Document Understanding - XML Extraction?

Help Activities
1

Hi all,

With the document understanding activities - when extracting data (in the DataExtraction Scope), there is a ‘Form Extractor’, and ‘Regex Based Extractor’ (etc), however, is there an equiverlant activity for XML based extraction that can be installed?

2

Currently the docu is not listing a XML extractor
let us know your details. Maybe we can help to setup a custom XML extraction approach that can be integrated within the flow

3

Is it something that we could do in a custom code block (or similar)?

Currently, we have a process which extracts fields from an xml document by supplying 2 elements of an xml path (a top level ‘parent’ field and a lower level ‘child’ field - for example, “Customer” and then “FirstName”).

4

Do your xml files have a consistent structure? Or they are completely different from one each other. In case they have a common structure you may use XML activities that come with UiPath.WebAPI.Activities package.

Otherwise if they don’t have a common structure you may treat them as text and try to extract data using RegEx or searching keywords in string.

Hope it helps!

5

They do have a consistant structure (for the most part). However, what we’re trying to do is process different instruction-documents from different clients. Some are XML documents (where we need to extract XML data), some are PDF files (which we use regex for)…

6

So XML Activities should be useful in your case

7

we prefer to process XML with XML Tools / Api

in case of only 2 elements are to retrieve and the XML element names are uniqu within the document, A regex approach can be checked.

Otherwise we would implement a custom XML Extractor step (do have the feeling that it not only a few lines within an invoke code) by:

  • define the document model as usual
  • define the XML Extractor config (e.g. which field will have which XPATH)
  • extract the the values, driven by the Document Type and it fields, using the Extractor config
  • manipulate / modify the ExtractionResult
    Document Understanding

just to shortlist some essential building blocks