Update to 1.8.5

.synk

@@ -0,0 +1,4 @@
+# .snyk file
+ignore:
+  certifi@MPL-2.0:
+    - "*"

examples_v2/abac_token_parameters.py

@@ -18,7 +18,7 @@
 ts: TSRestApiV2 = TSRestApiV2(server_url=server)
 
 # Simple function for translating {key: List} arguments into the full ABAC syntax for Full Access Token
-# parameters : {name: value ... }
+# parameters : {name: [value] ... }
 # filters: {{name}: [{values}... }
 def create_abac_section(parameters, filters, persist_all=False):
 

examples_v2/set_obj_id.py

@@ -22,7 +22,6 @@
     print(e)
     print(e.response.content)
     exit()
-
 def create_obj_id_update_request(guid: str, obj_id: str):
     update_req = {
         "headers_update":
@@ -39,6 +38,7 @@ def create_obj_id_update_request(guid: str, obj_id: str):
     }
     return update_req
 
+
 # { 'guid' : 'obj_id' }
 def create_multi_obj_id_update_request(guid_obj_id_map: Dict):
     update_req = {
@@ -98,16 +98,23 @@ def export_tml_with_obj_id(guid:Optional[str] = None,
         "include_obj_id": True
     }
 
-
     yaml_tml = ts.metadata_tml_export(metadata_ids=[guid], edoc_format='YAML',
                                       export_options=exp_opt)
 
+    # Get obj_id from the TML
+    lines = yaml_tml[0]['edoc'].splitlines()
+    if obj_id is None:
+        if lines[0].find('obj_id: ') != -1:
+            obj_id = lines[0].replace('obj_id: ', "")
+
+    obj_type = lines[1].replace(":", "")
+
     if save_to_disk is True:
         print(yaml_tml[0]['edoc'])
         print("-------")
 
         # Save the file with {obj_id}.{type}.{tml}
-        filename = "{}.table.tml".format(obj_id)
+        filename = "{}.{}.tml".format(obj_id, obj_type)
         with open(file=filename, mode='w') as f:
             f.write(yaml_tml[0]['edoc'])
 

