␈rief a class for the definition of a complete recognizer More...

#include <linguisticProcessing/core/Automaton/recognizer.h>

Inheritance diagram for Lima::LinguisticProcessing::Automaton::Recognizer:

Classes
class	CompareTriggerRule

class	TriggerRule

Public Member Functions
	Recognizer ()

	Recognizer (const Recognizer &)

virtual	~Recognizer ()

Recognizer &	operator= (const Recognizer &)

void	init (Common::XMLConfigurationFiles::GroupConfigurationStructure &unitConfiguration, Manager *manager) override
	initialize with parameters from configuration file.

uint64_t	numberOfRules () const

uint64_t	addRuleInStorage (Rule *rule)
	add a rule in the storage zone

uint64_t	addRule (TransitionUnit trigger, Rule rule)
	add a rule in the recognizer : add the rule in the storage zone, and associate the trigger to the rule find if the trigger already exists in the trigger list : if it is the case, add the rule in rules associated to this trigger, else add a new trigger (copy of the one in argument) and associate the rule to it.

void	addRule (TransitionUnit *trigger, const uint64_t index)
	associate a known rule to a trigger (the rule is identified by its index in the storage zone)

uint64_t	apply (const LinguisticAnalysisStructure::AnalysisGraph &graph, const LinguisticGraphVertex &begin, const LinguisticGraphVertex &end, AnalysisContent &analysis, std::vector< RecognizerMatch > &result, bool testAllVertices=false, bool stopAtFirstSuccess=true, bool onlyOneSuccessPerType=false, bool returnAtFirstSuccess=false, bool applySameRuleWhileSuccess=false) const
	apply the recognizer on a graph

uint64_t	apply (const LinguisticAnalysisStructure::AnalysisGraph &graph, const LinguisticGraphVertex &begin, const LinguisticGraphVertex &end, const LinguisticGraphVertex &upstreamBound, const LinguisticGraphVertex &downstreamBound, AnalysisContent &analysis, std::vector< RecognizerMatch > &result, bool testAllVertices, bool stopAtFirstSuccess, bool onlyOneSuccessPerType, bool returnAtFirstSuccess, bool applySameRuleWhileSuccess) const
	apply the recognizer on a graph between two nodes and search between two others.

uint64_t	testOnVertex (const LinguisticAnalysisStructure::AnalysisGraph &graph, LinguisticGraphVertex &current, const LinguisticGraphVertex &begin, const LinguisticGraphVertex &end, AnalysisContent &analysis, std::vector< RecognizerMatch > &result, bool stopAtFirstSuccess=true, bool onlyOneSuccessPerType=false, bool applySameRuleWhileSuccess=false) const
	test the recognizer on a given vertex : check if this vertex is a trigger and if a rule applies, add the result to the vector given as parameter (takes the first rule which gives a match)

uint64_t	resolveOverlappingEntities (std::vector< RecognizerMatch > &listEntities, const OverlapResolutionStrategy &strategy=DEFAULT_OVERLAP_STRATEGY) const
	resolve the problem of overlapping entities in the list of entities : when two entities are overlaping, only one is kept

uint64_t	testSetOfRules (const TransitionUnit &trigger, const SetOfRules &rules, const LinguisticAnalysisStructure::AnalysisGraph &graph, const LinguisticGraphVertex &position, const LinguisticGraphVertex &begin, const LinguisticGraphVertex &end, AnalysisContent &analysis, std::vector< RecognizerMatch > &match, std::set< Common::MediaticData::EntityType > *forbiddenTypes=0, bool stopAtFirstSuccess=true, bool onlyOneSuccessPerType=false, bool applySameRuleWhileSuccess=false) const
	TODO: toBeDeleted Parse tokens paths from the trigger point.

void	listTriggers () const

void	initializeSearchStructure ()

Public Member Functions inherited from Lima::LinguisticProcessing::AbstractResource
	AbstractResource (QObject *parent=0)

virtual	~AbstractResource ()

	AbstractResource (const AbstractResource &)

Public Member Functions inherited from Lima::LinguisticProcessing::AbstractResourceBase
	AbstractResourceBase (QObject *parent=0)

virtual	~AbstractResourceBase ()

	AbstractResourceBase (const AbstractResourceBase &)

QString	getResourceFileName (const QString &paramName)

