From 264c9bc043ba2ed3089b4c281bbeedb177cac3c2 Mon Sep 17 00:00:00 2001
From: "Michael[tm] Smith" <mike@w3.org>
Date: Mon, 2 Jan 2012 16:12:51 +0900
Subject: [PATCH] HTML IDs can contain anything except whitespace.

Introduced TY_(IsHTMLSpace)(uint c), which checks to see if c is one of the
chars that the HTML spec (and browsers) treat as a space in attribute
values: 0x020 (space), 0x009 (tab), 0x00a (LF), 0x00c (FF), or 0x00d (CF).
Can't use ANSI C isspace(int c) here because like standard functions for
many other langs, it also treats 0x00b as a space.
---
 src/attrs.c | 5 +----
 src/lexer.c | 5 +++++
 src/lexer.h | 1 +
 3 files changed, 7 insertions(+), 4 deletions(-)

diff --git a/src/attrs.c b/src/attrs.c
index 85d0751..84f9060 100644
--- a/src/attrs.c
+++ b/src/attrs.c
@@ -1367,11 +1367,8 @@ Bool TY_(IsValidHTMLID)(ctmbstr id)
     if (!s)
         return no;
 
-    if (!TY_(IsLetter)(*s++))
-        return no;
-
     while (*s)
-        if (!TY_(IsNamechar)(*s++))
+        if (TY_(IsHTMLSpace)(*s++))
             return no;
 
     return yes;
diff --git a/src/lexer.c b/src/lexer.c
index 522fcdc..04a139c 100644
--- a/src/lexer.c
+++ b/src/lexer.c
@@ -230,6 +230,11 @@ Bool TY_(IsLetter)(uint c)
     return (map & letter)!=0;
 }
 
+Bool TY_(IsHTMLSpace)(uint c)
+{
+    return c == 0x020 || c == 0x009 || c == 0x00a || c == 0x00c || c == 0x00d;
+}
+
 Bool TY_(IsNamechar)(uint c)
 {
     uint map = MAP(c);
diff --git a/src/lexer.h b/src/lexer.h
index 43b9b4a..d4d2958 100644
--- a/src/lexer.h
+++ b/src/lexer.h
@@ -416,6 +416,7 @@ void TY_(ConstrainVersion)( TidyDocImpl* doc, uint vers );
 Bool TY_(IsWhite)(uint c);
 Bool TY_(IsDigit)(uint c);
 Bool TY_(IsLetter)(uint c);
+Bool TY_(IsHTMLSpace)(uint c);
 Bool TY_(IsNewline)(uint c);
 Bool TY_(IsNamechar)(uint c);
 Bool TY_(IsXMLLetter)(uint c);