def __init__(self, config, datasetExpr, datasetNick, datasetID = 0):
DataProvider.__init__(self, config, datasetExpr, datasetNick, datasetID)
(self._path, self._events, selist) = utils.optSplit(datasetExpr, '|@')
self._selist = utils.parseList(selist, delimeter = ',', onEmpty = None)
if not (self._path and self._events):
raise ConfigError('Invalid dataset expression!\nCorrect: /local/path/to/file|events[@SE1,SE2]')
def __init__(self, config, datasetExpr, datasetNick=None):
ds_config = config.changeView(viewClass='TaggedConfigView',
addNames=[md5_hex(datasetExpr)])
if os.path.isdir(datasetExpr):
scan_pipeline = ['OutputDirsFromWork']
ds_config.set('source directory', datasetExpr)
datasetExpr = os.path.join(datasetExpr, 'work.conf')
else:
scan_pipeline = ['OutputDirsFromConfig', 'MetadataFromTask']
datasetExpr, selector = utils.optSplit(datasetExpr, '%')
ds_config.set('source config', datasetExpr)
ds_config.set('source job selector', selector)
ext_config = create_config(datasetExpr)
ext_task_name = ext_config.changeView(setSections=['global']).get(
['module', 'task'])
if 'ParaMod' in ext_task_name: # handle old config files
ext_task_name = ext_config.changeView(
setSections=['ParaMod']).get('module')
ext_task_cls = Plugin.getClass(ext_task_name)
for ext_task_cls in Plugin.getClass(ext_task_name).iterClassBases():
try:
scan_holder = GCProviderSetup.getClass('GCProviderSetup_' +
ext_task_cls.__name__)
except PluginError:
continue
scan_pipeline += scan_holder.scan_pipeline
break
ScanProviderBase.__init__(self, ds_config, datasetExpr, datasetNick,
scan_pipeline)
def __init__(self, config, datasetExpr, datasetNick = None):
DataProvider.__init__(self, config, datasetExpr, datasetNick)
(self._path, self._events, selist) = utils.optSplit(datasetExpr, '|@')
self._selist = parseList(selist, ',') or None
if not (self._path and self._events):
raise ConfigError('Invalid dataset expression!\nCorrect: /local/path/to/file|events[@SE1,SE2]')
def __init__(self, config, datasetExpr, datasetNick = None):
self._changeTrigger = triggerResync(['datasets', 'parameters'])
self._lumi_filter = config.getLookup('lumi filter', {}, parser = parseLumiFilter, strfun = strLumi, onChange = self._changeTrigger)
if not self._lumi_filter.empty():
config.set('dataset processor', 'LumiDataProcessor', '+=')
DataProvider.__init__(self, config, datasetExpr, datasetNick)
# LumiDataProcessor instantiated in DataProcessor.__ini__ will set lumi metadata as well
self._lumi_query = config.getBool('lumi metadata', not self._lumi_filter.empty(), onChange = self._changeTrigger)
config.set('phedex sites matcher mode', 'shell', '?=')
# PhEDex blacklist: 'T1_*_Disk nodes allow user jobs - other T1's dont!
self._phedexFilter = config.getFilter('phedex sites', '-* T1_*_Disk T2_* T3_*',
defaultMatcher = 'blackwhite', defaultFilter = 'strict', onChange = self._changeTrigger)
self._onlyComplete = config.getBool('only complete sites', True, onChange = self._changeTrigger)
self._locationFormat = config.getEnum('location format', CMSLocationFormat, CMSLocationFormat.hostname, onChange = self._changeTrigger)
self._pjrc = JSONRestClient(url = 'https://cmsweb.cern.ch/phedex/datasvc/json/prod/blockreplicas')
self._sitedb = SiteDB()
(self._datasetPath, self._datasetInstance, self._datasetBlock) = optSplit(datasetExpr, '@#')
instance_default = config.get('dbs instance', '', onChange = self._changeTrigger)
self._datasetInstance = self._datasetInstance or instance_default
if not self._datasetInstance:
self._datasetInstance = 'prod/global'
elif '/' not in self._datasetInstance:
self._datasetInstance = 'prod/%s' % self._datasetInstance
self._datasetBlock = self._datasetBlock or 'all'
self.onlyValid = config.getBool('only valid', True, onChange = self._changeTrigger)
def __init__(self, config, datasetExpr, datasetNick=None, datasetID=0):
self._lumi_filter = parseLumiFilter(config.get('lumi filter', ''))
if self._lumi_filter:
config.set('dataset processor', 'LumiDataProcessor', '+=')
DataProvider.__init__(self, config, datasetExpr, datasetNick,
datasetID)
