Zapier Tables and Python code: How can I read data?

Hi Troy,

Thanks for your reply.

What is that format, though? It seems to be line-based, with fields separated by newlines, and then a double-newline between each table row. The field lines are then in the format “name: value”, where value is not JSON. I think it’s maybe using Python’s repr to write these values, which is an odd choice. (repr’s output isn’t reliably deserializable the way JSON is.)

Thanks for the suggestion, but that sounds way more complicated than it should be. I wanted to avoid having to parse the string, but even that is a lot simpler than having to split every zap that needs to access a table into two zaps. I already have five zaps using this same code, and I really don’t want to turn that into ten. (I’m actually also trying to find a way to condense them down to a single zap.)

Sure:

I’m mapping “Data” from that action to the key “table_data” in Python:

And here’s the Python code to read it:

table_data = input_data.get('table_data')

table_data ends up being equal to:

"created_at: 2023-07-17T22:30:02.524000+00:00\ndata: {'f1': 'foo-bar', 'f2': 'hello, world!'}\nedited_at: 2023-07-17T22:30:16.747954+00:00\nerrors: {}\nid: 01H5JZK50W6G8HDJ280H68B98V\nschema_revision_id: 5\n\ncreated_at: 2023-07-17T18:12:48.231000+00:00\ndata: {'f1': 'foo-bar', 'f2': 'howdy, folks.'}\nedited_at: 2023-07-17T22:30:18.750945+00:00\nerrors: {}\nid: 01H5JGW4F7ZFJMC1TNYYZNMYDA\nschema_revision_id: 5"

Or if I print it (easier to read):

created_at: 2023-07-17T22:30:02.524000+00:00
data: {'f1': 'foo-bar', 'f2': 'hello, world!'}
edited_at: 2023-07-17T22:30:16.747954+00:00
errors: {}
id: 01H5JZK50W6G8HDJ280H68B98V
schema_revision_id: 5

created_at: 2023-07-17T18:12:48.231000+00:00
data: {'f1': 'foo-bar', 'f2': 'howdy, folks.'}
edited_at: 2023-07-17T22:30:18.750945+00:00
errors: {}
id: 01H5JGW4F7ZFJMC1TNYYZNMYDA
schema_revision_id: 5

Each of the fields from Zapier Tables will be it’s own variable (as an array) that you can map into the Code step.

 

How?

The options I see for mapping the data I need into Python from the Tables action are:

1. Data, which looks like this:

and despite saying “No data” in the mapping grid, contains the string I’ve been talking about in this post with all of the data from the table (including metadata), but in an odd format.

2. A column. The column I need is called “regex”, and looks like:

this value is not an array, but is instead a single string with all of the column’s values concatenated together with commas in between. In this example, I get “hello, world!,howdy, folks.”. Because it’s a single string, I can’t tell the difference between the row separator and a literal comma in the data.

Some more info, in case it matters: in my Tables action is a “Find Records (Output as Line Items) in Zapier Tables” action. I see there’s also a “Find or Create Record” option, but it looks like that only returns one record, while I need a list of records.

I’ve already written the code to parse it, it’s just that this is code that seems like it shouldn’t have to be written in the first place.

That is, why is Zapier cramming the table data into a string instead of passing structured data? And if it must pass use a string, why not use JSON or something else that’s designed for parsing, instead of an ad hoc format?

Frankly, this seems like a bug, or at the very least an omission, in the way Zapier Tables and Code by Zapier interoperate.