Parsing a string with "boundaries"

Hello.

I have had trouble coming up with a grammar for parsing a string that has a specific char at its “boundaries” in a greater path-like structure but which may also contain the boundary char internally.

So if the boundary char is * I need to pick out *foobar* in e.g. /foo/*foobar*/bar. My problem is that the *foobar* string may also contain the “boundary char” e.g. *foo*bar* so I cant use a negative char range for the parsing of the token between the /s.

Is it possible to do this with a PEG? And apologies if this is not the place for questions like this.

[jonathanbp]

Great! Thanks a lot for your help, @Mingun and @hildjj - it is much appreciated.

[hildjj]

Taking mingun’s grammar as a start, I came up with this, which gets all of the edge cases I could think of:

path = "/"? head:segment tail:("/" @segment)* "/"? { return [head, ...tail] }

segment
  = special
  / normal
  
special = "*" text:$special_char+ "*" { return { type: "special", text } }
normal = text:$[^/]+ { return { type: "normal", text } }

special_char
  = [^*]      // Any non-* is valid
  / "*" &[^/] // Following a * with anything but a / makes it a part of the text