Cross-datasource union and comparison：SPL Lightweight Multisource Mixed Computation Practices #5

Cross-base mixed-source computation is necessary when data of the same structure is stored in different databases annually. Performing data union is generally similar across various storage systems, with only data retrieval methods varying. This example demonstrates setting up connections to two databases, dba and dbb. Mixed-computation involves uniting data from both tables before performing calculations. The `@x` option closes database connections after queries, and data is unioned using the `|` symbol. Aggregation can then be performed on the combined dataset. Duplicate data handling is crucial, with `group@1` used to discard duplicates after sorting by a key. Mixed-source computations enable data comparison tasks like finding common or unique records between databases. Full joins, intersections, and differences can be achieved using specific functions. For large datasets that exceed memory, the SPL cursor mechanism is employed for mixed-source computations. Cursors can be concatenated for direct processing or merged using functions like `CS.merge()` which offers options for union, intersection, and difference. Computations on cursors begin at the final aggregation step. To handle multiple operations on a single cursor traversal, SPL's cursor reusing (channel) mechanism is utilized. Results can be exported to files if they are too large for memory. SPL facilitates cross-database and cross-source computing goals, and its source code is available on GitHub.