I really like the introduction of the new JSON reading feature in Alteryx 10, but I have a problem with it.
Previously, I used to input JSON files by reading them in as simple CSV files and by adjusting some settings (delimiter set to "\0"; first row contains field names set to NO) and then later parsing the fields by using the JSON parser tool. Before using the JSON parser, I also tend run a RecordID on the dataset, so that each inputted file will carry a uniqe ID that I can use to match up with the parsed JSON data records at a later stage. Although I considered this as workaround solution, it did the job just fine.
With the new JSON reader feature I thought I could make this process slightly more efficient by eliminating the need to fiddle with CSV settings and having to use the JSON parser. However, I encountered a problem. When I read in multiple JSON files from a single location (which is normally the case with my workflows) by using the JSON reader, the records from the files are parsed into two fields automatically and there is no option to RecordID the originating files. This means that if I read in multiple files with the JSON reader, I lose the ability to track back JSON elements to their original JSON file. As this is a crucial issue for me, at the moment I cannot utlise the JSON reader as it is.
Can anyone suggest a solution? If there is none, perhaps it could be a good feature to add to the JSON reader?
