From e35f4502383e368747d5f2bd8bcb9ed872315029 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sebasti=C3=A1n=20Mon=C3=ADa?= Date: Tue, 8 Oct 2024 23:26:42 -0400 Subject: [PATCH] Add customization to let EWW guess content-type if needed (bug#73133) --- lisp/net/eww.el | 39 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 38 insertions(+), 1 deletion(-) diff --git a/lisp/net/eww.el b/lisp/net/eww.el index b5d2f20781a..147982057c5 100644 --- a/lisp/net/eww.el +++ b/lisp/net/eww.el @@ -108,6 +108,19 @@ eww-suggest-uris eww-current-url eww-bookmark-urls)) +(defcustom eww-guess-content-type-functions + '(eww--html-if-doctype) + "List of functions used to guess a page's content-type. +These are only used when the page does not have a valid Content-Type +header. Functions are called in order, until one of them returns the +value to be used as Content-Type. They receive two parameters: an alist +of headers, and the buffer that holds the complete response. If the +list is exhausted, eww assumes \"text/plain\" so the user can see the +markup." + :version "31.1" + :group 'eww + :type '(repeat function)) + (defcustom eww-bookmarks-directory user-emacs-directory "Directory where bookmark files will be stored." :version "25.1" @@ -630,6 +643,30 @@ eww-html-p (member content-type '("text/html" "application/xhtml+xml"))) +(defun eww--guess-content-type (headers response-buffer) + "Use HEADERS and RESPONSE-BUFFER to guess the Content-Type. +Will call each function in `eww-guess-content-type-functions', until one +of them returns a value. This mechanism is used only if there isn't a +valid Content-Type header. If none of the functions can guess, return +\"text/plain\", so at least the mark up is displayed." + (or (run-hook-with-args-until-success + 'eww-guess-content-type-functions headers response-buffer) + "text/plain")) + +(defun eww--html-if-doctype (_headers response-buffer) + "Return \"text/html\" if RESPONSE-BUFFER has an HTML doctype declaration. +HEADERS is unused." + ;; https://html.spec.whatwg.org/multipage/syntax.html#the-doctype + (let ((case-fold-search t) + (target + "\\|system +\\(\\\"\\|'\\)+about:legacy-compat\\)")) + (with-current-buffer response-buffer + (goto-char (point-min)) + ;; match basic and also legacy variants as + ;; specified in link above + (when (re-search-forward target nil t) + "text/html")))) + (defun eww--rename-buffer () "Rename the current EWW buffer. The renaming scheme is performed in accordance with @@ -659,7 +696,7 @@ eww-render (content-type (mail-header-parse-content-type (if (zerop (length (cdr (assoc "content-type" headers)))) - "text/plain" + (eww--guess-content-type headers (current-buffer)) (cdr (assoc "content-type" headers))))) (charset (intern (downcase -- 2.43.0