You are viewing limited content. For full access, please sign in.

Question

Question

Possible enhancement: Add sub-classes for each document classification in Quick Fields

Quick Fields
Updated May 25, 2022
asked on May 23, 2022 • Show version history

Use case:  We have a hiring packet with multiple forms.  The first form will have a QR code on it with the employee number_doc type ("Hire Packet")_Year in the QR code.  We will need to scan the packet, store the metadata from the QR code, then split the packet so each form is stored in its own folder by employee number and year. 

It appears that the only way this can currently be accomplished is to have two scanning sessions.  The first session is used to assign the metadata to the document and store the document in a new folder.  The second session would then be able to split the new document into multiple documents based on OCR rules, and store the new documents into their new locations by employee number and year that are retrieved from the metadata fields.

Request:  Allow one session to have a document class with multiple sub-classes.  The primary class would read the QR code and identify the document as a hire packet.  The sub-classes would be used to split the document into new documents based on OCR rules but would also be able to assign field values from the metadata in the parent doc's QR code.

 

0 0

Answer

SELECTED ANSWER
replied on May 24, 2022

So, you have a session that already identifies and breaks up the documents in each packet, correct?

And the first document in each packet has a QR code (that you can use to identify that it's a new document, probably along with other criteria).

You can then use a Pattern Matching activity in the Page Processing section of that first document class to parse apart the QR barcode value into the the "employee number" and "year".  This gives you the two tokens that you're using for setting document storage location (and possibly setting them in metadata fields as well).

In that same Page Processing section of the first document class, you then add two Token Collector processes, one each configured to "collect" the "employee number" and "year" tokens.

In the Page Processing section of every-document-class-except-the-first, you then add two Token Retriever processes, one each linked to the Token Collector processes you added in the first document class.  This is how you can pass token values between documents in Quick Fields, which is what you want to do--you want to pass the QR code data to each document in the packet.

When the second packet is being processed, it's form will again get identified as the first document class.  The pattern matching activity will create tokens with new values.  And the Token Collector processes will then update the values to reflect that this is the second packet.  All subsequent documents (until a new packet) will then use these new values.

Each document class sets its own storage location (and field values) and that storage location (and field values) can use tokens generated from the Page Processing section, including from the Token Retriever processes.

Does that make sense?

1 0

Replies

replied on May 23, 2022

Unfortunately, this would be incredibly complex to do.  You have the right approach, I believe: use two different sessions.  If you have Quick Fields Agent, you can automate this process so that the user doesn't have to manually open the different sessions.

Alternatively, depending on the specifics of your situation, it may be possible that you could skip the first session identification phase by storing the result of the QR code in a Token Collector process and then using Token Retriever to retrieve the result for subsequent documents in the same packet.

1 0
View 4 previous replies
replied on May 24, 2022 • Show version history

I don't think Quick Fields Agent would be useful, as I have tested using it with other processes and I have experienced frequent issues with lost connection errors.  My VAR suggested running the agent on a different device than the main LF server but I have yet to try this.

As for the Token Collector process, what are your thoughts exactly on  how to implement this?  Are you thinking that I could collect the tokens in Pre-Classification Processing and then use them in the document classification? 

Thanks for the input.

Edit:  I cannot find either Token Collector or Token Retriever in the processes available in either Pre-Classification or First Page Identification activities.  Can you provide more information on how I would do this?

Thanks.

 

0 0
SELECTED ANSWER
replied on May 24, 2022

So, you have a session that already identifies and breaks up the documents in each packet, correct?

And the first document in each packet has a QR code (that you can use to identify that it's a new document, probably along with other criteria).

You can then use a Pattern Matching activity in the Page Processing section of that first document class to parse apart the QR barcode value into the the "employee number" and "year".  This gives you the two tokens that you're using for setting document storage location (and possibly setting them in metadata fields as well).

In that same Page Processing section of the first document class, you then add two Token Collector processes, one each configured to "collect" the "employee number" and "year" tokens.

In the Page Processing section of every-document-class-except-the-first, you then add two Token Retriever processes, one each linked to the Token Collector processes you added in the first document class.  This is how you can pass token values between documents in Quick Fields, which is what you want to do--you want to pass the QR code data to each document in the packet.

When the second packet is being processed, it's form will again get identified as the first document class.  The pattern matching activity will create tokens with new values.  And the Token Collector processes will then update the values to reflect that this is the second packet.  All subsequent documents (until a new packet) will then use these new values.

Each document class sets its own storage location (and field values) and that storage location (and field values) can use tokens generated from the Page Processing section, including from the Token Retriever processes.

Does that make sense?

1 0
replied on May 25, 2022 • Show version history

Yes, that makes sense.  So, using Pattern Matching in the first document class that has the QR Code will create/collect the tokens for that QR Code; this will be the token collector.  Then, in each subsequent document class I will use the Assign Token Value to retrieve the token token values from the Pattern Matching activity and this will be the token retriever.  Is that correct?  I guess I was just a bit confused by the terminology.

I will try this out.  Thanks.

Edit:  The Assign Token Value activity doesn't really behave as expected (compared to Workflow).  I can use it to create token values for the QR Code using the Split function but I don't have the option on the following document classes to retrieve the token values from the first document class; I only have the option to create new tokens.  Pattern Matching doesn't seem to apply here, since the Split function in Assign Token Value gets me what I need to create the tokens.  Now I'm just trying to figure out how to retrieve the token values on subsequent doc classes...

0 0
replied on May 25, 2022

No, you wouldn't use Assign Token Valuess. You would use Token Retriever in the subsequent document classes.

0 0
replied on May 25, 2022

I looked again and still cannot find "Token Retriever".

0 0
replied on May 25, 2022

Token Collector and Token Retriever are processes you add to your document classes.

In your screenshot, you have First Page Identifications selected in your Session Tree.  You need to select Page Processing for the document class.

Using an Assign Token Values process with the split function is a great alternative to using the Pattern Matching activity.  But you need to add two Token Collector processes *after* that to "publish" the token for other document classes to use. 

1 0
replied on May 25, 2022

Thanks for the clarification!  I obviously didn't even look there. 

0 0
You are not allowed to follow up in this post.

Sign in to reply to this post.

Details
Related Posts
Loading ...