Merge 'fix: cmp integer/float' from Jean Arhancet

Closes https://github.com/penberg/limbo/issues/339

Reviewed-by: Lauri Virtanen <pere-altea@hotmail.com>

Closes #340

core/types.rs

@@ -92,7 +92,7 @@ impl PartialOrd<OwnedValue> for OwnedValue {
                 int_left.partial_cmp(int_right)
             }
             (OwnedValue::Integer(int_left), OwnedValue::Float(float_right)) => {
-                float_right.partial_cmp(&(*int_left as f64))
+                (*int_left as f64).partial_cmp(float_right)
             }
             (OwnedValue::Float(float_left), OwnedValue::Integer(int_right)) => {
                 float_left.partial_cmp(&(*int_right as f64))

testing/scalar-functions.test

@@ -295,6 +295,18 @@ do_execsql_test max-str {
   select max('b','a','z')
 } {z}
 
+do_execsql_test min-int-float {
+  select min(1,5.0)
+} {1}
+
+do_execsql_test max-int-float {
+  select max(1,5.0)
+} {5.0}
+
+do_execsql_test min-float-int {
+  select min(5.0,1)
+} {1}
+
 do_execsql_test max-null {
   select max(null,null)
 } {}