[
Lists Home |
Date Index |
Thread Index
]
- To: Uche Ogbuji <Uche.Ogbuji@fourthought.com>
- Subject: Re: [xml-dev] SAX and parallel processing
- From: Alan Gutierrez <alan-xml-dev@engrm.com>
- Date: Sat, 1 Jan 2005 15:42:47 -0500
- Cc: xml-dev@lists.xml.org
- In-reply-to: <1104611359.3038.148.camel@borgia>
- Mail-followup-to: Uche Ogbuji <Uche.Ogbuji@fourthought.com>,xml-dev@lists.xml.org
- References: <1104455192.3038.5.camel@borgia> <200412310131.52268.miles@milessabin.com> <1104460276.3038.23.camel@borgia> <41D4CC8C.1080200@objfac.com> <1104465543.3038.28.camel@borgia> <20041231165744.GA20756@maribor.izzy.net> <75cb920c04123110151af471f9@mail.gmail.com> <41D5DAD0.7000909@objfac.com> <20041231234449.GA21911@maribor.izzy.net> <1104611359.3038.148.camel@borgia>
- User-agent: Mutt/1.4.1i
* Uche Ogbuji <Uche.Ogbuji@fourthought.com> [2005-01-01 15:29]:
>
> On Fri, 2004-12-31 at 18:44 -0500, Alan Gutierrez wrote:
> > The characters event is interesting, becuase it is an index into
> > the parse buffer (in theory, and on Xerces indeed), but a
> > characters evet is only ever at the top of the stack. I only
> > ever need one.
>
> Interestingly enough, this is precisely one of the aspects of
> Python/SAX that is incompatible to Java (as I intimated earlier).
> in Python/SAX, we decided to go with the (in our opinions) much
> less tangled approach of making the parameter from characters
> events an actual text object, rather than a set of offsets.
>
> I know the original SAX idea was optimization, but I do think this is
> exactly one of those areas where perhaps (IMO) premature optimization
> ends up limiting design evolution, and I also think that it interferes
> with the "Simple" part.
>
> Not a problem in Python/SAX.
I've placed the characters event in a Characters object, which
will convert to string as part of toString, and cache the result.
Both worlds.
In my model the data is called a Lexeme, and it is separate from
and Event.
--
Alan Gutierrez - alan@engrm.com
|