Public Member Functions inherited from Lima::InitializableObject< AbstractResource, ResourceInitializationParameters >
	InitializableObject ()

	InitializableObject (const InitializableObject &object)

InitializableObject &	operator= (const InitializableObject &object)

virtual	~InitializableObject ()

const std::string &	getId () const
	get the object id

void	setId (const std::string &id)
	set the object id

Protected Member Functions
void	clearSearchStructure ()

void	findNextSetOfRules (const LinguisticAnalysisStructure::AnalysisGraph &graph, LinguisticGraphVertex &vertex, AnalysisContent &analysis, const LinguisticAnalysisStructure::Token token, const LinguisticAnalysisStructure::MorphoSyntacticData data, std::vector< TriggerRule * > &matchingSetOfRules) const

void	init ()

void	copy (const Recognizer &a)

void	freeMem ()

void	setNormalizedForm (const LimaString &norm, RecognizerMatch &match) const

uint64_t	findRuleIndex (Rule *) const

Protected Member Functions inherited from Lima::LinguisticProcessing::AbstractResourceBase
LimaFileSystemWatcher &	resourceFileWatcher ()
	Accessor to the file watcher used to register resource files to be watched for changes.

Protected Attributes
std::vector< TriggerRule >	m_rules

SetOfRules	m_ruleStorage

MediaId	m_language

AutomatonControlParams	m_automatonControlParams
	parameters to control the search of the automaton

std::string	m_filename

TransitionSearchStructure< TriggerRule >	m_searchStructure

Protected Attributes inherited from Lima::LinguisticProcessing::AbstractResourceBase
AbstractResourceBasePrivate *	m_d

Protected Attributes inherited from Lima::InitializableObject< AbstractResource, ResourceInitializationParameters >
std::string	m_id

Friends
class	AutomatonReader

class	AutomatonWriter

LIMA_AUTOMATON_EXPORT std::ostream &	operator<< (std::ostream &, const Recognizer &)

Additional Inherited Members
Public Types inherited from Lima::InitializableObject< AbstractResource, ResourceInitializationParameters >
typedef InitializableObjectManager< AbstractResource, ResourceInitializationParameters >	Manager
	Manager is the type of the Manager associated to the initializableObject.

typedef InitializableObjectFactory< AbstractResource >	Factory
	Factory is the type of the Factory associated to the initializableObject.

Signals inherited from Lima::LinguisticProcessing::AbstractResourceBase
void	resourceFileChanged (const QString &path)
	Signal triggered whenever one of the registered files changes on disk.

Detailed Description

␈rief a class for the definition of a complete recognizer

this class is used for the description of a complete recognizer, i.e. a set of rules

Definition at line 77 of file recognizer.h.

Constructor & Destructor Documentation

◆ Recognizer() [1/2]

Lima::LinguisticProcessing::Automaton::Recognizer::Recognizer ( )

Definition at line 71 of file recognizer.cpp.

◆ Recognizer() [2/2]

Lima::LinguisticProcessing::Automaton::Recognizer::Recognizer ( const Recognizer & r )

Definition at line 82 of file recognizer.cpp.

◆ ~Recognizer()

Lima::LinguisticProcessing::Automaton::Recognizer::~Recognizer ( )

virtual

Definition at line 99 of file recognizer.cpp.

Member Function Documentation

◆ addRule() [1/2]

void Lima::LinguisticProcessing::Automaton::Recognizer::addRule	(	TransitionUnit *	trigger,
		const uint64_t	index
	)

associate a known rule to a trigger (the rule is identified by its index in the storage zone)

Parameters

trigger	the trigger
index	the index of the rule in the storage zone

Definition at line 1062 of file recognizer.cpp.

◆ addRule() [2/2]

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::addRule	(	TransitionUnit *	trigger,
		Rule *	rule
	)

add a rule in the recognizer : add the rule in the storage zone, and associate the trigger to the rule find if the trigger already exists in the trigger list : if it is the case, add the rule in rules associated to this trigger, else add a new trigger (copy of the one in argument) and associate the rule to it.

Parameters

trigger	the trigger of the rule
rule	the rule

Returns: the index of the rule in the storage zone

Definition at line 1043 of file recognizer.cpp.

◆ addRuleInStorage()

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::addRuleInStorage ( Rule * rule )

add a rule in the storage zone

Parameters

rule	the rule to add

Returns: the index of the rule in the storage zone

Definition at line 1035 of file recognizer.cpp.

