RuleBasedTransliterator. This set encodes
* the transliteration in one direction from one set of characters or short
* strings to another. A RuleBasedTransliterator consists of up to
* two such sets, one for the forward direction, and one for the reverse.
*
* A TransliterationRuleSet has one important operation, that of
* finding a matching rule at a given point in the text. This is accomplished
* by the findMatch() method.
*
* @author Alan Liu
*/
class TransliterationRuleSet {
/**
* Vector of rules, in the order added.
*/
UVector rules;
/**
* Length of the longest preceding context
*/
int32_t maxContextLength;
public:
/**
* Construct a new empty rule set.
*/
TransliterationRuleSet();
/**
* Return the maximum context length.
* @return the length of the longest preceding context.
*/
virtual int32_t getMaximumContextLength() const;
/**
* Add a rule to this set. Rules are added in order, and order is
* significant.
*
*
Once freeze() is called, this method must not be called.
* @param rule the rule to add
*/
virtual void addRule(TransliterationRule* adoptedRule,
UErrorCode& status);
/**
* Free up space. Once this method is called, addRule() must NOT
* be called again.
*/
virtual void freeze();
/**
* Attempt to find a matching rule at the specified point in the text. The
* text being matched occupies a virtual buffer consisting of the contents
* of result concatenated to a substring of text.
* The substring is specified by start and limit.
* The value of cursor is an index into this virtual buffer,
* from 0 to the length of the buffer. In terms of the parameters,
* cursor must be between 0 and result.length() + limit -
* start.
* @param text the untranslated text
* @param start the beginning index, inclusive; 0 <= start
* <= limit.
* @param limit the ending index, exclusive; start <= limit
* <= text.length().
* @param result tranlated text
* @param cursor position at which to translate next, an offset into result.
* If greater than or equal to result.length(), represents offset start +
* cursor - result.length() into text.
* @param data a dictionary mapping variables to the sets they
* represent (maps Character to UnicodeSet)
* @param filter the filter. Any character for which
* filter.isIn() returns false will not be
* altered by this transliterator. If filter is
* null then no filtering is applied.
* @return the matching rule, or null if none found.
*/
virtual TransliterationRule* findMatch(const UnicodeString& text,
int32_t start, int32_t limit,
const UnicodeString& result,
int32_t cursor,
const TransliterationRuleData& data,
const UnicodeFilter* filter) const;
/**
* Attempt to find a matching rule at the specified point in the text.
* @param text the text, both translated and untranslated
* @param start the beginning index, inclusive; 0 <= start
* <= limit.
* @param limit the ending index, exclusive; start <= limit
* <= text.length().
* @param cursor position at which to translate next, representing offset
* into text. This value must be between start and
* limit.
* @param data a dictionary mapping variables to the sets they
* represent (maps Character to UnicodeSet)
* @param filter the filter. Any character for which
* filter.isIn() returns false will not be
* altered by this transliterator. If filter is
* null then no filtering is applied.
* @return the matching rule, or null if none found.
*/
virtual TransliterationRule* findMatch(const Replaceable& text,
int32_t start, int32_t limit,
int32_t cursor,
const TransliterationRuleData& data,
const UnicodeFilter* filter) const;
/**
* Attempt to find a matching rule at the specified point in the text.
* Unlike findMatch(), this method does an incremental match.
* An incremental match requires that there be no partial matches that might
* pre-empt the full match that is found. If there are partial matches,
* then null is returned. A non-null result indicates that a full match has
* been found, and that it cannot be pre-empted by a partial match
* regardless of what additional text is added to the translation buffer.
* @param text the text, both translated and untranslated
* @param start the beginning index, inclusive; 0 <= start
* <= limit.
* @param limit the ending index, exclusive; start <= limit
* <= text.length().
* @param cursor position at which to translate next, representing offset
* into text. This value must be between start and
* limit.
* @param data a dictionary mapping variables to the sets they
* represent (maps Character to UnicodeSet)
* @param partial output parameter. partial[0] is set to
* true if a partial match is returned.
* @param filter the filter. Any character for which
* filter.isIn() returns false will not be
* altered by this transliterator. If filter is
* null then no filtering is applied.
* @return the matching rule, or null if none found, or if the text buffer
* does not have enough text yet to unambiguously match a rule.
*/
virtual TransliterationRule* findIncrementalMatch(const Replaceable& text,
int32_t start,
int32_t limit, int32_t cursor,
const TransliterationRuleData& data,
bool_t& isPartial,
const UnicodeFilter* filter) const;
};
#endif