Small speedup of construction of uninitialized QByteArrays

Directly construct uninitialized QByteArrays of required size
instead of default-constructing one and resizing it afterwards.

Reviewed-by: mariusSO

1 files changed, 2 insertions, 4 deletions

diff --git a/src/corelib/codecs/qisciicodec.cpp b/src/corelib/codecs/qisciicodec.cpp
index a33fd0dd75..c05431318e 100644
--- a/src/corelib/codecs/qisciicodec.cpp
+++ b/src/corelib/codecs/qisciicodec.cpp
@@ -187,8 +187,7 @@ QByteArray QIsciiCodec::convertFromUnicode(const QChar *uc, int len, ConverterSt
     }
     int invalid = 0;
 
-    QByteArray result;
-    result.resize(2*len); //worst case
+    QByteArray result(2 * len, Qt::Uninitialized); //worst case
 
     uchar *ch = reinterpret_cast<uchar *>(result.data());
 
@@ -250,8 +249,7 @@ QString QIsciiCodec::convertToUnicode(const char* chars, int len, ConverterState
         halant = state->state_data[0];
     }
 
-    QString result;
-    result.resize(len);
+    QString result(len, Qt::Uninitialized);
     QChar *uc = result.data();
 
     const int base = codecs[idx].base;