◆ apply() [1/2]

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::apply	(	const LinguisticAnalysisStructure::AnalysisGraph &	graph,
		const LinguisticGraphVertex &	begin,
		const LinguisticGraphVertex &	end,
		AnalysisContent &	analysis,
		std::vector< RecognizerMatch > &	result,
		bool	testAllVertices = `false`,
		bool	stopAtFirstSuccess = `true`,
		bool	onlyOneSuccessPerType = `false`,
		bool	returnAtFirstSuccess = `false`,
		bool	applySameRuleWhileSuccess = `false`
	)		const

apply the recognizer on a graph

Parameters

graph	the graph
begin	the first node to test (is also limit of the left context)
end	the last node to test (is also limit of the right context)
analysis	the content of the analysis (may be modified)
result	the results
testAllVertices	if true, test all vertices, otherwise, skip recognized expressions
stopAtFirstSuccess	if true, stop testing rules on the current node after one rule succeeded
onlyOneSuccessPerType	if true, stop testing rules with same type as a previously successful rule (only used if stopAtFirstSuccess is false)
returnAtFirstSuccess	if true, abort the search as soon a rule is successful (if true, stopAtFirstSuccess will be set to true)
applySameRuleWhileSuccess	if true, when a rule succeeds, retry to apply it on same vertex until the rule does not apply (use with care: setting this argument to true may cause loops if* rules are not well written). Will not apply if stopAtFirstSuccess.

Returns: number of rules successfuly applied

Definition at line 615 of file recognizer.cpp.

◆ apply() [2/2]

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::apply	(	const LinguisticAnalysisStructure::AnalysisGraph &	graph,
		const LinguisticGraphVertex &	begin,
		const LinguisticGraphVertex &	end,
		const LinguisticGraphVertex &	upstreamBound,
		const LinguisticGraphVertex &	downstreamBound,
		AnalysisContent &	analysis,
		std::vector< RecognizerMatch > &	result,
		bool	testAllVertices,
		bool	stopAtFirstSuccess,
		bool	onlyOneSuccessPerType,
		bool	returnAtFirstSuccess,
		bool	applySameRuleWhileSuccess
	)		const

apply the recognizer on a graph between two nodes and search between two others.

precondition: [begin, end] included in [upstreamBound,downstreamBound]

Parameters

graph	the graph
begin	the first node to test (is not the limit of the left context)
end	the last node to test (is not the limit of the right context)
upstreamBound	the limit of the left context)
downstreamBound	the limit of the right context)
analysis	the content of the analysis (may be modified)
result	the results
testAllVertices	if true, test all vertices, otherwise, skip recognized expressions
stopAtFirstSuccess	if true, stop testing rules after one rule succeeded
onlyOneSuccessPerType	if true, stop testing rules with same type as a previously successful rule (only used if stopAtFirstSuccess is false)
returnAtFirstSuccess	if true, abort the search as soon a rule is successful (if true, stopAtFirstSuccess will be set to true)
applySameRuleWhileSuccess	if true, when a rule succeeds, retry to apply it on same vertex until the rule does not apply (use with care: setting this argument to true may cause loops if* rules are not well written). Will not apply if stopAtFirstSuccess.

Returns: number of rules successfuly applied

Definition at line 643 of file recognizer.cpp.

◆ clearSearchStructure()

void Lima::LinguisticProcessing::Automaton::Recognizer::clearSearchStructure ( )

protected

Definition at line 1028 of file recognizer.cpp.

◆ copy()

void Lima::LinguisticProcessing::Automaton::Recognizer::copy ( const Recognizer & a )

protected

Definition at line 250 of file recognizer.cpp.

◆ findNextSetOfRules()

void Lima::LinguisticProcessing::Automaton::Recognizer::findNextSetOfRules	(	const LinguisticAnalysisStructure::AnalysisGraph &	graph,
		LinguisticGraphVertex &	vertex,
		AnalysisContent &	analysis,
		const LinguisticAnalysisStructure::Token *	token,
		const LinguisticAnalysisStructure::MorphoSyntacticData *	data,
		std::vector< TriggerRule * > &	matchingSetOfRules
	)		const

protected

Definition at line 972 of file recognizer.cpp.

◆ findRuleIndex()

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::findRuleIndex ( Rule * r ) const

protected

Definition at line 1086 of file recognizer.cpp.

◆ freeMem()

void Lima::LinguisticProcessing::Automaton::Recognizer::freeMem ( )

protected

Definition at line 276 of file recognizer.cpp.

◆ init() [1/2]

void Lima::LinguisticProcessing::Automaton::Recognizer::init ( )

protected

Definition at line 242 of file recognizer.cpp.

◆ init() [2/2]

void Lima::LinguisticProcessing::Automaton::Recognizer::init	(	Common::XMLConfigurationFiles::GroupConfigurationStructure &	unitConfiguration,
		Manager *	manager
	)

overridevirtual

initialize with parameters from configuration file.

Parameters

unitConfiguration	@IN : <group> tag in xml configuration file that contains parameters to initialize the object.
manager	@IN : manager that asked for initialization and carries parameters.

Exceptions

InvalidConfiguration when parameters are invalids.

Implements Lima::LinguisticProcessing::AbstractResource.

◆ initializeSearchStructure()

void Lima::LinguisticProcessing::Automaton::Recognizer::initializeSearchStructure ( )

Definition at line 1018 of file recognizer.cpp.

◆ listTriggers()

void Lima::LinguisticProcessing::Automaton::Recognizer::listTriggers ( ) const

Definition at line 1101 of file recognizer.cpp.

◆ numberOfRules()

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::numberOfRules ( ) const

inline

Definition at line 373 of file recognizer.h.

◆ operator=()

Recognizer & Lima::LinguisticProcessing::Automaton::Recognizer::operator= ( const Recognizer & r )

Definition at line 108 of file recognizer.cpp.

◆ resolveOverlappingEntities()

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::resolveOverlappingEntities	(	std::vector< RecognizerMatch > &	listEntities,
		const OverlapResolutionStrategy &	strategy = `DEFAULT_OVERLAP_STRATEGY`
	)		const

resolve the problem of overlapping entities in the list of entities : when two entities are overlaping, only one is kept

Parameters

strategy the strategy to use : IGNORE_FIRST will simply ignore the first entity IGNORE_SECOND will simply ignore the second entity IGNORE_SMALLEST will ignore the entity that covers the less words

Returns: the number of overlapping entities removed

Definition at line 890 of file recognizer.cpp.

◆ setNormalizedForm()

void Lima::LinguisticProcessing::Automaton::Recognizer::setNormalizedForm	(	const LimaString &	norm,
		RecognizerMatch &	match
	)		const

protected

Definition at line 594 of file recognizer.cpp.

◆ testOnVertex()

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::testOnVertex	(	const LinguisticAnalysisStructure::AnalysisGraph &	graph,
		LinguisticGraphVertex &	current,
		const LinguisticGraphVertex &	begin,
		const LinguisticGraphVertex &	end,
		AnalysisContent &	analysis,
		std::vector< RecognizerMatch > &	result,
		bool	stopAtFirstSuccess = `true`,
		bool	onlyOneSuccessPerType = `false`,
		bool	applySameRuleWhileSuccess = `false`
	)		const

test the recognizer on a given vertex : check if this vertex is a trigger and if a rule applies, add the result to the vector given as parameter (takes the first rule which gives a match)

Parameters

graph	the graph
current	the current vertex to test
begin	the limit on the left of the current vertex
end	the limit on the right of the current vertex
result	the vector in which the result is stored

Returns: true if a match was found, false otherwise

Definition at line 814 of file recognizer.cpp.

◆ testSetOfRules()

uint64_t Lima::LinguisticProcessing::Automaton::Recognizer::testSetOfRules	(	const TransitionUnit &	trigger,
		const SetOfRules &	rules,
		const LinguisticAnalysisStructure::AnalysisGraph &	graph,
		const LinguisticGraphVertex &	position,
		const LinguisticGraphVertex &	begin,
		const LinguisticGraphVertex &	end,
		AnalysisContent &	analysis,
		std::vector< RecognizerMatch > &	match,
		std::set< Common::MediaticData::EntityType > *	forbiddenTypes = `0`,
		bool	stopAtFirstSuccess = `true`,
		bool	onlyOneSuccessPerType = `false`,
		bool	applySameRuleWhileSuccess = `false`
	)		const

TODO: toBeDeleted Parse tokens paths from the trigger point.

