You could try the XPath node to pull out the body, then String Splitter to split it on the ‘,’ deliminator. You can probably then do something with Row Splitter to pull out the first row. Use a Transform/RowId (to get the single column into the RowIDs) and Transform again on the first row, which contains the column headings, and then use a Table Validator node (not sure which of the 2 versions you need) to apply the column headings to the table containing the data rows
I was just wondering why this data is wrapped into an HTML structure? According to your content type in the title, this should be plain text, or CSV to be precise.
Might it be, that you connected an HtmlParser to the HttpRetriever which tries to parse the plain text into an HTML table? Instead you could rather parse the result directly: Connect a HttpResultDataExtractor to the HttpRetriever and you’ll get a plain string column which you can process further.