# [pdftex] insert a pdf comment

Hartmut Henkel hartmut_henkel at gmx.de
Mon Jun 12 00:19:10 CEST 2006

```On Sun, 11 Jun 2006, ivo welch wrote:

> how do I insert a comment into the pdf file.
>   \pdfliteral{
>    \% this is a comment
>   }
>
> does not work, because the \ makes it into the pdf file, and
> presumably wreaks havoc.  omitting it won't help, either.  probably
> attempt.}.

e. g.

\pdfcompresslevel=0
A%
{\catcode`\%=11\pdfliteral direct {% a comment}}%
B
\bye

> some more minor suggestions to the pdf team as I am reading the docs:
>
> I believe most \pdf* variables are expanded at compile time.  for
> example, \pdfuniformdeviate is such a variable.  if pdf had a random
> number generator build in, it could be executed at pdf display time.

PDF doesn't offer that, afaik.

> if some variables have actions at pdf-run time, maybe they should have
> a slightly different name syntax (e.g., \Pdf... instead of \pdf).
> so, are there any variables that the pdf interpreter makes available?
> A pdfxform seems to be one such creature.
>
> I am of course very pleased that the base set of compile-time variable
> is becoming richer in pdftex.  Of course, far and away my biggest hope
> is that we will get a \pdfexec that is nothing but a clean
> "\write18+\input" combination, but without the need to write a file
> and then reparse it back.  (I hope taco hoekwater will figure out how
> to do this; he noted he had an idea after my pdftex feature
> suggestion.)
>
> I also just noted a new suggestion report that it should be possible
> to define a \pdfxform that does not suffer compression (i.e., escape
> the \pdfcompresslevel), perhaps \pdfxform*.

should be possible (but needs time to implement).

> This would allow me to create the equivalent of a post-latex but
> pre-run pdf variable (sed 's/myxformstring/withsomethingquick/'), and
> at very low cost---as long as I make sure I do not change the byte
> count.  (I hope standard pdf does not keep a checksum that I would
> also have to change.)
>