lisp/mail/unrmail.el

   1 ;;; unrmail.el --- convert Rmail Babyl files to mbox files
   2
   3 ;; Copyright (C) 1992, 2001-2013 Free Software Foundation, Inc.
   4
   5 ;; Maintainer: FSF
   6 ;; Keywords: mail
   7
   8 ;; This file is part of GNU Emacs.
   9
  10 ;; GNU Emacs is free software: you can redistribute it and/or modify
  11 ;; it under the terms of the GNU General Public License as published by
  12 ;; the Free Software Foundation, either version 3 of the License, or
  13 ;; (at your option) any later version.
  14
  15 ;; GNU Emacs is distributed in the hope that it will be useful,
  16 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 ;; GNU General Public License for more details.
  19
  20 ;; You should have received a copy of the GNU General Public License
  21 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  22
  23 ;;; Commentary:
  24
  25 ;;; Code:
  26
  27 ;;;###autoload
  28 (defun batch-unrmail ()
  29   "Convert old-style Rmail Babyl files to mbox format.
  30 Specify the input Rmail Babyl file names as command line arguments.
  31 For each Rmail file, the corresponding output file name
  32 is made by adding `.mail' at the end.
  33 For example, invoke `emacs -batch -f batch-unrmail RMAIL'."
  34   (if (not noninteractive)
  35       (error "`batch-unrmail' is to be used only with -batch"))
  36   (let ((error nil))
  37     (while command-line-args-left
  38       (or (unrmail (car command-line-args-left)
  39                    (concat (car command-line-args-left) ".mail"))
  40           (setq error t))
  41       (setq command-line-args-left (cdr command-line-args-left)))
  42     (message "Done")
  43     (kill-emacs (if error 1 0))))
  44
  45 (declare-function mail-mbox-from "mail-utils" ())
  46 (defvar rmime-magic-string)             ; in rmime.el, if you have it
  47
  48 (defcustom unrmail-mbox-format 'mboxrd
  49   "The mbox format that `unrmail' should produce.
  50 These formats separate messages using lines that start with \"From \".
  51 Therefore any lines in the message bodies that start with \"From \"
  52 must be quoted.  The `mboxo' format just prepends a \">\" to such lines.
  53 This is not reversible, because given a line starting with \">From \" in
  54 an mboxo file, it is not possible to know whether the original had a \">\"
  55 or not.  The `mxbord' format avoids this by also quoting \">From \" as
  56 \">>From \", and so on.  For this reason, mboxrd is recommended.
  57
  58 See also `rmail-mbox-format'."
  59   :type '(choice (const mboxrd)
  60                  (const mboxro))
  61   :version "24.4"
  62   :group 'rmail-files)
  63
  64 ;;;###autoload
  65 (defun unrmail (file to-file)
  66   "Convert old-style Rmail Babyl file FILE to mbox format file TO-FILE.
  67 The variable `unrmail-mbox-format' controls which mbox format to use."
  68   (interactive "fUnrmail (babyl file): \nFUnrmail into (new mailbox file): ")
  69   (with-temp-buffer
  70     ;; Read in the old Rmail file with no decoding.
  71     (let ((coding-system-for-read 'raw-text))
  72       (insert-file-contents file))
  73     ;; But make it multibyte.
  74     (set-buffer-multibyte t)
  75     (setq buffer-file-coding-system 'raw-text-unix)
  76
  77     (if (not (looking-at "BABYL OPTIONS"))
  78         (error "This file is not in Babyl format"))
  79
  80     ;; Decode the file contents just as Rmail did.
  81     (let ((coding-system rmail-file-coding-system)
  82           from to)
  83       (goto-char (point-min))
  84       (search-forward "\n\^_" nil t)    ; Skip BABYL header.
  85       (setq from (point))
  86       (goto-char (point-max))
  87       (search-backward "\n\^_" from 'mv)
  88       (if (= from (setq to (point)))
  89           (error "The input file contains no messages"))
  90       (unless (and coding-system
  91                    (coding-system-p coding-system))
  92         (setq coding-system
  93               ;; Emacs 21.1 and later writes RMAIL files in emacs-mule, but
  94               ;; earlier versions did that with the current buffer's encoding.
  95               ;; So we want to favor detection of emacs-mule (whose normal
  96               ;; priority is quite low), but still allow detection of other
  97               ;; encodings if emacs-mule won't fit.  The call to
  98               ;; detect-coding-with-priority below achieves that.
  99               (car (detect-coding-with-priority
 100                     from to
 101                     '((coding-category-emacs-mule . emacs-mule))))))
 102       (unless (memq coding-system
 103                     '(undecided undecided-unix))
 104         (set-buffer-modified-p t)       ; avoid locking when decoding
 105         (let ((buffer-undo-list t))
 106           (decode-coding-region from to coding-system))
 107         (setq coding-system last-coding-system-used))
 108
 109       (setq buffer-file-coding-system nil)
 110
 111       ;; We currently don't use this value, but maybe we should.
 112       (setq save-buffer-coding-system
 113             (or coding-system 'undecided)))
 114
 115     ;; Default the directory of TO-FILE based on where FILE is.
 116     (setq to-file (expand-file-name to-file default-directory))
 117     (condition-case ()
 118         (delete-file to-file)
 119       (file-error nil))
 120     (message "Writing messages to %s..." to-file)
 121     (goto-char (point-min))
 122
 123     (let ((temp-buffer (get-buffer-create " unrmail"))
 124           (from-buffer (current-buffer)))
 125
 126       ;; Process the messages one by one.
