return API response from set_with_dataframe about gspread-dataframe

[robin900]

Do you need the api response captured at runtime, or do you just need to see the response json as a log message? Because the response json is logged to the logger named gspread_dataframe at level DEBUG; set the logging level to DEBUG for the logger name or all loggers, and you will see the response json in the log output.

If you want the api response returned at runtime, tell me more about why you'd want it, and we can consider whether to change set_with_dataframe to return the api response object...

[robin900]

@franciscothebrady An example debug log message from set_with_dataframe shows what the API response json would contain:

{
  'spreadsheetId': '1P3rdCCxfO760TJdE-cbi0k_yy9vmC-joapjuGw9vNjc', 
  'updatedRange': 'wksht_int_test!A1:T20', 
  'updatedRows': 20, 
  'updatedColumns': 20,
  'updatedCells': 400
}

If you just want to see this in your logs, enable logging at level DEBUG either throughout your program or for the logger logging.getLogger('gspread_dataframe') and all of its relevant handlers. (Configuring DEBUG-level logging in Python can be tricky; let me know if I can help.)

But if you really need the API response to be returned by set_with_dataframe so that you have it at runtime, please, tell me more.

[franciscothebrady]

I literally just finished adding a logfile and some setup for this. I appreciate the added context and info about the logging.getLogger(). I'm running now and will update with results.

The context here is that I'm running this weekly to update some data for dashboards, and I want confirmation that the dataframe being pushed to gdrive is complete. These aren't huge sheets (2 - <10k rows), so I shouldn't be running up against any limits, but last week we had some odd results and it looks like one sheet wasn't completely updated. It sounds like if the full API response is returned that will confirm at least updatedRows, which might be enough. Thanks for you help.

[robin900]

In my experience, the Sheets API is well-behaved around updates and error reporting. Either your cell updates to a worksheet succeed completely, or the API responds with a 4xx or 5xx status code and your Python API client raises an exception. If you are seeing "incomplete updates" to a worksheet without an exception raised by set_with_dataframe, I would be shocked if the cause lies somewhere in gspread_dataframe or the Sheets API.

Let me know how this resolves for you; I'll keep the issue open in the meantime.

[franciscothebrady]

I think the log file is showing everything I needed. 200's all the way through. I'm leaning toward this being a further upstream issue, but I think I do need to build in some kind of exception handling if the API response isn't 200. I'm not a python expert yet. I think closing this issue for now is fine.