;;; semantic-grammar.wy -- LALR grammar of Semantic input grammars
;;
;; Copyright (C) 2002, 2003, 2004 David Ponce
;;
;; Author: David Ponce <david@dponce.com>
;; Maintainer: David Ponce <david@dponce.com>
;; Created: 26 Aug 2002
;; Keywords: syntax
;; X-RCS: $Id: semantic-grammar.wy,v 1.16 2005/09/30 20:20:27 zappo Exp $
;;
;; This file is not part of GNU Emacs.
;;
;; This program is free software; you can redistribute it and/or
;; modify it under the terms of the GNU General Public License as
;; published by the Free Software Foundation; either version 2, or (at
;; your option) any later version.
;;
;; This software is distributed in the hope that it will be useful,
;; but WITHOUT ANY WARRANTY; without even the implied warranty of
;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
;; General Public License for more details.
;;
;; You should have received a copy of the GNU General Public License
;; along with GNU Emacs; see the file COPYING.  If not, write to the
;; Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
;; Boston, MA 02110-1301, USA.

%{
  ;; Current parsed nonterminal name.
  (defvar semantic-grammar-wy--nterm nil)
  ;; Index of rule in a nonterminal clause.
  (defvar semantic-grammar-wy--rindx nil)
  }

;;%package semantic-grammar-wy

%languagemode wy-mode

;; Main
%start grammar
;; Reparse
%start prologue epilogue declaration nonterminal rule
;; EXPANDFULL
%start put_names put_values use_names

;; Keywords
%type    <keyword>
%keyword DEFAULT-PREC    "%default-prec"
%keyword NO-DEFAULT-PREC "%no-default-prec"
%keyword KEYWORD         "%keyword"
%keyword LANGUAGEMODE    "%languagemode"
%keyword LEFT            "%left"
%keyword NONASSOC        "%nonassoc"
%keyword PACKAGE         "%package"
%keyword PREC            "%prec"
%keyword PUT             "%put"
%keyword QUOTEMODE       "%quotemode"
%keyword RIGHT           "%right"
%keyword SCOPESTART      "%scopestart"
%keyword START           "%start"
%keyword TOKEN           "%token"
%keyword TYPE            "%type"
%keyword USE-MACROS      "%use-macros"

;; Literals
%type  <string>
%token <string>      STRING

%type  <symbol>      syntax ":?\\(\\sw\\|\\s_\\)+"
%token <symbol>      SYMBOL
%token <symbol>      PERCENT_PERCENT "\\`%%\\'"

%type  <char>        syntax semantic-grammar-lex-c-char-re
%token <char>        CHARACTER

%type  <qlist>       matchdatatype sexp syntax "\\s'\\s-*("
%token <qlist>       PREFIXED_LIST

%type  <sexp>        matchdatatype sexp syntax "\\="
%token <sexp>        SEXP

;; Don't generate these analyzers which needs special handling code.
%token <code>        PROLOGUE "%{...%}"
%token <code>        EPILOGUE "%%...EOF"

;; Blocks & Parenthesis
%type  <block>
%token <block>       PAREN_BLOCK "(LPAREN RPAREN)"
%token <block>       BRACE_BLOCK "(LBRACE RBRACE)"
%token <open-paren>  LPAREN      "("
%token <close-paren> RPAREN      ")"
%token <open-paren>  LBRACE      "{"
%token <close-paren> RBRACE      "}"

;; Punctuations
%type  <punctuation>
%token <punctuation> COLON       ":"
%token <punctuation> SEMI        ";"
%token <punctuation> OR          "|"
%token <punctuation> LT          "<"
%token <punctuation> GT          ">"

%%

grammar:
    prologue
  | epilogue
  | declaration
  | nonterminal
  | PERCENT_PERCENT
  ;

;;; Prologue/Epilogue
;;
prologue:
    PROLOGUE
    (CODE-TAG "prologue" nil)
  ;

epilogue:
    EPILOGUE
    (CODE-TAG "epilogue" nil)
  ;

