X-Git-Url: https://git.saurik.com/apple/icu.git/blobdiff_plain/f3c0d7a59d99c2a94c6b8822291f0e42be3773c9..340931cb2e044a2141d11567dd0f782524e32994:/icuSources/common/unicode/utf16.h

diff --git a/icuSources/common/unicode/utf16.h b/icuSources/common/unicode/utf16.h
index 9cc73f7b..f3bdb823 100644
--- a/icuSources/common/unicode/utf16.h
+++ b/icuSources/common/unicode/utf16.h
@@ -163,7 +163,7 @@
  * @see U16_GET
  * @stable ICU 2.4
  */
-#define U16_GET_UNSAFE(s, i, c) { \
+#define U16_GET_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
     (c)=(s)[i]; \
     if(U16_IS_SURROGATE(c)) { \
         if(U16_IS_SURROGATE_LEAD(c)) { \
@@ -172,7 +172,7 @@
             (c)=U16_GET_SUPPLEMENTARY((s)[(i)-1], (c)); \
         } \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Get a code point from a string at a random-access offset,
@@ -185,8 +185,8 @@
  *
  * The length can be negative for a NUL-terminated string.
  *
- * If the offset points to a single, unpaired surrogate, then that itself
- * will be returned as the code point.
+ * If the offset points to a single, unpaired surrogate, then
+ * c is set to that unpaired surrogate.
  * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT.
  *
  * @param s const UChar * string
@@ -197,21 +197,64 @@
  * @see U16_GET_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_GET(s, start, i, length, c) { \
+#define U16_GET(s, start, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+    if(U16_IS_SURROGATE(c)) { \
+        uint16_t __c2 = 0; \
+        if(U16_IS_SURROGATE_LEAD(c)) { \
+            if((i)+1!=(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \
+                (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
+            } \
+        } else { \
+            if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+                (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The offset may point to either the lead or trail surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the adjacent matching surrogate as well.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * If the offset points to a single, unpaired surrogate, then
+ * c is set to U+FFFD.
+ * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT_OR_FFFD.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<=i<length
+ * @param length string length
+ * @param c output UChar32 variable
+ * @see U16_GET_UNSAFE
+ * @stable ICU 60
+ */
+#define U16_GET_OR_FFFD(s, start, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
     (c)=(s)[i]; \
     if(U16_IS_SURROGATE(c)) { \
-        uint16_t __c2; \
+        uint16_t __c2 = 0; \
         if(U16_IS_SURROGATE_LEAD(c)) { \
             if((i)+1!=(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \
                 (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
+            } else { \
+                (c)=0xfffd; \
             } \
         } else { \
             if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
                 (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+            } else { \
+                (c)=0xfffd; \
             } \
         } \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /* definitions with forward iteration --------------------------------------- */
 
@@ -234,12 +277,12 @@
  * @see U16_NEXT
  * @stable ICU 2.4
  */
-#define U16_NEXT_UNSAFE(s, i, c) { \
+#define U16_NEXT_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
     (c)=(s)[(i)++]; \
     if(U16_IS_LEAD(c)) { \
         (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)++]); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Get a code point from a string at a code point boundary offset,
@@ -253,8 +296,7 @@
  * for a supplementary code point, in which case the macro will read
  * the following trail surrogate as well.
  * If the offset points to a trail surrogate or
- * to a single, unpaired lead surrogate, then that itself
- * will be returned as the code point.
+ * to a single, unpaired lead surrogate, then c is set to that unpaired surrogate.
  *
  * @param s const UChar * string
  * @param i string offset, must be i<length
@@ -263,16 +305,50 @@
  * @see U16_NEXT_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_NEXT(s, i, length, c) { \
+#define U16_NEXT(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
     (c)=(s)[(i)++]; \
     if(U16_IS_LEAD(c)) { \
-        uint16_t __c2; \
+        uint16_t __c2 = 0; \
         if((i)!=(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \
             ++(i); \
             (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
         } \
     } \
-}
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * The offset may point to the lead surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the following trail surrogate as well.
+ * If the offset points to a trail surrogate or
+ * to a single, unpaired lead surrogate, then c is set to U+FFFD.
