take a shortcut on common separating ASCII characters

Change-Id: I991801b9a0b7be6024612c8f29262f2ebd8e6a64

1 files changed, 17 insertions, 3 deletions

diff --git a/svl/source/numbers/zforfind.cxx b/svl/source/numbers/zforfind.cxx
index fef69c5e2d8c..038597043f30 100644
--- a/svl/source/numbers/zforfind.cxx
+++ b/svl/source/numbers/zforfind.cxx
@@ -459,14 +459,28 @@ bool ImpSvNumberInputScan::StringContainsWord( const OUString& rWhat,
          * form? */
 
         // Check simple ASCII first before invoking i18n or anything else.
-        if (rtl::isAsciiAlphanumeric( rString[nPos] ))
+        const sal_Unicode c = rString[nPos];
+
+        // Common separating ASCII characters in date context.
+        switch (c)
+        {
+            case ' ':
+            case '-':
+            case '.':
+            case '/':
+                return true;
+            default:
+                ;   // nothing
+        }
+
+        if (rtl::isAsciiAlphanumeric( c ))
             return false;   // Alpha or numeric is not word gap.
 
         sal_Int32 nIndex = nPos;
-        const sal_uInt32 c = rString.iterateCodePoints( &nIndex);
+        const sal_uInt32 uc = rString.iterateCodePoints( &nIndex);
         if (nPos+1 < nIndex)
             return true;    // Surrogate, assume these to be new words.
-        (void)c;
+        (void)uc;
 
         const sal_Int32 nType = pFormatter->GetCharClass()->getCharacterType( rString, nPos);
         using namespace ::com::sun::star::i18n;