Skip to content

Improve the scalaExpr parser #2

@domna

Description

@domna

The current scalaExpr Regex has the form [^%]* and is automatically converted into a parser which is used in the grammer to parse the file into blocks (whether it is a prolog or a scala part). But the problem with this Regex is, that it parses everything but the % character. This was a workaround for me that the Regex stops consuming everything in my file and recognizes the prolog block of the file.
The problem is that if there is any % character (even in strings) in the scala definiton above the prolog block the result is a parser error.

So there must be a way to stop the regex consuming everything and also have the possibility to write every character in the scala def above.

Metadata

Metadata

Assignees

Labels

No labels
No labels

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions