, sub in enumerate(subsets):
j = addProxy(self.createSubjob(job))
j.inputdata = job.inputdata
subLines = '\n'.join(sub)
from Ganga.GPIDev.Lib.File import FileBuffer
thiscsv = patterncsv % s
# Save in the main job's inputdir now, then the file will be moved to
# the inputdir of each subjobs.
job.getInputWorkspace().writefile(FileBuffer(thiscsv, subLines),
executable=0)
j.application.csvfile = os.path.join(job.inputdir, thiscsv)
j.application.outputfile = patternout % s
# Prepare the output filenames which must be unique
subjobs.append(stripProxy(j))
return subjobs
def split(self, job):
subjobs = []
filenames = job.inputdata.get_dataset_filenames()
subsets = []
# Less files than number of jobs wanted => easy
logger.info('Creating %d subjobs ...', len(filenames))
for nb in range(len(filenames)):
j = addProxy(self.createSubjob(job))
j.inputdata.set_dataset_filenames([filenames[nb]])
subjobs.append(stripProxy(j))
return subjobs
def loadObject(filename):
'''
These are complimentary functions to export/load which are already exported to
the GPI from Ganga.GPIDev.Persistency. The difference being that these functions will
export the objects using the pickle persistency format rather than a Ganga streaming
(human readable) format.
'''
import os
import pickle
import traceback
try:
with open(os.path.expandvars(os.path.expanduser(filename)), 'rb') as f:
r = pickle.load(f)
except:
logger.error("Problem when loading file '%s': %s" % (filename, traceback.format_exc()))
else:
return addProxy(r)
def split(self,job):
subjobs = []
filenames = job.inputdata.get_dataset_filenames()
subsets = []
# Less files than number of jobs wanted => easy
logger.info('Creating %d subjobs ...',len(filenames))
for nb in range(len(filenames)):
j = addProxy(self.createSubjob(job))
j.inputdata.set_dataset_filenames([filenames[nb]])
subjobs.append(stripProxy(j))
return subjobs
def split(self, job):
subjobs = []
filenames = job.inputdata.get_dataset_filenames()
logger.info('Creating %d subjobs ...', self.nbjobs)
if self.nbjobs < 1:
raise Exception('Number of nbjobs not set properly.')
subsets = []
# Less files than number of jobs wanted => easy
if len(filenames) < self.nbjobs:
for f in filenames:
subsets.append([f])
else:
isPerfectSplit = (len(filenames) % self.nbjobs) == 0
if isPerfectSplit:
# If the number of input files is divisible by nbjobs
# then all subjobs have the same number of input files
nbfulljobs = self.nbjobs
else:
# Otherwise all subjobs have the same number of input files
# except the last subjob which has less
nbfulljobs = self.nbjobs - 1
persub = len(filenames) / nbfulljobs
for nb in range(nbfulljobs):
Low = nb * persub
High = (nb + 1) * persub
subsets.append(filenames[Low:High])
if not isPerfectSplit:
subsets.append(filenames[High:])
for sub in subsets:
j = addProxy(self.createSubjob(job))
j.inputdata.set_dataset_filenames(sub)
subjobs.append(stripProxy(j))
return subjobs
def split(self,job):
subjobs = []
filenames = job.inputdata.get_dataset_filenames()
logger.info('Creating %d subjobs ...',self.nbjobs)
if self.nbjobs < 1:
raise Exception('Number of nbjobs not set properly.')
subsets = []
# Less files than number of jobs wanted => easy
if len(filenames) < self.nbjobs:
for f in filenames:
subsets.append([f])
else:
isPerfectSplit = (len(filenames) % self.nbjobs) == 0
if isPerfectSplit:
# If the number of input files is divisible by nbjobs
# then all subjobs have the same number of input files
nbfulljobs = self.nbjobs
else:
# Otherwise all subjobs have the same number of input files
# except the last subjob which has less
nbfulljobs = self.nbjobs - 1
persub = len(filenames) / nbfulljobs
for nb in range(nbfulljobs):
Low = nb*persub
High = (nb+1)*persub
subsets.append(filenames[Low:High])
if not isPerfectSplit:
subsets.append(filenames[High:])
for sub in subsets:
j = addProxy(self.createSubjob(job))
j.inputdata.set_dataset_filenames(sub)
subjobs.append(stripProxy(j))
return subjobs
def split(self,job):
import os
subjobs = []
subsets = splitCSVFile(job.application.csvfile, self.nbevents)
# Less files than number of jobs wanted => easy
logger.info('Creating %d subjobs ...',len(allLines))
# Base for the naming of each subjob's CSV file
tmpname = os.path.basename(incsvfile)
if len(tmpname.split('.')) > 1:
patterncsv = '.'.join(tmpname.split('.')[0:-1])+"_sub%d."+ tmpname.split('.')[-1]
else:
patterncsv = tmpname+"_sub%d"
# Base for the naming of each subjob's output file
tmpname = os.path.basename(job.application.outputfile)
if len(tmpname.split('.')) > 1:
patternout = '.'.join(tmpname.split('.')[0:-1])+"_sub%d."+ tmpname.split('.')[-1]
else:
patternout = tmpname+"_sub%d"
for s,sub in enumerate(subsets):
j = addProxy(self.createSubjob(job))
j.inputdata = job.inputdata
subLines = '\n'.join(sub)
