RE: A simple Annotation System

Just for the record* : I hacked together a browser-interface annotation
system prototype using JScript on IIS which is written up in "XML
Application Development with MSXML 4.0" (Wrox Press). It uses (a subset of)
the Annotea RDF language for data.

As you might imagine, the biggest problem was the creation of XPointers from
non-XPointer capable browsers such as IE. My crude but surprisingly usable
solution was to have the user copy & paste the section of text in the page
they wished to annotate into a HTML form (the form appears in the browser
below the source page). This string is then matched against the original
text to locate it and with a bit of jiggerypokery the XPath + position
obtained. If there is more than one matching block of text, the user is
given the choice of the alternatives.

To use MSXML on arbitrary HTML pages the source documents were
(automatically) run through a COM implementation of HTML Tidy to produce
XML.

Once the XPointer was obtained, generated the RDF syntax was straightforward
using regular DOM techniques (there wasn't any need in this app for the true
RDF model).

Going the other way, to display the annotated pages the original page is
read and chopped up and the necessary HTML elements inserted to provide an
icon and link to the comment, this augmented version being presented to the
user's browser. This was absolutely usable.

The source should be on Wrox's site if anyone's interested, if not, mail me.

Cheers,
Danny.

* Royalties? What could you be suggesting..?

-----------

http://dannyayers.com

"The lyf so short, the craft so long to lerne." - Chaucer



>-----Original Message-----
>From: www-annotation-request@w3.org
>[mailto:www-annotation-request@w3.org]On Behalf Of Charles
>McCathieNevile
>Sent: 03 December 2002 12:15
>To: Laurent Denoue
>Cc: www-annotation@w3.org; Venkatasubramani S
>Subject: RE: A simple Annotation System
>
>
>
>On Wed, 27 Nov 2002, Laurent Denoue wrote:
>
>>I also have a few questions:
>
>- how did you manage to annotate PDF documents? Do you simply annotate the
>whole document or can you select some text as well?
>
>- can you annotate part of a picture or only the whole picture? How do the
>annotations appear in the original document?
>
>The annotea protocol, by using Xpointer can address any kind of
>XML document.
>So it can point into part of an SVG image, for example. Nick Kew
>and Jim Ley
>did some work on pointing into non XML documents using what they called a
>fuzzy pointer - essentially it relies on having some way of addressing part
>of an object.
>
>There are people working on how to point to a part of a flash presentation,
>and Adobe have been working on tagged PDF which should provide a
>way to point
>towards parts of a PDF document, although I am hardly an expert there.
>
>CHeers
>
>Chaals
>

Received on Friday, 14 February 2003 13:55:10 UTC