# PhEDex blacklist: 'T1_DE_KIT', 'T1_US_FNAL' and '*_Disk' allow user jobs - other T1's dont!
self._lumi_query = config.getBool('lumi metadata',
self._lumi_filter != [])
self._phedexFilter = config.getFilter('phedex sites',
'-T3_US_FNALLPC',
defaultMatcher='blackwhite',
defaultFilter='weak')
self._phedexT1Filter = config.getFilter('phedex t1 accept',
'T1_DE_KIT T1_US_FNAL',
defaultMatcher='blackwhite',
defaultFilter='weak')
self._phedexT1Mode = config.get('phedex t1 mode', 'disk').lower()
self.onlyComplete = config.getBool('only complete sites', True)
self._locationFormat = config.getEnum('location format',
CMSLocationFormat,
CMSLocationFormat.hostname)
(self._datasetPath, self._url,
self._datasetBlock) = utils.optSplit(datasetExpr, '@#')
self._url = self._url or config.get('dbs instance', '')
self._datasetBlock = self._datasetBlock or 'all'
self.onlyValid = config.getBool('only valid', True)
def __init__(self, baseClass, value, config, tags, inherit, defaultName, pluginPaths):
(self._baseClass, self._config, self._tags, self._inherit, self._pluginPaths) = \
(baseClass, config, tags, inherit, pluginPaths)
(self._instClassName, self._instName) = utils.optSplit(value, ':')
if self._instName == '':
if not defaultName:
self._instName = self._instClassName.split('.')[-1] # Default: (non fully qualified) class name as instance name
else:
self._instName = defaultName
def _setup(self, setup_key, setup_mod):
if setup_key:
(delim, ds, de) = utils.optSplit(setup_key, '::')
modifier = identity
if setup_mod and (setup_mod.strip() != 'value'):
try:
modifier = eval('lambda value: ' + setup_mod) # pylint:disable=eval-used
except Exception:
raise ConfigError('Unable to parse delimeter modifier %r' %
setup_mod)
return (delim, parseStr(ds, int), parseStr(de, int), modifier)
def __init__(self, config, datasetExpr, datasetNick=None, datasetID=0):
changeTrigger = triggerResync(['datasets', 'parameters'])
self._lumi_filter = config.getLookup('lumi filter', {},
parser=parseLumiFilter,
strfun=strLumi,
onChange=changeTrigger)
if not self._lumi_filter.empty():
config.set('dataset processor', 'LumiDataProcessor', '+=')
DataProvider.__init__(self, config, datasetExpr, datasetNick,
datasetID)
# LumiDataProcessor instantiated in DataProcessor.__ini__ will set lumi metadata as well
self._lumi_query = config.getBool('lumi metadata',
not self._lumi_filter.empty(),
onChange=changeTrigger)
# PhEDex blacklist: 'T1_DE_KIT', 'T1_US_FNAL' and '*_Disk' allow user jobs - other T1's dont!
self._phedexFilter = config.getFilter('phedex sites',
'-T3_US_FNALLPC',
defaultMatcher='blackwhite',
defaultFilter='weak',
onChange=changeTrigger)
self._phedexT1Filter = config.getFilter('phedex t1 accept',
'T1_DE_KIT T1_US_FNAL',
defaultMatcher='blackwhite',
defaultFilter='weak',
onChange=changeTrigger)
self._phedexT1Mode = config.getEnum('phedex t1 mode',
PhedexT1Mode,
PhedexT1Mode.disk,
onChange=changeTrigger)
self.onlyComplete = config.getBool('only complete sites',
True,
onChange=changeTrigger)
self._locationFormat = config.getEnum('location format',
CMSLocationFormat,
CMSLocationFormat.hostname,
onChange=changeTrigger)
self._pjrc = JSONRestClient(
url='https://cmsweb.cern.ch/phedex/datasvc/json/prod/blockreplicas'
)
(self._datasetPath, self._url,
self._datasetBlock) = optSplit(datasetExpr, '@#')
self._url = self._url or config.get('dbs instance', '')
self._datasetBlock = self._datasetBlock or 'all'
self.onlyValid = config.getBool('only valid',
True,
onChange=changeTrigger)
def __init__(self, config, datasetExpr, datasetNick=None):
self._changeTrigger = triggerResync(['datasets', 'parameters'])
self._lumi_filter = config.getLookup('lumi filter', {},
parser=parseLumiFilter,
strfun=strLumi,
onChange=self._changeTrigger)
if not self._lumi_filter.empty():
config.set('dataset processor', 'LumiDataProcessor', '+=')
DataProvider.__init__(self, config, datasetExpr, datasetNick)
# LumiDataProcessor instantiated in DataProcessor.__ini__ will set lumi metadata as well
self._lumi_query = config.getBool('lumi metadata',
not self._lumi_filter.empty(),
onChange=self._changeTrigger)
config.set('phedex sites matcher mode', 'shell', '?=')
# PhEDex blacklist: 'T1_*_Disk nodes allow user jobs - other T1's dont!
self._phedexFilter = config.getFilter('phedex sites',
'-* T1_*_Disk T2_* T3_*',
defaultMatcher='blackwhite',
defaultFilter='strict',
onChange=self._changeTrigger)
self._onlyComplete = config.getBool('only complete sites',
True,
onChange=self._changeTrigger)
self._locationFormat = config.getEnum('location format',
CMSLocationFormat,
CMSLocationFormat.hostname,
onChange=self._changeTrigger)
self._pjrc = JSONRestClient(
url='https://cmsweb.cern.ch/phedex/datasvc/json/prod/blockreplicas'
)
self._sitedb = SiteDB()
(self._datasetPath, self._datasetInstance,
self._datasetBlock) = optSplit(datasetExpr, '@#')
instance_default = config.get('dbs instance',
'',
onChange=self._changeTrigger)
self._datasetInstance = self._datasetInstance or instance_default
if not self._datasetInstance:
self._datasetInstance = 'prod/global'
elif '/' not in self._datasetInstance:
self._datasetInstance = 'prod/%s' % self._datasetInstance
self._datasetBlock = self._datasetBlock or 'all'
self.onlyValid = config.getBool('only valid',
True,
onChange=self._changeTrigger)
def __init__(self, config, datasetExpr, datasetNick = None, datasetID = 0):
DataProvider.__init__(self, config, datasetExpr, datasetNick, datasetID)
self._CommonPrefix = max(self.enumValues) + 1
self._CommonMetadata = max(self.enumValues) + 2
self._handleEntry = {
'events': (DataProvider.NEntries, int, 'block entry counter'),
'id': (DataProvider.DatasetID, int, 'dataset ID'),
'metadata': (DataProvider.Metadata, parseJSON, 'metadata description'),
'metadata common': (self._CommonMetadata, parseJSON, 'common metadata'),
'nickname': (DataProvider.Nickname, str, 'dataset nickname'),
'prefix': (self._CommonPrefix, str, 'common prefix'),
'se list': (DataProvider.Locations, lambda value: parseList(value, ','), 'block location'),
}
(path, self._forcePrefix, self._filter) = utils.optSplit(datasetExpr, '@%')
self._filename = config.resolvePath(path, True, 'Error resolving dataset file: %s' % path)
def __init__(self, config, datasetExpr, datasetNick = None, datasetID = 0):
if os.path.isdir(datasetExpr):
GCProvider.stageDir[None] = ['OutputDirsFromWork']
config.set('source directory', datasetExpr)
datasetExpr = os.path.join(datasetExpr, 'work.conf')
else:
GCProvider.stageDir[None] = ['OutputDirsFromConfig', 'MetadataFromTask']
datasetExpr, selector = utils.optSplit(datasetExpr, '%')
config.set('source config', datasetExpr)
config.set('source job selector', selector)
extConfig = createConfig(datasetExpr)
extModule = extConfig.changeView(setSections = ['global']).get(['task', 'module'])
if 'ParaMod' in extModule: # handle old config files
extModule = extConfig.changeView(setSections = ['ParaMod']).get('module')
sGet = lambda scannerDict: scannerDict.get(None) + scannerDict.get(extModule, [])
sList = sGet(GCProvider.stageDir) + ['JobInfoFromOutputDir', 'FilesFromJobInfo'] + sGet(GCProvider.stageFile) + ['DetermineEvents', 'AddFilePrefix']
ScanProviderBase.__init__(self, config, sList, datasetNick, datasetID)
def __init__(self, config, datasetExpr, datasetNick = None):
DataProvider.__init__(self, config, datasetExpr, datasetNick)
self._CommonPrefix = max(self.enumValues) + 1
self._CommonMetadata = max(self.enumValues) + 2
self._handleEntry = {
'events': (DataProvider.NEntries, int, 'block entry counter'),
