[fix](insert) fix be core dump due to using the wrong schema to create column writers

[luozenglin]

Proposed changes

Issue Number: close #16658

Problem summary

Status SegmentWriter::_create_writers_with_block(
        const vectorized::Block* block,
        std::function<Status(uint32_t, const TabletColumn&)> create_column_writer) {
    ...
    for (size_t i = 0; i < block->columns(); ++i) {
        const auto& column_type_name = block->get_by_position(i);
        auto idx = _tablet_schema->field_index(column_type_name.name);
        if (idx >= 0) {
            RETURN_IF_ERROR(create_column_writer(i, _tablet_schema->column(idx)));
        } 
    ...
    }
    return Status::OK();
}

In qt_sql1 of test_insert.groovy, the schema of block in the above code is k6, k3, k1, sum(k1), while _tablet_schema is k1, k2, k3, k4. k3 in block and k3 in _tablet_schema have the same column name but not the corresponding column, which causes an error.

Checklist(Required)

• Does it affect the original behavior
• Has unit tests been added
• Has document been added or modified
• Does it need to update dependencies
• Is this PR support rollback (If NO, please explain WHY)

Further comments

If this is a relatively large or complex change, kick off the discussion at dev@doris.apache.org by explaining why you chose the solution you did and what alternatives you considered, etc...

[github-actions]

clang-tidy review says "All clean, LGTM! 👍"

[github-actions]

clang-tidy review says "All clean, LGTM! 👍"

[hello-stephen]

TeamCity pipeline, clickbench performance test result:
the sum of best hot time: 36.14 seconds
stream load tsv: 486 seconds loaded 74807831229 Bytes, about 146 MB/s
stream load json: 38 seconds loaded 2358488459 Bytes, about 59 MB/s
stream load orc: 68 seconds loaded 1101869774 Bytes, about 15 MB/s
stream load parquet: 28 seconds loaded 861443392 Bytes, about 29 MB/s
https://doris-community-test-1308700295.cos.ap-hongkong.myqcloud.com/tmp/20230216035451_clickbench_pr_97424.html