lisp/gnus/gnus-html.el

   1 ;;; gnus-html.el --- Quoted-Printable functions
   2
   3 ;; Copyright (C) 2010  Free Software Foundation, Inc.
   4
   5 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
   6 ;; Keywords: html, web
   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 ;; The idea is to provide a simple, fast and pretty minimal way to
  26 ;; render HTML (including links and images) in a buffer, based on an
  27 ;; external HTML renderer (i.e., w3m).
  28
  29 ;;; Code:
  30
  31 (defcustom gnus-html-cache-directory (nnheader-concat gnus-directory "html-cache/")
  32   "Where Gnus will cache images it downloads from the web."
  33   :group 'gnus-art
  34   :type 'directory)
  35
  36 (defcustom gnus-html-cache-size 500000000
  37   "The size of the Gnus image cache."
  38   :group 'gnus-art
  39   :type 'integer)
  40
  41 (defcustom gnus-html-frame-width 70
  42   "What width to use when rendering HTML."
  43   :group 'gnus-art
  44   :type 'integer)
  45
  46 ;;;###autoload
  47 (defun gnus-article-html (handle)
  48   (let ((article-buffer (current-buffer)))
  49     (save-restriction
  50       (narrow-to-region (point) (point))
  51       (save-excursion
  52         (mm-with-part handle
  53           (let* ((coding-system-for-read 'utf-8)
  54                  (coding-system-for-write 'utf-8)
  55                  (default-process-coding-system
  56                    (cons coding-system-for-read coding-system-for-write)))
  57             (call-process-region (point-min) (point-max)
  58                                  "w3m"
  59                                  nil article-buffer nil
  60                                  "-halfdump"
  61                                  "-no-cookie"
  62                                  "-O" "UTF-8"
  63                                  "-o" "ext_halfdump=1"
  64                                  "-t" (format "%s" tab-width)
  65                                  "-cols" (format "%s" gnus-html-frame-width)
  66                                  "-o" "display_image=off"
  67                                  "-T" "text/html"))))
  68       (gnus-html-wash-tags))))
  69
  70 (defun gnus-html-wash-tags ()
  71   (let (tag parameters string start end images)
  72     (mm-url-decode-entities)
  73     (goto-char (point-min))
  74     (while (re-search-forward "<\\([^ />]+\\)\\([^>]*\\)>" nil t)
  75       (setq tag (match-string 1)
  76             parameters (match-string 2)
  77             start (match-beginning 0))
  78       (when (plusp (length parameters))
  79         (set-text-properties 0 (1- (length parameters)) nil parameters))
  80       (delete-region start (point))
  81       (when (search-forward (concat "</" tag ">") nil t)
  82         (delete-region (match-beginning 0) (match-end 0)))
  83       (setq end (point))
  84       (cond
  85        ;; Fetch and insert a picture.
  86        ((equal tag "img_alt")
  87         (when (string-match "src=\"\\([^\"]+\\)" parameters)
  88           (setq parameters (match-string 1 parameters))
  89           (when (or (null mm-w3m-safe-url-regexp)
  90                     (string-match mm-w3m-safe-url-regexp parameters))
  91             (let ((file (gnus-html-image-id parameters)))
  92               (if (file-exists-p file)
  93                   ;; It's already cached, so just insert it.
  94                   (when (gnus-html-put-image file (point))
  95                     ;; Delete the ALT text.
  96                     (delete-region start end))
  97                 ;; We don't have it, so schedule it for fetching
  98                 ;; asynchronously.
  99                 (push (list parameters
 100                             (set-marker (make-marker) start)
 101                             (point-marker))
 102                       images))))))
 103        ;; Add a link.
 104        ((equal tag "a")
 105         (when (string-match "href=\"\\([^\"]+\\)" parameters)
 106           (setq parameters (match-string 1 parameters))
 107           (gnus-article-add-button start end
 108                                    'browse-url parameters
 109                                    parameters)
 110           (let ((overlay (gnus-make-overlay start end)))
 111             (gnus-overlay-put overlay 'evaporate t)
 112             (gnus-overlay-put overlay 'gnus-button-url parameters)
 113             (when gnus-article-mouse-face
 114               (gnus-overlay-put overlay 'mouse-face gnus-article-mouse-face)))))
 115        ;; Whatever.  Just ignore the tag.
 116        (t
 117         ))
 118       (goto-char start))
 119     (goto-char (point-min))
 120     ;; The output from -halfdump isn't totally regular, so strip
 121     ;; off any </pre_int>s that were left over.
