I have purchase order data set and line items are being extracted from po in abbyy. I want to count 'po line items'in PO data set which has exact line items from abbyy extracted. How to match both and count so as to export mismatch as true or false.
Match Purchase order count to Line items count
Was this article helpful?
0 out of 0 found this helpful
Comments
6 comments
Hello,
Please correct me if I am wrong, but you need to compare 2 tables and be able to tell the difference?
If so, the issue is way to complex than it seems. You need to be able to detect if the table rows have been swapped and simultaneously compare data inside the cells (that can be also recognized with mistakes).
This is a lot of coding. Just to point you out to the right direction:
https://social.msdn.microsoft.com/Forums/vstudio/en-US/6770d8c7-6135-4d01-990a-ad9146fcc170/c-code-to-compare-two-tables-and-display-the-mismatches?forum=csharpgeneral
Please consider contacting ABBYY Professional Services for this task.
Hello,
The problem is in comparing the individual rows. How do you know if the rows below are the same?
1. "Item1", "Item2", "Item3,4"
2. "Item1", "Item 2", "Item3.4"
Lets assume the following:
1) The order of the rows is exactly as in document (By my experience this assumption will not always be true)
2) The recognition quality is excellent (This assumption will most likely fail due to scanning quality and recognition of punctuation: "." can be recognized as ",", One space symbol can be recognized as 2. Even in the example you have provided there are "Item2" and "Item 2" records, that are different)
Under this 2 assumptions you can use the code provided by the link in my previous post (if query is empty - return true, otherwise - false), but you need to write the converter of the table from the FlexiCapture format to DataTable.
If you do not assume the above 2 statements, this could take hours and hours of coding, as I have wrote before, and will depend on a typical document recognition errors in you project.
The problem is way out of the scope of this forum.
Alexey
Suppose the PO DATASET has table is like this
SRNO PO NO ITEM1 ITEM2 ITEM3
1 1 ABC XYZ POR
2 1 JKD OED LKJ
3 1 NJR SED KOP
and Abbyy extracted on data form is same as above with additional row ie
SRNO PO NO ITEM1 ITEM2 ITEM3
1 1 ABC XYZ POR
2 1 JKD OED LKJ
3 1 NJR SED KOP
4 1 WER PLZ DO
for same PO NO 1 in Data set had 3 rows and for same "PO NO" which is 1 , line items extracted has 4 rows.
I just want to export a mismatch field as "false" in abbyy as for same "PO NO" - 1 there is a difference in both tables. PO Data set is in background.
consider that we have document is recognized correctly and operator has made correction if any. So now no error or mistakes but everything is same only additional row is there. What is the way to accomplish this. Can we count both no of lines and give mismatch?
Hello,
I have understood the premise of the problem. We have encountered the exact same problem in one of the projects and we have decided against implementing the solution due to the reasons above.
This will take couple of hours for me to write this program and this program will work only for ideal data with no recognition errors.
The problem is out of the scope of this forum.
Please consider contacting ABBYY Professional Services for this task.
Thanks, but please let me know if you get any solution to accomplish this, as we just want to export a field in Abbyy as "false"
Please sign in to leave a comment.