from Ganga.GPIDev.Lib.File import FileBuffer
thiscsv = patterncsv % s
# Save in the main job's inputdir now, then the file will be moved to
# the inputdir of each subjobs.
job.getInputWorkspace().writefile(FileBuffer(thiscsv,subLines),executable=0)
j.application.csvfile = os.path.join(job.inputdir,thiscsv)
j.application.outputfile = patternout % s
# Prepare the output filenames which must be unique
subjobs.append(stripProxy(j))
return subjobs
def loadObject(filename):
'''
These are complimentary functions to export/load which are already exported to
the GPI from Ganga.GPIDev.Persistency. The difference being that these functions will
export the objects using the pickle persistency format rather than a Ganga streaming
(human readable) format.
'''
import os
import pickle
import traceback
try:
with open(os.path.expandvars(os.path.expanduser(filename)), 'rb') as f:
r = pickle.load(f)
except:
logger.error("Problem when loading file '%s': %s" %
(filename, traceback.format_exc()))
else:
return addProxy(r)
def exportToGPI(name, _object, doc_section, docstring=None):
'''
Make object available publicly as "name" in Ganga.GPI module. Add automatic documentation to gangadoc system.
"doc_section" specifies how the object should be documented.
If docstring is specified then use it to document the object (only use for "Objects" section). Otherwise use __doc__ (via pydoc utilities).
FIXME: if you try to export the object instance, you should import it with fully qualified path, e.g.
import X.Y.Z
X.Y.Z.object = object
exportToGPI("obj",X.Y.Z.object,"Objects")
It has been observed that doing exportToGPI("obj",object,"Objects") may not work. To be understood.
'''
if isType(_object, GangaObject):
setattr(Ganga.GPI, name, addProxy(_object))
else:
setattr(Ganga.GPI, name, _object)
adddoc(name, getattr(Ganga.GPI, name), doc_section, docstring)
def exportToGPI(name, _object, doc_section, docstring=None):
'''
Make object available publicly as "name" in Ganga.GPI module. Add automatic documentation to gangadoc system.
"doc_section" specifies how the object should be documented.
If docstring is specified then use it to document the object (only use for "Objects" section). Otherwise use __doc__ (via pydoc utilities).
FIXME: if you try to export the object instance, you should import it with fully qualified path, e.g.
import X.Y.Z
X.Y.Z.object = object
exportToGPI("obj",X.Y.Z.object,"Objects")
It has been observed that doing exportToGPI("obj",object,"Objects") may not work. To be understood.
'''
if isType(_object, GangaObject):
setattr(Ganga.GPI, name, addProxy(_object))
else:
setattr(Ganga.GPI, name, _object)
adddoc(name, getattr(Ganga.GPI, name), doc_section, docstring)
def setUp(self):
super(TestGangaList, self).setUp()
self.plain1 = [self._makeRandomTFile() for _ in range(15)]
self.plain2 = [self._makeRandomTFile() for _ in range(10)]
self.proxied1 = GangaList()
self.proxied1.extend(self.plain1[:])
self.proxied2 = GangaList()
self.proxied2.extend(self.plain2[:])
t = TFile()
real_t = stripProxy(t)
new_proxy_t = addProxy(real_t)
#hopefully_t = stripProxy(new_proxy_t)
#assert real_t is hopefully_t
assert t is new_proxy_t
self.assertEqual(len(getProxyAttr(self.proxied1, '_list')), len(self.plain1), "Something's wrong with construction")
self.assertEqual(len(getProxyAttr(self.proxied2, '_list')), len(self.plain2), "Something's wrong with construction")
def setUp(self):
super(TestGangaList, self).setUp()
self.plain1 = [self._makeRandomTFile() for _ in range(15)]
self.plain2 = [self._makeRandomTFile() for _ in range(10)]
self.proxied1 = GangaList()
self.proxied1.extend(self.plain1[:])
self.proxied2 = GangaList()
self.proxied2.extend(self.plain2[:])
t = TFile()
real_t = stripProxy(t)
new_proxy_t = addProxy(real_t)
#hopefully_t = stripProxy(new_proxy_t)
#assert real_t is hopefully_t
assert t is new_proxy_t
self.assertEqual(len(getProxyAttr(self.proxied1, '_list')), len(self.plain1), "Something's wrong with construction")
self.assertEqual(len(getProxyAttr(self.proxied2, '_list')), len(self.plain2), "Something's wrong with construction")
def getDataset(self):
'''Gets the dataset from the bookkeeping for current dict.'''
