I'm trying to implement an interpreter for a language that is first parsed by TatSu and then interpreted. I'd like to deliver to my end-users the functionality that my interpreter reports as many potential errors contained in the source of the language at once as possible, including FailedParse errors raised by the underlying TatSu parser.
The only way I know of is that TatSu generates only a single FailedParse exception and then stops parsing.
Is there a way to wrap the TatSu parser in such a way that it resumes parsing the source and reports any further potential syntax errors in the source?