[xy] Fix spark dataframe read error.

[wangxiaoyou1993]

Description

Fix spark dataframe read error.
Do not write json file in spark data directory to avoid read error

java.lang.RuntimeException: file:/home/src/mage_data/.../output_0/type.json is not a Parquet file. 
Expected magic number at tail, but found [109, 101, 34, 125]

How Has This Been Tested?

• Tested with local spark docker image

Checklist

• The PR is tagged with proper labels (bug, enhancement, feature, documentation)
• I have performed a self-review of my own code
• I have added unit tests that prove my fix is effective or that my feature works
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• If new documentation has been added, relative paths have been added to the appropriate section of docs/mint.json

cc: