view lisp/progmodes/ebnf-iso.el @ 51151:fe11e703042b

Summary: MIME support added for e-mail processing that skips encoded regions. Allow user to skip saving Fcc messages with large attachments. Fixed region skipping bug with multi-line comments - e.g. tex $ regions spanning multiple lines. Added support for postscript and uuencoded regions. Redundant dictionary file names purged. Dictionary definition field name changed from "Character Set" to "Coding System". Fixed bug in reloading dictionaries. Modified headers to reflect new version. XEmacs menu now adds customize item. (ispell-check-version): No longer an aliased function. Returns library path if not called interactively. Variable `temporary-file-directory' protected if not loaded. (check-ispell-version): Now the alias for `ispell-check-version'. (ispell-message-fcc-skip): New variable that determines if and when to query about saving Fcc copy of message if an attachment is large. (ispell-skip-html): Declared buffer-local. (ispell-local-dictionary-alist): Docstring expanded. Tag name changed from "Character Set" to "Coding System". (ispell-dictionary-alist-1): Removed redundant command-line option to load brasileiro, british, and castellano dictionary files. (ispell-dictionary-alist-2): Removed redundant command-line option to load czech dictionary file. (ispell-dictionary-alist-3): Moved francais-tex here. (ispell-dictionary-alist-4): Removed german and german8 dictionaries. The deutsch ones are the correct definitions. `nederlands' and `nederlands8' dictionaries moved here. (ispell-dictionary-alist-5): `polish' and `portugues' dictionaries moved here. Removed redundant command-line option to `norsk' and `portugues'. (ispell-dictionary-alist-6): Removed redundant command-line option to load `russian' and `slovak' dictionary files. (ispell-dictionary-alist): Tag name changed from "Character Set" to "Coding System". (ispell-version): Updated to 3.6. (ispell-library-directory): Calls non-deprecated function. (ispell-valid-dictionary-list): New function returning all valid dictionaries on machine. (ispell-checking-message): Documentation string improved. (ispell-skip-region-alist): Added uuencoded and postscript region skipping. Improved http/e-mail/file regexp to not match `/.\w'. (ispell-html-skip-alists): New variable for html region support. (ispell-send-string): Removed redundant xemacs check. (ispell-word): Fix spelling error in documentation string, added extent information to support highlighting in ispell-minor-mode. (ispell-command-loop): Disable horizontal scrollbar in XEmacs choices buffer. (ispell-show-choices): Directly select `choices-window'. (ispell-help): Use default buffer size for electric help. (ispell-adjusted-window-height): Correct for xemacs detection. (ispell-start-process): Don't double specify dictionary file name. (ispell-init-process): Set `ispell-library-path' each call. (ispell-change-dictionary): Now only completes valid dictionaries. (ispell-region): Add support for MIME region skipping and Fcc message query for large attachments. (ispell-begin-skip-region-regexp): Add documentation string. Added message support and cleaned up code for generic and html regions. (ispell-begin-skip-region): Function is now requires alist argument. (ispell-begin-tex-skip-regexp): Added comments and support improved html and message regions. (ispell-skip-region-list): New function for MIME and region skipping. (ispell-tex-arg-end): Add documentation string. (ispell-ignore-fcc): New function to query saving Fcc message. (ispell-skip-region): Calculate alist for key match dynamically, html skipping pushed to alists. (ispell-get-line): Add support for multi-line comment regions. (ispell): Check that variables to continue spelling are bound. (ispell-message-text-end): Postscript and uuencoded regions now supported as MIME regions, rather than as end-of-message region. (ispell-mime-multipartp): New function supporting MIME. (ispell-mime-skip-part): New function supporting MIME. (ispell-message): Add MIME support. (ispell-buffer-local-parsing): Variable `ispell-skip-html' now local. (ispell-buffer-local-dict): Fixed bug for detecting and reloading new dictionary.
author Stefan Monnier <monnier@iro.umontreal.ca>
date Thu, 22 May 2003 21:34:00 +0000
parents cc820064216c
children 2d0c2b7dbf3f d7ddb3e565de
line wrap: on
line source

;;; ebnf-iso.el --- parser for ISO EBNF

;; Copyright (C) 1999, 2000, 2001 Free Software Foundation, Inc.

