[EPIC] Support converting Exprs and LogicalPlans --> SQL Strings

[alamb]

This ticket tracks various tickets related to converting Exprs and LogicalPlans to/from SQL strings

• Support parsing SQL strings to Exprs #8736
• Support LogicalPlan --> SQL String #8661
• Complete support for Expr --> String  #9726
• Basic support for converting Expr to SQL string #9495
• Add an example of how to use the SQL parser/unparser API #10524
• Add an example of how to convert LogicalPlan to/from SQL Strings #10550
• Make SQL strings generated from Exprs "prettier" #10557
• Make SQL strings generated from Exprs even "prettier" #10633

[devinjdangelo]

I think that we now have a solid enough foundation on these tasks that it would be a good first issue for a new contributor (#9495 especially). An individual or small number of logical Exprs could be handled in a stand alone PR with a test or two. I think this project is a great way to get more familiar with DataFusion's core data structures and optimization algorithms without already having expertise in database internals.

[alamb]

@devinjdangelo do you think we should file explicit follow on tickets, or do you think the existing PRs / documentation is clear enough about the ask (for someone who has no experience with DataFusion)?

I can file tickets if you prefer

[devinjdangelo]

@alamb I don't think we should file a ticket for every remaining Expr variant, but we could edit the description of #9495 to list the remaining unimplemented variants and make clear that we are looking for help for anyone interested to implement even just 1 at a time in a PR + write a test.

Here is the remaining list:

• InList
• ScalarFunction
• Between
• Case
• WindowFunction
• Like
• ScalarVariable
• SimilarTo
• Not
• IsNotNull
• IsTrue
• IsFalse
• IsUnknown
• IsNotTrue
• IsNotUnknown
• Negative
• GetIndexedField
• TryCast
• Sort
• Exists
• Wildcard
• GroupingSet
• Placeholder
• OuterReferenceColumn
• Unnest

[alamb]

@alamb I don't think we should file a ticket for every remaining Expr variant, but we could edit the description of #9495 to list the remaining unimplemented variants and make clear that we are looking for help for anyone interested to implement even just 1 at a time in a PR + write a test.

Thank yoU @devinjdangelo -- I filed #9726