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