examples_v2/spotter_apis.py

@@ -0,0 +1,198 @@
+# This is a simple example of using the AI endpoints with Python.
+# Thanks to Bill Back for the initial example
+import requests
+from requests.exceptions import HTTPError
+import tomllib
+
+from thoughtspot_rest_api_v1 import *
+
+#
+# Basic pattern for using Spotter / AI REST APIs:
+# 1. Create Conversation with Model (formerly Worksheet) GUID to the Create Conversation endpoint
+# 2. Provide Conversation ID and the same Model GUID to the Converse endpoint, along with a natural language question
+# 3. Converse returns TML Search Tokens, which then you have to send into the /searchdata endpoint to get the data itself
+#
+
+with open("config.toml", "rb") as config_file:
+    config = tomllib.load(config_file)
+
+# Assign config to local values for ease of use.
+
+# Login info
+TS_URL = config["config"]["TS_URL"]
+USERNAME = config["config"]["USERNAME"]
+PASSWORD = config["config"]["PASSWORD"]
+ORG_ID = config["config"]["ORG_ID"]
+
+# GUID for the model to use.
+MODEL_GUID = '{model_guid}'
+
+
+def create_client() -> TSRestApiV2:
+    """
+    Creates a new API client.
+    :return: A TSRestApiV2 client instance for making calls.
+    """
+    ts: TSRestApiV2 = TSRestApiV2(server_url=TS_URL)
+
+    try:
+        auth_token_response = ts.auth_token_full(username=USERNAME, password=PASSWORD,
+                                                 org_id=ORG_ID, validity_time_in_sec=36000)
+        ts.bearer_token = auth_token_response['token']
+
+    except HTTPError as error:
+        _bail_with_error(error)
+
+    print('Created a client')
+    return ts
+
+
+def do_single_call(ts: TSRestApiV2) -> None:
+    """
+    Tests single call to get an answer via an API.
+    :param ts: A TSRestApiV2 client instance for making calls that has already been authenticated.
+    :return: None
+    """
+    print('Testing single call questions')
+
+    resp = ts.ai_answer_create(metadata_identifier=MODEL_GUID,
+                               query="give me a list of all the things I sold and how many of each")
+    tokens = resp['tokens']
+
+    search_data_resp = call_search_data_api(ts=ts, model_guid=MODEL_GUID, search_tokens=tokens)
+    print_search_data(search_data_resp)
+
+
+def do_conversation(ts: TSRestApiV2) -> None:
+    """
+    Tests having a data conversation.
+    :param ts: A TSRestApiV2 client instance for making calls that has already been authenticated.
+    :return: None
+    """
+    print('Testing a full conversation')
+
+    try:
+        conv_create_resp = ts.ai_conversation_create(metadata_identifier=MODEL_GUID)
+        conversation_id = conv_create_resp['conversation_identifier']
+
+        msg_1 = "show me the top 20 selling items for the west region"
+        resp = ts.ai_conversation_converse(conversation_identifier=conversation_id,
+                                           metadata_identifier=MODEL_GUID,
+                                           message=msg_1)
+
+        # The response has a list with one item (??) that has the details.
+        search_data_resp = call_search_data_api(ts=ts, model_guid=MODEL_GUID, search_tokens=resp[0]['tokens'])
+        print_search_data(search_data_resp)
+
+        # follow-up questions
+        msg_2 = "break these out by store"
+        resp = ts.ai_conversation_converse(conversation_identifier=conversation_id,
+                                           metadata_identifier=MODEL_GUID,
+                                           message=msg_2)
+
+        search_data_resp = call_search_data_api(ts=ts, model_guid=MODEL_GUID, search_tokens=resp[0]['tokens'])
+        print_search_data(search_data_resp)
+
+    except HTTPError as error:
+        _bail_with_error(error)
+
+# Untested at this time, future feature
+'''
+def do_decomposed_query(ts: TSRestApiV2) -> None:
+    """
+    Tests decomposed queries (whatever those are).
+    :param ts: A TSRestApiV2 client instance for making calls that has already been authenticated.
+    :return: None
+    """
+    print('Testing a decomposed query')
+
+    # conversation ID is optional in this call.
+    # conversation_id = ts.ai_conversation_create(metadata_identifier=RETAIL_SALES_WS)['conversation_identifier']
+
+    # First, let's just use generically against a liveboard.
+    endpoint = "ai/analytical-questions"  # endpoint for decomposing
+    resp = ts.post_request(endpoint=endpoint, request={
+        "liveboardIds": [
+            "3f5d2d4b-87da-4f59-a144-85d444eada18"
+        ]
+    })
+
+    print(resp)
+'''
+
+def _bail_with_error(error: requests.exceptions.HTTPError) -> None:
+    """
+    Prints info about the error and then exits.
+    :param error:
+    :return:
+    """
+    print(error)
+    print(error.response.content)
+
+
+def call_search_data_api(ts: TSRestApiV2, model_guid: str, search_tokens: str) -> List:
+    """
+    Uses the search data API to get the data and then prints the results.
+    :param ts: The TSRestApiV2 client instance for making calls that has already been authenticated.
+    :param model_guid: A valid data source GUID.
+    :param search_tokens: The search tokens to use.
+    :return: None
+    """
+    print(f'Searching data {search_tokens}')
+    resp = ts.searchdata(
+        {"logical_table_identifier": MODEL_GUID, "query_string": search_tokens, "record_size": 50})
+
+    return resp
+
+
+def print_search_data(search_data) -> None:
+    """
+    Prints the data from a 'searchdata' call.  This assumes it works.
+    :param search_data: The response which has contents with the actual data.
+    :return: None
+    """
+    # Extract the table contents from the API response
+    contents = search_data['contents'][0]
+    column_names = contents['column_names']
+    data_rows = contents['data_rows']
+
+    # Compute the maximum width for each column for proper alignment
+    widths = []
+    for i, header in enumerate(column_names):
+        col_width = max(len(str(header)), max(len(str(row[i])) for row in data_rows))
+        widths.append(col_width)
+
+    # Create the header row and a separator row
+    header_row = " | ".join(str(header).ljust(widths[i]) for i, header in enumerate(column_names))
+    separator = "-+-".join("-" * widths[i] for i in range(len(widths)))
+    table_width = len(header_row)
+
+    # Separator for readability.
+    print("")
+    print("-" * table_width)
+    print("")
+
+    # Print the table to stdout
+    print(header_row)
+    print(separator)
+    for row in data_rows:
+        print(" | ".join(str(cell).ljust(widths[i]) for i, cell in enumerate(row)))
+
+    # Separator for readability.
+    print("")
+    print("-" * table_width)
+    print("")
+
+
+if __name__ == "__main__":
+    print('Testing AI')
+
+    tsapi = create_client()
+
+    do_single_call(tsapi)
+
+    do_conversation(tsapi)
+
+    # do_decomposed_query(tsapi) # new feature not yet in this version.
+
+    print('Testing complete')

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = thoughtspot_rest_api_v1
-version = 1.8.4
+version = 1.8.5
 description = Library implementing the ThoughtSpot V1 REST API
 long_description = file: README.md
 long_description_content_type = text/markdown

src/thoughtspot_rest_api_v1/_version.py

@@ -1 +1 @@
-__version__ = '1.8.4'
+__version__ = '1.8.5'

src/thoughtspot_rest_api_v1/tsrestapiv2.py

@@ -706,6 +706,27 @@ def metadata_headers_update(self, request: Dict):
         endpoint = 'metadata/headers/update'
         return self.post_request(endpoint=endpoint, request=request)
 
+    def metadata_update_obj_id(self, new_obj_id: str, guid: Optional[str], current_obj_id: Optional[str],
+                               request_override: Optional[Dict] = None):
+        endpoint = 'metadata/update-obj-id'
+        if request_override is not None:
+            request = request_override
+        else:
+            request = {
+                "metadata": [
+                    {"new_obj_id" : new_obj_id}
+                ]
+            }
+            if guid is not None:
+                request["metadata"][0]["metadata_identifier"] = guid
+            elif current_obj_id is not None:
+                request["metadata"][0]["current_obj_id"] = current_obj_id
+            else:
+                raise Exception("Must provide one of guid or current_obj_id")
+
+        return self.post_request(endpoint=endpoint, request=request)
+
+
 #
 # /reports/ endpoints
 #