Inheritance diagram for python.ConfigBlock.ConfigBlock:

Collaboration diagram for python.ConfigBlock.ConfigBlock:

Public Member Functions
	__init__ (self)
	setBlockName (self, name)
	getBlockName (self)
	factoryName (self)
	setFactoryName (self, name)
	instanceName (self)
	isUsedForConfig (self, config)
	applyConfigOverrides (self, config)
	addDependency (self, dependencyName, required=True)
	hasDependencies (self)
	getDependencies (self)
	addOption (self, name, defaultValue, *, type, info='', noneAction='ignore', required=False, expertMode=None)
	setOptionValue (self, name, value)
	getOptionValue (self, name)
	getOptions (self)
	printOptions (self, verbose=False, width=60, indent=" ")
	hasOption (self, name)
	__eq__ (self, blockName)
	__str__ (self)
	get_instance_count (cls)
	checkExpertSettings (self, config)
	__new__ (cls, name, bases, dct)

Public Attributes
	onlyForDSIDs

Static Public Attributes
dict	instance_counts = {}

Protected Member Functions
	_is_expert_value (self, rule, value)

Protected Attributes
str	_blockName = ''
	_factoryName = None
list	_dependencies = []
dict	_options = {}
dict	_expertModeSettings = {}

Detailed Description

the base class for classes implementing individual blocks of
configuration

A configuration block is a sequence of one or more algorithms that
should always be scheduled together, e.g. the muon four momentum
corrections could be a single block, muon selection could then be
another block.  The blocks themselves generally have their own
configuration options/properties specific to the block, and will
perform a dynamic configuration based on those options as well as
the overall job.

The actual configuration of the algorithms in the block will
depend on what other blocks are scheduled before and afterwards,
most importantly some algorithms will introduce shallow copies
that subsequent algorithms will need to run on, and some
algorithms will add selection decorations that subquent algorithms
should use as preselections.

The algorithms get created in a multi-step process (that may be
extended in the future): As a first step each block retrieves
references to the containers it uses (essentially marking its spot
in the processing chain) and also registering any shallow copies
that will be made.  In the second/last step each block then
creates the fully configured algorithms.

One goal is that when the algorithms get created they will have
their final configuration and there needs to be no
meta-configuration data attached to the algorithms, essentially an
inversion of the approach in AnaAlgSequence in which the
algorithms got created first with associated meta-configuration
and then get modified in susequent configuration steps.

For now this is mostly an empty base class, but another goal of
this approach is to make it easier to build another configuration
layer on top of this one, and this class will likely be extended
and get data members at that point.

The child class needs to implement the method `makeAlgs` which is
given a single `ConfigAccumulator` type argument. This is meant to
create the sequence of algorithms that this block configures. This
is currently (28 Jul 2025) called twice and should do the same thing
during both calls, but the plan is to change that to a single call.

The child class should also implement the method `getInstanceName`
which should return a string that is used to distinguish between
multiple instances of the same block. This is used to append the
instance name to the names of all algorithms created by this block,
and may in the future also be used to distinguish between multiple
instances of the block.

Definition at line 92 of file ConfigBlock.py.

Constructor & Destructor Documentation

◆ init()

python.ConfigBlock.ConfigBlock.__init__ ( self )

Definition at line 147 of file ConfigBlock.py.

