We have a Multi-page Layout for document which may have 2 or three pages. There is just one document section created and FLL is loaded against it. Headers are found on first page and Business data is to be read from second or third page.
Error with few documents having 2 pages : When we run the batch on certain documents , it throws following error : Incorrect page order and/or page set. go to Document in thumbnail view and correct document page order.
In such cases, it detects first page correctly but it does not detect anything from second page .
This is happening with 20% documents . Rest 80% documents are successfully processed.
If we load such cases in Flexilayout Studio for debugging, it works like charm . It also works fine if I load the same document against Document definition in Project setup station.
There are many possible explanations but in my experience the first thing to check would be if you're using footers, that the footer element isn't being 'found' on the first page. If it is found on the first page, select something unique for the footer that will only ever appear on the last page. This would be done through the flexilayout which of course would then have to be updated inside the document definition.
Update : I added a footer as well . This footer has just one element (Barcode type) . Footer Group element (Optional) with inner element Barcode as mandatory .
The footer is identified correctly in Flexilayout studio but on few documents it fails in Project setup station workflow and only first page elements are extracted nothing is read from second/third page.
I am open for suggestions .
Update : Finally found the root cause of this issue and a workaround as well. I am still looking for a permanent fix.
Classification technique : I am using a custom classification script for first page using FCTools.ClassifyPage in combination with Before Matching event to force match all pages with the document definition.
Issue at recognition stage: It was again trying to classify second and third page using classification batch , in case second page was matched with some different class then Recognition stage was failing completely for that page.
Paradox : FCTools.ClassifyPage uses classification batch for classifying first page and if I disable Classify Page at recognition stage then FCTool.ClassifyPage also fails to classify.
Tuning Classifier: I have made classifier rules more strong so that subsequent pages are not classified incorrectly at recognition.
Workaround : See the screenshot attached .
Step 1 : Select the classifier batch
Step 2: Select the Script check box.
This way FCTools.ClassifyPage keeps using classifier batch 16 and Recognition stage does not try to override my classification done in before matching .
Working for now.
So question...If you're only using headers, is that header element found on each page? When you use headers with no footer, a new document is assumed each time the header element is found.
I have been facing the same issue. Can you please share the code with me?
Please sign in to leave a comment.