Handling dynamic number of tags in XML to CSV conversion

[farooquedharwad]

Hello there,
I’m new to Knime. I was trying to convert one XML file to CSV file. I’m using XML Reader, then XPath to configure the tags and then using CSV Writer to create a CSV file.
I’m facing problem in case of dynamic number of tags in XML file. Say, in 1st sub-tag of root node, there are 2 Ids used, and in second sub-tag of root node, 3 or 4 Ids are used. These sub-tags have same names. When tried to use Multiple rows/Multiple Columns/Collection Cell, i’m not getting what I wanted exactly. Below is an example of XML input and expected output in CSV:

XML Input:
    <root>
      <subtag1>
        <id>1</id>
        <value>Apple</value>
        <id>2</id>
        <value>Mango</value>
      </subtag1>
      <subtag2>
        <id>1</id>
        <value>Banana</value>
        <id>2</id>
        <value>Orange</value>
        <id>3</id>
        <value>Kiwi</value>
      </subtag2>
      <subtag3>
        <id>1</id>
        <value>Grapes</value>
      </subtag3>
    </root>


Expected CSV Output:
1|2,Apple|Mango
1|2|3,Banana|Orange|Kiwi
1,Grapes

Any workaround to achieve same? Please suggest.
Thanks in advance.

[armingrudd]

Hi @farooquedharwad and welcome to the KNIME forum,

You can first use an XPath node to extract each “subtag” as a node cell value and then extract "id"s and "value"s of each “subtag” with another XPath node. Then you can use the GroupBy node to aggregate the results.