diff --git a/eeschema/lib_pin.cpp b/eeschema/lib_pin.cpp
index da454d85aa..e4b221f3bd 100644
--- a/eeschema/lib_pin.cpp
+++ b/eeschema/lib_pin.cpp
@@ -721,7 +721,6 @@ bool LIB_PIN::Save( OUTPUTFORMATTER& aFormatter )
 
 bool LIB_PIN::Load( LINE_READER& aLineReader, wxString& aErrorMsg )
 {
-    int  i, j;
     char pinAttrs[64];
     char pinOrient[64];
     char pinType[64];
@@ -738,11 +737,11 @@ bool LIB_PIN::Load( LINE_READER& aLineReader, wxString& aErrorMsg )
     // the full line starts by "X ". The pin data starts at line + 2.
     wxString utf8line = FROM_UTF8( aLineReader.Line() + 2 );
 	wxStringTokenizer tokenizer( utf8line, wxT(" \n\r" ) );
-    i = tokenizer.CountTokens();
+    int prms_count = tokenizer.CountTokens();
 
-    if( i < 11 )
+    if( prms_count < 11 )
     {
-        aErrorMsg.Printf( wxT( "pin had %d parameters of the required 11 or 12" ), i );
+        aErrorMsg.Printf( wxT( "pin had %d parameters of the required 11 or 12" ), prms_count );
         return false;
     }
 
@@ -767,10 +766,15 @@ bool LIB_PIN::Load( LINE_READER& aLineReader, wxString& aErrorMsg )
     strncpy( line, TO_UTF8( tmp ), len );
     line[len] = 0;
 
-    sscanf( line, "%d %d %d %63s %d %d %d %d %63s %63s",
+    int cnt = sscanf( line, "%d %d %d %63s %d %d %d %d %63s %63s",
                 &m_position.x, &m_position.y, &m_length, pinOrient, &m_numTextSize,
                 &m_nameTextSize, &m_Unit, &m_Convert, pinType, pinAttrs );
 
+    if( cnt != (prms_count - 2) )
+    {
+        aErrorMsg.Printf( wxT( "pin parameters read issue" ) );
+        return false;
+    }
 
     m_orientation = pinOrient[0] & 255;
 
@@ -825,9 +829,9 @@ bool LIB_PIN::Load( LINE_READER& aLineReader, wxString& aErrorMsg )
         return false;
     }
 
-    if( i == 12 )       /* Special Symbol defined */
+    if( prms_count >= 12 )       /* Special Symbol defined */
     {
-        for( j = strlen( pinAttrs ); j > 0; )
+        for( int j = strlen( pinAttrs ); j > 0; )
         {
             switch( pinAttrs[--j] )
             {
diff --git a/eeschema/sch_component.cpp b/eeschema/sch_component.cpp
index 938b4222d7..35c60c658d 100644
--- a/eeschema/sch_component.cpp
+++ b/eeschema/sch_component.cpp
@@ -1130,8 +1130,11 @@ bool SCH_COMPONENT::Save( FILE* f ) const
 
 bool SCH_COMPONENT::Load( LINE_READER& aLine, wxString& aErrorMsg )
 {
+    // Remark: avoid using sscanf to read texts entered by user
+    // which are UTF8 encoded, because sscanf does not work well on Windows
+    // with some UTF8 values.
     int         ii;
-    char        name1[256], name2[256],
+    char        name1[256],
                 char1[256], char2[256], char3[256];
     int         newfmt = 0;
     char*       ptcar;
@@ -1148,7 +1151,20 @@ bool SCH_COMPONENT::Load( LINE_READER& aLine, wxString& aErrorMsg )
             return true;
     }
 
-    if( sscanf( &line[1], "%255s %255s", name1, name2 ) != 2 )
+    // Parse the first line of description:
+    // like "L partname ref" (for instance "L 74LS00 U4"
+    // They are UTF8 texts, so do not use sscanf
+
+    line += 1;
+
+    if( *line == ' ' )
+        line++;
+
+    // line points the first parameter
+    wxString buffer( FROM_UTF8( line ) );
+    wxStringTokenizer tokenizer( buffer, wxT( " \r\n" ) );
+
+    if( tokenizer.CountTokens() < 2 )
     {
         aErrorMsg.Printf( wxT( "Eeschema component description error at line %d, aborted" ),
                           aLine.LineNumber() );
@@ -1156,18 +1172,15 @@ bool SCH_COMPONENT::Load( LINE_READER& aLine, wxString& aErrorMsg )
         return false;
     }
 
-    if( strcmp( name1, NULL_STRING ) != 0 )
-    {
-        for( ii = 0; ii < (int) strlen( name1 ); ii++ )
-        {
-            if( name1[ii] == '~' )
-                name1[ii] = ' ';
-        }
+    wxString partname = tokenizer.NextToken();
+    partname.Replace( wxT("~"), wxT(" ") );  // all spaces were replaced by ~ in files.
 
-        SetPartName( FROM_UTF8( name1 ) );
+    if( partname != NULL_STRING )
+    {
+        SetPartName( partname );
 
         if( !newfmt )
-            GetField( VALUE )->SetText( FROM_UTF8( name1 ) );
+            GetField( VALUE )->SetText( partname );
     }
     else
     {
@@ -1177,48 +1190,35 @@ bool SCH_COMPONENT::Load( LINE_READER& aLine, wxString& aErrorMsg )
         GetField( VALUE )->SetVisible( false );
     }
 
-    if( strcmp( name2, NULL_STRING ) != 0 )
+    wxString reference = tokenizer.NextToken();
+    reference.Replace( wxT("~"), wxT(" ") );  // all spaces were replaced by ~ in files.
+    reference.Trim( true );
+    reference.Trim( false );
+
+    if( reference != NULL_STRING )
     {
-        bool isDigit = false;
-
-        for( ii = 0; ii < (int) strlen( name2 ); ii++ )
+        wxString prefix = reference;
+        // Build reference prefix from the actual reference by removing trailing digits
+        // (Perhaps outdated code, only for very old schematic files)
+        while( prefix.Length() )
         {
-            if( name2[ii] == '~' )
-                name2[ii] = ' ';
+            if( ( prefix.Last() < '0' || prefix.Last() > '9') && prefix.Last() != '?' )
+                break;
 
-            // get RefBase from this, too. store in name1.
-            if( name2[ii] >= '0' && name2[ii] <= '9' )
-            {
-                isDigit   = true;
-                name1[ii] = 0;  //null-terminate.
-            }
-
-            if( !isDigit )
-            {
-                name1[ii] = name2[ii];
-            }
+            prefix.RemoveLast();
         }
 
-        name1[ii] = 0; //just in case
-        int  jj;
+        // Avoid a prefix containing trailing/leading spaces
+        prefix.Trim( true );
+        prefix.Trim( false );
 
-        for( jj = 0; jj<ii && name1[jj] == ' '; jj++ )
-            ;
-
-        if( jj == ii )
-        {
-            // blank string.
+        if( prefix.IsEmpty() )
             m_prefix = wxT( "U" );
-        }
         else
-        {
-            m_prefix = FROM_UTF8( &name1[jj] );
-
-            //printf("prefix: %s\n", TO_UTF8(component->m_prefix));
-        }
+            m_prefix = prefix;
 
         if( !newfmt )
-            GetField( REFERENCE )->SetText( FROM_UTF8( name2 ) );
+            GetField( REFERENCE )->SetText( reference );
     }
     else
     {