;; Author: Vinicius Jose Latorre <vinicius@cpqd.com.br>
;; Maintainer: Vinicius Jose Latorre <vinicius@cpqd.com.br>
;; Keywords: wp, ebnf, PostScript
;; Time-stamp: <2003-02-10 10:26:32 jbarranquero>
;; Version: 1.6

;; This file is part of GNU Emacs.

;; GNU Emacs 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.

;; GNU Emacs 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., 59 Temple Place - Suite 330,
;; Boston, MA 02111-1307, USA.

;;; Commentary:

;; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;;
;; This is part of ebnf2ps package.
;;
;; This package defines a parser for ISO EBNF.
;;
;; See ebnf2ps.el for documentation.
;;
;;
;; ISO EBNF Syntax
;; ---------------
;;
;;	See the URL:
;;	`http://www.cl.cam.ac.uk/~mgk25/iso-ebnf.html'
;;	("International Standard of the ISO EBNF Notation").
;;
;;
;; ISO EBNF = syntax rule, {syntax rule};
;;
;; syntax rule = meta identifier, '=', definition list, ';';
;;
;; definition list = single definition, {'|', single definition};
;;
;; single definition = term, {',', term};
;;
;; term = factor, ['-', exception];
;;
;; exception = factor (* without <meta identifier> *);
;;
;; factor = [integer, '*'], primary;
;;
;; primary = optional sequence | repeated sequence | special sequence
;;         | grouped sequence | meta identifier | terminal string
;;         | empty;
;;
;; empty = ;
;;
;; optional sequence = '[', definition list, ']';
;;
;; repeated sequence = '{', definition list, '}';
;;
;; grouped sequence = '(', definition list, ')';
;;
;; terminal string = "'", character - "'", {character - "'"}, "'"
;;                 | '"', character - '"', {character - '"'}, '"';
;;
;; special sequence = '?', {character - '?'}, '?';
;;
;; meta identifier = letter, { letter | decimal digit | ' ' };
;;
;; integer = decimal digit, {decimal digit};
;;
;; comment = '(*', {comment symbol}, '*)';
;;
;; comment symbol = comment (* <== NESTED COMMENT *)
;;                | terminal string | special sequence | character;
;;
;; letter = ? A-Z a-z ?;
;;
;; decimal digit = ? 0-9 ?;
;;
;; character = letter | decimal digit
;;           | ',' | '=' | '|' | '/' | '!' | '*' | '(' | ')' | '[' | ']' | '{'
;;           | '}' | "'" | '"' | '?' | '-' | ';' | '.' | ' ' | ':' | '+' | '_'
;;           | '%' | '@' | '&' | '#' | '$' | '<' | '>' | '\' | '^' | '`' | '~';
;;
;;
;; There is also the following alternative representation:
;;
;; STANDARD   ALTERNATIVE
;;    |    ==>   / or !
;;    [    ==>   (/
;;    ]    ==>   /)
;;    {    ==>   (:
;;    }    ==>   :)
;;    ;    ==>   .
;;
;;
;; Differences Between ISO EBNF And ebnf2ps ISO EBNF
;; -------------------------------------------------
;;
;; ISO EBNF accepts the characters given by <character> production above,
;; HORIZONTAL TAB (^I), VERTICAL TAB (^K), NEWLINE (^J or ^M) and FORM FEED
;; (^L), any other characters are illegal.  But ebnf2ps accepts also the
;; european 8-bit accentuated characters (from \240 to \377).
;;
;;
;; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

;;; Code:


(require 'ebnf-otz)


(defvar ebnf-iso-lex nil
  "Value returned by `ebnf-iso-lex' function.")


