You'll need to add indexing to the application for PDF's. If you have an acrobat license you can use the graphical indexer otherwise you'll need to use arspdump to map the text in the pdf (short syntax arspdump -f <input pdf> -o <outputfile> -p <# of pages to index>>. What you'll end up with is coordinates that contain the text in the doc.
In the view tab set datatype PDF if you have an Acrobat license otherwise use datatype of user defined with file extension of pdf.
For Indexer information use indexer type of PDF, it is basically the same as with other indexing except you use coordinates from arspdump instead of rows/columns:
Sample that will index the CMOD readme:
COORDINATES=IN
TRIGGER1=UL(3.30,1.10),LR(3.78,1.28),*,'README'
FIELD1=UL(1.85,1.59),LR(2.25,01.77),0,(TRIGGER=1,BASE=0)
FIELD2=UL(2.99,1.26),LR(4.01,1.44),0,(TRIGGER=1,BASE=0)
INDEX1='Doc_No',FIELD1,(TYPE=GROUP)
INDEX2='Doc_Date',FIELD2,(TYPE=GROUP)
INDEXSTARTBY=2
The PDF indexing section of the Indexing reference manual will give you a more detailed description of available options.