gistfile1.clj

;Hrm... I actually tried a similar approach before but:

(def identifier2
  (around
   whitespaces
   (choice
    [(*> (string "#")
         (<* number (string "#")))
     (*> (string "#")
         (<* number (string "=")))])))

(z/parse-once identifier2 "#234=")
#zetta.core.ResultFailure{:remainder (\=), :stack [], :msg "Failed reading: take-with"}
parse> (z/parse-once identifier2 "#234#")
#zetta.core.ResultDone{:remainder (), :result 234}

;I think i probably should read some attoparsec tutorials -- I'm pretty ignorant of the basics so I'm not ;sure where to begin fixing (i.e. maybe I've just built things wrong or loaded something incorrectly...)

If you change the order on the choice vector, the = works but the # doesn't...

uhmm it seems we have an issue here with the buffer not having the content back after going to the next choice :-/

Need to figure out why that's happening

(def id3 
  (around whitespaces (*> (char \#) 
                                          (<* number (choice [(char \=) 
                                                                           char \#)])))))

Try this one out

Yep that parses for me now. Am I able to distinguish between which choice is made with this approach? (i.e I want #'s to mean reference ids and ='s to mean id definitions)

I guess the trick would be not to throw away the parsed = and #, but throw the number and the matched char to a higher function with <$> ?