PyFlink中通过TableEnvironment.explain_sq查看执行计划的步骤是什么？

== Abstract Syntax Tree ==
LogicalSink(table=[default_catalog.default_database.my_sink], fields=[EXPR$0])
+- LogicalProject(EXPR$0=[sub_string($0, 2, 4)])
   +- LogicalTableScan(table=[[default_catalog, default_database, Unregistered_TableSource_1143388267, source: [PythonInputFormatTableSource(a)]]])

== Optimized Logical Plan ==
Sink(table=[default_catalog.default_database.my_sink], fields=[EXPR$0])
+- PythonCalc(select=[sub_string(a, 2, 4) AS EXPR$0])
   +- LegacyTableSourceScan(table=[[default_catalog, default_database, Unregistered_TableSource_1143388267, source: [PythonInputFormatTableSource(a)]]], fields=[a])

== Physical Execution Plan ==
Stage 1 : Data Source
    content : Source: PythonInputFormatTableSource(a)

    Stage 2 : Operator
        content : SourceConversion(table=[default_catalog.default_database.Unregistered_TableSource_1143388267, source: [PythonInputFormatTableSource(a)]], fields=[a])
        ship_strategy : FORWARD

        Stage 3 : Operator
            content : StreamExecPythonCalc
            ship_strategy : FORWARD

            Stage 4 : Data Sink
                content : Sink: Sink(table=[default_catalog.default_database.my_sink], fields=[EXPR$0])
                ship_strategy : FORWARD