Support writing numerics with larger scale than precision (#98)

In parquet, we cannot write numeric with larger scale than precision. But postgres allows it. We can adjust the numeric typmod in that case.

src/pgrx_tests/copy_type_roundtrip.rs

@@ -987,6 +987,27 @@ mod tests {
         test_table.assert_expected_and_result_rows();
     }
 
+    #[cfg(feature = "pg14")]
+    #[pg_test]
+    #[should_panic = "NUMERIC scale 8 must be between 0 and precision 5"]
+    fn test_numeric_with_larger_scale() {
+        let test_table = TestTable::<AnyNumeric>::new("numeric(5, 8)".into());
+        test_table.insert(
+            "INSERT INTO test_expected (a) VALUES (0.00012345), (0.00012340), (-0.00012345), (-0.00012340), (null), (0);",
+        );
+        test_table.assert_expected_and_result_rows();
+    }
+
+    #[cfg(not(feature = "pg14"))]
+    #[pg_test]
+    fn test_numeric_with_larger_scale() {
+        let test_table = TestTable::<AnyNumeric>::new("numeric(5, 8)".into());
+        test_table.insert(
+            "INSERT INTO test_expected (a) VALUES (0.00012345), (0.00012340), (-0.00012345), (-0.00012340), (null), (0);",
+        );
+        test_table.assert_expected_and_result_rows();
+    }
+
     #[pg_test]
     #[should_panic = "Special numeric values like NaN, Inf, -Inf are not allowed"]
     fn test_numeric_nan() {

src/type_compat/pg_arrow_type_conversions.rs

@@ -322,9 +322,19 @@ pub(crate) fn extract_precision_and_scale_from_numeric_typmod(typmod: i32) -> (u
     let mut precision = extract_precision_from_numeric_typmod(typmod);
     let mut scale = extract_scale_from_numeric_typmod(typmod);
 
-    // Even if PG allows negative scale, arrow does not. We adjust precision by adding scale to it.
+    // Even if PG allows negative scale, arrow does not.
+    // We adjust precision by adding scale to it and set scale to 0.
+    // e.g. "123.34::numeric(3,-2)" becomes "100::numeric(5,0)"
     if scale < 0 {
-        adjust_precision_and_scale_if_negative_scale(&mut precision, &mut scale);
+        precision += scale.abs();
+        scale = 0;
+    }
+
+    // Even if PG allows scale to be greater than precision, arrow does not.
+    // We set precision to the same value as scale.
+    // e.g. "0.0023::numeric(2,4)" becomes "0.0023::numeric(4,4)"
+    if scale > precision {
+        precision = scale;
     }
 
     debug_assert!(precision >= 0);
@@ -345,15 +355,6 @@ fn extract_scale_from_numeric_typmod(typmod: i32) -> i32 {
     (((typmod - pg_sys::VARHDRSZ as i32) & 0x7ff) ^ 1024) - 1024
 }
 
-// adjust_precision_and_scale_if_negative_scale adjusts precision and scale if scale is negative.
-// Even if PG allows negative scale, arrow does not. We adjust precision by adding scale to it.
-fn adjust_precision_and_scale_if_negative_scale(precision: &mut i32, scale: &mut i32) {
-    if *scale < 0 {
-        *precision += scale.abs();
-        *scale = 0;
-    }
-}
-
 fn is_unbounded_numeric_typmod(typmod: i32) -> bool {
     typmod == -1
 }