;;; Declarations
;;
declaration:
    decl
    (eval $1)
  ;

decl:
    default_prec_decl
  | no_default_prec_decl
  | languagemode_decl
  | package_decl
  | precedence_decl
  | put_decl
  | quotemode_decl
  | scopestart_decl
  | start_decl
  | keyword_decl
  | token_decl
  | type_decl
  | use_macros_decl
  ;

default_prec_decl:
    DEFAULT-PREC
    `(TAG "default-prec" 'assoc :value '("t"))
  ;

no_default_prec_decl:
    NO-DEFAULT-PREC
    `(TAG "default-prec" 'assoc :value '("nil"))
  ;

languagemode_decl:
    LANGUAGEMODE symbols
    `(TAG ',(car $2) 'languagemode :rest ',(cdr $2))
  ;

package_decl:
    PACKAGE SYMBOL
    `(PACKAGE-TAG ',$2 nil)
  ;

precedence_decl:
    associativity token_type_opt items
    `(TAG ',$1 'assoc :type ',$2 :value ',$3)
  ;

associativity:
    LEFT
    (progn "left")
  | RIGHT
    (progn "right")
  | NONASSOC
    (progn "nonassoc")
  ;

put_decl:
    PUT put_name put_value
    `(TAG ',$2 'put :value ',(list $3))
  | PUT put_name put_value_list
    `(TAG ',$2 'put :value ',$3)
  | PUT put_name_list put_value
    `(TAG ',(car $2) 'put :rest ',(cdr $2) :value ',(list $3))
  | PUT put_name_list put_value_list
    `(TAG ',(car $2) 'put :rest ',(cdr $2) :value ',$3)
  ;

put_name_list:
    BRACE_BLOCK
    (mapcar 'semantic-tag-name (EXPANDFULL $1 put_names))
  ;

put_names:
    LBRACE
    ()
  | RBRACE
    ()
  | put_name
 ;; Must return a list of Semantic tags to EXPANDFULL!
    (TAG $1 'put-name)
  ;

put_name:
    SYMBOL
  | token_type
  ;

put_value_list:
    BRACE_BLOCK
    (mapcar 'semantic-tag-code-detail (EXPANDFULL $1 put_values))
  ;

put_values:
    LBRACE
    ()
  | RBRACE
    ()
  | put_value
 ;; Must return a list of Semantic tags to EXPANDFULL!
    (CODE-TAG "put-value" $1)
  ;

put_value:
    SYMBOL any_value
    (cons $1 $2)
  ;

scopestart_decl:
    SCOPESTART SYMBOL
    `(TAG ',$2 'scopestart)
  ;

quotemode_decl:
    QUOTEMODE SYMBOL
    `(TAG ',$2 'quotemode)
  ;

start_decl:
    START symbols
    `(TAG ',(car $2) 'start :rest ',(cdr $2))
  ;

keyword_decl:
    KEYWORD SYMBOL string_value
    `(TAG ',$2 'keyword :value ',$3)
  ;

token_decl:
    TOKEN token_type_opt SYMBOL string_value
    `(TAG ',$3 ',(if $2 'token 'keyword) :type ',$2 :value ',$4)
  | TOKEN token_type_opt symbols
    `(TAG ',(car $3) 'token :type ',$2 :rest ',(cdr $3))
  ;

token_type_opt:
 ;; EMPTY
  | token_type
  ;

token_type:
    LT SYMBOL GT
    (progn $2)
  ;

type_decl:
    TYPE token_type plist_opt
    `(TAG ',$2 'type :value ',$3)
  ;

plist_opt:
 ;;EMPTY
  | plist
  ;

plist:
    plist put_value
    (append (list $2) $1)
  | put_value
    (list $1)
  ;

use_name_list:
    BRACE_BLOCK
    (mapcar 'semantic-tag-name (EXPANDFULL $1 use_names))
  ;

use_names:
    LBRACE
    ()
  | RBRACE
    ()
  | SYMBOL
 ;; Must return a list of Semantic tags to EXPANDFULL!
    (TAG $1 'use-name)
  ;

use_macros_decl:
    USE-MACROS SYMBOL use_name_list
    `(TAG "macro" 'macro :type ',$2 :value ',$3)
  ;

string_value:
    STRING
    (read $1)
  ;

