[code-reflection] RFR: Fix test failure after PR/51

[Paul Sandoz]

On Fri, 26 Apr 2024 17:46:53 GMT, Maurizio Cimadamore <mcimadamore at openjdk.org> wrote:

> A test is failing after my push of https://git.openjdk.org/babylon/pull/51
> 
> The problem is that I've tweaked the logic for parsing type-variables, so that now we check for a starting `#` in the core type factory.
> But the `#` char is not copied into the identifier, hence the issue.
> 
> Unfortunately sometimes tests seem to pass even though they are run after a `make images`. The problem is that after changes in the `j.l.reflect.code` packages, only `java.base` is rebuilt, and the `jdk.compiler` copy is not. This means that, unless one does `make images` all the time, some tests might pass (or fail) spuriously. We might need to look at the build again.

Opps i should have spotted that.

-------------

Marked as reviewed by psandoz (Lead).

PR Review: https://git.openjdk.org/babylon/pull/63#pullrequestreview-2025561066