(locale-charset-to-coding-system):

Don't rely on nil being a coding system.

diff --git a/lisp/international/mule-cmds.el b/lisp/international/mule-cmds.el
index f29e11b8e2639c52b47a1d0fc45215bd85d2f587..fb918094d4c521d7a833e10c89035a9d1321e9e7 100644 (file)
--- a/lisp/international/mule-cmds.el
+++ b/lisp/international/mule-cmds.el
@@ -1981,6 +1981,13 @@ of `buffer-file-coding-system' set by this function."
     ;; http://www.din.de/gremien/nas/nabd/iso3166ma/codlstp1/en_listp1.html.
     ;; CODESET and MODIFIER are implementation-dependent.
 
+     ;; jasonr comments: MS Windows uses three letter codes for
+     ;; languages instead of the two letter ISO codes that POSIX
+     ;; uses. In most cases the first two letters are the same, so
+     ;; most of the regexps in locale-language-names work. Japanese
+     ;; and Chinese are exceptions, which are listed in the
+     ;; non-standard section at the bottom of locale-language-names.
+
     ; aa Afar
     ; ab Abkhazian
     ("af" . "Latin-1") ; Afrikaans
@@ -2235,7 +2242,7 @@ is returned.  Thus, for instance, if charset \"ISO8859-2\",
          (if (locale-charset-match-p charset (caar cs))
              (setq c (intern (caar cs)))
            (pop cs)))
-       (coding-system-base c))))
+       (if c (coding-system-base c)))))
 
 ;; Fixme: This ought to deal with the territory part of the locale
 ;; too, for setting things such as calendar holidays, ps-print paper