SEO, MS Word, Compact HTML, and object/data

We have twenty-or-so MS Word 2000 documents that we want to display on
our website.

What we did was convert the MS Word documents to Compact HTML. We then
display a document via an
  <object data="/doc/somedoc.htm" height='100%' id='xyz' width='100%'>

This all works great except for a bit of a fly in the ointment.

Doing an SEO (Search Engine Optimization) analysis shows that a least
one analyzer does not analyze the contents of "/doc/somedoc.htm".

I guess it is reasonable not to count the contents of the document
pointed to because it might not even be owned by the displaying page.

- - - -

But these MS Word 2000 are _ours_. So does anyone know of a way to
automatically convert the htm file produced so that I can render it
rather than refer to the document via object/data?

We have twenty-or-so MS Word 2000 documents that we want to display on
our website.

What we did was convert the MS Word documents to Compact HTML. We then
display a document via an
<object data="/doc/somedoc.htm" height='100%' id='xyz' width='100%'>

This all works great except for a bit of a fly in the ointment.

Doing an SEO (Search Engine Optimization) analysis shows that a least
one analyzer does not analyze the contents of "/doc/somedoc.htm".

I guess it is reasonable not to count the contents of the document
pointed to because it might not even be owned by the displaying page.

- - - -

But these MS Word 2000 are _ours_. So does anyone know of a way to
automatically convert the htm file produced so that I can render it
rather than refer to the document via object/data?

If these documents are all alike in internal structure, you could write a little script using Nokogiri to capture only the id="whatever" node containing the page content, and then write that back out as a sort of partial. OR suck it into an ActiveRecord object and persist it in your database.

require 'rubygems'
require 'nokogiri'
require 'fileutils'

fp = '/path/to/your/file'
#if your starting document is well-formed
doc = Nokogiri::XML(File.read(fp))
#otherwise
#doc = Nokogiri::HTML(File.read(fp))
div = doc.at_css('#someDiv')
output = div.to_xhtml

Walter