allow runes to appear in the path portion of a url.  don't bother
with the % escaping, since existing code is incorrect, and shouldn't
be applied blindly to a url, and many servers can handle unescaped
runes in the url.  so while this isn't according-to-hoyle correct, it
should be an improvement.

Reference: /n/atom/patch/applied/webfsrunes
Date: Thu May 22 20:35:36 CES 2014
Signed-off-by: quanstro@quanstro.net

--- /sys/src/cmd/webfs/fns.h	Thu May 22 20:34:05 2014
+++ /sys/src/cmd/webfs/fns.h	Thu May 22 20:34:05 2014
@@ -49,8 +49,6 @@
 void		rewriteurl(Url*);
 int		seturlquery(Url*, char*);
 Url*		copyurl(Url*);
-char*	escapeurl(char*, int(*)(int));
-char*	unescapeurl(char*);
 void		initurl(void);
 
 /* util.c */
--- /sys/src/cmd/webfs/url.c	Thu May 22 20:34:05 2014
+++ /sys/src/cmd/webfs/url.c	Thu May 22 20:34:05 2014
@@ -111,12 +111,13 @@
  */
 
 /* RE character-class components -- these go in brackets */
+#define	RUNES			" -􏿿"
 #define UNWISE		"\\[\\]|\\\\^{}`"
 #define PUNCT			"\\-_.!~*'()"
 #define RES			";/?:@&=+$,"
 #define ALNUM		"a-zA-Z0-9"
 #define HEX			"0-9a-fA-F"
-#define UNRES			ALNUM PUNCT UNWISE
+#define UNRES			ALNUM PUNCT UNWISE RUNES
 
 /* RE components; _N => has N parenthesized subexpressions when expanded */
 #define ESCAPED_1			"(%[" HEX "][" HEX "])"
@@ -528,7 +529,7 @@
 	 * Newlines are not valid in a URI, but regexp(2) treats them specially 
 	 * so it's best to make sure there are none before proceeding.
 	 */
-	if(strchr(url, '\n')){
+	if(strchr(url, '\n') != nil){
 		werrstr("newline in URI");
 		return -1;
 	}
@@ -542,7 +543,7 @@
 	 * will still get by, but that's legitimate, and if it ends up causing
 	 * a NUL then someone is unescaping too many times.
 	 */
-	if(strstr(url, "%00")){
+	if(strstr(url, "%00") != nil){
 		werrstr("escaped NUL in URI");
 		return -1;
 	}
@@ -1017,77 +1018,3 @@
 	}
 	return v;
 }
-
-static int
-dhex(char c)
-{
-	if('0' <= c && c <= '9')
-		return c-'0';
-	if('a' <= c && c <= 'f')
-		return c-'a'+10;
-	if('A' <= c && c <= 'F')
-		return c-'A'+10;
-	return 0;
-}
-
-char*
-escapeurl(char *s, int (*needesc)(int))
-{
-	int n;
-	char *t, *u;
-	Rune r;
-	static char *hex = "0123456789abcdef";
-
-	n = 0;
-	for(t=s; *t; t++)
-		if((*needesc)(*t))
-			n++;
-
-	u = emalloc(strlen(s)+2*n+1);
-	t = u;
-	for(; *s; s++){
-		s += chartorune(&r, s);
-		if(r >= 0xFF){
-			werrstr("URLs cannot contain Runes > 0xFF");
-			free(t);
-			return nil;
-		}
-		if((*needesc)(r)){
-			*u++ = '%';
-			*u++ = hex[(r>>4)&0xF];
-			*u++ = hex[r&0xF];
-		}else
-			*u++ = r;
-	}
-	*u = '\0';
-	return t;
-}
-
-char*
-unescapeurl(char *s)
-{
-	char *r, *w;
-	Rune rune;
-
-	s = estrdup(s);
-	for(r=w=s; *r; r++){
-		if(*r=='%'){
-			r++;
-			if(!isxdigit(r[0]) || !isxdigit(r[1])){
-				werrstr("bad escape sequence '%.3s' in URL", r);
-				return nil;
-			}
-			if(r[0]=='0' && r[2]=='0'){
-				werrstr("escaped NUL in URL");
-				return nil;
-			}
-			rune = (dhex(r[0])<<4)|dhex(r[1]);	/* latin1 */
-			w += runetochar(w, &rune);
-			r += 2;
-		}else
-			*w++ = *r;
-	}
-	*w = '\0';
-	return s;
-}
-