-;;; UTF-7 - A Mail-Safe Transformation Format of Unicode - RFC 2152
-;;; This is a transformation format of Unicode that contains only 7-bit
-;;; ASCII octets and is intended to be readable by humans in the limiting
-;;; case that the document consists of characters from the US-ASCII
-;;; repertoire.
-;;; In short, runs of characters outside US-ASCII are encoded as base64
-;;; inside delimiters.
-;;; A variation of UTF-7 is specified in IMAP 4rev1 (RFC 2060) as the way
-;;; to represent characters outside US-ASCII in mailbox names in IMAP.
-;;; This library supports both variants, but the IMAP variation was the
-;;; reason I wrote it.
-;;; The routines convert UTF-7 -> UTF-16 (16 bit encoding of Unicode)
-;;; -> current character set, and vice versa.
-;;; However, until Emacs supports Unicode, the only Emacs character set
-;;; supported here is ISO-8859.1, which can trivially be converted to/from
-;;; Unicode.
-;;; When decoding results in a character outside the Emacs character set,
-;;; an error is thrown. It is up to the application to recover.
+
+;; UTF-7 - A Mail-Safe Transformation Format of Unicode - RFC 2152
+;; This is a transformation format of Unicode that contains only 7-bit
+;; ASCII octets and is intended to be readable by humans in the limiting
+;; case that the document consists of characters from the US-ASCII
+;; repertoire.
+;; In short, runs of characters outside US-ASCII are encoded as base64
+;; inside delimiters.
+;; A variation of UTF-7 is specified in IMAP 4rev1 (RFC 2060) as the way
+;; to represent characters outside US-ASCII in mailbox names in IMAP.
+;; This library supports both variants, but the IMAP variation was the
+;; reason I wrote it.
+;; The routines convert UTF-7 -> UTF-16 (16 bit encoding of Unicode)
+;; -> current character set, and vice versa.
+;; However, until Emacs supports Unicode, the only Emacs character set
+;; supported here is ISO-8859.1, which can trivially be converted to/from
+;; Unicode.
+;; When decoding results in a character outside the Emacs character set,
+;; an error is thrown. It is up to the application to recover.
+
+;; UTF-7 should be done by providing a coding system. Mule-UCS does
+;; already, but I don't know if it does the IMAP version and it's not
+;; clear whether that should really be a coding system. The UTF-16
+;; part of the conversion can be done with coding systems available
+;; with Mule-UCS or some versions of Emacs. Unfortunately these were
+;; done wrongly (regarding handling of byte-order marks and how the
+;; variants were named), so we don't have a consistent name for the
+;; necessary coding system. The code below doesn't seem to DTRT
+;; generally. E.g.:
+;;
+;; (utf7-encode "a+£")
+;; => "a+ACsAow-"
+;;
+;; $ echo "a+£"|iconv -f iso-8859-1 -t utf-7
+;; a+-+AKM
+;;
+;; -- fx
+