Thanks for the pointer. I was able to fix the issue without using metdata sequencer and still using the XML Splitter. However, I have tested with using Meta data sequencer instead of XML splitter and the error still persists. The fix was not related to what you use for splitting but to do with the Capture Fields Generator plugin as I describe below.
I will explain the scenario in detail.
My xml data stream has a node that has the id of the document with the following structure.
I want to use the contents of the id node to identify the capture document in the database, so that when I go to Prefernces > PP Capture > Manage Documents, I see a list like as follows. (Please let me know if there is another better way to do this).
Here is my workflow that has removed the I/O errors but I have a new issue with printing the capture ready doc on the printer.
1 - Folder capture
2 - XML Splitter - It is set to split when condition node "<record>" is found.
3 - Create metadata - I have selected the appropriate ptk file.
4 - Capture Fields Generator - Here is where I have now selected "Continue Process with PDF (with new metadata)" instead of "Continue process with optimized PS (no metadata)". This is what fixed the errors as mentioned in my previous post. Also here is where in the "Document Title" field I am using the "<id>" field from the data stream through a data selection. This stores the capture documents as required and as mentioned above - CaptureDoc_001, CaptureDoc_002...etc.
5 - Unconditional Branch (go right) - Printer queue output (LPR pointing to the IP of my printer). I have selected passthrough mode. The new problem here is now that it gives PS errors in the printout. I assume it is because the printer is expecting PS and not an already ripped PDF?
6 - After the Branch (go down) - Send to folder and save as .pdf - this works without any problems.