+ * Decode an encoded URI component in place.
+ * <b>This function does NOT decode entire URIs.</b>
+ * It decodes components (e.g. host name, directory, file name).
+ * Decoded URIs do not exist in the real world (see RFC3986 §2.4).
+ * Complete URIs are always "encoded" (or they are syntaxically invalid).
+ *
+ * Note that URI encoding is different from Javascript escaping. Especially,
+ * white spaces and Unicode non-ASCII code points are encoded differently.
+ *
+ * \return psz on success, NULL if it was not properly encoded