Mishandling of document with only NULL bytes

[moorereason]

Given a TOML document of any(?) length made with only NULL bytes, toml11 returns a valid, empty document:

$ echo -n '\x00\x00\x00\x00' | check_toml_test
{}

I expect this input to be invalid.

AFAICT if there are any non-NULL bytes after the leading NULL bytes, the parser given an error as expected.

Found while doing differential fuzzing against go-toml/v2.

[ToruNiina]

Thank you for reporting! Historically I added this for the null terminated string, but currently we don't need this. I removed it.