 127       (while (re-search-forward "^\^_\^l" nil t)
 128         (let ((beg (point))
 129               (end (save-excursion
 130                      (if (re-search-forward "^\^_\\(\^l\\|\\'\\)" nil t)
 131                          (match-beginning 0)
 132                        (point-max))))
 133               (coding 'raw-text)
 134               label-line attrs keywords
 135               mail-from reformatted)
 136           (with-current-buffer temp-buffer
 137             (setq buffer-undo-list t)
 138             (erase-buffer)
 139             (setq buffer-file-coding-system coding)
 140             (insert-buffer-substring from-buffer beg end)
 141             (goto-char (point-min))
 142             (forward-line 1)
 143             ;; Record whether the header is reformatted.
 144             (setq reformatted (= (following-char) ?1))
 145
 146             ;; Collect the label line, then get the attributes
 147             ;; and the keywords from it.
 148             (setq label-line
 149                   (buffer-substring (point)
 150                                     (save-excursion (forward-line 1)
 151                                                     (point))))
 152             (re-search-forward ",, ?")
 153             (unless (eolp)
 154               (setq keywords
 155                     (buffer-substring (point)
 156                                       (progn (end-of-line)
 157                                              (1- (point)))))
 158               ;; Mbox rmail needs the spaces.  Bug#2303.
 159               ;;; (setq keywords
 160               ;;;           (replace-regexp-in-string ", " "," keywords))
 161               )
 162
 163             (setq attrs
 164                   (list
 165                    (if (string-match ", answered," label-line) ?A ?-)
 166                    (if (string-match ", deleted," label-line) ?D ?-)
 167                    (if (string-match ", edited," label-line) ?E ?-)
 168                    (if (string-match ", filed," label-line) ?F ?-)
 169                    (if (string-match ", retried," label-line) ?R ?-)
 170                    (if (string-match ", forwarded," label-line) ?S ?-)
 171                    (if (string-match ", unseen," label-line) ?U ?-)
 172                    (if (string-match ", resent," label-line) ?r ?-)))
 173
 174             ;; Delete the special Babyl lines at the start,
 175             ;; and the ***EOOH*** line, and the reformatted header if any.
 176             (goto-char (point-min))
 177             (if reformatted
 178                 (progn
 179                   (forward-line 2)
 180                   ;; Delete Summary-Line headers.
 181                   (let ((case-fold-search t))
 182                     (while (looking-at "Summary-Line:")
 183                       (forward-line 1)))
 184                   (delete-region (point-min) (point))
 185                   ;; Delete the old reformatted header.
 186                   (re-search-forward "^[*][*][*] EOOH [*][*][*]\n")
 187                   (forward-line -1)
 188                   (let ((start (point)))
 189                     (search-forward "\n\n")
 190                     (delete-region start (point))))
 191               ;; Not reformatted.  Delete the special
 192               ;; lines before the real header.
 193               (re-search-forward "^[*][*][*] EOOH [*][*][*]\n")
 194               (delete-region (point-min) (point)))
 195
 196             ;; Handle rmime formatting.
 197             (when (require 'rmime nil t)
 198               (let ((start (point)))
 199                 (while (search-forward rmime-magic-string nil t))
 200                 (delete-region start (point))))
 201
 202             ;; Some operations on the message header itself.
 203             (goto-char (point-min))
 204             (save-restriction
 205               (narrow-to-region
 206                (point-min)
 207                (save-excursion (search-forward "\n\n" nil 'move) (point)))
 208
 209               ;; Fetch or construct what we should use in the `From ' line.
 210               (setq mail-from (or (let ((from (mail-fetch-field "Mail-From")))
 211                                     ;; mail-mbox-from (below) returns a
 212                                     ;; string that ends in a newline, but
 213                                     ;; but mail-fetch-field does not, so
 214                                     ;; we append a newline here.
 215                                     (if from
 216                                         (format "%s\n" from)))
 217                                   (mail-mbox-from)))
 218
 219               ;; If the message specifies a coding system, use it.
 220               (let ((maybe-coding (mail-fetch-field "X-Coding-System")))
 221                 (if maybe-coding
 222                     (setq coding
 223                           ;; Force Unix EOLs.
 224                           (coding-system-change-eol-conversion
 225                            (intern maybe-coding) 0))
 226                   ;; If there's no X-Coding-System header, assume the
 227                   ;; message was never decoded.
 228                   (setq coding 'raw-text-unix)))
 229
 230               ;; Delete the Mail-From: header field if any.
 231               (when (re-search-forward "^Mail-from:" nil t)
 232                 (beginning-of-line)
 233                 (delete-region (point)
 234                                (progn (forward-line 1) (point)))))
 235
 236             (goto-char (point-min))
 237             ;; Insert the `From ' line.
 238             (insert mail-from)
 239             ;; Record the keywords and attributes in our special way.
 240             (insert "X-RMAIL-ATTRIBUTES: " (apply 'string attrs) "\n")
 241             (when keywords
 242               (insert "X-RMAIL-KEYWORDS: " keywords "\n"))
 243             ;; Convert From to >From, etc.
 244             (let ((case-fold-search nil)
 245                   (fromline (if (eq 'mboxrd unrmail-mbox-format)
 246                             "^>*From "
 247                           "^From ")))
 248               (while (re-search-forward fromline nil t)
 249                 (beginning-of-line)
 250                 (insert ?>)
 251                 (forward-line 1)))
 252             (goto-char (point-max))
 253             ;; Add terminator blank line to message.
 254             (insert "\n")
 255             ;; Write it to the output file, suitably encoded.
 256             (let ((coding-system-for-write coding))
 257               (write-region (point-min) (point-max) to-file t
 258                             'nomsg)))))
 259       (kill-buffer temp-buffer))
 260     (message "Writing messages to %s...done" to-file)))
 261
 262 (provide 'unrmail)
 263
 264 ;;; unrmail.el ends here