    def __init__ (self) :
        self._blockName = ''
        self._factoryName = None
        self._dependencies = []
        self._options = {} # used with block configuration to set arbitrary option
        self._expertModeSettings = {} # dictionary to track expert mode requirements for each option
        self.addOption('groupName', '', type=str,
            info=('Used to specify this block when setting an'
                ' option at an arbitrary location.'))
        self.addOption('skipOnData', False, type=bool,
            info=('User option to prevent the block from running'
                  ' on data. This only affects blocks that are'
                  ' intended to run on data.'))
        self.addOption('skipOnMC', False, type=bool,
            info=('User option to prevent the block from running'
                  ' on MC. This only affects blocks that are'
                  ' intended to run on MC.'))
        self.addOption('skipWithSystematics', False, type=bool,
            info=('User option to prevent the block from running with systematics.'))
        self.addOption('onlyForDSIDs', [], type=list,
            info=('Used to specify which MC DSIDs to allow this'
                  ' block to run on. Each element of the list'
                  ' can be a full DSID (e.g. 410470), or a regex'
                  ' (e.g. 410.* to select all 410xxx DSIDs, or'
                  ' ^(?!410) to veto them). An empty list means no'
                  ' DSID restriction.'))
        self.addOption('propertyOverrides', {}, type=dict,
            info=('EXPERT USE ONLY: A dictionary of properties to'
                  ' override at the end of configuration. This should'
                  ' take the form'
                  ' {"algName.toolName.propertyName": value, ...},'
                  ' without any automatically applied postfixes for'
                  ' the algorithm name. THIS IS MEANT TO BE EXPERT'
                  ' USAGE ONLY. Properties that need to be set by'
                  ' the user should be declared as options on the'
                  ' block itself. EXPERT USE ONLY!'),
                  expertMode=True)
        # Increment the instance count for the current class
        cls = type(self)  # Get the actual class of the instance (also derived!)
        if cls not in ConfigBlock.instance_counts:
            ConfigBlock.instance_counts[cls] = 0
        # Note: we do need to check in the call stack that we are
        # in a real makeConfig situation, and not e.g. printAlgs
        stack = inspect.stack()
        for frame_info in stack:
            # Get the class name (if any) from the frame
            parent_cls = frame_info.frame.f_locals.get('self', None)
            if parent_cls is None or not isinstance(parent_cls, ConfigBlock):
                # If the frame does not belong to an instance of ConfigBlock, it's an external caller
                if frame_info.function == "makeConfig":
                    ConfigBlock.instance_counts[cls] += 1
                    break
 
 

Member Function Documentation

◆ eq()

python.ConfigBlock.ConfigBlock.__eq__	(		self,
			blockName )

Implementation of == operator. Used for seaching configSeque.
E.g. if blockName in configSeq:

Definition at line 423 of file ConfigBlock.py.

    def __eq__(self, blockName):
        """
        Implementation of == operator. Used for seaching configSeque.
        E.g. if blockName in configSeq:
        """
        return self._blockName == blockName
 
 

◆ new()

python.ConfigBlock.BlockNameProcessorMeta.__new__	(	cls,
		name,
		bases,
		dct )

inherited

Definition at line 53 of file ConfigBlock.py.

    def __new__(cls, name, bases, dct):
        # Automatically apply alphanumeric-only decorator to 'instanceName()' method
        if 'instanceName' in dct and callable(dct['instanceName']):
            dct['instanceName'] = alphanumeric_block_name(dct['instanceName'])
        return super().__new__(cls, name, bases, dct)
 

◆ str()

python.ConfigBlock.ConfigBlock.__str__ ( self )

Definition at line 431 of file ConfigBlock.py.

    def __str__(self):
        return self._blockName if self._blockName else self.factoryName()

◆ _is_expert_value()

python.ConfigBlock.ConfigBlock._is_expert_value	(	self,
		rule,
		value )

protected

Check whether value matches an expert mode rule.
Rule can be:
- A literal (compared with ==)
- A callable predicate (called with value)
- A special marker string (common callable)

Definition at line 440 of file ConfigBlock.py.

    def _is_expert_value(self, rule, value):
        """
        Check whether value matches an expert mode rule.
        Rule can be:
        - A literal (compared with ==)
        - A callable predicate (called with value)
        - A special marker string (common callable)
        """
        if callable(rule):
            return rule(value)
 
        if isinstance(rule, str):
            if rule == "nonemptystring":
                return isinstance(value, str) and value != ""
            if rule == "nonemptylist":
                return isinstance(value, list) and value != []
            if rule == "positiveint":
                return isinstance(value, int) and value > 0
 
        # Fallback: direct value comparison
        return value == rule
 

◆ addDependency()

python.ConfigBlock.ConfigBlock.addDependency	(	self,
		dependencyName,
		required = True )

Add a dependency for the block. Dependency is corresponds to the
blockName of another block. If required is True, will throw an
error if dependency is not present; otherwise will move this
block after the required block. If required is False, will do
nothing if required block is not present; otherwise, it will
move block after required block.

