I likely need a nested IF Else RegEx series to incrementally concatenate/merge up to four strings currently in four columns (first series col 1, second series col 1 + col 2, third series col 1 + col 2 + col 3, forth series col 1+ col 2 + col 3 + col 4). Please see attached image (temp). This is the output of KNIME Term Neighborhood Extractor.
The Term Neighborhood Extractor creates these four columns in two directions: Right Neighbors and Left Neighbors.
The logic of the Right Neighbor matching is to first test the individual Term (col 1 (in the image col F)) for a text match (I will provide a library for string matching in a later step). If that single string matches a library then stop and place the string in the output column. If that fails then test col 1 + col 2 (first right neighbor (in attached image col J). If that combination matches a library then place concatenated string in the output column. If that fails then text col 1 + col 2 + col 3 for a text match. If that fails then cols 1+2+3+4. If all fails then terminate leaving a blank cell.
At the point of matching during a series the actual concatenated text is saved to the output cell. If no match then continue combining the next column.
For example, using the attached graphic: Row 8 (Right Neighbor) “mountain + views” = “mountain views”, which is tested against a library of terms. Row 11 (Right Neighbor) “upgrades” + “throughout” + “the” + “home” = “upgrades throughout the home” is grouped and tested against a library. Row 18 (Left Neighbor) “floors”… “gleaming” + “hardwood” + “floors” = “gleaming hardwood floors” and tested later against a dictionary.
The Left Neighbor matching is then similar but simply using different columns in a different order.
I’ve explored finding some videos on IF/THEN or IF/ELSE, but most are in Python, which appears to use different language expression than the KNIME Rule Engine needs for RegEx.
Thanks for your help.
