W3C home > Mailing lists > Public > www-validator@w3.org > February 2001

Re: Problem with parse tree output

From: Terje Bless <link@tss.no>
Date: Mon, 5 Feb 2001 14:55:31 +0100
To: Nick Kew <nick@webthing.com>
cc: Cliff Hudson <chronos@lightmail.com>, www-validator@w3.org
Message-ID: <20010205150042-r01010600-97402a5c@>
On 05.02.01 at 12:38, Nick Kew <nick@webthing.com> wrote:

>As I said in my response to the OP, the alternative, cleaner and much more
>efficient approach to this would be to hack the C++ code in SP.

I disagree. The output from SP is well defined -- de facto standard ESIS
format -- and mucking with it will not gain us anything. We'd still have to
parse it to create the complex datastructure and I don't even think the
parsing would be significantly easier with a different output format.

>learning curve for hacking SP is much steeper than the alternative basic
>Perl hackery, but there is now a development community doing that (
><URL:http://openjade.sourceforge.net/> ), as well as my own Code Valet
>which is a very specific hack to web-ify SP.

However, there are some features that can favourably be added at the SP
level. Error identifiers, for instance, would make all sorts of things
mouch easier. Other areas of improvement at the SP level would be better
calling conventions -- maybe even a Perl XS module interface! -- and better
control of the Catalog.
Received on Monday, 5 February 2001 09:00:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:17:28 UTC