Definition at line 295 of file ConfigBlock.py.

    def addDependency(self, dependencyName, required=True):
        """
        Add a dependency for the block. Dependency is corresponds to the
        blockName of another block. If required is True, will throw an
        error if dependency is not present; otherwise will move this
        block after the required block. If required is False, will do
        nothing if required block is not present; otherwise, it will
        move block after required block.
        """
        if not self.hasDependencies():
            # add option to block ignore dependencies
            self.addOption('ignoreDependencies', [], type=list,
                           info='List of dependencies defined in the ConfigBlock to ignore.')
        self._dependencies.append(ConfigBlockDependency(dependencyName, required))
 

◆ addOption()

python.ConfigBlock.ConfigBlock.addOption	(		self,
			name,
			defaultValue,
		*	,
			type,
			info = '',
			noneAction = 'ignore',
			required = False,
			expertMode = None )

declare the given option on the configuration block

This should only be called in the constructor of the
configuration block.

NOTE: The backend to option handling is slated to be replaced
at some point.  This particular function should essentially
stay the same, but some behavior may change.

Definition at line 318 of file ConfigBlock.py.

                   type, info='', noneAction='ignore', required=False, expertMode=None) :
        """declare the given option on the configuration block
 
        This should only be called in the constructor of the
        configuration block.
 
        NOTE: The backend to option handling is slated to be replaced
        at some point.  This particular function should essentially
        stay the same, but some behavior may change.
        """
        if name in self._options :
            raise KeyError (f'duplicate option: {name}')
        if type not in [str, bool, int, float, list, dict, None] :
            raise TypeError (f'unknown option type: {type}')
        noneActions = ['error', 'set', 'ignore']
        if noneAction not in noneActions :
            raise ValueError (f'invalid noneAction: {noneAction} [allowed values: {noneActions}]')
 
        # Store expert mode settings if provided
        if expertMode is not None:
            if expertMode is True:
                # in this case we will just check against the default value
                self._expertModeSettings[name] = True
            elif not isinstance(expertMode, list):
                raise TypeError (f'expertMode must be a list, got {type(expertMode)}')
            else:
                # here we will check against a list of custom values
                self._expertModeSettings[name] = expertMode
 
        setattr (self, name, defaultValue)
        self._options[name] = ConfigBlockOption(type=type, info=info,
            noneAction=noneAction, required=required, default=defaultValue)
 
 

◆ applyConfigOverrides()

python.ConfigBlock.ConfigBlock.applyConfigOverrides	(		self,
			config )

Apply any configuration overrides specified in the block's
`propertyOverrides` option. This is meant to be called at the
end of the configuration process, after all algorithms have been
created and configured.

Definition at line 270 of file ConfigBlock.py.

    def applyConfigOverrides(self, config):
        """
        Apply any configuration overrides specified in the block's
        `propertyOverrides` option. This is meant to be called at the
        end of the configuration process, after all algorithms have been
        created and configured.
        """
        for key, value in self.propertyOverrides.items():
            # Split the key into algorithm name, tool name, and property name
            parts = key.split('.')
            if len(parts) < 2:
                raise Exception(f"Invalid override key format: {key}")
            alg = config.getAlgorithm(parts[0])
            if alg is None:
                raise Exception(f"Algorithm {parts[0]} not found in config for override: {key}")
            for name in parts[1:-1]:
                # Navigate through tools if necessary
                if hasattr(alg, name):
                    alg = getattr(alg, name)
                else:
                    raise Exception(f"Tool {name} not found for override: {key}")
            # Set the property on the algorithm/tool. This is probably a
            # horrible hack, but `setattr` didn't work for me.
            alg.__setattr__(parts[-1], value)
 

◆ checkExpertSettings()

python.ConfigBlock.ConfigBlock.checkExpertSettings	(		self,
			config )

Check if any settings require expert mode and validate accordingly.
If any setting is set to a value that requires expert mode but we're
not in expert mode, raise an error.

