Why 10 lines?

[Juri Linkov]

sgml-html-meta-auto-coding-function has the hard-coded limit of 10 lines
to search for the HTML meta tag.  But HTML files can have the HTML meta tag
outside the 10-line limit.  For example, HTML files generated by livejournal
contain this tag on 11-th line, and Emacs fails to recognize the coding
of such HTML files.

I propose to limit the search for the HTML meta tag by the end of the
existing HTML header (by looking for </head>).  The limit of 10 lines
(or perhaps any slightly increased number) could be still applied only
for the case if there are no HTML header.

Index: lisp/international/mule.el
===================================================================
RCS file: /sources/emacs/emacs/lisp/international/mule.el,v
retrieving revision 1.236
diff -c -r1.236 mule.el
*** lisp/international/mule.el  24 May 2006 13:22:12 -0000      1.236
--- lisp/international/mule.el  1 Jun 2006 00:55:47 -0000
***************
*** 2253,2261 ****
    "If the buffer has an HTML meta tag, use it to determine encoding.
  This function is intended to be added to `auto-coding-functions'."
    (setq size (min (+ (point) size)
-                 ;; Only search forward 10 lines
                  (save-excursion
!                   (forward-line 10)
                    (point))))
    (when (and (search-forward "<html" size t)
             (re-search-forward 
"<meta\\s-+http-equiv=\"content-type\"\\s-+content=\"text/\\sw+;\\s-*charset=\\(.+?\\)\""
 size t))
--- 2257,2267 ----
    "If the buffer has an HTML meta tag, use it to determine encoding.
  This function is intended to be added to `auto-coding-functions'."
    (setq size (min (+ (point) size)
                  (save-excursion
!                   ;; Limit the search by the end of the HTML header
!                   (or (search-forward "</head>" size t)
!                       ;; In case of no header, search only 10 lines
!                       (forward-line 10))
                    (point))))
    (when (and (search-forward "<html" size t)
             (re-search-forward 
"<meta\\s-+http-equiv=\"content-type\"\\s-+content=\"text/\\sw+;\\s-*charset=\\(.+?\\)\""
 size t))

-- 
Juri Linkov
http://www.jurta.org/emacs/