+ *
+ * @param s const UChar * string
+ * @param i string offset, must be i<length
+ * @param length string length
+ * @param c output UChar32 variable
+ * @see U16_NEXT_UNSAFE
+ * @stable ICU 60
+ */
+#define U16_NEXT_OR_FFFD(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if(U16_IS_SURROGATE(c)) { \
+        uint16_t __c2 = 0; \
+        if(U16_IS_SURROGATE_LEAD(c) && (i)!=(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \
+            ++(i); \
+            (c)=U16_GET_SUPPLEMENTARY((c), __c2); \
+        } else { \
+            (c)=0xfffd; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Append a code point to a string, overwriting 1 or 2 code units.
@@ -287,14 +363,14 @@
  * @see U16_APPEND
  * @stable ICU 2.4
  */
-#define U16_APPEND_UNSAFE(s, i, c) { \
+#define U16_APPEND_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
     if((uint32_t)(c)<=0xffff) { \
         (s)[(i)++]=(uint16_t)(c); \
     } else { \
         (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \
         (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Append a code point to a string, overwriting 1 or 2 code units.
@@ -313,7 +389,7 @@
  * @see U16_APPEND_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_APPEND(s, i, capacity, c, isError) { \
+#define U16_APPEND(s, i, capacity, c, isError) UPRV_BLOCK_MACRO_BEGIN { \
     if((uint32_t)(c)<=0xffff) { \
         (s)[(i)++]=(uint16_t)(c); \
     } else if((uint32_t)(c)<=0x10ffff && (i)+1<(capacity)) { \
@@ -322,7 +398,7 @@
     } else /* c>0x10ffff or not enough space */ { \
         (isError)=TRUE; \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Advance the string offset from one code point boundary to the next.
@@ -334,11 +410,11 @@
  * @see U16_FWD_1
  * @stable ICU 2.4
  */
-#define U16_FWD_1_UNSAFE(s, i) { \
+#define U16_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
     if(U16_IS_LEAD((s)[(i)++])) { \
         ++(i); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Advance the string offset from one code point boundary to the next.
@@ -353,11 +429,11 @@
  * @see U16_FWD_1_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_FWD_1(s, i, length) { \
+#define U16_FWD_1(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \
     if(U16_IS_LEAD((s)[(i)++]) && (i)!=(length) && U16_IS_TRAIL((s)[i])) { \
         ++(i); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Advance the string offset from one code point boundary to the n-th next one,
@@ -371,13 +447,13 @@
  * @see U16_FWD_N
  * @stable ICU 2.4
  */
-#define U16_FWD_N_UNSAFE(s, i, n) { \
+#define U16_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
     int32_t __N=(n); \
     while(__N>0) { \
         U16_FWD_1_UNSAFE(s, i); \
         --__N; \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Advance the string offset from one code point boundary to the n-th next one,
@@ -394,13 +470,13 @@
  * @see U16_FWD_N_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_FWD_N(s, i, length, n) { \
+#define U16_FWD_N(s, i, length, n) UPRV_BLOCK_MACRO_BEGIN { \
     int32_t __N=(n); \
     while(__N>0 && ((i)<(length) || ((length)<0 && (s)[i]!=0))) { \
         U16_FWD_1(s, i, length); \
         --__N; \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Adjust a random-access offset to a code point boundary
@@ -415,11 +491,11 @@
  * @see U16_SET_CP_START
  * @stable ICU 2.4
  */
-#define U16_SET_CP_START_UNSAFE(s, i) { \
+#define U16_SET_CP_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
     if(U16_IS_TRAIL((s)[i])) { \
         --(i); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Adjust a random-access offset to a code point boundary
@@ -435,11 +511,11 @@
  * @see U16_SET_CP_START_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_SET_CP_START(s, start, i) { \
+#define U16_SET_CP_START(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
     if(U16_IS_TRAIL((s)[i]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \
         --(i); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /* definitions with backward iteration -------------------------------------- */
 
@@ -463,12 +539,12 @@
  * @see U16_PREV
  * @stable ICU 2.4
  */
-#define U16_PREV_UNSAFE(s, i, c) { \
+#define U16_PREV_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
     (c)=(s)[--(i)]; \
     if(U16_IS_TRAIL(c)) { \
         (c)=U16_GET_SUPPLEMENTARY((s)[--(i)], (c)); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Move the string offset from one code point boundary to the previous one
@@ -481,8 +557,7 @@
  * for a supplementary code point, then the macro will read
  * the preceding lead surrogate as well.