Definition at line 462 of file ConfigBlock.py.

    def checkExpertSettings(self, config):
        """
        Check if any settings require expert mode and validate accordingly.
        If any setting is set to a value that requires expert mode but we're
        not in expert mode, raise an error.
        """
        for option_name, expert_rule in self._expertModeSettings.items():
            current_value = self.getOptionValue(option_name)
            default_value = self._options[option_name].default
 
            if expert_rule is True:
                # Any deviation from the default requires expert mode
                if current_value != default_value:
                    warnings.warn(
                        f"Block '{self.factoryName()}' option '{option_name}' "
                        f"set to '{current_value}' (default '{default_value}'), "
                        f"requires expert mode.",
                        ExpertModeWarning, stacklevel=2
                    )
 
            else:  # it's a list of expert values/markers/predicates
                for ev in expert_rule:
                    if self._is_expert_value(ev, current_value):
                        warnings.warn(
                            f"Block '{self.factoryName()}' option '{option_name}' "
                            f"set to expert-only value '{current_value}'. "
                            f"Requires expert mode.",
                            ExpertModeWarning, stacklevel=2
                        )
        # All checks passed
        return

◆ factoryName()

python.ConfigBlock.ConfigBlock.factoryName ( self )

get the factory name for this block

This is mostly to give a reliable means of identifying the type
of block we have in error messages. This is meant to be
automatically set by the factory based on the requested block
name, but there are a number of fallbacks. It is best not to
assume a specific format, this is mostly meant to be used as an
identifier in output messages.

Definition at line 209 of file ConfigBlock.py.

    def factoryName(self):
        """get the factory name for this block
        
        This is mostly to give a reliable means of identifying the type
        of block we have in error messages. This is meant to be
        automatically set by the factory based on the requested block
        name, but there are a number of fallbacks. It is best not to
        assume a specific format, this is mostly meant to be used as an
        identifier in output messages.
        """
        if self._factoryName is not None and self._factoryName != '':
            return self._factoryName
        # If no factory name is set and the block has a name, use that
        if self._blockName is not None and self._blockName != '':
            return self._blockName
        # Use the class name as a fallback
        return self.__class__.__name__
 

◆ get_instance_count()

python.ConfigBlock.ConfigBlock.get_instance_count ( cls )

Definition at line 436 of file ConfigBlock.py.

    def get_instance_count(cls):
        # Access the current count for this class
        return ConfigBlock.instance_counts.get(cls, 0)
 

◆ getBlockName()

python.ConfigBlock.ConfigBlock.getBlockName ( self )

Get blockName

Definition at line 205 of file ConfigBlock.py.

    def getBlockName(self):
        """Get blockName"""
        return self._blockName
 

◆ getDependencies()

python.ConfigBlock.ConfigBlock.getDependencies ( self )

Return the list of dependencies.

Definition at line 314 of file ConfigBlock.py.

    def getDependencies(self):
        """Return the list of dependencies. """
        return self._dependencies
 

◆ getOptions()

python.ConfigBlock.ConfigBlock.getOptions ( self )

Return a copy of the options associated with the block

Definition at line 386 of file ConfigBlock.py.

    def getOptions(self):
        """Return a copy of the options associated with the block"""
        return self._options.copy()
 
 

◆ getOptionValue()

python.ConfigBlock.ConfigBlock.getOptionValue	(		self,
			name )

Returns config option value, if present; otherwise return None

Definition at line 380 of file ConfigBlock.py.

    def getOptionValue(self, name):
        """Returns config option value, if present; otherwise return None"""
        if name in self._options:
            return getattr(self, name)
 
 

◆ hasDependencies()

python.ConfigBlock.ConfigBlock.hasDependencies ( self )

Return True if there is a dependency.

