Use rtl functions instead of own surrogate checking/combining

Change-Id: I3eb05d8f5b0761bc3b672d4c855eb469f8cc1a29
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/127375
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>

13 files changed, 48 insertions, 115 deletions

diff --git a/sal/rtl/uri.cxx b/sal/rtl/uri.cxx
index 441a6c69f1e8..0c5479563405 100644
--- a/sal/rtl/uri.cxx
+++ b/sal/rtl/uri.cxx
@@ -219,21 +219,7 @@ namespace {
 
 void writeUcs4(rtl_uString ** pBuffer, sal_Int32 * pCapacity, sal_uInt32 nUtf32)
 {
-    assert(rtl::isUnicodeCodePoint(nUtf32));
-    if (nUtf32 <= 0xFFFF)
-    {
-        writeUnicode(pBuffer, pCapacity, static_cast< sal_Unicode >(nUtf32));
-    }
-    else
-    {
-        nUtf32 -= 0x10000;
-        writeUnicode(
-            pBuffer, pCapacity,
-            static_cast< sal_Unicode >(nUtf32 >> 10 | 0xD800));
-        writeUnicode(
-            pBuffer, pCapacity,
-            static_cast< sal_Unicode >((nUtf32 & 0x3FF) | 0xDC00));
-    }
+    rtl_uStringbuffer_insertUtf32(pBuffer, pCapacity, (*pBuffer)->length, nUtf32);
 }
 
 void writeEscapeOctet(rtl_uString ** pBuffer, sal_Int32 * pCapacity,
@@ -284,20 +270,7 @@ bool writeEscapeChar(rtl_uString ** pBuffer, sal_Int32 * pCapacity,
         rtl_UnicodeToTextConverter aConverter
             = rtl_createUnicodeToTextConverter(eCharset);
         sal_Unicode aSrc[2];
-        sal_Size nSrcSize;
-        if (nUtf32 <= 0xFFFF)
-        {
-            aSrc[0] = static_cast< sal_Unicode >(nUtf32);
-            nSrcSize = 1;
-        }
-        else
-        {
-            aSrc[0] = static_cast< sal_Unicode >(
-                ((nUtf32 - 0x10000) >> 10) | 0xD800);
-            aSrc[1] = static_cast< sal_Unicode >(
-                ((nUtf32 - 0x10000) & 0x3FF) | 0xDC00);
-            nSrcSize = 2;
-        }
+        sal_Size nSrcSize = rtl::splitSurrogates(nUtf32, aSrc);
 
         char aDst[32]; // FIXME  random value
         sal_uInt32 nInfo;
diff --git a/sal/rtl/ustrbuf.cxx b/sal/rtl/ustrbuf.cxx
index 53e178e5881a..7c5b187edbcb 100644
--- a/sal/rtl/ustrbuf.cxx
+++ b/sal/rtl/ustrbuf.cxx
@@ -174,17 +174,7 @@ void rtl_uStringbuffer_insertUtf32(
     SAL_THROW_EXTERN_C()
 {
     sal_Unicode buf[2];
-    sal_Int32 len;
-    OSL_ASSERT(rtl::isUnicodeScalarValue(c));
-    if (c <= 0xFFFF) {
-        buf[0] = static_cast<sal_Unicode>(c);
-        len = 1;
-    } else {
-        c -= 0x10000;
-        buf[0] = static_cast<sal_Unicode>((c >> 10) | 0xD800);
-        buf[1] = static_cast<sal_Unicode>((c & 0x3FF) | 0xDC00);
-        len = 2;
-    }
+    sal_Int32 len = rtl::splitSurrogates(c, buf);
     rtl_uStringbuffer_insert(pThis, capacity, offset, buf, len);
 }
 
