ICU-11886 unsafe-backward set, review comments

. remove #if 0's and stray comments
. cleanup UnicodeSet docs (deserialize does not freeze)
. remove dead code in collationdatareader.cpp

X-SVN-Rev: 37960

diff --git a/icu4c/source/common/unicode/uniset.h b/icu4c/source/common/unicode/uniset.h
index e12fe2944572b11272a6d32ee618750040cc0ac6..0a3aeb79782cfccffe1c922fc16b79a9f653aaef 100644 (file)
--- a/icu4c/source/common/unicode/uniset.h
+++ b/icu4c/source/common/unicode/uniset.h
@@ -379,7 +379,6 @@ public:
 
     /**
      * Constructs a set from the output of serialize().
-     * The resulting set will be frozen.
      *
      * @param buffer the 16 bit array
      * @param bufferLen the original length returned from serialize()
@@ -391,7 +390,7 @@ public:
     UnicodeSet(const uint16_t buffer[], int32_t bufferLen,
                ESerialization serialization, UErrorCode &status);
 #endif
-    
+
     /**
      * Constructs a set from the given pattern.  See the class
      * description for the syntax of the pattern language.

diff --git a/icu4c/source/common/uniset.cpp b/icu4c/source/common/uniset.cpp
index 260c17168b9b8ad16200c83a1580746e86b2c2f1..a305ad41292a7b04a1f9695f419d94f9af3e8db8 100644 (file)
--- a/icu4c/source/common/uniset.cpp
+++ b/icu4c/source/common/uniset.cpp
@@ -1474,7 +1474,6 @@ UnicodeSet& UnicodeSet::compact() {
 
 /**
  * Deserialize constructor.
- * Result will be frozen.
  */
 UnicodeSet::UnicodeSet(const uint16_t data[], int32_t dataLen, ESerialization serialization, UErrorCode &ec)
   : len(1), capacity(1+START_EXTRA), list(0), bmpSet(0), buffer(0),
@@ -1499,7 +1498,7 @@ UnicodeSet::UnicodeSet(const uint16_t data[], int32_t dataLen, ESerialization se
     setToBogus();
     return;
   }
-  
+
   // bmp?
   int32_t headerSize = ((data[0]&0x8000)) ?2:1;
   int32_t bmpLength = (headerSize==1)?data[0]:data[1];
@@ -1522,15 +1521,6 @@ UnicodeSet::UnicodeSet(const uint16_t data[], int32_t dataLen, ESerialization se
     printf("<<16@%d[%d] %X\n", i+headerSize, i, list[i]);
 #endif
   }
-  /*
-  if(bmpLength>0) {
-    bmpSet= new BMPSet(list, bmpLength);
-    if(bmpSet == NULL) {
-      ec = U_MEMORY_ALLOCATION_ERROR;
-      setToBogus();
-      return;
-    }
-    }*/
   // copy smp
   for(i=bmpLength;i<len;i++) {
     list[i] = ((UChar32)data[headerSize+bmpLength+(i-bmpLength)*2+0] << 16) +

diff --git a/icu4c/source/i18n/collationdatareader.cpp b/icu4c/source/i18n/collationdatareader.cpp
index 63be9c91064388b7c138df0114038621cd12fb33..f7098f984199dda286b5843270e3d9c0f5554247 100644 (file)
--- a/icu4c/source/i18n/collationdatareader.cpp
+++ b/icu4c/source/i18n/collationdatareader.cpp
@@ -25,18 +25,12 @@
 #include "collationrootelements.h"
 #include "collationsettings.h"
 #include "collationtailoring.h"
+#include "collunsafe.h"
 #include "normalizer2impl.h"
 #include "uassert.h"
 #include "ucmndata.h"
 #include "utrie2.h"
 
-// #if U_HAVE_COLLUNSAFE
-#if 1
-#include "collunsafe.h"
-
-#endif
-// #end
-
 U_NAMESPACE_BEGIN
 
