[FLINK-14500] Support Flink Python User-Defined Stateless Function for Table - Phase 2

• Type: New Feature
• Status: Open
• Priority: Minor
• Resolution: Unresolved
• Fix Version/s: 2.0.0
• Component/s: API / Python
• Labels: auto-deprioritized-major auto-unassigned
• Assignee: Unassigned
• Reporter: Dian Fu
• Created: 23/Oct/19 07:07
• Updated: 10/Jul/24 06:57

Description

This is the umbrella Jira which tracks the functionalities of "Python User-Defined Stateless Function for Table" which are planned to be supported in 1.11, such as docker mode support, user-defined metrics support, arrow support, etc.

Issue Links

• relates to: FLINK-14013 Support Flink Python User-Defined Stateless Function for Table

Sub-Tasks

• Support to run the Python worker in docker mode: Open
• Support accessing job parameters in Python user-defined functions: Closed
• Support use-defined metrics in Python user-defined functions: Closed
• User Apache Arrow as the serializer for data transmission between Java operator and Python harness: Closed
• Optimize for Python UDFs with all parameters are constant values: Open
• Support to configure the log level of the Python user-defined functions: Open
• The close method does not work for Python UserDefinedFunction: Closed
• Support python UDFs with constant value of time interval types: Open
• Add Python Table Function Runner And Operator: Closed
• Introduce Python Physical Correlate RelNodes which are containers for Python TableFunction: Closed
• Add Python building blocks to make sure the basic functionality of Python TableFunction could work: Closed
• Add rules to push down the Java Calls contained in Python Correlate node: Closed
• Add rules to transpose the join condition of Python Correlate node: Closed
• Add Python UDTF doc: Closed
• Support Python UDTF in old planner under batch mode: Closed
• Support Python UDTF in blink planner under batch mode: Closed