diff --git a/sal/textenc/convertbig5hkscs.cxx b/sal/textenc/convertbig5hkscs.cxx
index 77484666982e..bf3a2e2cf59d 100644
--- a/sal/textenc/convertbig5hkscs.cxx
+++ b/sal/textenc/convertbig5hkscs.cxx
@@ -131,7 +131,7 @@ sal_Size ImplConvertBig5HkscsToUnicode(void const * pData,
                         nUnicode = pBig5Data[nRow].mpToUniTrailTab[nChar - n];
                         if (nUnicode == 0)
                             nUnicode = 0xFFFF;
-                        assert(!ImplIsHighSurrogate(nUnicode));
+                        assert(!rtl::isHighSurrogate(nUnicode));
                     }
                 }
                 if (nUnicode == 0xFFFF)
@@ -192,11 +192,11 @@ sal_Size ImplConvertBig5HkscsToUnicode(void const * pData,
                         }
                         ++p;
                     }
-                    assert(!ImplIsHighSurrogate(nUnicode));
+                    assert(!rtl::isHighSurrogate(nUnicode));
                 }
                 if (nUnicode == 0xFFFF)
                     goto bad_input;
-                if (ImplIsHighSurrogate(nUnicode))
+                if (rtl::isHighSurrogate(nUnicode))
                     if (pDestBufEnd - pDestBufPtr >= 2)
                     {
                         nOffset += nLast - nFirst + 1;
@@ -329,19 +329,19 @@ sal_Size ImplConvertUnicodeToBig5Hkscs(void const * pData,
         sal_uInt32 nChar = *pSrcBuf++;
         if (nHighSurrogate == 0)
         {
-            if (ImplIsHighSurrogate(nChar))
+            if (rtl::isHighSurrogate(nChar))
             {
                 nHighSurrogate = static_cast<sal_Unicode>(nChar);
                 continue;
             }
-            else if (ImplIsLowSurrogate(nChar))
+            else if (rtl::isLowSurrogate(nChar))
             {
                 bUndefined = false;
                 goto bad_input;
             }
         }
-        else if (ImplIsLowSurrogate(nChar))
-            nChar = ImplCombineSurrogates(nHighSurrogate, nChar);
+        else if (rtl::isLowSurrogate(nChar))
+            nChar = rtl::combineSurrogates(nHighSurrogate, nChar);
         else
         {
             bUndefined = false;
diff --git a/sal/textenc/converteuctw.cxx b/sal/textenc/converteuctw.cxx
index edb3c07fa934..8684b643c66e 100644
--- a/sal/textenc/converteuctw.cxx
+++ b/sal/textenc/converteuctw.cxx
@@ -206,7 +206,7 @@ sal_Size ImplConvertEucTwToUnicode(void const * pData,
                             = pCns116431992Data[nOffset + (nChar - nFirst)];
                         if (nUnicode == 0xFFFF)
                             goto bad_input;
-                        else if (ImplIsHighSurrogate(nUnicode))
+                        else if (rtl::isHighSurrogate(nUnicode))
                             if (pDestBufEnd - pDestBufPtr >= 2)
                             {
                                 nOffset += nLast - nFirst + 1;
@@ -340,19 +340,19 @@ sal_Size ImplConvertUnicodeToEucTw(void const * pData,
         sal_uInt32 nChar = *pSrcBuf++;
         if (nHighSurrogate == 0)
         {
-            if (ImplIsHighSurrogate(nChar))
+            if (rtl::isHighSurrogate(nChar))
             {
                 nHighSurrogate = static_cast<sal_Unicode>(nChar);
                 continue;
             }
-            else if (ImplIsLowSurrogate(nChar))
+            else if (rtl::isLowSurrogate(nChar))
             {
                 bUndefined = false;
                 goto bad_input;
             }
         }
-        else if (ImplIsLowSurrogate(nChar))
-            nChar = ImplCombineSurrogates(nHighSurrogate, nChar);
+        else if (rtl::isLowSurrogate(nChar))
+            nChar = rtl::combineSurrogates(nHighSurrogate, nChar);
         else
         {
             bUndefined = false;
diff --git a/sal/textenc/convertgb18030.cxx b/sal/textenc/convertgb18030.cxx
index 88f5a999d87d..d7963762fa2b 100644
--- a/sal/textenc/convertgb18030.cxx
+++ b/sal/textenc/convertgb18030.cxx
@@ -172,10 +172,7 @@ sal_Size ImplConvertGb18030ToUnicode(void const * pData,
                     if (pDestBufEnd - pDestBufPtr >= 2)
                     {
                         nCode -= 189000 - 0x10000;
-                        *pDestBufPtr++
-                            = static_cast<sal_Unicode>(ImplGetHighSurrogate(nCode));
-                        *pDestBufPtr++
-                            = static_cast<sal_Unicode>(ImplGetLowSurrogate(nCode));
+                        pDestBufPtr += rtl::splitSurrogates(nCode, pDestBufPtr);
                         startOfCurrentChar = nConverted + 1;
                     }
                     else
@@ -330,19 +327,19 @@ sal_Size ImplConvertUnicodeToGb18030(void const * pData,
         sal_uInt32 nChar = *pSrcBuf++;
         if (nHighSurrogate == 0)
         {
-            if (ImplIsHighSurrogate(nChar))
+            if (rtl::isHighSurrogate(nChar))
             {
                 nHighSurrogate = static_cast<sal_Unicode>(nChar);
                 continue;
             }
-            else if (ImplIsLowSurrogate(nChar))
+            else if (rtl::isLowSurrogate(nChar))
             {
                 bUndefined = false;
                 goto bad_input;
             }
         }
-        else if (ImplIsLowSurrogate(nChar))
-            nChar = ImplCombineSurrogates(nHighSurrogate, nChar);
+        else if (rtl::isLowSurrogate(nChar))
+            nChar = rtl::combineSurrogates(nHighSurrogate, nChar);
         else
         {
             bUndefined = false;
diff --git a/sal/textenc/convertisciidevangari.cxx b/sal/textenc/convertisciidevangari.cxx
index 759328565690..75fcadcf7fe4 100644
--- a/sal/textenc/convertisciidevangari.cxx
+++ b/sal/textenc/convertisciidevangari.cxx
@@ -266,20 +266,20 @@ sal_Size UnicodeToIsciiDevanagari::convert(sal_Unicode const* pSrcBuf, sal_Size
         char cSpecialChar = 0;
         if (cHighSurrogate == 0)
         {
-            if (ImplIsHighSurrogate(c))
