Home | History | Annotate | Download | only in gennorm2 
      1 //  2017 and later: Unicode, Inc. and others.
      2 // License & terms of use: http://www.unicode.org/copyright.html
      3 
      4 // extradata.h
      5 // created: 2017jun04 Markus W. Scherer
      6 // (pulled out of n2builder.cpp)
      7 
      8 // Write mappings and compositions in compact form for Normalizer2 "extra data",
      9 // the data that does not fit into the trie itself.
     10 
     11 #ifndef __EXTRADATA_H__
     12 #define __EXTRADATA_H__
     13 
     14 #include "unicode/utypes.h"
     15 
     16 #if !UCONFIG_NO_NORMALIZATION
     17 
     18 #include "unicode/errorcode.h"
     19 #include "unicode/unistr.h"
     20 #include "unicode/utf16.h"
     21 #include "hash.h"
     22 #include "norms.h"
     23 #include "toolutil.h"
     24 #include "utrie2.h"
     25 #include "uvectr32.h"
     26 
     27 U_NAMESPACE_BEGIN
     28 
     29 class ExtraData : public Norms::Enumerator {
     30 public:
     31     ExtraData(Norms &n, UBool fast);
     32 
     33     void rangeHandler(UChar32 start, UChar32 end, Norm &norm) U_OVERRIDE;
     34 
     35     UnicodeString maybeYesCompositions;
     36     UnicodeString yesYesCompositions;
     37     UnicodeString yesNoMappingsAndCompositions;
     38     UnicodeString yesNoMappingsOnly;
     39     UnicodeString noNoMappingsCompYes;
     40     UnicodeString noNoMappingsCompBoundaryBefore;
     41     UnicodeString noNoMappingsCompNoMaybeCC;
     42     UnicodeString noNoMappingsEmpty;
     43 
     44 private:
     45     /**
     46      * Requires norm.hasMapping().
     47      * Returns the offset of the "first unit" from the beginning of the extraData for c.
     48      * That is the same as the length of the optional data
     49      * for the raw mapping and the ccc/lccc word.
     50      */
     51     int32_t writeMapping(UChar32 c, const Norm &norm, UnicodeString &dataString);
     52     int32_t writeNoNoMapping(UChar32 c, const Norm &norm,
     53                              UnicodeString &dataString, Hashtable &previousMappings);
     54     UBool setNoNoDelta(UChar32 c, Norm &norm) const;
     55     /** Requires norm.compositions!=nullptr. */
     56     void writeCompositions(UChar32 c, const Norm &norm, UnicodeString &dataString);
     57     void writeExtraData(UChar32 c, Norm &norm);
     58 
     59     UBool optimizeFast;
     60     Hashtable previousNoNoMappingsCompYes;  // If constructed in runtime code, pass in UErrorCode.
     61     Hashtable previousNoNoMappingsCompBoundaryBefore;
     62     Hashtable previousNoNoMappingsCompNoMaybeCC;
     63     Hashtable previousNoNoMappingsEmpty;
     64 };
     65 
     66 U_NAMESPACE_END
     67 
     68 #endif // #if !UCONFIG_NO_NORMALIZATION
     69 
     70 #endif  // __EXTRADATA_H__
     71