fix(parser): preserve quoted-key text in redefinition error messages

[SAY-5]

What does this change do?

Fixes the corrupted quoted-key string that appears in redefinition error messages. A key written as ["foo"] previously produced an error of the form cannot redefine existing table '"fofoo"'. With this change the diagnostic shows "foo", as written in the source.

Is it related to an exisiting bug report or feature request?

Fixes #300.

Pre-merge checklist

• I've read CONTRIBUTING.md
• I've rebased my changes against the current HEAD of origin/master (if necessary)
• I've added new test cases to verify my change
• I've regenerated toml.hpp (how-to)
• I've updated any affected documentation
• I've rebuilt and run the tests with at least one of:
  • Clang 8 or higher
  • GCC 8 or higher
  • MSVC 19.20 (Visual Studio 2019) or higher
• I've added my name to the list of contributors in README.md

[marzer]

This fix looks very LLM-coded. The way the comments are written, and the fact that my PR template was annihilated wholesale, yet it's requirements were still referenced.

... Am I right? If so, please remove the LLM junk comments, restore the PR description, and complete the template by hand.

[SAY-5]

Acknowledged, closing. Sorry for the noise.

[marzer]

@SAY-5 to be clear, I don't have a problem with you using an LLM to arrive at a fix or whatever, just don't blindly go "cursor go brrrrr" in submitting it.

If you want to contribute your fix while also observing project norms I'll happily accept it.

[SAY-5]

@marzer apologies for the noise earlier. Reopened with three things addressed:

1. Trimmed the two LLM-style narrative comments in parser.inl down to a single line near the snapshot, since the variable name and the surrounding go_back(2u) already explain the rollback.
2. Trimmed the matching comments in the regenerated toml.hpp.
3. Restored the PR description so it follows the template verbatim and ticked the boxes that actually apply.

The fix itself is unchanged. The PR currently lacks (a) updating README.md contributors and (b) verifying on Clang/MSVC. Happy to add the README entry on request, and to add those compiler check-marks once I can verify on those toolchains.

[marzer]

Thanks, I appreciate it.

• I'm ok with CI hitting the other toolchains, so long as you've hit it on at least one.
• Adding to the contributor list is optional, of course :). I added that because I wanted to ensure people remembered to record their work here, and it would survive git. Totally OK if you don't want to do that, no dramas.

I'll look at the rest now.

[marzer]

Looks good, happy to merge pending any last-second editorial changes.