[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Novel use of .char
From: |
Deri |
Subject: |
Re: Novel use of .char |
Date: |
Wed, 18 Dec 2024 15:56:12 +0000 |
On Wednesday, 18 December 2024 12:36:13 GMT Tadziu Hoffmann wrote:
> > I wrote "pdf: pdfpic" to mimic the behaviour of .PSPIC,
> > render image down from current point, [...]
> > The above difference means this works for pdf:-
> >
> > .char \[gnu] \v'-9p'\X'pdf: pdfpic GNU-head-small.pdf -L 10p'\h'10p'
> > A GNU head \[gnu] image.
>
> Thanks for the feedback! This solution is perfectly workable.
> However, I had to change the -9p to -8.8p (which is the
> aspect ratio of the image times 10p) to get the baseline of
> the text to the right of the image to be aligned with the
> text on the left.
>
> Interestingly, I also get a vertical shift if I wrap the sequence
> \v'...'\X'...' in \Z'...' which is supposed to restore the
> horizontal and vertical position after outputting the contents.
> Does that mean that the pdfpic special does not restore the
> graphics state (with q/Q) after inserting the image file?
> In that case there might be a discrepancy between what troff
> and gropdf consider to be the current output position.
Hi Tadziu,
Mea culpa (again!) although I was aware of this faux pas for a while. I
mistakenly add the height of the image to gropdf's Y position after rendering
the image. While this means that the sequence - up height of image : image :
right width of image - as in Peter's code, works, it is clearly "wrong", as
your \Z experiment shows. If I corrected this the sequence required would
become - up height of image : image : down height of image : right width of
image.
I have not done anything yet, (a) no one has reported it as bug, and (b) I
hate changing existing behaviour which may affect users documents. Given what
I discovered yesterday regarding grops rendering up from the base line and
this aberrant behaviour of mine, I am gradually coming round to the view that
a new "pdf: inline" (I did not like "pdf: pdfpicup", sounds like some sort of
road vehicle!) is required, which would sit the image on the baseline as "ps:
import" and leave the Y position well alone!! I would appreciate guidance on
whether this is a good idea or not.
The reason why this behaviour has not been reported sooner is probably because
.PDFPIC includes a break after the image so the grout file includes new H and
V commands which reset the position for gropdf.
Cheers
Deri
- Re: Novel use of .char, (continued)
- Re: Novel use of .char, G. Branden Robinson, 2024/12/15
- Re: Novel use of .char, Deri, 2024/12/16
- Re: Novel use of .char, G. Branden Robinson, 2024/12/17
- Re: Novel use of .char, Peter Schaffter, 2024/12/17
- Re: Novel use of .char, G. Branden Robinson, 2024/12/17
- Re: Novel use of .char, Peter Schaffter, 2024/12/17
- Re: Novel use of .char, Tadziu Hoffmann, 2024/12/17
- Re: Novel use of .char, Peter Schaffter, 2024/12/17
- Re: Novel use of .char, Deri, 2024/12/18
- Re: Novel use of .char, Tadziu Hoffmann, 2024/12/18
- Re: Novel use of .char,
Deri <=
- "pdf: pdfpic" and "pdf: import" (was: Novel use of .char), G. Branden Robinson, 2024/12/18
- Re: "pdf: pdfpic" and "pdf: import" (was: Novel use of .char), Deri, 2024/12/18
- Re: "pdf: pdfpic" and "pdf: import" (was: Novel use of .char), G. Branden Robinson, 2024/12/18
- Re: Novel use of .char, Tadziu Hoffmann, 2024/12/17
- Re: Novel use of .char, G. Branden Robinson, 2024/12/17
- Re: Novel use of .char, Deri, 2024/12/17
- synchronous and asynchronous grout (was: Novel use of .char), G. Branden Robinson, 2024/12/17
- Re: synchronous and asynchronous grout (was: Novel use of .char), Deri, 2024/12/18
- Re: synchronous and asynchronous grout (was: Novel use of .char), G. Branden Robinson, 2024/12/18
- Re: synchronous and asynchronous grout (was: Novel use of .char), Deri, 2024/12/18