+            if (rtl::isHighSurrogate(c))
             {
                 cHighSurrogate = static_cast< sal_Unicode >(c);
                 continue;
             }
-            else if (ImplIsLowSurrogate(c))
+            else if (rtl::isLowSurrogate(c))
             {
                 bUndefined = false;
                 goto bad_input;
             }
         }
-        else if (ImplIsLowSurrogate(c))
+        else if (rtl::isLowSurrogate(c))
         {
-            c = ImplCombineSurrogates(cHighSurrogate, c);
+            c = rtl::combineSurrogates(cHighSurrogate, c);
         }
         else
         {
diff --git a/sal/textenc/convertiso2022cn.cxx b/sal/textenc/convertiso2022cn.cxx
index 9e89c27486db..60c2b57436ff 100644
--- a/sal/textenc/convertiso2022cn.cxx
+++ b/sal/textenc/convertiso2022cn.cxx
@@ -329,7 +329,7 @@ sal_Size ImplConvertIso2022CnToUnicode(void const * pData,
                             = pCns116431992Data[nOffset + (nChar - nFirst)];
                         if (nUnicode == 0xFFFF)
                             goto bad_input;
-                        else if (ImplIsHighSurrogate(nUnicode))
+                        else if (rtl::isHighSurrogate(nUnicode))
                             if (pDestBufEnd - pDestBufPtr >= 2)
                             {
                                 nOffset += nLast - nFirst + 1;
@@ -556,19 +556,19 @@ sal_Size ImplConvertUnicodeToIso2022Cn(void const * pData,
         sal_uInt32 nChar = *pSrcBuf++;
         if (nHighSurrogate == 0)
         {
-            if (ImplIsHighSurrogate(nChar))
+            if (rtl::isHighSurrogate(nChar))
             {
                 nHighSurrogate = static_cast<sal_Unicode>(nChar);
                 continue;
             }
-            else if (ImplIsLowSurrogate(nChar))
+            else if (rtl::isLowSurrogate(nChar))
             {
                 bUndefined = false;
                 goto bad_input;
             }
         }
-        else if (ImplIsLowSurrogate(nChar))
-            nChar = ImplCombineSurrogates(nHighSurrogate, nChar);
+        else if (rtl::isLowSurrogate(nChar))
+            nChar = rtl::combineSurrogates(nHighSurrogate, nChar);
         else
         {
             bUndefined = false;
diff --git a/sal/textenc/convertiso2022jp.cxx b/sal/textenc/convertiso2022jp.cxx
index 4024653f04e5..935bc8515d77 100644
--- a/sal/textenc/convertiso2022jp.cxx
+++ b/sal/textenc/convertiso2022jp.cxx
@@ -375,19 +375,19 @@ sal_Size ImplConvertUnicodeToIso2022Jp(void const * pData,
         sal_uInt32 nChar = *pSrcBuf++;
         if (nHighSurrogate == 0)
         {
-            if (ImplIsHighSurrogate(nChar))
+            if (rtl::isHighSurrogate(nChar))
             {
                 nHighSurrogate = static_cast<sal_Unicode>(nChar);
                 continue;
             }
-            else if (ImplIsLowSurrogate(nChar))
+            else if (rtl::isLowSurrogate(nChar))
             {
                 bUndefined = false;
                 goto bad_input;
             }
         }
-        else if (ImplIsLowSurrogate(nChar))
-            nChar = ImplCombineSurrogates(nHighSurrogate, nChar);
+        else if (rtl::isLowSurrogate(nChar))
+            nChar = rtl::combineSurrogates(nHighSurrogate, nChar);
         else
         {
             bUndefined = false;
diff --git a/sal/textenc/convertiso2022kr.cxx b/sal/textenc/convertiso2022kr.cxx
index 5c7971ba5af6..5aea5c66c877 100644
--- a/sal/textenc/convertiso2022kr.cxx
+++ b/sal/textenc/convertiso2022kr.cxx
@@ -353,19 +353,19 @@ sal_Size ImplConvertUnicodeToIso2022Kr(void const * pData,
             sal_uInt32 nChar = *pSrcBuf++;
             if (nHighSurrogate == 0)
             {
-                if (ImplIsHighSurrogate(nChar))
+                if (rtl::isHighSurrogate(nChar))
                 {
                     nHighSurrogate = static_cast<sal_Unicode>(nChar);
                     continue;
                 }
-                else if (ImplIsLowSurrogate(nChar))
+                else if (rtl::isLowSurrogate(nChar))
                 {
                     bUndefined = false;
                     goto bad_input;
                 }
             }
-            else if (ImplIsLowSurrogate(nChar))
-                nChar = ImplCombineSurrogates(nHighSurrogate, nChar);
+            else if (rtl::isLowSurrogate(nChar))
+                nChar = rtl::combineSurrogates(nHighSurrogate, nChar);
             else
             {
                 bUndefined = false;
diff --git a/sal/textenc/convertsinglebytetobmpunicode.cxx b/sal/textenc/convertsinglebytetobmpunicode.cxx
index b948ba0fbc5c..43c2b9cf39ce 100644
--- a/sal/textenc/convertsinglebytetobmpunicode.cxx
+++ b/sal/textenc/convertsinglebytetobmpunicode.cxx
@@ -111,17 +111,17 @@ sal_Size rtl_textenc_convertBmpUnicodeToSingleByte(
         bool undefined = true;
         sal_uInt32 c = *srcBuf++;
         if (highSurrogate == 0) {
-            if (ImplIsHighSurrogate(c)) {
+            if (rtl::isHighSurrogate(c)) {
                 highSurrogate = static_cast< sal_Unicode >(c);
                 continue;
             }
-            else if (ImplIsLowSurrogate(c))
+            else if (rtl::isLowSurrogate(c))
             {
                 undefined = false;
                 goto bad_input;
             }
-        } else if (ImplIsLowSurrogate(c)) {
-            c = ImplCombineSurrogates(highSurrogate, c);
+        } else if (rtl::isLowSurrogate(c)) {
+            c = rtl::combineSurrogates(highSurrogate, c);
         } else {
             undefined = false;
             goto bad_input;
diff --git a/sal/textenc/handleundefinedunicodetotextchar.cxx b/sal/textenc/handleundefinedunicodetotextchar.cxx
index 320562495e97..76aed03e1881 100644
--- a/sal/textenc/handleundefinedunicodetotextchar.cxx
+++ b/sal/textenc/handleundefinedunicodetotextchar.cxx
@@ -106,7 +106,7 @@ bool sal::detail::textenc::handleUndefinedUnicodeToTextChar(
 
