Converting RTE to PNSI

Problem Statement

PostgreSQL version 13(pg 13) uses PNSI instead of RTE in many functions. PNSI also exist in some of functions in PG11 and PG12 along with RTE as well. What we want is, find a way or research about converting RTE to PNSI in PG11 and PG12 version. If we are successful, multiple versions of PostgreSQL can be closer to each other. Which can help us in the development of the versions.

RTE

RangeTableEntry, or RTE for short, is a key idea in PostgreSQL query processing. Within a query, it represents tables or subqueries. These entities' names, aliases, columns, join criteria, and access methods are all important metadata that RTE keeps. Planning and running queries require this information.

PNSI

Another key idea in PostgreSQL query processing is the ParseNameSpaceItem (PNSI). It represents a variety of items mentioned in a query, including as tables, columns, functions, and more. These objects' names, kinds, aliases, and other pertinent metadata are recorded by PNSI. PNSI makes ensuring that object referencing is right, handles conflicts, and upholds query accuracy during query processing and analysis.

Relation

PNSI encapsulates RTE as a part of its overall structure. When a table or subquery is referenced in a query, an RTE is created to represent that specific object. This RTE is then included within the PNSI structure, which manages the namespace information of the entire query. In this way, RTEs are encompassed within the broader PNSI framework.

Feasibility of replacement

According to my understanding, as RTE is a part in PNSI we can access and work with the RTE-related information within its structure. PNSI serves as a higher-level container that manages the namespace information, including RTEs representing tables and subqueries. This allows for a unified approach to handle various objects referenced in a query, while still providing access to the specific metadata and optimizations associated with RTE.
So, while RTE and PNSI have their own roles and functionalities, PNSI provides a way to encompass and work with RTE-specific parts as needed. This encapsulation allows for a more comprehensive and flexible query processing framework, enabling efficient object referencing, conflict resolution, and optimized query planning.

Conclusion

In conclusion, the conversion of RangeTableEntry (RTE) to ParseNameSpaceItem (PNSI) in PostgreSQL versions 11 and 12 is a promising avenue for bringing multiple versions of PostgreSQL closer together. This transformation allows for a unified approach to query processing, enabling efficient object referencing, conflict resolution, and optimized query planning, while maintaining access to RTE-specific metadata and optimizations.