Use GraphQL to Return Each Workbook, Extract, Viewers*

I've found the querying (or the back-end data) a bit clunky, so I gave up trying to get everything at once, and just created my own data source from multiple queries, relating a data source list to a workbook list. It might not give you everything you want, but maybe the queries themselves will give you some help moving toward a query that works for you:

// Workbooks, just paste the query string into GraphQL if you want to see the output
query = """
query workbooks {
  workbooks {
    name
    id
    projectName
    owner {
      name
      username
      email
    }
    embeddedDatasources {
      id
    }    
  }
}
"""

In Python, to get it into a 'readable' dataframe format with one row for each data source ID:

// All of the imports I have, though probably not all are used in the code I've pasted:
import pandas as pd
import tableauserverclient as TSC
import json
from tqdm import trange
from IPython.display import display, clear_output
 
import requests
import re
from unidecode import unidecode
 
 
 
dfw = pd.json_normalize(ts._server.metadata.query(query)["data"]["workbooks"])
dfw = dfw.explode('embeddedDatasources').reset_index(drop=True)
dfw["embeddedDatasources"] = dfw["embeddedDatasources"].apply(lambda x: x["id"] if isinstance(x,dict) else None)
 
# Make a dictionary to put the workbook ID in the datasource, instead of the other way round
ds_to_wb = dict(dfw[["Datasource ID","Workbook ID"]].dropna().values)

A parameterised query to get the fields in each data source (you could just omit the fields if you just want data source names). I haven't put in the subsequent stuff around getting field comments etc. and calculation formulas:

query = """
query datasource ($ds: ID) {
  datasources  (filter: {id: $ds} ) {
        id
        name
        fields {
          id  
        }
  }
}
"""

Then run that query for each datasource ID:

ws_list = list(dfw["Datasource ID"].dropna().unique())
 
dfd_list = []
 
for x in trange(len(ws_list)):
    ds = ws_list[x]
    data = ts._server.metadata.query(query, variables={"ds": ds} )
    df = pd.json_normalize(data["data"]["datasources"])
    dfd_list.append(df) 
 
dfd = pd.concat(dfd_list)
dfd = dfd.explode('fields').reset_index(drop=True)
dict_expanded = pd.json_normalize(dfd['fields'])
dfd = pd.concat([dfd.drop(columns=['fields']), dict_expanded], axis=1)
 
dfd.columns = ["Datasource ID","Datasource Name","Field ID"]
 
dfd["Workbook ID"] = dfd["Datasource ID"].map(ds_to_wb)