Is it possible in Ghostscript to add watermark to

2019-01-19 00:32发布

I convert PDF -> many JPEG and many JPEG -> many PDF using ghostscript. I need to add watermark text on every converted JPEG (PDF) page. Is it possible using only Ghostscript and PostScript?

The only way I found:

gswin32c -q -sDEVICE=pdfwrite -dBATCH -dNOPAUSE -sOutputFile=output.pdf watermark.ps input.pdf

But this will insert watermark.ps watermark on first separate page in output.pdf.

Can I do this on output PDF pages directly?

Can I do this on output JPEG pages directly?

<<
   /BeginPage
   { gsave
       /Helvetica_Bold 120 selectfont
       .85 setgray 130 70 moveto 50 rotate (Sample) show
     grestore
   } bind
>> setpagedevice

If I use /EndPage instead of /BeginPage - it says setpagedevice is not applicable...

How to remake this script for /EndPage?

3条回答
放我归山
2楼-- · 2019-01-19 01:03

The accepted answer was inserting pages for me. The pages were blank aside from the watermark. If you run into this try adding the 2eq bit here

<<
   /EndPage
   {
     2 eq { pop false }
     {
         gsave      
         /Helvetica_Bold 120 selectfont
         .85 setgray 130 70 moveto 50 rotate (Sample) show
         grestore
         true
     } ifelse
   } bind
>> setpagedevice

I found the following site that pointed me in the correct direction

http://habjan.blogspot.com/2013/10/how-to-programmatically-add-watermark.html

Here's the calling syntax where the above file is saved as watermark.ps and gswin32c references the ghostscript exe

gswin32c -dBATCH -dNOPAUSE -q -sDEVICE=pdfwrite -sOutputFile=watermarked.pdf watermark.ps original.pdf
查看更多
做个烂人
3楼-- · 2019-01-19 01:06

Bit too big for a comment, so I've added a new answer. The EndPage procedure (see page 441 of the PostScript Language Reference Manual) takes two additional parameters on the stack, a count of pages emitted so far, and a reason code.

You can use the count of pages to do interesting things like duplexing, or only marking even pages or whatever, but I assume in this case you don't want it, so you just 'pop' it from the stack.

The reason code tells you why the page is being emitted, again you probably don't care so you just pop the value.

Finally the EndPage must return a boolean value to the interpreter saying whether or not to transmit the page (this allows you to do other interesting things, like only printing the first 10 pages and so on).

So you need to initially remove two values, execute your code and return a boolean. Pretty trivial:

<<
   /EndPage
   { pop pop %% *BEFORE* gsave as that puts a gsave object on the stack
     gsave
     /Helvetica_Bold 120 selectfont
     .85 setgray 130 70 moveto 50 rotate (Sample) show
     grestore
     true %% transmit the page, set to false to not transmit the page
   } bind
>> setpagedevice
查看更多
做自己的国王
4楼-- · 2019-01-19 01:06

I don't know what you mean by 'directly'. Its possible, as you have found, to have a PostScript interpreter do many kinds of things on a per-page basis. PostScript is a programming language after all.

I would suggest that the /BeginPage and/or /EndPage procedures in the page device dictionary would be the place to start. These allow you to execute arbitrary PostScript at the start or end of every page.

If you define a /BeginPage procedure then it will be executed before any marking operations from the input program, if you define a /EndPage then it will be executed after the marking operations from the input program (on a page by page basis(.

This allows you to have your own marks lie 'under' or 'over' the marks from the program.

查看更多
登录 后发表回答