support count(*)/count(1)

Changelog category (leave one):

• New Feature

Changelog entry (a user-readable short description of the changes that goes to CHANGELOG.md):
relate to issue #45

Read from file sources or java batch iterator will fail with empty columns to read. This happends when running the following queries.

select count(1) from t;
select sum(1) from t;
select 1 from t;

In this case, the rows number is the only information to need. To solve this, we try to build a block with an virtual column which indicates the rows number, and this virtual column will not be really used.

In SubstraitFileSource, if the file implement getTotalRows, we build blocks with a constant column whose total size is equal to getTotalRows's result.

In SourceFromJavaIter, we also return blocks with a constant column whose size is equal to input block's rows.

For merge tree, we try to read a least one column instead of the original empty header.

Information about CI checks: https://clickhouse.tech/docs/en/development/continuous-integration/