michael@0: /*
michael@0: **********************************************************************
michael@0: * Copyright (C) 1999-2007, International Business Machines Corporation
michael@0: * and others. All Rights Reserved.
michael@0: **********************************************************************
michael@0: *   Date        Name        Description
michael@0: *   11/17/99    aliu        Creation.
michael@0: **********************************************************************
michael@0: */
michael@0: #ifndef RBT_DATA_H
michael@0: #define RBT_DATA_H
michael@0: 
michael@0: #include "unicode/utypes.h"
michael@0: #include "unicode/uclean.h"
michael@0: 
michael@0: #if !UCONFIG_NO_TRANSLITERATION
michael@0: 
michael@0: #include "unicode/uobject.h"
michael@0: #include "rbt_set.h"
michael@0: #include "hash.h"
michael@0: 
michael@0: U_NAMESPACE_BEGIN
michael@0: 
michael@0: class UnicodeFunctor;
michael@0: class UnicodeMatcher;
michael@0: class UnicodeReplacer;
michael@0: 
michael@0: /**
michael@0:  * The rule data for a RuleBasedTransliterators.  RBT objects hold
michael@0:  * a const pointer to a TRD object that they do not own.  TRD objects
michael@0:  * are essentially the parsed rules in compact, usable form.  The
michael@0:  * TRD objects themselves are held for the life of the process in
michael@0:  * a static cache owned by Transliterator.
michael@0:  *
michael@0:  * This class' API is a little asymmetric.  There is a method to
michael@0:  * define a variable, but no way to define a set.  This is because the
michael@0:  * sets are defined by the parser in a UVector, and the vector is
michael@0:  * copied into a fixed-size array here.  Once this is done, no new
michael@0:  * sets may be defined.  In practice, there is no need to do so, since
michael@0:  * generating the data and using it are discrete phases.  When there
michael@0:  * is a need to access the set data during the parse phase, another
michael@0:  * data structure handles this.  See the parsing code for more
michael@0:  * details.
michael@0:  */
michael@0: class TransliterationRuleData : public UMemory {
michael@0: 
michael@0: public:
michael@0: 
michael@0:     // PUBLIC DATA MEMBERS
michael@0: 
michael@0:     /**
michael@0:      * Rule table.  May be empty.
michael@0:      */
michael@0:     TransliterationRuleSet ruleSet;
michael@0: 
michael@0:     /**
michael@0:      * Map variable name (String) to variable (UnicodeString).  A variable name
michael@0:      * corresponds to zero or more characters, stored in a UnicodeString in
michael@0:      * this hash.  One or more of these chars may also correspond to a
michael@0:      * UnicodeMatcher, in which case the character in the UnicodeString in this hash is
michael@0:      * a stand-in: it is an index for a secondary lookup in
michael@0:      * data.variables.  The stand-in also represents the UnicodeMatcher in
michael@0:      * the stored rules.
michael@0:      */
michael@0:     Hashtable variableNames;
michael@0: 
michael@0:     /**
michael@0:      * Map category variable (UChar) to set (UnicodeFunctor).
michael@0:      * Variables that correspond to a set of characters are mapped
michael@0:      * from variable name to a stand-in character in data.variableNames.
michael@0:      * The stand-in then serves as a key in this hash to lookup the
michael@0:      * actual UnicodeFunctor object.  In addition, the stand-in is
michael@0:      * stored in the rule text to represent the set of characters.
michael@0:      * variables[i] represents character (variablesBase + i).
michael@0:      */
michael@0:     UnicodeFunctor** variables;
michael@0: 
michael@0:     /**
michael@0:      * Flag that indicates whether the variables are owned (if a single
michael@0:      * call to Transliterator::createFromRules() produces a CompoundTransliterator
michael@0:      * with more than one RuleBasedTransliterator as children, they all share
michael@0:      * the same variables list, so only the first one is considered to own
michael@0:      * the variables)
michael@0:      */
michael@0:     UBool variablesAreOwned;
michael@0: 
michael@0:     /**
michael@0:      * The character that represents variables[0].  Characters
michael@0:      * variablesBase through variablesBase +
michael@0:      * variablesLength - 1 represent UnicodeFunctor objects.
michael@0:      */
michael@0:     UChar variablesBase;
michael@0: 
michael@0:     /**
michael@0:      * The length of variables.
michael@0:      */
michael@0:     int32_t variablesLength;
michael@0: 
michael@0: public:
michael@0: 
michael@0:     /**
michael@0:      * Constructor
michael@0:      * @param status Output param set to success/failure code on exit.
michael@0:      */
michael@0:     TransliterationRuleData(UErrorCode& status);
michael@0: 
michael@0:     /**
michael@0:      * Copy Constructor
michael@0:      */
michael@0:     TransliterationRuleData(const TransliterationRuleData&);
michael@0: 
michael@0:     /**
michael@0:      * destructor
michael@0:      */
michael@0:     ~TransliterationRuleData();
michael@0: 
michael@0:     /**
michael@0:      * Given a stand-in character, return the UnicodeFunctor that it
michael@0:      * represents, or NULL if it doesn't represent anything.
michael@0:      * @param standIn    the given stand-in character.
michael@0:      * @return           the UnicodeFunctor that 'standIn' represents
michael@0:      */
michael@0:     UnicodeFunctor* lookup(UChar32 standIn) const;
michael@0: 
michael@0:     /**
michael@0:      * Given a stand-in character, return the UnicodeMatcher that it
michael@0:      * represents, or NULL if it doesn't represent anything or if it
michael@0:      * represents something that is not a matcher.
michael@0:      * @param standIn    the given stand-in character.
michael@0:      * @return           return the UnicodeMatcher that 'standIn' represents
michael@0:      */
michael@0:     UnicodeMatcher* lookupMatcher(UChar32 standIn) const;
michael@0: 
michael@0:     /**
michael@0:      * Given a stand-in character, return the UnicodeReplacer that it
michael@0:      * represents, or NULL if it doesn't represent anything or if it
michael@0:      * represents something that is not a replacer.
michael@0:      * @param standIn    the given stand-in character.
michael@0:      * @return           return the UnicodeReplacer that 'standIn' represents
michael@0:      */
michael@0:     UnicodeReplacer* lookupReplacer(UChar32 standIn) const;
michael@0: 
michael@0: 
michael@0: private:
michael@0:     TransliterationRuleData &operator=(const TransliterationRuleData &other); // forbid copying of this class
michael@0: };
michael@0: 
michael@0: U_NAMESPACE_END
michael@0: 
michael@0: #endif /* #if !UCONFIG_NO_TRANSLITERATION */
michael@0: 
michael@0: #endif