'id': (None, None, 'dataset ID'), # legacy key - skip
'metadata': (DataProvider.Metadata, parseJSON, 'metadata description'),
'metadata common': (self._CommonMetadata, parseJSON, 'common metadata'),
'nickname': (DataProvider.Nickname, str, 'dataset nickname'),
'prefix': (self._CommonPrefix, str, 'common prefix'),
'se list': (DataProvider.Locations, lambda value: parseList(value, ','), 'block location'),
}
(path, self._forcePrefix, self._filter) = utils.optSplit(datasetExpr, '@%')
self._filename = config.resolvePath(path, True, 'Error resolving dataset file: %s' % path)
def __init__(self, config, datasetExpr, datasetNick, datasetID = 0):
if os.path.isdir(datasetExpr):
GCProvider.stageDir[None] = ['OutputDirsFromWork']
config.set('source directory', datasetExpr)
datasetExpr = os.path.join(datasetExpr, 'work.conf')
else:
GCProvider.stageDir[None] = ['OutputDirsFromConfig', 'MetadataFromTask']
datasetExpr, selector = utils.optSplit(datasetExpr, '%')
config.set('source config', datasetExpr)
config.set('source job selector', selector)
extConfig = Config([DefaultFilesConfigFiller(), FileConfigFiller([datasetExpr])], datasetExpr)
extModule = extConfig.get('global', ['task', 'module'])
if 'ParaMod' in extModule:
extModule = extConfig.get('ParaMod', 'module')
sGet = lambda scannerDict: scannerDict.get(None) + scannerDict.get(extModule, [])
sList = sGet(GCProvider.stageDir) + ['FilesFromJobInfo'] + sGet(GCProvider.stageFile) + ['DetermineEvents', 'AddFilePrefix']
ScanProviderBase.__init__(self, config, sList, datasetNick, datasetID)
def __init__(self, config, datasetExpr, datasetNick = None, datasetID = 0):
self._lumi_filter = parseLumiFilter(config.get('lumi filter', ''))
if self._lumi_filter:
config.set('dataset processor', 'LumiDataProcessor', '+=')
DataProvider.__init__(self, config, datasetExpr, datasetNick, datasetID)
# PhEDex blacklist: 'T1_DE_KIT', 'T1_US_FNAL' and '*_Disk' allow user jobs - other T1's dont!
self._lumi_query = config.getBool('lumi metadata', self._lumi_filter != [])
self._phedexFilter = config.getFilter('phedex sites', '-T3_US_FNALLPC',
defaultMatcher = 'blackwhite', defaultFilter = 'weak')
self._phedexT1Filter = config.getFilter('phedex t1 accept', 'T1_DE_KIT T1_US_FNAL',
defaultMatcher = 'blackwhite', defaultFilter = 'weak')
self._phedexT1Mode = config.get('phedex t1 mode', 'disk').lower()
self.onlyComplete = config.getBool('only complete sites', True)
self._locationFormat = config.getEnum('location format', CMSLocationFormat, CMSLocationFormat.hostname)
(self._datasetPath, self._url, self._datasetBlock) = utils.optSplit(datasetExpr, '@#')
self._url = self._url or config.get('dbs instance', '')
self._datasetBlock = self._datasetBlock or 'all'
self.onlyValid = config.getBool('only valid', True)
def __init__(self, config, datasetExpr, datasetNick = None, datasetID = 0):
changeTrigger = triggerResync(['datasets', 'parameters'])
self._lumi_filter = config.getLookup('lumi filter', {}, parser = parseLumiFilter, strfun = strLumi, onChange = changeTrigger)
if not self._lumi_filter.empty():
config.set('dataset processor', 'LumiDataProcessor', '+=')
DataProvider.__init__(self, config, datasetExpr, datasetNick, datasetID)
# LumiDataProcessor instantiated in DataProcessor.__ini__ will set lumi metadata as well
self._lumi_query = config.getBool('lumi metadata', not self._lumi_filter.empty(), onChange = changeTrigger)
