[CH] Bug fix set empty_as_default for nullable<String> type field

### Backend

CH (ClickHouse)

### Bug description

If we set the config `spark.gluten.sql.text.input.empty.as.default` as `true`, then when enconter a empty string for field, it would parse the text input field as NULL, and this is different from spark, as spark would parse it as empty string

### Spark version

Spark-3.3.x

### Spark configurations

_No response_

### System information

_No response_

### Relevant logs

_No response_

No source file, test, or entry point is named. Reproduce the ClickHouse text-input case with spark.gluten.sql.text.input.empty.as.default=true and a nullable field, then trace the parsing path; done means an empty input remains an empty string, matching Spark 3.3.x rather than becoming NULL.

Written by the indexing model from the issue text.