if not self.dict:
return None
cmd = 'bkQueryDict(%s)' % self.dict
result = get_result(cmd, 'BK query error.', 'BK query error.')
files = []
value = result['Value']
if 'LFNs' in value:
files = value['LFNs']
if not type(files) is list:
if 'LFNs' in files: # i.e. a dict of LFN:Metadata
files = files['LFNs'].keys()
from GangaDirac.Lib.Files.DiracFile import DiracFile
this_list = [DiracFile(lfn=f) for f in files]
from GangaLHCb.Lib.LHCbDataset import LHCbDataset
ds = LHCbDataset(files=this_list, fromRef=True)
return addProxy(ds)
def getDataset(self):
'''Gets the dataset from the bookkeeping for current dict.'''
if not self.dict:
return None
cmd = 'bkQueryDict(%s)' % self.dict
result = get_result(cmd, 'BK query error.', 'BK query error.')
files = []
value = result['Value']
if 'LFNs' in value:
files = value['LFNs']
if not type(files) is list:
if 'LFNs' in files: # i.e. a dict of LFN:Metadata
files = files['LFNs'].keys()
from GangaDirac.Lib.Files.DiracFile import DiracFile
this_list = [DiracFile(lfn=f) for f in files]
from GangaLHCb.Lib.LHCbDataset import LHCbDataset
ds = LHCbDataset(files=this_list, fromRef=True)
return addProxy(ds)
def split(self, job):
subjobs = []
filenames = job.inputdata.get_dataset_filenames()
if self.nbfiles < 1:
raise Exception('Number of nbfiles not set properly.')
subsets = splitNbInputFile(filenames, self.nbfiles)
logger.info('Creating %d subjobs ...', len(subjobs))
for sub in subsets:
j = addProxy(self.createSubjob(job))
j.inputdata.set_dataset_filenames(sub)
subjobs.append(stripProxy(j))
return subjobs
def split(self,job):
subjobs = []
filenames = job.inputdata.get_dataset_filenames()
if self.nbfiles < 1:
raise Exception('Number of nbfiles not set properly.')
subsets = splitNbInputFile(filenames, self.nbfiles)
logger.info('Creating %d subjobs ...',len(subjobs))
for sub in subsets:
j = addProxy(self.createSubjob(job))
j.inputdata.set_dataset_filenames(sub)
subjobs.append(stripProxy(j))
return subjobs
def getDataset(self):
"""Gets the dataset from the bookkeeping for current dict."""
if not self.dict:
return None
cmd = "bkQueryDict(%s)" % self.dict
result = get_result(cmd, "BK query error.", "BK query error.")
files = []
value = result["Value"]
if "LFNs" in value:
files = value["LFNs"]
if not type(files) is list:
if "LFNs" in files: # i.e. a dict of LFN:Metadata
files = files["LFNs"].keys()
from GangaDirac.Lib.Files.DiracFile import DiracFile
this_list = [DiracFile(lfn=_file) for _file in files]
from GangaLHCb.Lib.LHCbDataset import LHCbDataset
ds = LHCbDataset(this_list)
return addProxy(ds)
def next(self):
return addProxy(next(self.it))
def _export_pop(self, index=-1):
self.checkReadOnly()
return addProxy(self.pop(index))
class TestSharedFileGetPut(TestMassStorageGetPut):
"""Testing the get/put/copyTo methods of SharedFile"""
fileClass = addProxy(SharedFile)
class TestMassStorageGetPut(GangaUnitTest):
"""Testing the get/put/copyTo methods of MassStorage"""
_temp_files = []
_managed_files = []
# Num of sj in tests
sj_len = 3
fileClass = addProxy(MassStorageFile)
# Where on local storage we want to have our 'MassStorage solution'
outputFilePath = '/tmp/Test' + _getName(fileClass) + 'GetPut'
# This sets up a MassStorageConfiguration which works by placing a file on local storage somewhere we can test using standard tools
MassStorageTestConfig = {
'defaultProtocol': 'file://',
'fileExtensions': [''],
'uploadOptions': {
'path': outputFilePath,
'cp_cmd': 'cp',
'ls_cmd': 'ls',
'mkdir_cmd': 'mkdir -p'
},
'backendPostprocess': {
'LSF': 'client',
'LCG': 'client',
'ARC': 'client',
'Dirac': 'client',
'PBS': 'client',
'Interactive': 'client',
'Local': 'client',
'CREAM': 'client'
}
}
def setUp(self):
"""
Configure the MassStorageFile for the test
"""
extra_opts = [('PollThread', 'autostart', 'False'),
('Local', 'remove_workdir', 'False'),
('TestingFramework', 'AutoCleanup', 'False'),
