Hi @Jake_Hoover - the two methods I talked about were very much in the no-code/low-code space and accessible to all. But, if you’re already planning to use the API for loading the data, then I would definitely recommend this as the tool to connect your 4 data sources. Definitely possible to have a process the finds records in table B and connects them to the matching records in table A.

On the issue of 4 tables or 1 mega-table, my instinct is to go for 4 “data loading” tables and create a 5th table that connects the records, creates the views and so on and is the “working” table. But that is based on a preference for keeping the loaded data “as is” in case of the need for audit or issue resolution - obviously, I don’t have any understanding of your data, set-up, working practices etc, which would influence the approach.

Yep - not seen too much of that either. Think there’s a lot of potential to make bigger apps in Airtable if you use the API and the new scripting block.

JB

Hi @JonathanBowen, thanks so much for the quick response! Much appreciated.

I took a cursory look over the API, and did not see anything recorded on how we could leverage the Linking feature of Airtable. Do you have any other background/directions you could point me in?

Re: 4 tables vs. 1 mega-table, thank you for the insight! I agree, conceptually it is far more clean. Creating a 5th table to pull all of the data together is definitely the goal-- and then as you mentioned, creating multiple views from the 5th ‘master’ table.

Bonus question, would we be able to use Airtable to parse values from a single string/field into multiple fields? An example application would be taking a URL with various UTM parameters, and auto-generating various fields

We post to a field called ‘Entry URL’ - airtable.com?utm_source=linkedin&utm_medium=cpc&utm_campaign=better-than-spreadsheets

We then have other fields’ Entry URL Source’, ‘Entry URL Medium’, etc. which would parse the URL to identify the ‘Entry URL Source’ is “linkedin”, Entry URL Medium’ as “cpc”, etc. We could also manipulate this data before we post to Airtable, but this type of application would serve a few other uses cases in addition to the example above.

Yes, you can do this (to some degree, depending upon your scenario) with Airtable formula fields. Better would be the new Airtable scripting block - see here for example.

But even better, as you’re planning to use the API to load the data, I think is to do this on import and make it an integral part of the data loading process (even if you keep the original field data for reference alongside the split fields).

I’ll post again on the API/linking question.

JB

Hi @Jake_Hoover - so to show a simple example, let’s say you’ve got this set-up:

Table1 - a data loading table


Master - the master table

What I want to do is have the ID in Master populated with the common ID (across all data loading tables) then have linked fields which hold the same IDs, but as links to Table 1 (and other data loading tables).

So the end result should look like this:

Populating this manually in the base, then looking at the API docs for the base gives me this to create a record:

curl -v -X POST https://api.airtable.com/v0/YOUR_APP_ID/Master \


  -H "Authorization: Bearer YOUR_API_KEY" \


  -H "Content-Type: application/json" \


  --data '{


  "fields": {


    "Name": "9999",


    "Table1 Link": k


      "recqKN5YMTsnqfFr8"


    ]


  }


}'

So the “Table1 Link” field is an array with the record ID from table 1 (it would be an array of length 1 as we only want one record in table 1 to link with a record in Master - but it still needs to be an array).

I tried this out with a simple Python script and it all works fine. There’s obviously some things you would need to think about, e.g load Table 1 into Master, then Table 2 etc. but you need to check if the Master ID already exists, in which case update otherwise create, and so on, but all doable.

Hope this helps!

JB