Class analyzer is a meta class that just calls all modules in FreeLing in the right order. More...

#include <analyzer.h>

Collaboration diagram for freeling::analyzer:

Classes
class	analyzer_config_options
	Class analyzer::config_options contains the configuration options that define which modules are active and which configuration files are loaded for each of them at construction time. More...
class	analyzer_invoke_options
	Class analyzer::invoke_options contains the options that define the behaviour of each module in the analyze on the next analysis. More...
Public Types
typedef analyzer_config_options	config_options
typedef analyzer_invoke_options	invoke_options
Public Member Functions
	analyzer (const config_options &cfg)
void	set_current_invoke_options (const invoke_options &opt, bool check=true)
const invoke_options &	get_current_invoke_options () const
	~analyzer ()
void	analyze (document &doc) const
	analyze further levels on a partially analyzed document
void	analyze (std::list< sentence > &ls) const
	analyze further levels on partially analyzed sentences
void	analyze (const wstring &text, document &doc, bool parag=false) const
	analyze text as a whole document
void	analyze (const wstring &text, std::list< sentence > &ls, bool flush=false)
	Analyze text as a partial document.
void	flush_buffer (std::list< sentence > &ls)
void	reset_offset ()
Private Member Functions
template<class T >
void	do_analysis (T &doc) const
	analyze further levels on a partially analyzed document
void	tokenize_split (const std::wstring &text, std::list< sentence > &ls, unsigned long &offs, std::list< word > &av, unsigned long &nsent, bool flush, splitter::session_id sp_ses) const
Private Attributes
tokenizer *	tk
splitter *	sp
maco *	morfo
nec *	neclass
senses *	sens
ukb *	dsb
POS_tagger *	hmm
POS_tagger *	relax
phonetics *	phon
chart_parser *	parser
dep_txala *	deptxala
dep_treeler *	deptreeler
relaxcor *	corfc
semgraph_extract *	sge
analyzer_invoke_options	current_invoke_options
splitter::session_id	sp_id
unsigned long	offs
unsigned long	nsentence
std::list< word >	tokens

Detailed Description

Class analyzer is a meta class that just calls all modules in FreeLing in the right order.

Its construction options allow to instantiate different kinds of analysis pipelines, and or different languages. Also, invocation options may be altered at each call, tuning the analysis to each particular sentence or document needs. For a finer control, underlying modules should be called directly.

Member Typedef Documentation

typedef analyzer_config_options freeling::analyzer::config_options

typedef analyzer_invoke_options freeling::analyzer::invoke_options

Constructor & Destructor Documentation

freeling::analyzer::analyzer ( const config_options & cfg )

--------------------------------------------- Create analyzers defined by config_options. ---------------------------------------------

freeling::analyzer::~analyzer ( )

Member Function Documentation

void freeling::analyzer::analyze ( document & doc ) const

analyze further levels on a partially analyzed document

References freeling::COREF, and freeling::SEMGRAPH.

void freeling::analyzer::analyze ( std::list< sentence > & ls ) const

analyze further levels on partially analyzed sentences

void freeling::analyzer::analyze	(	const wstring &	text,
		document &	doc,
		bool	parag = `false`
	)		const

analyze text as a whole document

void freeling::analyzer::analyze	(	const wstring &	text,
		std::list< sentence > &	ls,
		bool	flush = `false`
	)

Analyze text as a partial document.

Retain incomplete sentences in buffer in case next call completes them (except if flush==true)

template<class T >

void freeling::analyzer::do_analysis ( T & doc ) const [private]

analyze further levels on a partially analyzed document

References freeling::COREF, freeling::DEP, freeling::HMM, freeling::MORFO, freeling::NO_WSD, freeling::PARSED, freeling::RELAX, freeling::SHALLOW, freeling::TAGGED, freeling::TREELER, freeling::TXALA, and freeling::UKB.

void freeling::analyzer::flush_buffer ( std::list< sentence > & ls )

const analyzer::invoke_options & freeling::analyzer::get_current_invoke_options ( ) const

void freeling::analyzer::reset_offset ( )

void freeling::analyzer::set_current_invoke_options	(	const invoke_options &	opt,
		bool	check = `true`
	)

void freeling::analyzer::tokenize_split	(	const std::wstring &	text,
		std::list< sentence > &	ls,
		unsigned long &	offs,
		std::list< word > &	av,
		unsigned long &	nsent,
		bool	flush,
		splitter::session_id	sp_ses
	)		const `[private]`