So : firstly you should probably avoid directly reading from full paths in your PRPC activities (or indeed any web application); if you really do need to read from a particular file location ('outside' of the PRPC Web App) - PRPC has the concept of a FileListener which can monitor designated directories for files.

Another option: if you can copy your PDF to the 'ServiceExport' directory in your PRPC system  (this is where uploads/downloads of things like ZIP files / RAP files take place) - this is located in your 'PegaTemp' directory like this:

[Path to 'PegaTemp']/StaticContent/global/ServiceExport

You can read files from this directory using the Pega Engine API; for instance the following Java Activity Step will read a PDF from the directory and extract the text from it (using 'PDFBox').

try { 
    PRInputStream pris=new PRInputStream( new PRFile( PDFPath) );
    org.apache.pdfbox.pdmodel.PDDocument doc=org.apache.pdfbox.pdmodel.PDDocument.load( pris );
    org.apache.pdfbox.text.PDFTextStripper pdTextStripper=new org.apache.pdfbox.text.PDFTextStripper();
    text=pdTextStripper.getText( doc );
}
catch(Exception e) { 
    throw new PRRuntimeException(e);
}

The Local Variable 'PDFPath' is first set up (using a 'Property-Set' method) like this:

Property-Set local.PDFPath <- .pxServiceExportPath + Param.pdf_filename [ StepPage:pxProcess ]

Then you can view the contents of the Local Variable 'text' using a 'Show-Property'.

Show-Property Local.text

Here's some screenshots of such an example Activity:

And here's it running (using this example PDF) :

(The Browser shows this text all squashed together; but in fact (View Source shows this) the text *does* have preserved linebreaks etc):

This was done on PRPC73: if you have an earlier version of PRPC - you might find you need to adjust the package names.

Check 'pr_engineclasses' table if unsure; eg:

select 
      pzpackage,pzclass from pr_engineclasses 
where 
       lower(pzjar) like '%pdfbox%' 
and    lower(pzclass) like '%stripper%';