# PhEDex blacklist: 'T1_DE_KIT', 'T1_US_FNAL' and '*_Disk' allow user jobs - other T1's dont!
self._phedexFilter = config.getFilter('phedex sites', '-T3_US_FNALLPC',
defaultMatcher = 'blackwhite', defaultFilter = 'weak', onChange = changeTrigger)
self._phedexT1Filter = config.getFilter('phedex t1 accept', 'T1_DE_KIT T1_US_FNAL',
defaultMatcher = 'blackwhite', defaultFilter = 'weak', onChange = changeTrigger)
self._phedexT1Mode = config.getEnum('phedex t1 mode', PhedexT1Mode, PhedexT1Mode.disk, onChange = changeTrigger)
self.onlyComplete = config.getBool('only complete sites', True, onChange = changeTrigger)
self._locationFormat = config.getEnum('location format', CMSLocationFormat, CMSLocationFormat.hostname, onChange = changeTrigger)
self._pjrc = JSONRestClient(url = 'https://cmsweb.cern.ch/phedex/datasvc/json/prod/blockreplicas')
(self._datasetPath, self._url, self._datasetBlock) = optSplit(datasetExpr, '@#')
self._url = self._url or config.get('dbs instance', '')
self._datasetBlock = self._datasetBlock or 'all'
self.onlyValid = config.getBool('only valid', True, onChange = changeTrigger)
def __init__(self, config, datasetExpr, datasetNick = None):
ds_config = config.changeView(viewClass = 'TaggedConfigView', addNames = [md5_hex(datasetExpr)])
if os.path.isdir(datasetExpr):
scan_pipeline = ['OutputDirsFromWork']
ds_config.set('source directory', datasetExpr)
datasetExpr = os.path.join(datasetExpr, 'work.conf')
else:
scan_pipeline = ['OutputDirsFromConfig', 'MetadataFromTask']
datasetExpr, selector = utils.optSplit(datasetExpr, '%')
ds_config.set('source config', datasetExpr)
ds_config.set('source job selector', selector)
ext_config = create_config(datasetExpr)
ext_task_name = ext_config.changeView(setSections = ['global']).get(['module', 'task'])
if 'ParaMod' in ext_task_name: # handle old config files
ext_task_name = ext_config.changeView(setSections = ['ParaMod']).get('module')
ext_task_cls = Plugin.getClass(ext_task_name)
for ext_task_cls in Plugin.getClass(ext_task_name).iterClassBases():
try:
scan_holder = GCProviderSetup.getClass('GCProviderSetup_' + ext_task_cls.__name__)
except PluginError:
continue
scan_pipeline += scan_holder.scan_pipeline
break
ScanProviderBase.__init__(self, ds_config, datasetExpr, datasetNick, scan_pipeline)
def __init__(self, config, datasetExpr, datasetNick, datasetID = 0):
DataProvider.__init__(self, config, datasetExpr, datasetNick, datasetID)
# PhEDex blacklist: 'T1_DE_KIT', 'T1_US_FNAL' and '*_Disk' allow user jobs - other T1's dont!
self.phedexBL = config.getList('phedex sites', ['-T3_US_FNALLPC'])
self.phedexWL = config.getList('phedex t1 accept', ['T1_DE_KIT', 'T1_US_FNAL'])
self.phedexT1 = config.get('phedex t1 mode', 'disk').lower()
self.onlyComplete = config.getBool('only complete sites', True)
self.locationFormat = config.get('location format', 'hostname').lower() # hostname or sitedb
if self.locationFormat not in ['hostname', 'sitedb', 'both']:
raise ConfigError('Invalid location format: %s' % self.locationFormat)
(self.datasetPath, self.url, self.datasetBlock) = utils.optSplit(datasetExpr, '@#')
self.url = QM(self.url, self.url, config.get('dbs instance', ''))
self.datasetBlock = QM(self.datasetBlock, self.datasetBlock, 'all')
self.includeLumi = config.getBool('keep lumi metadata', False)
self.onlyValid = config.getBool('only valid', True)
self.checkUnique = config.getBool('check unique', True)
# This works in tandem with active task module (cmssy.py supports only [section] lumi filter!)
self.selectedLumis = parseLumiFilter(config.get('lumi filter', ''))
if self.selectedLumis:
utils.vprint('Runs/lumi section filter enabled! (%d entries)' % len(self.selectedLumis), -1, once = True)
utils.vprint('\tThe following runs and lumi sections are selected:', 1, once = True)
utils.vprint('\t' + utils.wrapList(formatLumi(self.selectedLumis), 65, ',\n\t'), 1, once = True)
def splitParse(opt):
(delim, ds, de) = utils.optSplit(opt, '::')
return (delim, parseStr(ds, int), parseStr(de, int))
def splitParse(opt): (delim, ds, de) = utils.optSplit(opt, '::') return (delim, parseStr(ds, int), parseStr(de, int))
def splitParse(opt): (delim, ds, de) = utils.optSplit(opt, '::') return (delim, utils.parseInt(ds), utils.parseInt(de))
def __init__(self, config, datasetExpr, datasetNick, datasetID = 0): DataProvider.__init__(self, config, datasetExpr, datasetNick, datasetID) (path, self._forcePrefix, self._filter) = utils.optSplit(datasetExpr, '@%') self._filename = config.resolvePath(path, True, 'Error resolving dataset file: %s' % path)
