Don't unescape %00

This patch prevents %00 from being unescaped inside Wget because
unescaping it effectively truncates the resulting C string.  This is
not a grave bug, but it results in counter-intuitive behavior, such
as:

$ wget http://host/foo%00bar
--15:52:34--  http://host/foo%00bar
           => `foo'
...
$ wget http://host/%00
--15:52:51--  http://host/%00
           => `index.html'

Rather than truncating the string I think it's more better to just
leave %00 as-is.  With the below patch the above result in:

$ wget http://host/foo%00bar
--15:54:21--  http://host/foo%00bar
           => `foo%00bar'
...
$ wget http://host/%00
--15:54:36--  http://host/%00
           => `%00'


2005-05-05  Hrvoje Niksic  <[email protected]>

	* url.c (url_unescape): Don't unescape %00, it effectively
	truncates the string.

Index: src/url.c
===================================================================
RCS file: /pack/anoncvs/wget/src/url.c,v
retrieving revision 1.121
diff -u -r1.121 url.c
--- src/url.c	2005/05/03 15:24:30	1.121
+++ src/url.c	2005/05/05 13:56:18
@@ -175,10 +175,16 @@
 	}
       else
 	{
+	  char c;
 	  /* Do nothing if '%' is not followed by two hex digits. */
 	  if (!h[1] || !h[2] || !(ISXDIGIT (h[1]) && ISXDIGIT (h[2])))
 	    goto copychar;
-	  *t = X2DIGITS_TO_NUM (h[1], h[2]);
+	  c = X2DIGITS_TO_NUM (h[1], h[2]);
+	  /* Don't unescape %00 because there is no way to insert it
+	     into a C string without effectively truncating it. */
+	  if (c == '\0')
+	    goto copychar;
+	  *t = c;
 	  h += 2;
 	}
     }