     /* Surrogates Characters should result in */
     /* one replacement character */
-    if (ImplIsHighSurrogate(c))
+    if (rtl::isHighSurrogate(c))
     {
         if ( ((*ppSrcBuf) + 1) == pEndSrcBuf )
         {
@@ -115,7 +115,7 @@ bool sal::detail::textenc::handleUndefinedUnicodeToTextChar(
         }
 
         c = *((*ppSrcBuf)+1);
-        if (ImplIsLowSurrogate(c))
+        if (rtl::isLowSurrogate(c))
             (*ppSrcBuf)++;
         else
         {
diff --git a/sal/textenc/tcvtutf8.cxx b/sal/textenc/tcvtutf8.cxx
index ca29156c418f..f210b654d57f 100644
--- a/sal/textenc/tcvtutf8.cxx
+++ b/sal/textenc/tcvtutf8.cxx
@@ -199,10 +199,7 @@ sal_Size ImplConvertUtf8ToUnicode(
                 else
                     goto no_output;
             else if (pDestBufEnd - pDestBufPtr >= 2)
-            {
-                *pDestBufPtr++ = static_cast<sal_Unicode>(ImplGetHighSurrogate(nUtf32));
-                *pDestBufPtr++ = static_cast<sal_Unicode>(ImplGetLowSurrogate(nUtf32));
-            }
+                pDestBufPtr += rtl::splitSurrogates(nUtf32, pDestBufPtr);
             else
                 goto no_output;
         }
@@ -349,18 +346,18 @@ sal_Size ImplConvertUnicodeToUtf8(
         sal_uInt32 nChar = *pSrcBufPtr++;
         if (nHighSurrogate == 0)
         {
-            if (ImplIsHighSurrogate(nChar) && !bJavaUtf8)
+            if (rtl::isHighSurrogate(nChar) && !bJavaUtf8)
             {
                 nHighSurrogate = static_cast<sal_Unicode>(nChar);
                 continue;
             }
-            else if (ImplIsLowSurrogate(nChar) && !bJavaUtf8)
+            else if (rtl::isLowSurrogate(nChar) && !bJavaUtf8)
             {
                 goto bad_input;
             }
         }
-        else if (ImplIsLowSurrogate(nChar) && !bJavaUtf8)
-            nChar = ImplCombineSurrogates(nHighSurrogate, nChar);
+        else if (rtl::isLowSurrogate(nChar) && !bJavaUtf8)
+            nChar = rtl::combineSurrogates(nHighSurrogate, nChar);
         else
             goto bad_input;
 
