Pep and Nom

the ℕ𝕠𝕞 delimiter register

The (grammar) parse-token delimiter register in the ℙ𝕖𝕡 machine

The delimiter register determines what character will be used for delimiting parse tokens on the stack when using the “push” and “pop ” commands. This can be set with the delim command. The default parse-token delimiter is the '*' asterisk character, but this really a matter of personal preference.

    # fragment
    begin { delim '/'; }
    read; "(",")" { put; clear; add "bracket/"; push; }  
  

The delimiter character will often (almost always) be set in a 'begin' block. I normally use the “*” character, but “/” or “;” might be good options. Apart from the parse-token delimiter character, any character (including spaces) may be used in parse tokens.

    # script fragment
    "noun phrase*verb phrase*" {
      clear; add "sentence*"; push; .reparse 
    }
  

The ℕ𝕠𝕞 script below prints only capitalized words, one per line.

    # use [A-Z] instead of [:alpha:] for good old ascii
    # [:upper:] should be nicely unicode when translated to 
    # go, java, python, etc
    read; [:upper:] { 
      while [:upper:]; put; clear; add "cap words*"; push; 
    }
    clear;
  parse>
    pop;pop; 
    "cap words*cap words*" {
      # compile (concatenate) the text attributes
      clear; get; add "\n"; ++; get; --; put;
      clear; add "cap words*"; push; .reparse
    }
    push; push;
    (eof) {
      pop; add "\n"; get; add "\n"; print; quit;
    }
  

notes