(defconst ebnf-no-meta-identifier nil)


;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;; Syntactic analyzer


;;; ISO EBNF = syntax rule, {syntax rule};

(defun ebnf-iso-parser (start)
  "ISO EBNF parser."
  (let ((total (+ (- ebnf-limit start) 1))
	(bias (1- start))
	(origin (point))
	syntax-list token rule)
    (goto-char start)
    (setq token (ebnf-iso-lex))
    (and (eq token 'end-of-input)
	 (error "Invalid ISO EBNF file format"))
    (while (not (eq token 'end-of-input))
      (ebnf-message-float
       "Parsing...%s%%"
       (/ (* (- (point) bias) 100.0) total))
      (setq token (ebnf-iso-syntax-rule token)
	    rule  (cdr token)
	    token (car token))
      (or (ebnf-add-empty-rule-list rule)
	  (setq syntax-list (cons rule syntax-list))))
    (goto-char origin)
    syntax-list))


;;; syntax rule = meta identifier, '=', definition list, ';';

(defun ebnf-iso-syntax-rule (token)
  (let ((header ebnf-iso-lex)
	(action ebnf-action)
	body)
    (setq ebnf-action nil)
    (or (eq token 'non-terminal)
	(error "Invalid meta identifier syntax rule"))
    (or (eq (ebnf-iso-lex) 'equal)
	(error "Invalid syntax rule: missing `='"))
    (setq body (ebnf-iso-definition-list))
    (or (eq (car body) 'period)
	(error "Invalid syntax rule: missing `;' or `.'"))
    (setq body (cdr body))
    (ebnf-eps-add-production header)
    (cons (ebnf-iso-lex)
	  (ebnf-make-production header body action))))


;;; definition list = single definition, {'|', single definition};

(defun ebnf-iso-definition-list ()
  (let (body sequence)
    (while (eq (car (setq sequence (ebnf-iso-single-definition)))
	       'alternative)
      (setq sequence (cdr sequence)
	    body     (cons sequence body)))
    (ebnf-token-alternative body sequence)))


;;; single definition = term, {',', term};

(defun ebnf-iso-single-definition ()
  (let (token seq term)
    (while (and (setq term  (ebnf-iso-term (ebnf-iso-lex))
		      token (car term)
		      term  (cdr term))
		(eq token 'catenate))
      (setq seq (cons term seq)))
    (cons token
	  (cond
	   ;; null sequence
	   ((null seq)
	    term)
	   ;; sequence with only one element
	   ((and (null term) (= (length seq) 1))
	    (car seq))
	   ;; a real sequence
	   (t
	    (ebnf-make-sequence (nreverse (cons term seq))))
	   ))))


;;; term = factor, ['-', exception];
;;;
;;; exception = factor (* without <meta identifier> *);

(defun ebnf-iso-term (token)
  (let ((factor (ebnf-iso-factor token)))
    (if (not (eq (car factor) 'except))
	;; factor
	factor
      ;; factor - exception
      (let ((ebnf-no-meta-identifier t))
	(ebnf-token-except (cdr factor) (ebnf-iso-factor (ebnf-iso-lex)))))))


;;; factor = [integer, '*'], primary;

(defun ebnf-iso-factor (token)
  (if (eq token 'integer)
      (let ((times ebnf-iso-lex))
	(or (eq (ebnf-iso-lex) 'repeat)
	    (error "Missing `*'"))
	(ebnf-token-repeat times (ebnf-iso-primary (ebnf-iso-lex))))
    (ebnf-iso-primary token)))


;;; primary = optional sequence | repeated sequence | special sequence
;;;         | grouped sequence | meta identifier | terminal string
;;;         | empty;
;;;
;;; empty = ;
;;;
;;; optional sequence = '[', definition list, ']';
;;;
;;; repeated sequence = '{', definition list, '}';
;;;
;;; grouped sequence = '(', definition list, ')';
;;;
;;; terminal string = "'", character - "'", {character - "'"}, "'"
;;;                 | '"', character - '"', {character - '"'}, '"';
;;;
;;; special sequence = '?', {character - '?'}, '?';
;;;
;;; meta identifier = letter, {letter | decimal digit};

