Fixed issue 14 (http://code.google.com/p/httrack/issues/detail?id=14) related to the way non-ascii characters are being decoded

Rationale:
  * inside URI
    * non-ascii characters are read with the page encoding, and transformed into UTF-8
    * url-escaped %xx are considered utf-8 sequences to be decoded, unless they form invalis sequences (in such case we left them as-is)
    * html entities (names, or decimal/hex) are decoded as utf-8 characters
  * inside query string
    * non-ascii characters are read as binary, and escaped using %xx
    * url-escaped %xx are left unless not harmful (alphanum, for example)
    * html entities (names, or decimal/hex) are decoded as utf-8 characters and encoded back to the page encoding (possibly using %xx)
  * inside hostnames
    * non-ascii characters are encoded using IDNA

Example:
  * are equivalent in a iso-8859-1 page: http://foo/café.html http://foo/caf%c3%a9.html http://caf&#a9;.html

0 files changed, 0 insertions, 0 deletions