(*
    ||M||  This file is part of HELM, an Hypertextual, Electronic        
    ||A||  Library of Mathematics, developed at the Computer Science     
    ||T||  Department, University of Bologna, Italy.                     
    ||I||                                                                
    ||T||  HELM is free software; you can redistribute it and/or         
    ||A||  modify it under the terms of the GNU General Public License   
    \   /  version 2 or (at your option) any later version.              
     \ /   This software is distributed as is, NO WARRANTY.              
      V_______________________________________________________________ *)

{ 
   module L  = Log
   module G  = Options
   module AP = AutParser

   let level = 0
   
   let out s = if !G.debug_lexer then L.warn level s else ()

(* This turns an Automath identifier into an XML nmtoken *)
   let quote id =
      let l = String.length id in
      let rec aux i =
         if i < l then begin
            if id.[i] = '\'' || id.[i] = '`' then id.[i] <- '_';
	    aux (succ i)
         end else
	    id
      in
      aux 0
}

let LC  = ['#' '%']
let OC  = "{"
let CC  = "}"
let SPC = [' ' '\t' '\n']+
let NL  = "\n"
let ID  = ['0'-'9' 'A'-'Z' 'a'-'z' '_' '\'' '`']+

rule line_comment = parse
   | NL  { () }
   | OC  { block_comment lexbuf; line_comment lexbuf }
   | _   { line_comment lexbuf }
   | eof { () } 
and block_comment = parse
   | CC  { () }
   | OC  { block_comment lexbuf; block_comment lexbuf }
   | LC  { line_comment lexbuf; block_comment lexbuf  }
   | _   { block_comment lexbuf }
and token = parse
   | SPC      { token lexbuf } 
   | LC       { line_comment lexbuf; token lexbuf  }
   | OC       { block_comment lexbuf; token lexbuf }
   | "_E"     { out "E"; AP.E         }
   | "'_E'"   { out "E"; AP.E         }
   | "---"    { out "EB"; AP.EB       }
   | "'eb'"   { out "EB"; AP.EB       }
   | "EB"     { out "EB"; AP.EB       }
   | "--"     { out "EXIT"; AP.EXIT   }
   | "PN"     { out "PN"; AP.PN       }
   | "'pn'"   { out "PN"; AP.PN       }
   | "PRIM"   { out "PN"; AP.PN       }
   | "'prim'" { out "PN"; AP.PN       }
   | "???"    { out "PN"; AP.PN       }
   | "PROP"   { out "PROP"; AP.PROP   }
   | "'prop'" { out "PROP"; AP.PROP   }
   | "TYPE"   { out "TYPE"; AP.TYPE   }
   | "'type'" { out "TYPE"; AP.TYPE   }
   | ID       { out "ID"; 
                   let s = Lexing.lexeme lexbuf in
                   if !G.unquote then AP.IDENT s else AP.IDENT (quote s)
              }
   | ":="     { out "DEF"; AP.DEF     }
   | "("      { out "OP"; AP.OP       }
   | ")"      { out "CP"; AP.CP       }
   | "["      { out "OB"; AP.OB       }
   | "]"      { out "CB"; AP.CB       }
   | "<"      { out "OA"; AP.OA       }
   | ">"      { out "CA"; AP.CA       }
   | "@"      { out "AT"; AP.AT       }
   | "~"      { out "TD"; AP.TD       }
   | "\""     { out "QT"; AP.QT       }
   | ":"      { out "CN"; AP.CN       }   
   | ","      { out "CM"; AP.CM       }
   | ";"      { out "SC"; AP.SC       }
   | "."      { out "FS"; AP.FS       }   
   | "+"      { out "PLUS"; AP.PLUS   }
   | "-"      { out "MINUS"; AP.MINUS }
   | "*"      { out "TIMES"; AP.TIMES }
   | "="      { out "DEF"; AP.DEF     }
   | eof      { out "EOF"; AP.EOF     }