(defun ebnf-iso-primary (token)
  (let ((primary
	 (cond
	  ;; terminal string
	  ((eq token 'terminal)
	   (ebnf-make-terminal ebnf-iso-lex))
	  ;; meta identifier
	  ((eq token 'non-terminal)
	   (ebnf-make-non-terminal ebnf-iso-lex))
	  ;; special sequence
	  ((eq token 'special)
	   (ebnf-make-special ebnf-iso-lex))
	  ;; grouped sequence
	  ((eq token 'begin-group)
	   (let ((body (ebnf-iso-definition-list)))
	     (or (eq (car body) 'end-group)
		 (error "Missing `)'"))
	     (cdr body)))
	  ;; optional sequence
	  ((eq token 'begin-optional)
	   (let ((body (ebnf-iso-definition-list)))
	     (or (eq (car body) 'end-optional)
		 (error "Missing `]' or `/)'"))
	     (ebnf-token-optional (cdr body))))
	  ;; repeated sequence
	  ((eq token 'begin-zero-or-more)
	   (let* ((body   (ebnf-iso-definition-list))
		  (repeat (cdr body)))
	     (or (eq (car body) 'end-zero-or-more)
		 (error "Missing `}' or `:)'"))
	     (ebnf-make-zero-or-more repeat)))
	  ;; empty
	  (t
	   nil)
	  )))
    (cons (if primary
	      (ebnf-iso-lex)
	    token)
	  primary)))


;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;; Lexical analyzer


(defconst ebnf-iso-token-table
  ;; control character & 8-bit character are set to `error'
  (let ((table (make-vector 256 'error))
	(char ?\040))
    ;; printable character
    (while (< char ?\060)
      (aset table char 'character)
      (setq char (1+ char)))
    ;; digits:
    (while (< char ?\072)
      (aset table char 'integer)
      (setq char (1+ char)))
    (while (< char ?\101)
      (aset table char 'character)
      (setq char (1+ char)))
    ;; upper case letters:
    (while (< char ?\133)
      (aset table char 'non-terminal)
      (setq char (1+ char)))
    (while (< char ?\141)
      (aset table char 'character)
      (setq char (1+ char)))
    ;; lower case letters:
    (while (< char ?\173)
      (aset table char 'non-terminal)
      (setq char (1+ char)))
    (while (< char ?\177)
      (aset table char 'character)
      (setq char (1+ char)))
    ;; European 8-bit accentuated characters:
    (setq char ?\240)
    (while (< char ?\400)
      (aset table char 'non-terminal)
      (setq char (1+ char)))
    ;; Override space characters:
    (aset table ?\013 'space)		; [VT] vertical tab
    (aset table ?\n   'space)		; [NL] linefeed
    (aset table ?\r   'space)		; [CR] carriage return
    (aset table ?\t   'space)		; [HT] horizontal tab
    (aset table ?\    'space)		; [SP] space
    ;; Override form feed character:
    (aset table ?\f 'form-feed)		; [FF] form feed
    ;; Override other lexical characters:
    (aset table ?\" 'double-terminal)
    (aset table ?\' 'single-terminal)
    (aset table ?\? 'special)
    (aset table ?*  'repeat)
    (aset table ?,  'catenate)
    (aset table ?-  'except)
    (aset table ?=  'equal)
    (aset table ?\) 'end-group)
    table)
  "Vector used to map characters to a lexical token.")


