bug#16225: 24.3.50; [PATCH] eww: machinery to set character encoding.

[Kenjiro NAKAYAMA]

> However, the implementation doesn't seem ideal.  The best way to handle
> this would be to have the `E' command prompt for the charset, and then
> re-render the page immediately without setting any global variables.

Thank you Lars,

I fixed it. Can you please review again? I send the patch again.

Signed-off-by: Kenjiro NAKAYAMA <nakayamakenjiro@gmail.com>

        * net/eww.el (eww-display-html,eww-display-raw): Change to
          enable to set encoding type.
          (eww-mode-map): New key map and easy-menu to set encoding type.
          (eww-set-character-encoding): New funtion to set encoding type.

---
 lisp/net/eww.el | 43 ++++++++++++++++++++++++++++++-------------
 1 file changed, 30 insertions(+), 13 deletions(-)

diff --git a/lisp/net/eww.el b/lisp/net/eww.el
index 02c93a0..8fbf94d 100644
--- a/lisp/net/eww.el
+++ b/lisp/net/eww.el
@@ -172,7 +172,7 @@ word(s) will be searched for via `eww-search-prefix'."
                    "/")
               (expand-file-name file))))
 
-(defun eww-render (status url &optional point)
+(defun eww-render (status url &optional point encode)
   (let ((redirect (plist-get status :redirect)))
     (when redirect
       (setq url redirect)))
@@ -192,7 +192,7 @@ word(s) will be searched for via `eww-search-prefix'."
                    (or (cdr (assq 'charset (cdr content-type)))
                        (eww-detect-charset (equal (car content-type)
                                                   "text/html"))
-                       "utf8"))))
+                       "utf-8"))))
         (data-buffer (current-buffer)))
     (unwind-protect
        (progn
@@ -203,12 +203,12 @@ word(s) will be searched for via `eww-search-prefix'."
                                  (car content-type)))
             (eww-browse-with-external-browser url))
           ((equal (car content-type) "text/html")
-           (eww-display-html charset url nil point))
+           (eww-display-html charset url nil point encode))
           ((string-match-p "\\`image/" (car content-type))
            (eww-display-image)
            (eww-update-header-line-format))
           (t
-           (eww-display-raw)
+           (eww-display-raw encode)
            (eww-update-header-line-format)))
          (setq eww-current-url url
                eww-history-position 0))
@@ -243,12 +243,12 @@ word(s) will be searched for via `eww-search-prefix'."
 (declare-function libxml-parse-html-region "xml.c"
                  (start end &optional base-url))
 
-(defun eww-display-html (charset url &optional document point)
+(defun eww-display-html (charset url &optional document point encode)
   (or (fboundp 'libxml-parse-html-region)
       (error "This function requires Emacs to be compiled with libxml2"))
-  (unless (eq charset 'utf8)
+  (unless (eq charset encode)
     (condition-case nil
-       (decode-coding-region (point) (point-max) charset)
+       (decode-coding-region (point) (point-max) encode)
       (coding-system-error nil)))
   (let ((document
         (or document
@@ -363,11 +363,16 @@ word(s) will be searched for via `eww-search-prefix'."
                                  (list :background (car new-colors))
                                  t))))))
 
-(defun eww-display-raw ()
+(defun eww-display-raw (&optional encode)
   (let ((data (buffer-substring (point) (point-max))))
     (eww-setup-buffer)
     (let ((inhibit-read-only t))
-      (insert data))
+      (insert data)
+      (unless (eq encode 'utf-8)
+       (encode-coding-region (point-min) (1+ (length data)) 'utf-8)
+       (condition-case nil
+           (decode-coding-region (point-min) (1+ (length data)) encode)
+         (coding-system-error nil))))
     (goto-char (point-min))))
 
 (defun eww-display-image ()
@@ -420,6 +425,7 @@ word(s) will be searched for via `eww-search-prefix'."
     (define-key map "C" 'url-cookie-list)
     (define-key map "v" 'eww-view-source)
     (define-key map "H" 'eww-list-histories)
+    (define-key map "E" 'eww-set-character-encoding)
 
     (define-key map "b" 'eww-add-bookmark)
     (define-key map "B" 'eww-list-bookmarks)
@@ -442,7 +448,8 @@ word(s) will be searched for via `eww-search-prefix'."
        ["List histories" eww-list-histories t]
        ["Add bookmark" eww-add-bookmark t]
        ["List bookmarks" eww-list-bookmarks t]
-       ["List cookies" url-cookie-list t]))
+       ["List cookies" url-cookie-list t]
+       ["Character Encoding" eww-set-character-encoding]))
     map))
 
 (defvar eww-tool-bar-map
@@ -552,11 +559,11 @@ appears in a <link> or <a> tag."
        (eww-browse-url (shr-expand-url best-url eww-current-url))
       (user-error "No `top' for this page"))))
 
-(defun eww-reload ()
+(defun eww-reload (&optional encode)
   "Reload the current page."
   (interactive)
   (url-retrieve eww-current-url 'eww-render
-               (list eww-current-url (point))))
+               (list eww-current-url (point) encode)))
 
 ;; Form support.
 
@@ -1032,7 +1039,7 @@ If EXTERNAL, browse the URL using `shr-external-browser'."
      ((and (url-target (url-generic-parse-url url))
           (eww-same-page-p url eww-current-url))
       (eww-save-history)
-      (eww-display-html 'utf8 url eww-current-dom))
+      (eww-display-html 'utf-8 url eww-current-dom))
      (t
       (eww-browse-url url)))))
 
@@ -1083,6 +1090,16 @@ Differences in #targets are ignored."
        (setq count (1+ count)))
       (expand-file-name file directory)))
 
+(defun eww-set-character-encoding (encode)
+  "Set character encoding."
+  (interactive "sSet Character Encoding (default utf-8): ")
+  (cond ((zerop (length encode))
+        (eww-reload 'utf-8))
+       (t
+        (if (not (coding-system-p (intern encode)))
+            (user-error "Invalid encodeing type.")
+          (eww-reload (intern encode))))))
+
 ;;; Bookmarks code
 
 (defvar eww-bookmarks nil)
-- 
1.8.3.1

Regards,

Kenjiro

larsi@gnus.org writes:

> Kenjiro NAKAYAMA <nakayamakenjiro@gmail.com> writes:
>
>> This report includes a patch to the bug. Please, review and install it
>> to the official tree if appreciated.
>>
>> The user can't change encoding type. This patch is to fix it.
>
> Allowing the user to alter the encoding seems like a good idea.
> However, the implementation doesn't seem ideal.  The best way to handle
> this would be to have the `E' command prompt for the charset, and then
> re-render the page immediately without setting any global variables.