Inconsistent exception types in code "ᵀrue=1"

[Locked-chess-official]

Bug report

Bug description:

It got ValueError in other place when compile without traceback(I think it should be SyntaxError). In IDLE when you run this code it tips SyntaxError (but in compile and exec it can only be caught with ValueError).

Due to that it raised in compile, using SyntaxError to describe it is better.

CPython versions tested on:

CPython main branch

Operating systems tested on:

No response

Linked PRs

• gh-145675: Raise SyntaxError when constants are used as an identifier #145690

[aisk]

I'm not sure whether we should change it to raise SyntaxError, or just fix the invalid title in IDLE, So I'll tag them both 😂, feel free to change it.

[encukou]

I think this should be SyntaxError -- raising ValueError at compile time is weird.

It looks like there was an assumption that this can only happen when creating an AST node object "manually"?

[StanFromIreland]

ValueError was implemented intentionally: d9b0b07

It does make more sense for it to be a SyntaxError, cc @ZeroIntensity @pablogsal I'll also open a PR to change it.

[terryjreedy]

The non-GUI part of IDLE's Shell is a subclass of code.interactive_interpreter, which used codeop functions to compile code. Code run from an IDLE editor is compiled slightly differently. The two may differ in details. And IDLE may, and I think, possibly should, turn all compile errors into SyntaxErrors if not already. The differences seem more confusing than helpful, especially for beginners.