Add a way to use an external command to download HTML in eww

* doc/misc/eww.texi (Advanced): Document it.

* lisp/net/eww.el (eww-retrieve): New function.
(eww-reload): Use it.
(eww): Ditto.
(eww-retrieve-command): New variable.

3 files changed, 52 insertions, 2 deletions

diff --git a/doc/misc/eww.texi b/doc/misc/eww.texi
index 85be112402c..e814d0ac486 100644
--- a/doc/misc/eww.texi
+++ b/doc/misc/eww.texi
@@ -212,6 +212,23 @@ in an external browser by customizing
 @node Advanced
 @chapter Advanced
 
+@findex eww-retrieve-command
+  EWW normally uses @code{url-retrieve} to fetch the @acronym{HTML}
+before rendering it. It can sometimes be convenient to use an external
+program to do this, and @code{eww-retrieve-command} should then be a
+list that specifies a command and the parameters.  For instance, to
+use the Chromium browser, you could say something like this:
+
+@lisp
+(setq eww-retrieve-command
+      '("chromium" "--headless"
+                    "--virtual-time-budget=3000"
+                    "--dump-dom"))
+@end lisp
+
+The command should return the @acronym{HTML} on standard output, and
+the data should use @acronym{UTF-8} as the charset.
+
 @findex eww-view-source
 @kindex v
 @cindex Viewing Source
diff --git a/etc/NEWS b/etc/NEWS
index 8ff62b6dc05..ddc2fb9d600 100644
--- a/etc/NEWS
+++ b/etc/NEWS
@@ -809,6 +809,11 @@ background colors or transparency, such as xbm, pbm, svg, png and gif.
 ** EWW
 
 +++
+*** New variable 'eww-retrieve-command'.
+This can be used to download data via an external command.  If nil
+(the default), then 'url-retrieve' is used.
+
++++
 *** New Emacs command line convenience function.
 The 'eww-browse' command has been added, which allows you to register
 Emacs as a MIME handler for "text/x-uri", and will call 'eww' on the
diff --git a/lisp/net/eww.el b/lisp/net/eww.el
index 07aa48aeaee..bc23fb913f3 100644
--- a/lisp/net/eww.el
+++ b/lisp/net/eww.el
@@ -134,6 +134,15 @@ The string will be passed through `substitute-command-keys'."
   :type '(choice (const :tag "Unlimited" nil)
                  integer))
 
+(defcustom eww-retrieve-command nil
+  "Command to retrieve an URL via an external program.
+If nil, `url-retrieve' is used to download the data.  If non-nil,
+this should be a list where the first item is the program, and
+the rest are the arguments."
+  :version "28.1"
+  :type '(choice (const :tag "Use `url-retrieve'" nil)
+                 (list string)))
+
 (defcustom eww-use-external-browser-for-content-type
   "\\`\\(video/\\|audio/\\|application/ogg\\)"
   "Always use external browser for specified content-type."
@@ -346,9 +355,28 @@ killed after rendering."
         (let ((eww-buffer (current-buffer)))
           (with-current-buffer buffer
             (eww-render nil url nil eww-buffer)))
-      (url-retrieve url #'eww-render
+      (eww-retrieve url #'eww-render
                     (list url nil (current-buffer))))))
 
+(defun eww-retrieve (url callback cbargs)
+  (if (null eww-retrieve-command)
+      (url-retrieve url #'eww-render
+                    (list url nil (current-buffer)))
+    (let ((buffer (generate-new-buffer " *eww retrieve*")))
+      (with-current-buffer buffer
+        (set-buffer-multibyte nil)
+        (make-process
+         :name "*eww fetch*"
+         :buffer (current-buffer)
+         :stderr (get-buffer-create " *eww error*")
+         :command (append eww-retrieve-command (list url))
+         :sentinel (lambda (process _)
+                     (unless (process-live-p process)
+                       (with-current-buffer buffer
+                         (goto-char (point-min))
+                         (insert "Content-type: text/html; charset=utf-8\n\n")
+                         (apply #'funcall callback nil cbargs)))))))))
+
 (function-put 'eww 'browse-url-browser-kind 'internal)
 
 (defun eww--dwim-expand-url (url)
@@ -1117,7 +1145,7 @@ just re-display the HTML already fetched."
           (eww-display-html 'utf-8 url (plist-get eww-data :dom)
                             (point) (current-buffer)))
       (let ((url-mime-accept-string eww-accept-content-types))
-        (url-retrieve url #'eww-render
+        (eww-retrieve url #'eww-render
                       (list url (point) (current-buffer) encode))))))
 
 ;; Form support.