diff --git a/sal/textenc/unichars.hxx b/sal/textenc/unichars.hxx
index 8ca1021da005..e627d3c95c3e 100644
--- a/sal/textenc/unichars.hxx
+++ b/sal/textenc/unichars.hxx
@@ -31,34 +31,10 @@
 
 bool ImplIsControlOrFormat(sal_uInt32 nUtf32);
 
-// All code points that are high-surrogates, as of Unicode 3.1.1.
-inline bool ImplIsHighSurrogate(sal_uInt32 nUtf32) { return nUtf32 >= 0xD800 && nUtf32 <= 0xDBFF; }
-
-// All code points that are low-surrogates, as of Unicode 3.1.1.
-inline bool ImplIsLowSurrogate(sal_uInt32 nUtf32) { return nUtf32 >= 0xDC00 && nUtf32 <= 0xDFFF; }
-
 bool ImplIsPrivateUse(sal_uInt32 nUtf32);
 
 bool ImplIsZeroWidth(sal_uInt32 nUtf32);
 
-inline sal_uInt32 ImplGetHighSurrogate(sal_uInt32 nUtf32)
-{
-    assert(nUtf32 >= 0x10000);
-    return ((nUtf32 - 0x10000) >> 10) | 0xD800;
-}
-
-inline sal_uInt32 ImplGetLowSurrogate(sal_uInt32 nUtf32)
-{
-    assert(nUtf32 >= 0x10000);
-    return ((nUtf32 - 0x10000) & 0x3FF) | 0xDC00;
-}
-
-inline sal_uInt32 ImplCombineSurrogates(sal_uInt32 nHigh, sal_uInt32 nLow)
-{
-    assert(ImplIsHighSurrogate(nHigh) && ImplIsLowSurrogate(nLow));
-    return (((nHigh & 0x3FF) << 10) | (nLow & 0x3FF)) + 0x10000;
-}
-
 #endif
 
 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */