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 (eval-when-compile (require 'cl))
  32 (eval-when-compile (require 'mm-decode))
  33 (require 'mm-url)
  34
  35 (defcustom gnus-html-cache-directory (nnheader-concat gnus-directory "html-cache/")
  36   "Where Gnus will cache images it downloads from the web."
  37   :version "24.1"
  38   :group 'gnus-art
  39   :type 'directory)
  40
  41 (defcustom gnus-html-cache-size 500000000
  42   "The size of the Gnus image cache."
  43   :version "24.1"
  44   :group 'gnus-art
  45   :type 'integer)
  46
  47 (defcustom gnus-html-frame-width 70
  48   "What width to use when rendering HTML."
  49   :version "24.1"
  50   :group 'gnus-art
  51   :type 'integer)
  52
  53 (defcustom gnus-blocked-images "."
  54   "Images that have URLs matching this regexp will be blocked."
  55   :version "24.1"
  56   :group 'gnus-art
  57   :type 'regexp)
  58
  59 (defcustom gnus-max-image-proportion 0.7
  60   "How big pictures displayed are in relation to the window they're in.
  61 A value of 0.7 means that they are allowed to take up 70% of the
  62 width and height of the window.  If they are larger than this,
  63 and Emacs supports it, then the images will be rescaled down to
  64 fit these criteria."
  65   :version "24.1"
  66   :group 'gnus-art
  67   :type 'float)
  68
  69 ;;;###autoload
  70 (defun gnus-article-html (handle)
  71   (let ((article-buffer (current-buffer)))
  72     (save-restriction
  73       (narrow-to-region (point) (point))
  74       (save-excursion
  75         (mm-with-part handle
  76           (let* ((coding-system-for-read 'utf-8)
  77                  (coding-system-for-write 'utf-8)
  78                  (default-process-coding-system
  79                    (cons coding-system-for-read coding-system-for-write))
  80                  (charset (mail-content-type-get (mm-handle-type handle)
  81                                                  'charset)))
  82             (when (and charset
  83                        (setq charset (mm-charset-to-coding-system charset))
  84                        (not (eq charset 'ascii)))
  85               (mm-decode-coding-region (point-min) (point-max) charset))
  86             (call-process-region (point-min) (point-max)
  87                                  "w3m"
  88                                  nil article-buffer nil
  89                                  "-halfdump"
  90                                  "-no-cookie"
  91                                  "-I" "UTF-8"
  92                                  "-O" "UTF-8"
  93                                  "-o" "ext_halfdump=1"
  94                                  "-o" "pre_conv=1"
  95                                  "-t" (format "%s" tab-width)
  96                                  "-cols" (format "%s" gnus-html-frame-width)
  97                                  "-o" "display_image=off"
  98                                  "-T" "text/html"))))
  99       (gnus-html-wash-tags))))
 100
 101 (defvar gnus-article-mouse-face)
 102
 103 (defun gnus-html-wash-tags ()
 104   (let (tag parameters string start end images url)
 105     (mm-url-decode-entities)
 106     (goto-char (point-min))
 107     (while (re-search-forward "<\\([^ />]+\\)\\([^>]*\\)>" nil t)
 108       (setq tag (match-string 1)
 109             parameters (match-string 2)
 110             start (match-beginning 0))
 111       (when (plusp (length parameters))
 112         (set-text-properties 0 (1- (length parameters)) nil parameters))
 113       (delete-region start (point))
 114       (when (search-forward (concat "</" tag ">") nil t)
 115         (delete-region (match-beginning 0) (match-end 0)))
 116       (setq end (point))
 117       (cond
 118        ;; Fetch and insert a picture.
 119        ((equal tag "img_alt")
 120         (when (string-match "src=\"\\([^\"]+\\)" parameters)
 121           (setq url (match-string 1 parameters))
 122           (gnus-message 8 "Fetching image URL %s" url)
 123           (if (string-match "^cid:\\(.*\\)" url)
 124               ;; URLs with cid: have their content stashed in other
 125               ;; parts of the MIME structure, so just insert them
 126               ;; immediately.
 127               (let ((handle (mm-get-content-id
 128                              (setq url (match-string 1 url))))
 129                     image)
 130                 (when handle
 131                   (mm-with-part handle
 132                     (setq image (gnus-create-image (buffer-string)
 133                                                    nil t))))
 134                 (when image
 135                   (delete-region start end)
 136                   (gnus-put-image image)))
 137             ;; Normal, external URL.
 138             (when (or (null gnus-blocked-images)
 139                       (not (string-match gnus-blocked-images url)))
 140               (let ((file (gnus-html-image-id url)))
 141                 (if (file-exists-p file)
 142                     ;; It's already cached, so just insert it.
 143                     (when (gnus-html-put-image file (point))
 144                       ;; Delete the ALT text.
 145                       (delete-region start end))
 146                   ;; We don't have it, so schedule it for fetching
 147                   ;; asynchronously.
 148                   (push (list url
 149                               (set-marker (make-marker) start)
 150                               (point-marker))
 151                         images)))))))
 152        ;; Add a link.
 153        ((equal tag "a")
 154         (when (string-match "href=\"\\([^\"]+\\)" parameters)
 155           (setq url (match-string 1 parameters))
 156           (gnus-message 8 "Fetching link URL %s" url)
 157           (gnus-article-add-button start end
 158                                    'browse-url url
 159                                    url)
 160           (let ((overlay (gnus-make-overlay start end)))
 161             (gnus-overlay-put overlay 'evaporate t)
 162             (gnus-overlay-put overlay 'gnus-button-url url)
 163             (when gnus-article-mouse-face
 164               (gnus-overlay-put overlay 'mouse-face gnus-article-mouse-face)))))
 165        ;; The upper-case IMG_ALT is apparently just an artifact that
 166        ;; should be deleted.
 167        ((equal tag "IMG_ALT")
 168         (delete-region start end))
 169        ;; Whatever.  Just ignore the tag.
 170        (t
 171         ))
 172       (goto-char start))
 173     (goto-char (point-min))
 174     ;; The output from -halfdump isn't totally regular, so strip
 175     ;; off any </pre_int>s that were left over.
 176     (while (re-search-forward "</pre_int>" nil t)
 177       (replace-match "" t t))
 178     (when images
 179       (gnus-html-schedule-image-fetching (current-buffer) (nreverse images)))))
 180
 181 (defun gnus-html-schedule-image-fetching (buffer images)
 182   (gnus-message 8 "Scheduling image fetching in buffer %s, images %s" buffer images)
 183   (let* ((url (caar images))
 184          (process (start-process
 185                    "images" nil "curl"
 186                    "-s" "--create-dirs"
 187                    "--location"
 188                    "--max-time" "60"
 189                    "-o" (gnus-html-image-id url)
 190                    url)))
 191     (process-kill-without-query process)
 192     (set-process-sentinel process 'gnus-html-curl-sentinel)
 193     (gnus-set-process-plist process (list 'images images
 194                                           'buffer buffer))))
 195
 196 (defun gnus-html-image-id (url)
 197   (expand-file-name (sha1 url) gnus-html-cache-directory))
 198
 199 (defun gnus-html-curl-sentinel (process event)
 200   (when (string-match "finished" event)
 201     (let* ((images (gnus-process-get process 'images))
 202            (buffer (gnus-process-get process 'buffer))
 203            (spec (pop images))
 204            (file (gnus-html-image-id (car spec))))
 205       (when (and (buffer-live-p buffer)
 206                  ;; If the position of the marker is 1, then that
 207                  ;; means that the text it was in has been deleted;
 208                  ;; i.e., that the user has selected a different
 209                  ;; article before the image arrived.
 210                  (not (= (marker-position (cadr spec)) (point-min))))
 211         (with-current-buffer buffer
 212           (let ((inhibit-read-only t))
 213             (when (gnus-html-put-image file (cadr spec))
 214               (delete-region (1+ (cadr spec)) (caddr spec))))))
 215       (when images
 216         (gnus-html-schedule-image-fetching buffer images)))))
 217
 218 (defun gnus-html-put-image (file point)
 219   (when (display-graphic-p)
 220     (let ((image (ignore-errors
 221                    (gnus-create-image file))))
 222       (save-excursion
 223         (goto-char point)
 224         (if (and image
 225                  ;; Kludge to avoid displaying 30x30 gif images, which
 226                  ;; seems to be a signal of a broken image.
 227                  (not (and (listp image)
 228                            (eq (plist-get (cdr image) :type) 'gif)
 229                            (= (car (image-size image t)) 30)
 230                            (= (cdr (image-size image t)) 30))))
 231             (progn
 232               (gnus-put-image (gnus-html-rescale-image image))
 233               t)
 234           (when (fboundp 'find-image)
 235             (gnus-put-image (find-image
 236                              '((:type xpm :file "lock-broken.xpm")))))
 237           nil)))))
 238
 239 (defun gnus-html-rescale-image (image)
 240   (if (not (fboundp 'imagemagick-types))
 241       image
 242     (let* ((width (car (image-size image t)))
 243            (height (cdr (image-size image t)))
 244            (edges (window-pixel-edges))
 245            (window-width (truncate (* gnus-max-image-proportion
 246                                       (- (nth 2 edges) (nth 0 edges)))))
 247            (window-height (truncate (* gnus-max-image-proportion
 248                                        (- (nth 3 edges) (nth 1 edges)))))
 249            scaled-image)
 250       (when (> width window-width)
 251         (setq window-height (truncate (* window-height
 252                                          (/ (* 1.0 window-width) width)))))
 253       (if (> height window-height)
 254           (or (create-image file 'imagemagick nil
 255                             :height window-height)
 256               image)
 257         image))))
 258
 259 (defun gnus-html-prune-cache ()
 260   (let ((total-size 0)
 261         files)
 262     (dolist (file (directory-files gnus-html-cache-directory t nil t))
 263       (let ((attributes (file-attributes file)))
 264         (unless (nth 0 attributes)
 265           (incf total-size (nth 7 attributes))
 266           (push (list (time-to-seconds (nth 5 attributes))
 267                       (nth 7 attributes) file)
 268                 files))))
 269     (when (> total-size gnus-html-cache-size)
 270       (setq files (sort files (lambda (f1 f2)
 271                                 (< (car f1) (car f2)))))
 272       (dolist (file files)
 273         (when (> total-size gnus-html-cache-size)
 274           (decf total-size (cadr file))
 275           (delete-file (nth 2 file)))))))
 276
 277 ;;;###autoload
 278 (defun gnus-html-prefetch-images (summary)
 279   (let (blocked-images urls)
 280     (when (buffer-live-p summary)
 281       (with-current-buffer summary
 282         (setq blocked-images gnus-blocked-images))
 283       (save-match-data
 284         (while (re-search-forward "<img.*src=[\"']\\([^\"']+\\)" nil t)
 285           (let ((url (match-string 1)))
 286             (if (or (null blocked-images)
 287                     (not (string-match blocked-images url)))
 288                 (unless (file-exists-p (gnus-html-image-id url))
 289                   (push url urls)
 290                   (push (gnus-html-image-id url) urls)
 291                   (push "-o" urls))
 292               (gnus-message 8 "Image URL %s is blocked" url))))
 293         (let ((process
 294                (apply 'start-process
 295                       "images" nil "curl"
 296                       "-s" "--create-dirs"
 297                       "--location"
 298                       "--max-time" "60"
 299                       urls)))
 300           (process-kill-without-query process))))))
 301
 302 (provide 'gnus-html)
 303
 304 ;;; gnus-html.el ends here