Merge pull request #4 from Hugoberry/utf-8

Support for Wide-characters in column names

Author: Hugoberry

src/msolap_utils.cpp

@@ -3,23 +3,15 @@
 namespace duckdb {
 
 std::string MSOLAPUtils::SanitizeColumnName(const std::wstring &name) {
-    std::string result;
-    result.reserve(name.size());
-    
-    for (size_t i = 0; i < name.size(); i++) {
-        // Replace brackets with underscores as required
-        if (name[i] == L'[' || name[i] == L']') {
-            result += '_';
-        } else if (name[i] <= 255) {
-            // Convert to ASCII if possible
-            result += static_cast<char>(name[i]);
-        } else {
-            // Replace non-ASCII with underscore
-            result += '_';
+    std::wstring sanitized = name;
+    for (size_t i = 0; i < sanitized.size(); i++) {
+        if (sanitized[i] == L'[' || sanitized[i] == L']') {
+            sanitized[i] = L'_';
         }
     }
 
-    return result;
+    // Convert the entire string to UTF-8 properly
+    return WindowsUtil::UnicodeToUTF8(sanitized.c_str());
 }
 
 Value MSOLAPUtils::ConvertVariantToValue(VARIANT* pVar) {

test/sql/msolap_table_constructor.test

@@ -0,0 +1,25 @@
+# name: test/sql/msolap_table_constructor.test
+# description: test msolap extension with DAX table constructor
+# group: [msolap]
+
+# Require statement will ensure this test is run with the extension loaded
+require msolap
+
+# Test German special characters in column names and contents
+query II
+FROM msolap(
+    'Provider=MSOLAP;Data Source=localhost:54043;Catalog=53e9bf8c-d0e3-49c8-818c-3505753645bd',
+    'EVALUATE
+     DATATABLE( "🦆", STRING, "äöü", STRING, 
+        {    
+            {"Duck", "DB"},
+            {"Straße", "äöü"},
+            {"München", "Köln"},
+            {"Jürgen", "Größe"}
+        })'
+);
+----
+Straße	äöü
+München	Köln
+Jürgen	Größe
+