Definition at line 310 of file ConfigBlock.py.

    def hasDependencies(self):
        """Return True if there is a dependency."""
        return bool(self._dependencies)
 

◆ hasOption()

python.ConfigBlock.ConfigBlock.hasOption	(		self,
			name )

whether the configuration block has the given option

WARNING: The backend to option handling is slated to be
replaced at some point.  This particular function may change
behavior, interface or be removed/replaced entirely.

Definition at line 413 of file ConfigBlock.py.

    def hasOption (self, name) :
        """whether the configuration block has the given option
 
        WARNING: The backend to option handling is slated to be
        replaced at some point.  This particular function may change
        behavior, interface or be removed/replaced entirely.
        """
        return name in self._options
 
 

◆ instanceName()

python.ConfigBlock.ConfigBlock.instanceName ( self )

Get the name of the instance

The name of the instance is used to distinguish between multiple
instances of the same block. Most importantly, this will be
appended to the names of all algorithms created by this block.
This defaults to an empty string, but block implementations
should override it with an appropriate name based on identifying
options set on this instance. A typical example would be the
name of the (main) container, plus potentially the selection or
working point.

Ideally all blocks should override this method, but for backward
compatibility (28 Jul 25) it defaults to an empty string.

Definition at line 236 of file ConfigBlock.py.

    def instanceName(self):
        """Get the name of the instance
 
        The name of the instance is used to distinguish between multiple
        instances of the same block. Most importantly, this will be
        appended to the names of all algorithms created by this block.
        This defaults to an empty string, but block implementations
        should override it with an appropriate name based on identifying
        options set on this instance. A typical example would be the
        name of the (main) container, plus potentially the selection or
        working point.
 
        Ideally all blocks should override this method, but for backward
        compatibility (28 Jul 25) it defaults to an empty string.
        """
        return ''
 

◆ isUsedForConfig()

python.ConfigBlock.ConfigBlock.isUsedForConfig	(		self,
			config )

whether this block should be used for the given configuration

This is used by `ConfigSequence` to determine whether this block
should be included in the configuration.

Definition at line 253 of file ConfigBlock.py.

    def isUsedForConfig(self, config):
        """
        whether this block should be used for the given configuration
 
        This is used by `ConfigSequence` to determine whether this block
        should be included in the configuration.
        """
        if self.skipWithSystematics and not config.noSystematics():
            return False
        if self.skipOnData and config.dataType() is DataType.Data:
            return False
        if self.skipOnMC and config.dataType() is not DataType.Data:
            return False
        if not filter_dsids(self.onlyForDSIDs, config):
            return False
        return True
 

◆ printOptions()

python.ConfigBlock.ConfigBlock.printOptions	(	self,
		verbose = False,
		width = 60,
		indent = " " )

Prints options and their values

Definition at line 391 of file ConfigBlock.py.

    def printOptions(self, verbose=False, width=60, indent="    "):
        """
        Prints options and their values
        """
        def printWrap(text, width=60, indent="    "):
            wrapper = textwrap.TextWrapper(width=width, initial_indent=indent,
                subsequent_indent=indent)
            for line in wrapper.wrap(text=text):
                logCPAlgCfgBlock.info(line)
 
        for opt, vals in self.getOptions().items():
            if verbose:
                logCPAlgCfgBlock.info(indent + f"\033[4m{opt}\033[0m: {self.getOptionValue(opt)}")
                logCPAlgCfgBlock.info(indent*2 + f"\033[4mtype\033[0m: {vals.type}")
                logCPAlgCfgBlock.info(indent*2 + f"\033[4mdefault\033[0m: {vals.default}")
                logCPAlgCfgBlock.info(indent*2 + f"\033[4mrequired\033[0m: {vals.required}")
                logCPAlgCfgBlock.info(indent*2 + f"\033[4mnoneAction\033[0m: {vals.noneAction}")
                printWrap(f"\033[4minfo\033[0m: {vals.info}", indent=indent*2)
            else:
                logCPAlgCfgBlock.info(indent + f"{ opt}: {self.getOptionValue(opt)}")
 
 

◆ setBlockName()

python.ConfigBlock.ConfigBlock.setBlockName	(		self,
			name )

Set blockName

Definition at line 201 of file ConfigBlock.py.

    def setBlockName(self, name):
        """Set blockName"""
        self._blockName = name
 

◆ setFactoryName()

python.ConfigBlock.ConfigBlock.setFactoryName	(		self,
			name )

set the factory name for this block

This is meant to be called automatically by the factory based on
the requested block name. If you are creating a block without a factory,
you can call this method to set the factory name manually.