 namespace {
@@ -269,16 +263,7 @@ CollationDataReader::read(const CollationTailoring *base, const uint8_t *inBytes
             return;
         }
         if(baseData == NULL) {
-#if defined( COLLUNSAFE_COLL_VERSION )
-#if defined(COLLUNSAFE_RANGES)  /* slower but still an improvement*/
-          tailoring.unsafeBackwardSet = new UnicodeSet();
-
-          for(int32_t i=0;i<unsafe_rangeCount*2;i+=2) {
-            tailoring.unsafeBackwardSet->add(unsafe_ranges[i+0],unsafe_ranges[i+1]);
-          }
-          tailoring.unsafeBackwardSet->freeze();
-#elif defined (COLLUNSAFE_SERIALIZE)
-          /* faster */
+#if defined(COLLUNSAFE_COLL_VERSION) && defined (COLLUNSAFE_SERIALIZE)
           tailoring.unsafeBackwardSet = new UnicodeSet(unsafe_serializedData, unsafe_serializedCount, UnicodeSet::kSerialized, errorCode);
           if(tailoring.unsafeBackwardSet == NULL) {
             errorCode = U_MEMORY_ALLOCATION_ERROR;
@@ -286,10 +271,6 @@ CollationDataReader::read(const CollationTailoring *base, const uint8_t *inBytes
           } else if (U_FAILURE(errorCode)) {
             return;
           }
-#else
-#error no unsafe-backwards strategy chosen
-#endif
-
 #else
             // Create the unsafe-backward set for the root collator.
             // Include all non-zero combining marks and trail surrogates.
@@ -308,7 +289,7 @@ CollationDataReader::read(const CollationTailoring *base, const uint8_t *inBytes
                 return;
             }
             data->nfcImpl.addLcccChars(*tailoring.unsafeBackwardSet);
-#endif
+#endif // !COLLUNSAFE_SERIALIZE || !COLLUNSAFE_COLL_VERSION
         } else {
             // Clone the root collator's set contents.
             tailoring.unsafeBackwardSet = static_cast<UnicodeSet *>(

diff --git a/icu4c/source/test/intltest/usettest.cpp b/icu4c/source/test/intltest/usettest.cpp
index f38311237d11dfca85fb335f074461c76625f959..46e502ea826d58ff02eddcb669c32b595b85cd53 100644 (file)
--- a/icu4c/source/test/intltest/usettest.cpp
+++ b/icu4c/source/test/intltest/usettest.cpp
@@ -1966,17 +1966,6 @@ void UnicodeSetTest::checkSerializeRoundTrip(const UnicodeSet& t, UErrorCode &st
     errln("checkSerializeRoundTrip: error %s serializing buffer\n", u_errorName(status));
     return;
   }
-
-#if 0
-  UnicodeString pat; t.toPattern(pat, FALSE);
-  infoln(pat);
-  printf(" %d: ", len);
-  for(int i=0;i<len;i++) {
-    printf( " %04X ", serializeBuffer.getAlias()[i]);
-  }
-  printf("\n");
-#endif
-  
   UnicodeSet deserialized(serializeBuffer.getAlias(), len, UnicodeSet::kSerialized, status);
   if(U_FAILURE(status)) {
     errln("checkSerializeRoundTrip: error %s deserializing buffer: buf %p len %d, original %d\n", u_errorName(status), serializeBuffer.getAlias(), len, t.getRangeCount());

diff --git a/icu4c/source/tools/gencolusb/extract_unsafe_backwards.cpp b/icu4c/source/tools/gencolusb/extract_unsafe_backwards.cpp
index b680f79e6e0ba5a33ac343ba366eb9055419fc1b..72e55c6956593438f5c40e2ee05d5b811ba63675 100644 (file)
--- a/icu4c/source/tools/gencolusb/extract_unsafe_backwards.cpp
+++ b/icu4c/source/tools/gencolusb/extract_unsafe_backwards.cpp
@@ -15,9 +15,9 @@
 /**
  * Define the type of generator to use. Choose one.
  */
-#define SERIALIZE 1   //< Default: use UnicodeSet.serialize() (best, fast, requires new UnicodeSet c'tor)
-#define RANGES 0      //< Enumerate ranges (works, not as fast)
-#define PATTERN 0     //< Generate a UnicodeSet pattern (broken AND probably slow)
+#define SERIALIZE 1   //< Default: use UnicodeSet.serialize() and a new internal c'tor
+#define RANGES 0      //< Enumerate ranges (works, not as fast. No support in collationdatareader.cpp)
+#define PATTERN 0     //< Generate a UnicodeSet pattern (depends on #11891 AND probably slower. No support in collationdatareader.cpp)
 
 int main(int argc, const char *argv[]) {
     UErrorCode errorCode = U_ZERO_ERROR;
@@ -57,7 +57,7 @@ int main(int argc, const char *argv[]) {
 #endif
     
 #if PATTERN
-    fprintf(stderr,".. pattern\n");
+    fprintf(stderr,".. pattern. (Note: collationdatareader.cpp does not support this form also see #11891)\n");
     // attempt to use pattern
     
     UnicodeString pattern;
@@ -128,6 +128,7 @@ int main(int argc, const char *argv[]) {
 #endif
 
 #if RANGE
+    fprintf(stderr, "COLLUNSAFE_RANGE - no code support in collationdatareader.cpp for this\n");
     printf("#define COLLUNSAFE_RANGE 1\n");
     printf("static const int32_t unsafe_rangeCount = %d;\n", rangeCount);
     printf("static const UChar32 unsafe_ranges[%d] = { \n", rangeCount*2);