Parameters

triggerToken	: the token that triggered the rules
graph	: the graph of the analysis
current	: the vertex in the graph corresponding to the trigger
offset	: the offset of the set of rules to test
results	: the output

Returns: true if match, false otherwise bool parse(const LinguisticAnalysisStructure::Token& triggerToken, const LinguisticAnalysisStructure::AnalysisGraph& graph, const LinguisticGraphVertex& current, uint64_t offset, AnalysisContent& analysis, RecognizerMatch& result) const; test a given set of rules of the recognizer on a text, at a position given by a LinguisticAnalysisStructure::AnalysisGraph

Parameters

trigger	the TransitionUnit corresponding to the trigger
rules	the set of rules to test
position	the position in the text from with to test
match	the resulting match (if success)
forbiddenTypes	if not 0, contains the types currently not permitted for the current trigger (in case of negative rules); updated when leaving the function
stopAtFirstSuccess	if true, stop testing rules after one rule succeeded
applySameRuleWhileSuccess	if true, when a rule succeeds, retry to apply it on same vertex until the rule does not apply (use with care: setting this argument to true may cause loops if* rules are not well written). Will not apply if stopAtFirstSuccess.

Returns: true if there is a match : the reference arguments match and rule contain in this case the resulting match and the rule for wich the test has succeeded the match can be a MultipleMatch or a PIMultipleMatch

Definition at line 325 of file recognizer.cpp.

Friends And Related Symbol Documentation

◆ AutomatonReader

friend class AutomatonReader

friend

Definition at line 81 of file recognizer.h.

◆ AutomatonWriter

friend class AutomatonWriter

friend

Definition at line 82 of file recognizer.h.

◆ operator<<

LIMA_AUTOMATON_EXPORT std::ostream & operator<<	(	std::ostream &	os,
		const Recognizer &	r
	)

friend

Definition at line 1113 of file recognizer.cpp.

Member Data Documentation

◆ m_automatonControlParams

AutomatonControlParams Lima::LinguisticProcessing::Automaton::Recognizer::m_automatonControlParams

protected

parameters to control the search of the automaton

Definition at line 337 of file recognizer.h.

◆ m_filename

std::string Lima::LinguisticProcessing::Automaton::Recognizer::m_filename

protected

Definition at line 340 of file recognizer.h.

◆ m_language

MediaId Lima::LinguisticProcessing::Automaton::Recognizer::m_language

protected

Definition at line 334 of file recognizer.h.

◆ m_rules

std::vector<TriggerRule> Lima::LinguisticProcessing::Automaton::Recognizer::m_rules

protected

Definition at line 330 of file recognizer.h.

◆ m_ruleStorage

SetOfRules Lima::LinguisticProcessing::Automaton::Recognizer::m_ruleStorage

protected

Definition at line 331 of file recognizer.h.

◆ m_searchStructure

TransitionSearchStructure<TriggerRule> Lima::LinguisticProcessing::Automaton::Recognizer::m_searchStructure

protected

Definition at line 344 of file recognizer.h.

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

lima_linguisticprocessing/src/linguisticProcessing/core/Automaton/recognizer.h
lima_linguisticprocessing/src/linguisticProcessing/core/Automaton/recognizer.cpp

Classes

Public Member Functions

Protected Member Functions

Protected Attributes

Friends

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

◆ Recognizer() [1/2]

◆ Recognizer() [2/2]

◆ ~Recognizer()

Member Function Documentation

◆ addRule() [1/2]

◆ addRule() [2/2]

◆ addRuleInStorage()

◆ apply() [1/2]

◆ apply() [2/2]

◆ clearSearchStructure()

◆ copy()

◆ findNextSetOfRules()

◆ findRuleIndex()

◆ freeMem()

◆ init() [1/2]

◆ init() [2/2]

◆ initializeSearchStructure()

◆ listTriggers()

◆ numberOfRules()

◆ operator=()

◆ resolveOverlappingEntities()

◆ setNormalizedForm()

◆ testOnVertex()

◆ testSetOfRules()

Friends And Related Symbol Documentation

◆ AutomatonReader

◆ AutomatonWriter

◆ operator<<

Member Data Documentation

◆ m_automatonControlParams

◆ m_filename

◆ m_language

◆ m_rules

◆ m_ruleStorage

◆ m_searchStructure