(defun ebnf-iso-initialize ()
  "Initialize ISO EBNF token table."
  (if ebnf-iso-alternative-p
      ;; Override alternative lexical characters:
      (progn
	(aset ebnf-iso-token-table ?\( 'left-parenthesis)
	(aset ebnf-iso-token-table ?\[ 'character)
	(aset ebnf-iso-token-table ?\] 'character)
	(aset ebnf-iso-token-table ?\{ 'character)
	(aset ebnf-iso-token-table ?\} 'character)
	(aset ebnf-iso-token-table ?|  'character)
	(aset ebnf-iso-token-table ?\; 'character)
	(aset ebnf-iso-token-table ?/  'slash)
	(aset ebnf-iso-token-table ?!  'alternative)
	(aset ebnf-iso-token-table ?:  'colon)
	(aset ebnf-iso-token-table ?.  'period))
    ;; Override standard lexical characters:
    (aset ebnf-iso-token-table ?\( 'begin-parenthesis)
    (aset ebnf-iso-token-table ?\[ 'begin-optional)
    (aset ebnf-iso-token-table ?\] 'end-optional)
    (aset ebnf-iso-token-table ?\{ 'begin-zero-or-more)
    (aset ebnf-iso-token-table ?\} 'end-zero-or-more)
    (aset ebnf-iso-token-table ?|  'alternative)
    (aset ebnf-iso-token-table ?\; 'period)
    (aset ebnf-iso-token-table ?/  'character)
    (aset ebnf-iso-token-table ?!  'character)
    (aset ebnf-iso-token-table ?:  'character)
    (aset ebnf-iso-token-table ?.  'character)))


;; replace the range "\240-\377" (see `ebnf-range-regexp').
(defconst ebnf-iso-non-terminal-chars
  (ebnf-range-regexp " 0-9A-Za-z" ?\240 ?\377))


(defun ebnf-iso-lex ()
  "Lexical analyser for ISO EBNF.

Return a lexical token.

See documentation for variable `ebnf-iso-lex'."
  (if (>= (point) ebnf-limit)
      'end-of-input
    (let (token)
      ;; skip spaces and comments
      (while (if (> (following-char) 255)
		 (progn
		   (setq token 'error)
		   nil)
	       (setq token (aref ebnf-iso-token-table (following-char)))
	       (cond
		((eq token 'space)
		 (skip-chars-forward " \013\n\r\t" ebnf-limit)
		 (< (point) ebnf-limit))
		((or (eq token 'begin-parenthesis)
		     (eq token 'left-parenthesis))
		 (forward-char)
		 (if (/= (following-char) ?*)
		     ;; no comment
		     nil
		   ;; comment
		   (ebnf-iso-skip-comment)
		   t))
		((eq token 'form-feed)
		 (forward-char)
		 (setq ebnf-action 'form-feed))
		(t nil)
		)))
      (cond
       ;; end of input
       ((>= (point) ebnf-limit)
	'end-of-input)
       ;; error
       ((eq token 'error)
	(error "Illegal character"))
       ;; integer
       ((eq token 'integer)
	(setq ebnf-iso-lex (ebnf-buffer-substring "0-9"))
	'integer)
       ;; special: ?special?
       ((eq token 'special)
	(setq ebnf-iso-lex (concat "?"
				   (ebnf-string " ->@-~" ?\? "special")
				   "?"))
	'special)
       ;; terminal: "string"
       ((eq token 'double-terminal)
	(setq ebnf-iso-lex (ebnf-string " !#-~" ?\" "terminal"))
	'terminal)
       ;; terminal: 'string'
       ((eq token 'single-terminal)
	(setq ebnf-iso-lex (ebnf-string " -&(-~" ?\' "terminal"))
	'terminal)
       ;; non-terminal
       ((eq token 'non-terminal)
	(setq ebnf-iso-lex
	      (ebnf-iso-normalize
	       (ebnf-trim-right
		(ebnf-buffer-substring ebnf-iso-non-terminal-chars))))
	(and ebnf-no-meta-identifier
	     (error "Exception sequence should not contain a meta identifier"))
	'non-terminal)
       ;; begin optional, begin list or begin group
       ((eq token 'left-parenthesis)
	(forward-char)
	(cond ((= (following-char) ?/)
	       (forward-char)
	       'begin-optional)
	      ((= (following-char) ?:)
	       (forward-char)
	       'begin-zero-or-more)
	      (t
	       'begin-group)
	      ))
       ;; end optional or alternative
       ((eq token 'slash)
	(forward-char)
	(if (/= (following-char) ?\))
	    'alternative
	  (forward-char)
	  'end-optional))
       ;; end list
       ((eq token 'colon)
	(forward-char)
	(if (/= (following-char) ?\))
	    'character
	  (forward-char)
	  'end-zero-or-more))
       ;; begin group
       ((eq token 'begin-parenthesis)
	'begin-group)
       ;; miscellaneous
       (t
	(forward-char)
	token)
       ))))


