Working with PCL2PDF

[jeffs42885]

Greetings all,

I am trying to index a PDF using the OnDemand indexer (8.5.0.6) after it runs through the pcl2pdf transform. Currently we used Xenos to do this.

Looking through the REXX scripts that xenos uses, it looks like it is breaking the file (usually has about 10 or so participants) based on the social security number changing. I am no REXX expert but here is what I see


IF ssn <> previous_ssn THEN newdoc = YES

IF newdoc = YES THEN DO
        rc = dm_GenClose(gen_h)
        file_open = FALSE

Right now with PCL2PDF, I don't see a way to do this using the administrator client. I tried a few things. On the bottom of the page there is a field that is similar to this.




John Doe            Page  1  of   4                                            111SSN132                       FORM123  -  11252013 - 123456

Using the OnDemand indexer, I am triggering off the word page. and then using the appropriate fields for SSN, Contract, and date. It looks like there is a slight shift sometimes in the word PAGE which probably also means that the fields also shift their position on the page. I tried using arspdump to make sure that I was picking up the right coordinates, but the graphical indexer did not like that. I have a hand full of sample files and some of them are loading, some of them are not. Here is what I am using in the indexer.

COORDINATES=IN
TRIGGER1=UL(2.06,10.56),LR(2.67,11.00),*,'Page'
FIELD1=UL(4.64,10.60),LR(5.84,10.90),0,(TRIGGER=1,BASE=0)
FIELD2=UL(6.77,10.53),LR(7.27,11.00),0,(TRIGGER=1,BASE=0)
FIELD3=UL(7.40,10.52),LR(8.27,10.97),0,(TRIGGER=1,BASE=0)
INDEX1='rdate',FIELD2,(TYPE=GROUP)
INDEX2='ssn',FIELD1,(TYPE=GROUP)
INDEX3='contract',FIELD3,(TYPE=GROUP)
INDEXSTARTBY=2
RESTYPE=ALL

I also need to drop the last page, which I cannot figure out how to do in the admin tool, but xenos was doing it. Any help will be appreciated.

[pankaj.puranik]

Hi Jeff

Working with PDFs could be frustrating sometimes.
Specially when there are situations like you mentioned.
The most likely cause for the erros is that the coordinates are off only one some pages in the entire file.
You could check this by picking up the erroneous file and then running it througharspdoci manually.
THis will generate the index out file for you.

In the index out file check the values. You would most likely see the fields that have a wrong/bad value captured.
You may also check this using the arspdump as you already mentioned.

I am not sure how you would remove the last page except if you can write a program to chande the byte offset in the ind file before CMOD loads it.

[jeffs42885]

I figured this would be a bit tricky.

Kinda makes me wish that I could transform them to AFP, and then index them that way. The actual error I am getting is that the trigger is not found.