('Output', _getName(self.fileClass),
self.MassStorageTestConfig),
('Output', 'FailJobIfNoOutputMatched', 'True')]
super(TestMassStorageGetPut, self).setUp(extra_opts=extra_opts)
@staticmethod
def cleanUp():
""" Cleanup the current temp jobs """
from Ganga.GPI import jobs
for j in jobs:
shutil.rmtree(j.backend.workdir, ignore_errors=True)
j.remove()
@classmethod
def setUpClass(cls):
""" This creates a safe place to put the files into 'mass-storage' """
cls.outputFilePath = tempfile.mkdtemp()
cls.MassStorageTestConfig['uploadOptions']['path'] = cls.outputFilePath
@classmethod
def tearDownClass(cls):
""" Cleanup the current temp objects """
for file_ in cls._temp_files:
os.unlink(file_)
cls._temp_files = []
for file_ in cls._managed_files:
os.unlink(os.path.join(cls.outputFilePath, file_.namePattern))
cls._managed_files = []
shutil.rmtree(cls.outputFilePath, ignore_errors=True)
def test_a_test_put(self):
"""Test that a job can be submitted with inputfiles in the input"""
MassStorageFile = self.fileClass
_ext = '.root'
file_1 = generate_unique_temp_file(_ext)
file_2 = generate_unique_temp_file(_ext)
self._temp_files.append(file_1)
self._temp_files.append(file_2)
msf_1 = MassStorageFile(file_1)
msf_2 = MassStorageFile(file_2)
self._managed_files.append(msf_1)
self._managed_files.append(msf_2)
msf_1.put()
msf_2.put()
for file_ in [msf for msf in (msf_1, msf_2)]:
assert os.path.isfile(
os.path.join(self.outputFilePath, file_.namePattern))
file_.localDir = ''
assert file_.localDir == ''
def test_b_test_get(self):
"""Test that the files were made accessible to the WN area and collected as LocalFile objects in outputfiles"""
from Ganga.GPI import Job
tmpdir = tempfile.mkdtemp()
# Test in the case that the files don't have a parent or a localDir
for file_ in self._managed_files:
file_.localDir = ''
try:
assert file_.localDir == ''
file_.get()
print("Unexpected localDir: %s" % file_.localDir)
failed = False
except GangaException:
failed = True
assert failed
# Test in the case that the localDir has been set
for file_ in self._managed_files:
file_.localDir = tmpdir
print("localDir: %s" % file_.localDir)
file_.get()
assert os.path.isfile(os.path.join(tmpdir, file_.namePattern))
file_.localDir = ''
assert file_.localDir == ''
# Test in the case that the object is 'owned' by a Job
j = Job()
outputdir = stripProxy(j).getOutputWorkspace(create=True).getPath()
j.outputfiles = self._managed_files
for file_ in j.outputfiles:
assert stripProxy(file_).getJobObject() is stripProxy(j)
assert file_.localDir == ''
file_.get()
assert os.path.isfile(os.path.join(outputdir, file_.namePattern))
shutil.rmtree(tmpdir, ignore_errors=True)
self.cleanUp()
def test_c_test_copyTo(self):
""" Test the new copyTo interface"""
tmpdir = tempfile.mkdtemp()
for file_ in self._managed_files:
file_.localDir = ''
stripProxy(file_).copyTo(tmpdir)
assert os.path.isfile(os.path.join(tmpdir, file_.namePattern))
shutil.rmtree(tmpdir, ignore_errors=True)
def testDelItem(self):
"""Test __delitem__"""
for p in [addProxy(x) for x in self.plain1[:]]:
assert isProxy(p)
del self.proxied1[self.proxied1.index(p)]
from __future__ import absolute_import
try:
import unittest2 as unittest
except ImportError:
import unittest
import random
import string
from Ganga.GPIDev.Base.Proxy import addProxy, getProxyAttr, isProxy, isType, stripProxy
from Ganga.GPIDev.Lib.GangaList.GangaList import GangaList
GangaList = addProxy(GangaList)
from .TFile import TFile
TFile = addProxy(TFile)
# set the seed for repeatable tests
random.seed(666)
from Ganga.Utility.logging import getLogger
logger = getLogger(modulename=True)
class TestGangaList(unittest.TestCase):
def __init__(self, *args, **kwargs):
super(TestGangaList, self).__init__(*args, **kwargs)
self.plain1 = []
def testInsert(self):
t = addProxy(TFile(name='foo'))
self.proxied1.insert(8, t)
assert self.proxied1[8] == t
