feat(shannon): Reading and writing open table format

[ShannonBase]

To support reading and writing the open table format, such as parquets. which will enable Shannonbase lakehouse processing abilities. Nowdays, volume of data is getting larger, and the type of data is going to be more versatile, such as structured data, semi-structured data, unstructured data.

MySQL Heatwave, Databricks, etc. are all support data lakehouse.

[RingsC]

Using AFSL(Abstract File System Layer) or specific engine to operate on specific engine?

for example:
create table test (xxx) engine = innodb
ENGINE_ATTRIBUTE='{"file":
[{"region":"",
"namespace":"",
"bucket":"<bucket_name>",
"name":"<orders_file_location>"}]}';

[ShannonBase]

using create tablespace. When we ar creating a table space, we can specify the path of space file path. then using this tablespace to create a table.

by this, we can locate shannon rapid table to a NVME disk.

ref: https://dev.mysql.com/doc/refman/8.0/en/create-tablespace.html