lisp/international/latexenc.el

   1 ;;; latexenc.el --- guess correct coding system in LaTeX files -*-coding: utf-8 -*-
   2
   3 ;; Copyright (C) 2005-2017 Free Software Foundation, Inc.
   4
   5 ;; Author: Arne Jørgensen <arne@arnested.dk>
   6 ;; Keywords: mule, coding system, latex
   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 ;; This code tries to guess the correct coding system of a LaTeX file.
  26
  27 ;; First it searches for a \inputencoding{...} or
  28 ;; \usepackage[...]{inputenc} line in the file and looks up the ... in
  29 ;; `latex-inputenc-coding-alist' to find the corresponding coding
  30 ;; system.
  31
  32 ;; If this fails it will search for AUCTeX's TeX-master or tex-mode's
  33 ;; tex-main-file variable in the local variables section and visit
  34 ;; that file to get the coding system from the master file. This check
  35 ;; can be disabled by setting `latexenc-dont-use-TeX-master-flag' to
  36 ;; t.
  37
  38 ;; If we have still not found a coding system we will try to use the
  39 ;; standard tex-mode's `tex-guess-main-file' and get the coding system
  40 ;; from the main file. This check can be disabled by setting
  41 ;; `latexenc-dont-use-tex-guess-main-file-flag' to t.
  42
  43 ;; The functionality is enabled by adding the function
  44 ;; `latexenc-find-file-coding-system' to `file-coding-system-alist'
  45 ;; like this
  46
  47 ;; (add-to-list 'file-coding-system-alist
  48 ;;           '("\\.\\(tex\\|ltx\\|dtx\\|drv\\)\\'" . latexenc-find-file-coding-system))
  49
  50 ;;; Code:
  51
  52 ;;;###autoload
  53 (defcustom latex-inputenc-coding-alist
  54   (purecopy
  55   '(("ansinew" . windows-1252) ; MS Windows ANSI encoding, extension of Latin-1
  56     ("applemac" . mac-roman)
  57     ("ascii" . us-ascii)
  58     ("cp1250" . windows-1250) ; MS Windows encoding, codepage 1250
  59     ("cp1252" . windows-1252) ; synonym of ansinew
  60     ("cp1257" . cp1257)
  61     ("cp437de" . cp437) ; IBM code page 437 (German version): 225 is \ss
  62     ("cp437" . cp437) ; IBM code page 437: 225 is \beta
  63     ("cp850" . cp850) ; IBM code page 850
  64     ("cp852" . cp852) ; IBM code page 852
  65     ("cp858" . cp858) ; IBM code page 850 but with a euro symbol
  66     ("cp865" . cp865) ; IBM code page 865
  67     ("latin1" . iso-8859-1)
  68     ("latin2" . iso-8859-2)
  69     ("latin3" . iso-8859-3)
  70     ("latin4" . iso-8859-4)
  71     ("latin5" . iso-8859-5)
  72     ("latin9" . iso-8859-15)
  73     ;; ("latin10" . undecided)
  74     ;; ("macce" . undecided) ; Apple Central European
  75     ("next" . next) ; The Next encoding
  76     ("utf8" . utf-8)
  77     ("utf8x" . utf-8))) ; used by the Unicode LaTeX package
  78   "Mapping from LaTeX encodings in \"inputenc.sty\" to Emacs coding systems.
  79 LaTeX encodings are specified with \"\\usepackage[encoding]{inputenc}\".
  80 Used by the function `latexenc-find-file-coding-system'."
  81   :group 'files
  82   :group 'mule
  83   :type '(alist :key-type (string :tag "LaTeX input encoding")
  84                 :value-type (coding-system :tag "Coding system")))
  85
  86 ;;;###autoload
  87 (defun latexenc-inputenc-to-coding-system (inputenc)
  88   "Return the corresponding coding-system for the specified input encoding.
  89 Return nil if no matching coding system can be found."
  90   (cdr (assoc inputenc latex-inputenc-coding-alist)))
  91
  92 ;;;###autoload
  93 (defun latexenc-coding-system-to-inputenc (cs)
  94   "Return the corresponding input encoding for the specified coding system.
  95 Return nil if no matching input encoding can be found."
  96   (let (result)
  97     (catch 'result
  98       (dolist (elem latex-inputenc-coding-alist result)
  99         (let ((elem-cs (cdr elem)))
 100           (when (and (coding-system-p elem-cs)
 101                      (coding-system-p cs)
 102                      (eq (coding-system-base cs) (coding-system-base elem-cs)))
 103             (setq result (car elem))
 104             (throw 'result result)))))))
 105
 106 (defvar latexenc-dont-use-TeX-master-flag nil
 107   "Non-nil means don't follow TeX-master to find the coding system.")