;; Return a Lisp readable form
any_value:
    SYMBOL
  | STRING
  | PAREN_BLOCK
  | PREFIXED_LIST
  | SEXP
  ;

symbols:
    lifo_symbols
    (nreverse $1)
  ;

lifo_symbols:
    lifo_symbols SYMBOL
    (cons $2 $1)
  | SYMBOL
    (list $1)
  ;

;;; Grammar rules
;;
nonterminal:
    SYMBOL
    (setq semantic-grammar-wy--nterm $1
          semantic-grammar-wy--rindx 0)
    COLON rules SEMI
    (TAG $1 'nonterminal :children $4)
  ;

rules:
    lifo_rules
    (apply 'nconc (nreverse $1))
  ;

lifo_rules:
    lifo_rules OR rule
    (cons $3 $1)
  | rule
    (list $1)
  ;

rule:
    rhs
    (let* ((nterm semantic-grammar-wy--nterm)
           (rindx semantic-grammar-wy--rindx)
           (rhs   $1)
           comps prec action elt)
      (setq semantic-grammar-wy--rindx (1+ semantic-grammar-wy--rindx))
      (while rhs
        (setq elt (car rhs)
              rhs (cdr rhs))
        (cond
         ;; precedence level
         ((vectorp elt)
          (if prec
              (error "duplicate %%prec in `%s:%d' rule" nterm rindx))
          (setq prec (aref elt 0)))
         ;; action
         ((consp elt)
          ;; don't forget that rhs items are in reverse order, so
          ;; the end-of-rule semantic action is the first item.
          (if (or action comps)
              ;; a mid-rule action
              (setq comps (cons elt comps)
                    ;; keep rule and action index synchronized
                    semantic-grammar-wy--rindx
                    (1+ semantic-grammar-wy--rindx))
            ;; the end-of-rule action
            (setq action (car elt))))
         ;; item
         (t
          (setq comps (cons elt comps)))))
      (EXPANDTAG
       (TAG (format "%s:%d" nterm rindx) 'rule
            :type (if comps "group" "empty")
            :value comps :prec prec :expr action)))
  ;

rhs:
 ;; EMPTY
  | rhs item
    (cons $2 $1)
  | rhs action
    (cons (list $2) $1)
  | rhs PREC item
    (cons (vector $3) $1)
  ;

action:
    PAREN_BLOCK
  | PREFIXED_LIST
  | BRACE_BLOCK
    (format "(progn\n%s)"
            (let ((s $1))
              (if (string-match "^{[\r\n\t ]*" s)
                  (setq s (substring s (match-end 0))))
              (if (string-match "[\r\n\t ]*}$" s)
                  (setq s (substring s 0 (match-beginning 0))))
              s))
  ;

items:
    lifo_items
    (nreverse $1)
  ;

lifo_items:
    lifo_items item
    (cons $2 $1)
  | item
    (list $1)
  ;

item:
    SYMBOL
  | CHARACTER
  ;

%%

(define-lex semantic-grammar-lexer
  "Lexical analyzer that handles Semantic grammar buffers.
It ignores whitespaces, newlines and comments."
  semantic-lex-ignore-newline
  semantic-lex-ignore-whitespace
  ;; Must detect prologue/epilogue before other symbols/keywords!
  semantic-grammar-lex-prologue
  semantic-grammar-lex-epilogue
  semantic-grammar-wy--<keyword>-keyword-analyzer
  semantic-grammar-wy--<symbol>-regexp-analyzer
  semantic-grammar-wy--<char>-regexp-analyzer
  semantic-grammar-wy--<string>-sexp-analyzer
  ;; Must detect comments after strings because `comment-start-skip'
  ;; regexp match semicolons inside strings!
  semantic-lex-ignore-comments
  ;; Must detect prefixed list before punctuation because prefix chars
  ;; are also punctuations!
  semantic-grammar-wy--<qlist>-sexp-analyzer
  ;; Must detect punctuations after comments because the semicolon can
  ;; be a punctuation or a comment start!
  semantic-grammar-wy--<punctuation>-string-analyzer
  semantic-grammar-wy--<block>-block-analyzer
  semantic-grammar-wy--<sexp>-sexp-analyzer
  )

;;; semantic-grammar.wy ends here