Definition at line 227 of file ConfigBlock.py.

    def setFactoryName(self, name):
        """set the factory name for this block
        
        This is meant to be called automatically by the factory based on
        the requested block name. If you are creating a block without a factory,
        you can call this method to set the factory name manually.
        """
        self._factoryName = name
 

◆ setOptionValue()

python.ConfigBlock.ConfigBlock.setOptionValue	(	self,
		name,
		value )

set the given option on the configuration block

NOTE: The backend to option handling is slated to be replaced
at some point.  This particular function should essentially
stay the same, but some behavior may change.

Definition at line 353 of file ConfigBlock.py.

    def setOptionValue (self, name, value) :
        """set the given option on the configuration block
 
        NOTE: The backend to option handling is slated to be replaced
        at some point.  This particular function should essentially
        stay the same, but some behavior may change.
        """
 
        if name not in self._options :
            raise KeyError (f'unknown option "{name}" in block "{self.__class__.__name__}"')
        noneAction = self._options[name].noneAction
        if value is not None or noneAction == 'set' :
            # check type if specified
            optType = self._options[name].type
            # convert int to float to prevent crash
            if optType is float and type(value) is int:
                value = float(value)
            if optType is not None and optType != type(value):
                raise ValueError(f'{name} for block {self.__class__.__name__} should '
                    f'be of type {optType} not {type(value)}')
            setattr (self, name, value)
        elif noneAction == 'ignore' :
            pass
        elif noneAction == 'error' :
            raise ValueError (f'passed None for setting option {name} with noneAction=error')
 
 

Member Data Documentation

◆ _blockName

str python.ConfigBlock.ConfigBlock._blockName = ''

protected

Definition at line 148 of file ConfigBlock.py.

◆ _dependencies

python.ConfigBlock.ConfigBlock._dependencies = []

protected

Definition at line 150 of file ConfigBlock.py.

◆ _expertModeSettings

dict python.ConfigBlock.ConfigBlock._expertModeSettings = {}

protected

Definition at line 152 of file ConfigBlock.py.

◆ _factoryName

python.ConfigBlock.ConfigBlock._factoryName = None

protected

Definition at line 149 of file ConfigBlock.py.

◆ _options

dict python.ConfigBlock.ConfigBlock._options = {}

protected

Definition at line 151 of file ConfigBlock.py.

◆ instance_counts

dict python.ConfigBlock.ConfigBlock.instance_counts = {}

static

Definition at line 145 of file ConfigBlock.py.

◆ onlyForDSIDs

python.ConfigBlock.ConfigBlock.onlyForDSIDs

Definition at line 266 of file ConfigBlock.py.

The documentation for this class was generated from the following file:

ConfigBlock.py

Public Member Functions

Public Attributes

Static Public Attributes

Protected Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

◆ __init__()

Member Function Documentation

◆ __eq__()

◆ __new__()

◆ __str__()

◆ _is_expert_value()

◆ addDependency()

◆ addOption()

◆ applyConfigOverrides()

◆ checkExpertSettings()

◆ factoryName()

◆ get_instance_count()

◆ getBlockName()

◆ getDependencies()

◆ getOptions()

◆ getOptionValue()

◆ hasDependencies()

◆ hasOption()

◆ instanceName()

◆ isUsedForConfig()

◆ printOptions()

◆ setBlockName()

◆ setFactoryName()

◆ setOptionValue()

Member Data Documentation

◆ _blockName

◆ _dependencies

◆ _expertModeSettings

◆ _factoryName

◆ _options

◆ instance_counts

◆ onlyForDSIDs

◆ init()

◆ eq()

◆ new()

◆ str()