Loading parquet file using pipeline inserts date type values as 0000-00-00

Hi I am trying to load parquet file from s3 to single store , one of the filed in parquet has date datatype and value as 2020-07-07 [yyyy-mm-dd], while loading this parquet file into memsql i get values as 0000-00-00. I tried cast and to_date etc, nothing worked out. any help on this appreciated.

DROP PIPELINE IF EXISTS test_pipeline;

CREATE OR REPLACE PIPELINE test_pipeline AS
LOAD DATA S3 ‘test_folder/test.parquet/c=*/*.parquet’
CONFIG ‘{“region”: “xxx”}’
CREDENTIALS ‘{“aws_access_key_id”: “xxxx”, “aws_secret_access_key”: “xxxxx”}’
INTO TABLE test_tbl (a <- a,
b <- b,
@c <- c,
d <- d
)
FORMAT PARQUET
SET c = cast(@c as date)
;

This thread should be applicable: DATE format support while loading data from PARQUET file which located in S3

We convert Parquet int96 timestamps to a format directly compatible with e.g. DATETIME, but Parquet writers can also represent times as plain old integer parquet values representing offsets from 1970-01-01, and we don’t yet perform automatic conversion for those. The workaround is to use timestampadd() as described, comparing the integer values against the expected date to guess what unit the integer offset is in, and whether the time should be interpreted as UTC or as local timezone time.