  * If the offset is behind a lead surrogate or behind a single, unpaired
- * trail surrogate, then that itself
- * will be returned as the code point.
+ * trail surrogate, then c is set to that unpaired surrogate.
  *
  * @param s const UChar * string
  * @param start starting string offset (usually 0)
@@ -491,16 +566,49 @@
  * @see U16_PREV_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_PREV(s, start, i, c) { \
+#define U16_PREV(s, start, i, c) UPRV_BLOCK_MACRO_BEGIN { \
     (c)=(s)[--(i)]; \
     if(U16_IS_TRAIL(c)) { \
-        uint16_t __c2; \
+        uint16_t __c2 = 0; \
         if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
             --(i); \
             (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
         } \
     } \
-}
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a trail surrogate unit
+ * for a supplementary code point, then the macro will read
+ * the preceding lead surrogate as well.
+ * If the offset is behind a lead surrogate or behind a single, unpaired
+ * trail surrogate, then c is set to U+FFFD.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<i
+ * @param c output UChar32 variable
+ * @see U16_PREV_UNSAFE
+ * @stable ICU 60
+ */
+#define U16_PREV_OR_FFFD(s, start, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(U16_IS_SURROGATE(c)) { \
+        uint16_t __c2 = 0; \
+        if(U16_IS_SURROGATE_TRAIL(c) && (i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+            --(i); \
+            (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+        } else { \
+            (c)=0xfffd; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Move the string offset from one code point boundary to the previous one.
@@ -513,11 +621,11 @@
  * @see U16_BACK_1
  * @stable ICU 2.4
  */
-#define U16_BACK_1_UNSAFE(s, i) { \
+#define U16_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
     if(U16_IS_TRAIL((s)[--(i)])) { \
         --(i); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Move the string offset from one code point boundary to the previous one.
@@ -531,11 +639,11 @@
  * @see U16_BACK_1_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_BACK_1(s, start, i) { \
+#define U16_BACK_1(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
     if(U16_IS_TRAIL((s)[--(i)]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \
         --(i); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Move the string offset from one code point boundary to the n-th one before it,
@@ -550,13 +658,13 @@
  * @see U16_BACK_N
  * @stable ICU 2.4
  */
-#define U16_BACK_N_UNSAFE(s, i, n) { \
+#define U16_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
     int32_t __N=(n); \
     while(__N>0) { \
         U16_BACK_1_UNSAFE(s, i); \
         --__N; \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Move the string offset from one code point boundary to the n-th one before it,
@@ -572,13 +680,13 @@
  * @see U16_BACK_N_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_BACK_N(s, start, i, n) { \
+#define U16_BACK_N(s, start, i, n) UPRV_BLOCK_MACRO_BEGIN { \
     int32_t __N=(n); \
     while(__N>0 && (i)>(start)) { \
         U16_BACK_1(s, start, i); \
         --__N; \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Adjust a random-access offset to a code point boundary after a code point.
@@ -593,11 +701,11 @@
  * @see U16_SET_CP_LIMIT
  * @stable ICU 2.4
  */
-#define U16_SET_CP_LIMIT_UNSAFE(s, i) { \
+#define U16_SET_CP_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
     if(U16_IS_LEAD((s)[(i)-1])) { \
         ++(i); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 /**
  * Adjust a random-access offset to a code point boundary after a code point.
@@ -616,10 +724,10 @@
  * @see U16_SET_CP_LIMIT_UNSAFE
  * @stable ICU 2.4
  */
-#define U16_SET_CP_LIMIT(s, start, i, length) { \
+#define U16_SET_CP_LIMIT(s, start, i, length) UPRV_BLOCK_MACRO_BEGIN { \
     if((start)<(i) && ((i)<(length) || (length)<0) && U16_IS_LEAD((s)[(i)-1]) && U16_IS_TRAIL((s)[i])) { \
         ++(i); \
     } \
-}
+} UPRV_BLOCK_MACRO_END
 
 #endif