Where it is written
https://docs.godotengine.org/en/stable/tutorials/scripting/gdscript/warning_system.html
The names of the warnings to ignore match the names of the project settings. For example, to ignore the warning configured by the debug/gdscript/warnings/unused_variable project setting, use @warning_ignore("unused_variable").
What it would report
@warning_ignore("not_a_real_warning_name")
var y: int = 0A name that matches no warning silently suppresses nothing. The author believes
a warning is handled; it is not, and no diagnostic says so. gdck check
reports the file clean today.
Same silent-failure shape as #3 — an annotation that looks right, does nothing,
and is never complained about.
Applies to @warning_ignore, @warning_ignore_start and
@warning_ignore_restore.
Could it be fixed automatically?
Only when the name is a near-miss of a real one, and even then it is a
suggestion rather than a fix the author did not choose. Report only.
Where it would be wrong
This one carries a maintenance cost rather than a false-positive risk: it needs
the list of valid warning names, that list is Godot's and moves between
versions, and a project on a newer Godot than the list would see correct code
reported. That is worse than saying nothing.
So it only makes sense with a decision about where the list comes from and what
happens when it is out of date — erring toward silence on an unknown name is
not an option, since an unknown name is the whole point of the rule. Possibly
the list belongs behind a configured Godot version, defaulting to the newest
gdck knows and reporting nothing if the project declares something newer.
Lower priority than #3 for that reason.
Where it is written
https://docs.godotengine.org/en/stable/tutorials/scripting/gdscript/warning_system.html
What it would report
A name that matches no warning silently suppresses nothing. The author believes
a warning is handled; it is not, and no diagnostic says so.
gdck checkreports the file clean today.
Same silent-failure shape as #3 — an annotation that looks right, does nothing,
and is never complained about.
Applies to
@warning_ignore,@warning_ignore_startand@warning_ignore_restore.Could it be fixed automatically?
Only when the name is a near-miss of a real one, and even then it is a
suggestion rather than a fix the author did not choose. Report only.
Where it would be wrong
This one carries a maintenance cost rather than a false-positive risk: it needs
the list of valid warning names, that list is Godot's and moves between
versions, and a project on a newer Godot than the list would see correct code
reported. That is worse than saying nothing.
So it only makes sense with a decision about where the list comes from and what
happens when it is out of date — erring toward silence on an unknown name is
not an option, since an unknown name is the whole point of the rule. Possibly
the list belongs behind a configured Godot version, defaulting to the newest
gdckknows and reporting nothing if the project declares something newer.Lower priority than #3 for that reason.