Update ICU to 54.1 step 1

source/common/dictionarydata.cpp

 /*
 *******************************************************************************
-* Copyright (C) 2013, International Business Machines
+* Copyright (C) 2014, International Business Machines
 * Corporation and others.  All Rights Reserved.
 *******************************************************************************
 * dictionarydata.h
 *
 * created on: 2012may31
 * created by: Markus W. Scherer & Maxime Serrano
 */
 
 #include "dictionarydata.h"
 #include "unicode/ucharstrie.h"
@@ skipping 19 matching lines @@
 }
 
 UCharsDictionaryMatcher::~UCharsDictionaryMatcher() {
     udata_close(file);
 }
 
 int32_t UCharsDictionaryMatcher::getType() const {
     return DictionaryData::TRIE_TYPE_UCHARS;
 }
 
-int32_t UCharsDictionaryMatcher::matches(UText *text, int32_t maxLength, int32_t *lengths, int32_t &count, int32_t limit, int32_t *values) const {
+int32_t UCharsDictionaryMatcher::matches(UText *text, int32_t maxLength, int32_t limit,
+                            int32_t *lengths, int32_t *cpLengths, int32_t *values,
+                            int32_t *prefix) const {
+
     UCharsTrie uct(characters);
-    UChar32 c = utext_next32(text);
-    if (c < 0) {
-        return 0;
-    }
-    UStringTrieResult result = uct.first(c);
-    int32_t numChars = 1;
-    count = 0;
-    for (;;) {
+    int32_t startingTextIndex = utext_getNativeIndex(text);
+    int32_t wordCount = 0;
+    int32_t codePointsMatched = 0;
+
+    for (UChar32 c = utext_next32(text); c >= 0; c=utext_next32(text)) {
+        UStringTrieResult result = (codePointsMatched == 0) ? uct.first(c) : uct.next(c);
+        int32_t lengthMatched = utext_getNativeIndex(text) - startingTextIndex;
+        codePointsMatched += 1;
         if (USTRINGTRIE_HAS_VALUE(result)) {
-            if (count < limit) {
+            if (wordCount < limit) {
                 if (values != NULL) {
-                    values[count] = uct.getValue();
+                    values[wordCount] = uct.getValue();
                 }
-                lengths[count++] = numChars;
+                if (lengths != NULL) {
+                    lengths[wordCount] = lengthMatched;
+                }
+                if (cpLengths != NULL) {
+                    cpLengths[wordCount] = codePointsMatched;
+                }
+                ++wordCount;
             }
             if (result == USTRINGTRIE_FINAL_VALUE) {
                 break;
             }
         }
         else if (result == USTRINGTRIE_NO_MATCH) {
             break;
         }
-
-        // TODO: why do we have a text limit if the UText knows its length?
-        if (numChars >= maxLength) {
+        if (lengthMatched >= maxLength) {
             break;
         }
+    }
 
-        c = utext_next32(text);
-        if (c < 0) {
-            break;
-        }
-        ++numChars;
-        result = uct.next(c);
+    if (prefix != NULL) {
+        *prefix = codePointsMatched;
     }
-    return numChars;
+    return wordCount;
 }
 
 BytesDictionaryMatcher::~BytesDictionaryMatcher() {
     udata_close(file);
 }
 
 UChar32 BytesDictionaryMatcher::transform(UChar32 c) const {
     if ((transformConstant & DictionaryData::TRANSFORM_TYPE_MASK) == DictionaryData::TRANSFORM_TYPE_OFFSET) {
         if (c == 0x200D) {
             return 0xFF;
         } else if (c == 0x200C) {
             return 0xFE;
         }
         int32_t delta = c - (transformConstant & DictionaryData::TRANSFORM_OFFSET_MASK);
         if (delta < 0 || 0xFD < delta) {
             return U_SENTINEL;
         }
         return (UChar32)delta;
     }
     return c;
 }
 
 int32_t BytesDictionaryMatcher::getType() const {
     return DictionaryData::TRIE_TYPE_BYTES;
 }
 
-int32_t BytesDictionaryMatcher::matches(UText *text, int32_t maxLength, int32_t *lengths, int32_t &count, int32_t limit, int32_t *values) const {
+int32_t BytesDictionaryMatcher::matches(UText *text, int32_t maxLength, int32_t limit,
+                            int32_t *lengths, int32_t *cpLengths, int32_t *values,
+                            int32_t *prefix) const {
     BytesTrie bt(characters);
-    UChar32 c = utext_next32(text);
-    if (c < 0) {
-        return 0;
-    }
-    UStringTrieResult result = bt.first(transform(c));
-    int32_t numChars = 1;
-    count = 0;
-    for (;;) {
+    int32_t startingTextIndex = utext_getNativeIndex(text);
+    int32_t wordCount = 0;
+    int32_t codePointsMatched = 0;
+
+    for (UChar32 c = utext_next32(text); c >= 0; c=utext_next32(text)) {
+        UStringTrieResult result = (codePointsMatched == 0) ? bt.first(transform(c)) : bt.next(transform(c));
+        int32_t lengthMatched = utext_getNativeIndex(text) - startingTextIndex;
+        codePointsMatched += 1;
         if (USTRINGTRIE_HAS_VALUE(result)) {
-            if (count < limit) {
+            if (wordCount < limit) {
                 if (values != NULL) {
-                    values[count] = bt.getValue();
-            }
-                lengths[count++] = numChars;
+                    values[wordCount] = bt.getValue();
+                }
+                if (lengths != NULL) {
+                    lengths[wordCount] = lengthMatched;
+                }
+                if (cpLengths != NULL) {
+                    cpLengths[wordCount] = codePointsMatched;
+                }
+                ++wordCount;
             }
             if (result == USTRINGTRIE_FINAL_VALUE) {
                 break;
             }
         }
         else if (result == USTRINGTRIE_NO_MATCH) {
             break;
         }
-
-        // TODO: why do we have a text limit if the UText knows its length?
-        if (numChars >= maxLength) {
+        if (lengthMatched >= maxLength) {
             break;
         }
+    }
 
-        c = utext_next32(text);
-        if (c < 0) {
-            break;
-        }
-        ++numChars;
-        result = bt.next(transform(c));
+    if (prefix != NULL) {
+        *prefix = codePointsMatched;
     }
-    return numChars;
+    return wordCount;
 }
 
 
 U_NAMESPACE_END
 
 U_NAMESPACE_USE
 
 U_CAPI int32_t U_EXPORT2
 udict_swap(const UDataSwapper *ds, const void *inData, int32_t length,
            void *outData, UErrorCode *pErrorCode) {
@@ skipping 69 matching lines @@
         // but may be used later.
         offset = nextOffset;
         nextOffset = indexes[DictionaryData::IX_RESERVED2_OFFSET];
         offset = nextOffset;
         nextOffset = indexes[DictionaryData::IX_TOTAL_SIZE];
         offset = nextOffset;
     }
     return headerSize + size;
 }
 #endif