Re: [Groff] Re: begin page blues

[Miklos Somogyi]

On 24/02/2006, at 6:50 PM, Werner LEMBERG wrote:

> > > Well, the .devicem macro will be exactly the same as \Y.  To
> > > create similar code with grops you have to use the `ps: def ...'
> > > grops command to define a PS macro which is then executed with
> > > `ps: exec' -- I'm sure you know that :-)
> >
> > I wouldn't do that. I have thousands of lines of PS routines that I
> > cat together with the prologue. Troff works with the prologue, I
> > work with my library.  Independence is as easy as apple pie.
>
> Then simply use the -P option of grops (or the corresponding
> GROPS_PROLOGUE environment variable) to pass your own prologue file.

The `cat`ted file IS the new prologue. Having access to all of my 
routines,
I see no need for \X'ps: def ..' or `mdef' to add defintions to the 
original part
of the prologue.

> grops intentionally has the limitation that you can't insert arbitrary
> PS commands which aren't encapsulated.  I consider this a good thing,
> and I'm not going to change this.

I am not saying that you should, just the opposite. Grops encapsulates 
only
with a dictionary (that \X'ps: def ..' may bypass).
That's not much of a protection, save/restore would be better.
At least the reader needs to be reminded to use his/her own save/restore
if he/she is into more serious PS business with scalings, etc.

> >    E.g. you need groff's .so in the middle of ps code, 'cause things
> >    need to be initialized before the external code and things need
> >    to be performed after it finished executing.
>
> This doesn't work.  Instead, you have to use the `.trf' request -- to
> define a macro `x' which contains the contents of file `f', say
>
>   .ev 1
>   .di x
>   .  trf f
>   .di
>   .ev
>
> This can be then expanded with \\*f.  BTW, there is a mild limitation
> that only valid groff characters are accepted in the file passed as an
> argument to `.trf'.

Thank you. I guess you mean \\*x or .x
I've tried both and prepended the included file with \! and not.
Neither of these combinations went into the PS file as PS instructions, 
just
as text to be plotted or something to cause error exits.
What did I do wrong?

However, \X'ps: file ..' does work within a macro without prefixing the 
file
with anything. That's good, ... (I bet that many of you can continue 
this line).

Question: I'll certainly have DCT encoded images in an ASCII85 coat to 
be read-in.
These only contain printable characters but, alas, backslash is 
included.
Would I need to duplicate them? Other possible problems?

Miklos

> > > BTW, does SGI troff directly emit PostScript?  Otherwise all lines
> > > with `\!' are put into the intermediate output file -- this also
> > > needs an extension to the PS output device, starting and ending
> > > rough PS data with `%' and `.', respectively.
> >
> > But I would not think that it did.
>
> It *must* do something non-standard!  `\!', if not used within a
> diversion, emits its argument verbatim into the intermediate output
> file.  Probably SGI troff doesn't use intermediate output at all but
> directly produces PS output?  Anyway, its use of `\!' is not portable.
>
>     Werner