 122     (while (re-search-forward "</pre_int>" nil t)
 123       (replace-match "" t t))
 124     (when images
 125       (gnus-html-schedule-image-fetching (current-buffer) (nreverse images)))))
 126
 127 (defun gnus-html-schedule-image-fetching (buffer images)
 128   (let* ((url (caar images))
 129          (process (start-process
 130                    "images" nil "curl"
 131                    "-s" "--create-dirs"
 132                    "--location"
 133                    "--max-time" "60"
 134                    "-o" (gnus-html-image-id url)
 135                    url)))
 136     (process-kill-without-query process)
 137     (set-process-sentinel process 'gnus-html-curl-sentinel)
 138     (set-process-plist process (list 'images images
 139                                      'buffer buffer))))
 140
 141 (defun gnus-html-image-id (url)
 142   (expand-file-name (sha1 url) gnus-html-cache-directory))
 143
 144 (defun gnus-html-curl-sentinel (process event)
 145   (when (string-match "finished" event)
 146     (let* ((images (getf (process-plist process) 'images))
 147            (buffer (getf (process-plist process) 'buffer))
 148            (spec (pop images))
 149            (file (gnus-html-image-id (car spec))))
 150       (when (and (buffer-live-p buffer)
 151                  ;; If the position of the marker is 1, then that
 152                  ;; means that the text is was in has been deleted;
 153                  ;; i.e., that the user has selected a different
 154                  ;; article before the image arrived.
 155                  (not (= (marker-position (cadr spec)) 1)))
 156         (save-excursion
 157           (set-buffer buffer)
 158           (let ((buffer-read-only nil))
 159             (when (gnus-html-put-image file (cadr spec))
 160               (delete-region (1+ (cadr spec)) (caddr spec))))))
 161       (when images
 162         (gnus-html-schedule-image-fetching buffer images)))))
 163
 164 (defun gnus-html-put-image (file point)
 165   (let ((image (ignore-errors
 166                  (create-image file))))
 167     (if (and image
 168              ;; Kludge to avoid displaying 30x30 gif images, which
 169              ;; seems to be a signal of a broken image.
 170              (not (and (eq (getf (cdr image) :type) 'gif)
 171                        (= (car (image-size image t)) 30)
 172                        (= (cdr (image-size image t)) 30))))
 173         (progn
 174           (gnus-put-image image nil nil point)
 175           t)
 176       (when (fboundp 'find-image)
 177         (gnus-put-image (find-image '((:type xpm :file "lock-broken.xpm")))
 178                         nil nil point))
 179       nil)))
 180
 181 (defun gnus-html-prune-cache ()
 182   (let ((total-size 0)
 183         files)
 184     (dolist (file (directory-files gnus-html-cache-directory t nil t))
 185       (let ((attributes (file-attributes file)))
 186         (unless (nth 0 attributes)
 187           (incf total-size (nth 7 attributes))
 188           (push (list (time-to-seconds (nth 5 attributes))
 189                       (nth 7 attributes) file)
 190                 files))))
 191     (when (> total-size gnus-html-cache-size)
 192       (setq files (sort files (lambda (f1 f2)
 193                                 (< (car f1) (car f2)))))
 194       (dolist (file files)
 195         (when (> total-size gnus-html-cache-size)
 196           (decf total-size (cadr file))
 197           (delete-file (nth 2 file)))))))
 198
 199 ;;;###autoload
 200 (defun gnus-html-prefetch-images (summary)
 201   (let (safe-url-regexp urls)
 202     (when (buffer-live-p summary)
 203       (save-excursion
 204         (set-buffer summary)
 205         (setq safe-url-regexp mm-w3m-safe-url-regexp))
 206       (save-match-data
 207         (while (re-search-forward "<img.*src=[\"']\\([^\"']+\\)" nil t)
 208           (let ((url (match-string 1)))
 209             (when (or (null safe-url-regexp)
 210                       (string-match safe-url-regexp url))
 211               (unless (file-exists-p (gnus-html-image-id url))
 212                 (push url urls)
 213                 (push (gnus-html-image-id url) urls)
 214                 (push "-o" urls)))))
 215         (let ((process
 216                (apply 'start-process
 217                       "images" nil "curl"
 218                       "-s" "--create-dirs"
 219                       "--location"
 220                       "--max-time" "60"
 221                       urls)))
 222           (process-kill-without-query process))))))
 223
 224 (provide 'gnus-html)
 225
 226 ;;; gnus-html.el ends here