Re: Rtf -> Html & Html -> Rtf

Return to index

Author

Message

Sergey Tkachenko

Posted: 04/23/2003 18:06:15

A visual control (TRichView) has a very small overhead comparing with an

imaginary non-visual control storing RichView document (overhead is one

window handle, and may be several bytes of memory).

The main overhead is requirement to store a complete document in memory

(conversion rtf -> richview -> html and vice versa, instead of imaginary

direct conversion rtf -> html on the fly). But, as I said, in some cases it

is unavoidable (for example, all properties of paragraph bullet in rtf are

not known until this paragraph is read completely; another example is

tables).

If you do not use hyperlinks, you can do conversion without formatting

document (LoadRTF, without calling Format). This will greatly spead up the

conversion and will lower memory usage.

> > There is no non-visual component for storing RichView data

> (TRVReportHelper

> > creates invisible RichView), and converting RTF to RichView is

impossible

> > without accumulating data (it's impossible to convert RTF table to HTML

> > table until the whole table is read from RTF).

>

> So, just to confirm, are you saying that it is feasible/practical to have

a

> hidden control and just have an API wrapper around this; then pass an RTF

> string and get back a HTML one? (and visa versa?)

>

> Is there going to be a large overhead in this if the text is simple (like

> bold, italic, bullet points etc)?

>

>