RE: Colors in a file generated with putwin(3c)

[John S. Urban]

Thank you for your response and your consideration. 
 The biggest question that came to my mind is that, for my current purposes,
I want to capture the screen or pad definition exactly,
Including any customized colors and the background color. But does that mean
if I read the dump back in to another program
That it will have the side-effect of changing it'sy current init_pair(3c)
settings and/or background color mode if that is added to the dump?
  For my purposes that is what I want; but that might
Not be what others expect to have happen. But init_pair(3c) settings do not
appear to be set on a per-window basis or can just
Be thought of as applying to "stdsrc", I guess. So I understand my request
may not be as straight-forward as it appears because it could have
side-effects.
  Thanks again, and thanks for the support you provide ncurses(3c);
John S. Urban

-----Original Message-----
From: Thomas Dickey [mailto:address@hidden 
Sent: Monday, March 2, 2015 5:22 AM
To: John S. Urban
Cc: address@hidden
Subject: Re: Colors in a file generated with putwin(3c)

On Sat, Feb 28, 2015 at 10:06:00PM -0500, John S. Urban wrote:
> QUESTION:
> 
> Files generated with putwin(3c) do not appear to contain information 
> to reproduce the color pairs that were in use when the putwin(3c) call 
> was made. Is that correct?

That seems to be the case: as I read the code, it is writing only the text.
 
> BACKGROUND:
> 
> I made a routine to print pads or windows as pre-formatted HTML. It 
> works quite well for my purposes when called directly from an 
> ncurses(3c) application.  Well enough that others are interested in 
> using it with their applications. To make a long story short (my 
> programs are in Fortran, theirs are in C and
> 
> Python for starters) I thought I would make a small application that 
> read a
> putwin(3c) file  (thus having a way to generate screen-shots that was 
> language-independent and already part of the ncurses(3c) library) and 
> convert that to HTML.  It works well enough for everything but color 
> for those currently interested in it.  Am I doing something wrong or 
> is there no
> 
> 
> Color pair information in the dump files?   If not, is that a feature or a
> bug? 

I think it is a bug.  (I am puzzled why it was overlooked, aside from noting
that putwin is not often used...).

On the other hand, I do not like the implementation (but have had no
compelling reason to revise it), because it relies upon the binary layout of
the window structs.  For instance, you could not transfer between
normal/wide libraries.  (I'm not keen on starting a new development task at
this moment because I was trying to wrap up things for a release, but I'll
see what I can do).

--
Thomas E. Dickey <address@hidden> http://invisible-island.net
ftp://invisible-island.net