Make default numeric scale 9 (#86)

To give the integral part more room, we decided to make the default scale = 9
when the user does not specify precision and scale for numeric type.

Multiplication of 2 numeric doubles the result scale and gives the integral part
less digits. It is wiser to make it smaller but still useful number.

README.md

@@ -239,7 +239,7 @@ There is currently only one GUC parameter to enable/disable the `pg_parquet`:
 >    * `numeric(9 < P <= 18, S)` is represented as `INT64` with `DECIMAL` logical type
 >    * `numeric(18 < P <= 38, S)` is represented as `FIXED_LEN_BYTE_ARRAY(9-16)` with `DECIMAL` logical type
 >    * `numeric(38 < P, S)` is represented as `BYTE_ARRAY` with `STRING` logical type
->    * `numeric` is allowed by Postgres. (precision and scale not specified). These are represented by a default precision (38) and scale (16) instead of writing them as string. You get runtime error if your table tries to read or write a numeric value which is not allowed by the default precision and scale (22 integral digits before decimal point, 16 digits after decimal point).
+>    * `numeric` is allowed by Postgres. (precision and scale not specified). These are represented by a default precision (38) and scale (9) instead of writing them as string. You get runtime error if your table tries to read or write a numeric value which is not allowed by the default precision and scale (29 integral digits before decimal point, 9 digits after decimal point).
 > - (2) The `date` type is represented according to `Unix epoch` when writing to Parquet files. It is converted back according to `PostgreSQL epoch` when reading from Parquet files.
 > - (3) The `timestamptz` and `timetz` types are adjusted to `UTC` when writing to Parquet files. They are converted back with `UTC` timezone when reading from Parquet files.
 > - (4) The `geometry` type is represented as `BYTE_ARRAY` encoded as `WKB` when `postgis` extension is created. Otherwise, it is represented as `BYTE_ARRAY` with `STRING` logical type.

src/pgrx_tests/copy_type_roundtrip.rs

@@ -863,7 +863,7 @@ mod tests {
 
     #[pg_test]
     #[should_panic(
-        expected = "numeric value contains 23 digits before decimal point, which exceeds max allowed integral digits 22 during copy to parquet"
+        expected = "numeric value contains 30 digits before decimal point, which exceeds max allowed integral digits 29 during copy to parquet"
     )]
     fn test_invalid_unbounded_numeric_integral_digits() {
         let invalid_integral_digits =
@@ -879,7 +879,7 @@ mod tests {
 
     #[pg_test]
     #[should_panic(
-        expected = "numeric value contains 17 digits after decimal point, which exceeds max allowed decimal digits 16 during copy to parquet"
+        expected = "numeric value contains 10 digits after decimal point, which exceeds max allowed decimal digits 9 during copy to parquet"
     )]
     fn test_invalid_unbounded_numeric_decimal_digits() {
         let invalid_decimal_digits = DEFAULT_UNBOUNDED_NUMERIC_SCALE + 1;

src/type_compat/pg_arrow_type_conversions.rs

@@ -277,7 +277,7 @@ fn rescale_unbounded_numeric_or_error(
 
 const MAX_NUMERIC_PRECISION: u32 = 38;
 pub(crate) const DEFAULT_UNBOUNDED_NUMERIC_PRECISION: u32 = MAX_NUMERIC_PRECISION;
-pub(crate) const DEFAULT_UNBOUNDED_NUMERIC_SCALE: u32 = 16;
+pub(crate) const DEFAULT_UNBOUNDED_NUMERIC_SCALE: u32 = 9;
 pub(crate) const DEFAULT_UNBOUNDED_NUMERIC_MAX_INTEGRAL_DIGITS: u32 =
     DEFAULT_UNBOUNDED_NUMERIC_PRECISION - DEFAULT_UNBOUNDED_NUMERIC_SCALE;