DialogFlow detectIntentText

Question

I am using DF with Python API and here is the code and I can't use the detect intent text.

If I use the second line I get the next error: google.api_core.exceptions.PermissionDenied: 403 IAM permission 'dialogflow.sessions.detectIntent' on 'projects/newagent/agent' denied.

If I use the first one: google.api_core.exceptions.InvalidArgument: 400 Resource name 'projects/newagent/agent/environments/draft/users//agent/sessions/5276b6d4-a0b6-4e91-84d3-16512d1f3299' does not match 'projects//agent/environments//users//sessions/'.

I have enabled billing on Google Cloud and the user has Owner privileges. What is going wrong?

def detect_intent_texts(project_id, session_id, texts, language_code):
session_client = dialogflow_v2.SessionsClient()

#----------------------------------------------------------Lines that I talk about in the question---------------------------------------------------------------------------------------------------
#session = session_client.session_path(project_id, session_id)
session = "projects/newagent/agent/environments/draft/users/<user id>/sessions/6344a857-9de5-406c-ba0f-c71b7b3ffdba"
#----------------------------------------------------------Lines that I talk about in the question---------------------------------------------------------------------------------------------------

for text in texts:
    text_input = dialogflow_v2.types.TextInput(text=text, language_code=language_code)
    query_input = dialogflow_v2.types.QueryInput(text=text_input)
    response = session_client.detect_intent(session=session, query_input=query_input)

detect_intent_texts("newagent/agent/environments/draft/users/<User Number>",str(uuid.uuid4()),"Que tal?","es-ES")

dsesto dsesto · Accepted Answer · 2018-05-08T15:21:48

The Session ID should have the format projects/<Project ID>/agent/sessions/<Session ID> (being <Project ID> the ID of the GCP project where your agent is located and <Session ID> the ID you use for your ongoing session), as can be seen in this documentation page.

In your code I see that you are calling the detect_intent_texts() function like:

project_id = "newagent/agent/environments/draft/users/<User Number>"
session_id = str(uuid.uuid4())
texts = "Que tal?"
language_code = "es-ES"

I see two main errors here:

The Project ID has the wrong format, it should be the ID of your GCP project, which usually has a format like my-first-project or similar, and slashes / are not supported, so you are using a wrong Project ID.
The text should be a Python list of strings, like ["hello"] and not just "hello".

Just as an example, the following minimal code provides the result below:

import dialogflow

def detect_intent_texts(project_id, session_id, texts, language_code):
    session_client = dialogflow.SessionsClient()

    session = session_client.session_path(project_id, session_id)
    print('Session path: {}\n'.format(session))

    for text in texts:
        text_input = dialogflow.types.TextInput(text=text, language_code=language_code)
        query_input = dialogflow.types.QueryInput(text=text_input)
        response = session_client.detect_intent(session=session, query_input=query_input)

        print('Fulfillment text: {}\n'.format(response.query_result.fulfillment_text))

detect_intent_texts("my-project","abcd",["hello"],"en-US")

Result:

user@my-project:~/dialogflow$ python detect_intent_minimal.py
Session path: projects/my-project/agent/sessions/abcd
Fulfillment text: Hi!

Therefore I suspect that changing the project_id to its correct value and the texts to a list should solve your issues.

EDIT: I have been able to reproduce the issue that you are seeing with a 403 PermissionDenied message by using a Service Account without the required permissions.

In order to run intents in Dialogflow, you need to use a Service Account with one of the following roles:

Dialogflow API Admin and Dialogflow API Client can query for intents, and therefore, one of those is required in order to make the type of requests you are trying to do with your script.

I see you said that your user has owner privileges over the project. However, the issue may be that you are using a wrong service account. In order to set up authentication correctly, follow the steps detailed in the docs. In summary, you will have to create a Service Account with the right permissions, download its JSON key, and use it as an environment variable by running the command export GOOGLE_APPLICATION_CREDENTIALS="/path/to/your/key.json" wherever you are running the script.

DialogFlow detectIntentText

2 Answers