[whatwg] SGML declaration

[Ian Hickson]

On Thu, 15 Jul 2004, fantasai wrote:
> 
> Having looked a bit more carefully at the SGML declaration, I hafta say 
> I definitely have plans to make changes. I can make brackets valid in 
> IDs for you, isn't that just *spiffy*? :D (HTML4 itself adds : and _ to 
> the list, in case you were wondering.)
> 
> As for SHORTTAG, it's not going anywhere. It's needed for dropping 
> quotes and default values and attribute minimization.

|                  LCNMCHAR ".-_:[]"
|                  UCNMCHAR ".-_:[]"

Cool. How about just making any valid string a valid ID? I never 
understood why IDs were limited to only certain characters. 
Implementations seem pretty loose about this anyway. (Why is id="1" not 
valid, e.g.?)

|                  NET      "`"     -- change from "/" to avoid mixup 
|                                      with XML --

If NETENABL is NO (as it is in your decl now), do we need to change this? 
If we do need to change this, any chance we could make it a less arbitrary 
character, such as � ?

|     STARTTAG
|       EMPTY    NO  -- outlaws "<>" --

Good good.

|       UNCLOSED NO  -- outlaws "<foo" --

We should allow this, it's actually implemented.

|        NETENABL NO  -- outlaws "<p/text<em/more text/ nested/" --

Aww! But I wanted to use this!

|      ENDTAG
|        EMPTY    NO  -- outlaws "</>" --

Pity, but ok. (I actually use this in examples quite a bit, it saves a 
lot of typing. Not that anyone implements it.)

|        UNCLOSED NO  -- outlaws "</foo" --

Again, we should allow this, it's actually implemented.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'