Summary

• Account page decode memory before IChunkReader#nextPageData() in the Pipe scan parser.
• Add a memory-controlled chunk reader wrapper for normal and aligned multi-page chunks.
• Estimate page decompression plus BatchData/vector object memory and use it for splitting aligned chunks.
• Add a regression test for high-compression aligned multi-page chunks.

Tests

• mvn -pl iotdb-core/datanode spotless:apply
• git diff --check
• mvn -pl iotdb-core/datanode -DskipTests compile failed due to unrelated existing datanode compile errors in schema/query modules.