Can antlr be applied to decompiled code?

[kimbjunn]

I know that there are many cases where various grammatical errors actually exist in decompiled codes generated by various decompilers such as IDA or Ghidra.

Can Antlr4 successfully generate AST trees for decompiled codes despite such grammatical errors?

For example, decompiled codes generated by IDA often contain labels without colons, such as
LAB_12144
return 0;

, or codes generated by Ghidra often contain undefined ghidra-specific data types, such as
undefined1 u1;.

I wonder if it is possible to parse these cases without problems.