2010/11/24 Doug Blank <doug.blank@gmail.com>
On Tue, Nov 23, 2010 at 5:23 PM, Rob Healey <robhealey1@gmail.com> wrote:
> Greetings:
>
> Please correct me if I am incorrect in my understanding of a few things:
>
> 1) If all of the stand-alone pieces or sub-parts of NarWeb and will only be
> used to create web pages, then what is the purpose for needing and using
> HtmlDoc?  Gerald already created libhtml.py to parse through the created
> pages and write them out to XHtml...

There was a request to attempt to unify some things between the two
HTML gen systems, but as I mentioned, I'm not working on that level of
refactor.

But, it might be possible to include text reports using the HTMLDoc
backend. For example, it could be possible to have a link of of each
person's page to a statically created HTML text report.


Just to put the dot's on the i, HtmlDoc is the html implementation of our textreport doc backend. There is no reason for narweb to use that.
HtmlDoc uses libhtml, which narweb uses also. However, HtmlDoc does this via htmlbackend, which contains some functions that narweb might need or might want to extend (eg links in notes would be done by the backend, but then we need a way to know the page to link to).

So, the overlap is only in some specific functions. As the aim is different, some overlap probably cannot be avoided. The complicated stuff should be done only once though, which specifically means eg styled notes.

Benny