On Tue, Dec 07, 2021 at 06:49:11PM +0100, Hans Bezemer wrote:
> Hi Ralph,
>
> > > awk '/^Page size:/ {print ".nr pdf-wid (p;"$3")"; print ".nr pdf-ht
> > (p;"$5")"}'
> > .nr pdf-wid (p;595)
> > .nr pdf-ht (p;842)
> > $
> Thanks, this helped.
> Changed the appropriate lines into:
> . sy pdfinfo @$1 | \
> awk '/^Page size:/ {print ".nr pdf-wid (p;"$3")"; \
> print ".nr pdf-ht (p;"$5")"}' \
> > /tmp/pdfpic\n[$$]
>
> This seems to work fine.
> I've attached a patch file.
>
>
>
> Kind regards,
>
> Hans
> --- pdfpic.tmac.origin Tue Dec 7 18:28:07 2021
> +++ pdfpic.tmac Tue Dec 7 18:34:15 2021
> @@ -84,10 +84,8 @@
> .\" get image dimensions
> . ec @
> . sy pdfinfo @$1 | \
> -grep "Page *size" | \
> -sed -e 's/Page *size: *\\([[:digit:].]*\\) *x *\\([[:digit:].]*\\).*$/\
> -.nr pdf-wid (p;\\1)\\n\
Yeah, standard (well, BSD in this case, at least OpenBSD and macOS)
sed can't be expected to be able to insert a newline using "\n" in the
replacement text of the s/// command (that's a GNU sed extension). You
would have to use an escaped literal newline instead. Looking at the
code above, it seems as removing the "n" of "\n" would be enough as it
leaves "\\\" at the end of the line. I haven't tested that.
> -.nr pdf-ht (p;\\2)/' \
> +awk '/^Page size:/ {print ".nr pdf-wid (p;"$3")"; \
I'm noting that the original regular expression allows for zero or more
spaces between "Page" and "size", and also accepts zero or more spaces
after the "size:" substring. The awk code can't allow no spaces, as it
would offset the field numbering. I don't know if this is an issue or
not.
> +print ".nr pdf-ht (p;"$5")"}' \
> > /tmp/pdfpic\n[$$]
> . so /tmp/pdfpic\n[$$]
> . sy rm /tmp/pdfpic\n[$$]
--
Andreas (Kusalananda) Kähäri
SciLifeLab, NBIS, ICM
Uppsala University, Sweden
.
