Acrobat Catalog Problems and Workarounds
 Acrobat Catalog Problems and Workarounds
Cataloging Large Amounts of PDF Files - Must Modify Registry
Q: I have a client who is using Adobe Catalog to create an index of a large
number of PDF files. He has advised me that Catalog will not allow you to
index over 3GB of PDF files, or allow the acrobat index to become larger
than 150 MB. I have never heard of these limitations with Catalog. Is
anyone familiar with this, or is there something being done incorrectly?
A: Suggest that multiple indexes be done and then add those indexes to the
search option. If there are that many files, I would suggest a 3rd party
application that can handle the large amount of pdf files.
A: The root issue is that Catalog's default settings are such that there is an idividual file size limit. On the Mac this is easily reset on the Mac in the Edit > Preferences to accommodate larger files, on PCs you must modify the Registry.
I have no idea what happens on Unix. Details are in the Acrobat Catalog API Reference, mentioned below. I'm now cataloging large files and file sets just fine.
A: Up to version 3: there is an acrocat.ini file in the Windows folder which stores preferences (less convenient, but something we could live with).
In version 4: you need to add settings in the Windows registry, best done by importing them from a text file (see note below). Numeric values has to be entered in hex.
The information is concealed in the Acrobat Catalog API Reference, part of the Acrobat SDK, accessible through partners.adobe.com/asn/developer/acrosdk/main.html.
(Relevant portion is in pages 8-16).
It is a recommended practice to back up your registry before modifying it; back up your whole system too.
As the note in the CatAPI.pdf document warns: "Missing or incorrect values in the registry can render the Windows NT installation unusable".
Auto-attaching and Index Causes Problems between Platforms
We have had an "Index not found" error message when using Reader 3.x with a
catalog generated with 4.x - but not necessarily systematically.
Also auto-attaching an index can cause problems between platforms.
But the biggest pitfalls lie in using PC-generated indexes on UNIX platforms.
The following responses from Jon White (both to the forum and direct to me)
explain what you need to do. By following these instructions to the letter we
have been able to overcome our problems.
I can however state that auto-attaching is something that we have abandoned
for the time being pending further testing.
Unix Case Sensitivity and 8.3 format
We regularly create Catalog indexes for our UNIX customers with no problem,
but you do have to carefully observe a few rules:
1. If you're using WinNT, upgrade to Acrobat 4.0. Catalog 3.x
on NT has a casing problem that breaks indexes on UNIX. (Strangely
enough, indexes created on Win95 don't having this casing problem.)
2. All PDF filenames in the index must be lowercase and 8.3 format.
3. The INDEX.PDX filename and the INDEX folder name must be uppercase.
If you don't want to upgrade to Acrobat 4.0, you can write a script
that downcases all of the objects underneath the INDEX folder.
Limited Unix Support
Adobe has never made Catalog available on any UNIX platform, and with 4.0
they have further backed away from UNIX support by providing only the
Reader (no Exchange or Distiller) for UNIX platforms.
Catalog backward compatability Problem
If the users are using the latest version of 4.0 reader there should not be
a problem with using Catalog 4.0. If your users are using 3.0 you will have
a problem. I believe the problems lies with the fact that 3.0 is 16 bit and
4.0 is 32 bit
|