;; replace the range "\177-\237" (see `ebnf-range-regexp').
(defconst ebnf-iso-comment-chars
  (ebnf-range-regexp "^*(\000-\010\016-\037" ?\177 ?\237))


(defun ebnf-iso-skip-comment ()
  (forward-char)
  (cond
   ;; open EPS file
   ((and ebnf-eps-executing (= (following-char) ?\[))
    (ebnf-eps-add-context (ebnf-iso-eps-filename)))
   ;; close EPS file
   ((and ebnf-eps-executing (= (following-char) ?\]))
    (ebnf-eps-remove-context (ebnf-iso-eps-filename)))
   ;; any other action in comment
   (t
    (setq ebnf-action (aref ebnf-comment-table (following-char))))
   )
  (let ((pair 1))
    (while (> pair 0)
      (skip-chars-forward ebnf-iso-comment-chars ebnf-limit)
      (cond ((>= (point) ebnf-limit)
	     (error "Missing end of comment: `*)'"))
	    ((= (following-char) ?*)
	     (skip-chars-forward "*" ebnf-limit)
	     (when (= (following-char) ?\))
	       ;; end of comment
	       (forward-char)
	       (setq pair (1- pair))))
	    ((= (following-char) ?\()
	     (skip-chars-forward "(" ebnf-limit)
	     (when (= (following-char) ?*)
	       ;; beginning of comment
	       (forward-char)
	       (setq pair (1+ pair))))
	    (t
	     (error "Illegal character"))
	    ))))


(defun ebnf-iso-eps-filename ()
  (forward-char)
  (buffer-substring-no-properties
   (point)
   (let ((chars (concat ebnf-iso-comment-chars "\n"))
	 found)
     (while (not found)
       (skip-chars-forward chars ebnf-limit)
       (setq found
	     (cond ((>= (point) ebnf-limit)
		    (point))
		   ((= (following-char) ?*)
		    (skip-chars-forward "*" ebnf-limit)
		    (if (/= (following-char) ?\))
			nil
		      (backward-char)
		      (point)))
		   ((= (following-char) ?\()
		    (forward-char)
		    (if (/= (following-char) ?*)
			nil
		      (backward-char)
		      (point)))
		   (t
		    (point))
		   )))
     found)))


(defun ebnf-iso-normalize (str)
  (if (not ebnf-iso-normalize-p)
      str
    (let ((len (length str))
	  (stri 0)
	  (spaces 0))
      ;; count exceeding spaces
      (while (< stri len)
	(if (/= (aref str stri) ?\ )
	    (setq stri (1+ stri))
	  (setq stri (1+ stri))
	  (while (and (< stri len) (= (aref str stri) ?\ ))
	    (setq stri   (1+ stri)
		  spaces (1+ spaces)))))
      (if (zerop spaces)
	  ;; no exceeding space
	  str
	;; at least one exceeding space
	(let ((new (make-string (- len spaces) ?\ ))
	      (newi 0))
	  ;; eliminate exceeding spaces
	  (setq stri 0)
	  (while (> spaces 0)
	    (if (/= (aref str stri) ?\ )
		(progn
		  (aset new newi (aref str stri))
		  (setq stri (1+ stri)
			newi (1+ newi)))
	      (aset new newi (aref str stri))
	      (setq stri (1+ stri)
		    newi (1+ newi))
	      (while (and (> spaces 0) (= (aref str stri) ?\ ))
		(setq stri   (1+ stri)
		      spaces (1- spaces)))))
	  ;; remaining is normalized
	  (while (< stri len)
	    (aset new newi (aref str stri))
	    (setq stri (1+ stri)
		  newi (1+ newi)))
	  new)))))


;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;


(provide 'ebnf-iso)


;;; ebnf-iso.el ends here