 108
 109 (defvar latexenc-dont-use-tex-guess-main-file-flag nil
 110   "Non-nil means don't use tex-guessmain-file to find the coding system.")
 111
 112 ;;;###autoload
 113 (defun latexenc-find-file-coding-system (arg-list)
 114   "Determine the coding system of a LaTeX file if it uses \"inputenc.sty\".
 115 The mapping from LaTeX's \"inputenc.sty\" encoding names to Emacs
 116 coding system names is determined from `latex-inputenc-coding-alist'."
 117   (if (eq (car arg-list) 'insert-file-contents)
 118       (save-excursion
 119         ;; try to find the coding system in this file
 120         (goto-char (point-min))
 121         (if (catch 'cs
 122               (let ((case-fold-search nil))
 123                 (while (search-forward "inputenc" nil t)
 124                   (goto-char (match-beginning 0))
 125                   (beginning-of-line)
 126                   (if (or (looking-at "[^%\n]*\\\\usepackage\\[\\([^]]*\\)\\]{\\([^}]*,\\)?inputenc\\(,[^}]*\\)?}")
 127                           (looking-at "[^%\n]*\\\\inputencoding{\\([^}]*\\)}"))
 128                       (throw 'cs t)
 129                     (goto-char (match-end 0))))))
 130             (let* ((match (match-string 1))
 131                    (sym (or (latexenc-inputenc-to-coding-system match)
 132                             (intern match))))
 133               (cond
 134                ((coding-system-p sym) sym)
 135                ((and (require 'code-pages nil t) (coding-system-p sym)) sym)
 136                (t 'undecided)))
 137           ;; else try to find it in the master/main file
 138
 139           ;; Fixme: If the current file is in an archive (e.g. tar,
 140           ;; zip), we should find the master file in that archive.
 141           ;; But, that is not yet implemented.   -- K.Handa
 142           (let ((default-directory (if (stringp (nth 1 arg-list))
 143                                        (file-name-directory (nth 1 arg-list))
 144                                      default-directory))
 145                 latexenc-main-file)
 146             ;; Is there a TeX-master or tex-main-file in the local variables
 147             ;; section?
 148             (unless latexenc-dont-use-TeX-master-flag
 149               (goto-char (point-max))
 150               (search-backward "\n\^L" (max (- (point-max) 3000) (point-min))
 151                                'move)
 152               (search-forward "Local Variables:" nil t)
 153               (when (re-search-forward
 154                      "^%+ *\\(TeX-master\\|tex-main-file\\): *\"\\(.+\\)\""
 155                      nil t)
 156                 (let ((file (match-string 2)))
 157                   (dolist (ext `("" ,(if (boundp 'TeX-default-extension)
 158                                          (concat "." TeX-default-extension)
 159                                        "")
 160                                  ".tex" ".ltx" ".dtx" ".drv"))
 161                     (if (and (null latexenc-main-file) ;Stop at first.
 162                              (file-exists-p (concat file ext)))
 163                         (setq latexenc-main-file (concat file ext)))))))
 164             ;; try tex-modes tex-guess-main-file
 165             (when (and (not latexenc-dont-use-tex-guess-main-file-flag)
 166                        (not latexenc-main-file))
 167               ;; Use a separate `when' so the byte-compiler sees the fboundp.
 168               (when (fboundp 'tex-guess-main-file)
 169                 (let ((tex-start-of-header "\\\\document\\(style\\|class\\)"))
 170                   (setq latexenc-main-file (tex-guess-main-file)))))
 171             ;; if we found a master/main file get the coding system from it
 172             (if (and latexenc-main-file
 173                      (file-regular-p latexenc-main-file)
 174                      (file-readable-p latexenc-main-file))
 175                 (let* ((latexenc-dont-use-tex-guess-main-file-flag t)
 176                        (latexenc-dont-use-TeX-master-flag t)
 177                        (latexenc-main-buffer
 178                         (find-file-noselect latexenc-main-file t)))
 179                   (coding-system-base   ;Disregard the EOL part of the CS.
 180                    (with-current-buffer latexenc-main-buffer
 181                      (or coding-system-for-write buffer-file-coding-system
 182                          'undecided))))
 183               'undecided))))
 184     'undecided))
 185
 186 \f
 187 (provide 'latexenc)
 188
 189 ;;; latexenc.el ends here