Grid Extractor - How To Identify Row Headers

tcoates

Does someone have an example of how to configure a Grid extractor for a table?  I am trying to understand how to define the X and Y extractors.  

I am trying to improve a table extraction where values in some columns are missing (meaning there is nothing printed on the document).  I have made some of the data groups variable, but since I do not know which columns might be blank the row extractor does not seem to be the best option.

For example:
Column1     Column2     Column3     Column4
Date            Item             Cost            Comments
1/1/18          WidgetA       $23.00        Outdoor Rated
1/2/18                               $45.00       Not In Stock
1/3/18           Widget B                        Not for Sale
1/4/18           Widget C     -$10.00       Refund Amount
1/5/18           WidgetA      $23.00       
1/6/18            WidgetA     $23.00        Enjoy

Thanks

GrooperGuru

I'm glad you've brought up this discussion. This is a common problem in table extraction that are row-based logic is not designed to handle elegantly. There are a few options that MAY work in your scenario, though the configuration is pretty difficult to explain. The shortest answer is... this is a MAJOR enhancement in the upcoming version of Grooper (2.7). We will have new methods for capturing table data that are based on horizontal/vertical cross sectioning. These will use header extractors to understand the X position of columns. Then row extractors to find the Y location of the data. For now, I'm going to PM you a short powerpoint that explains these new features. We will soon have a full article in x Change, but it is a very big topic to cover.

For now, I may have some workarounds. Do you have time this morning for a 15 minute call?