IBM Watson Analytics EXPLORE/PREDICT Service Public APIs? - ibm-watson

I have just started to learn the Watson analytics from IBM Watson Analytics I have uploaded a csv file and used EXPLORE, PREDICT features. Is there anyway can I use these features using REST API? in Java. Is there any REST API available for this.? I don't know how to start exactly.

The REST APIs for Watson Analytics use OAuth2 for Authentication once you have a token you can upload data and navigate the folders. There isn't a way to create Explorations and Predictions currently. Below is a code snip from the Java sample showing how you would get the user's profile information in Java.
String apiURL = "https://" + WATSON_ANALYTICS_API_URL + WATSON_ANALYTICS_API_BASE_PATH + "/accounts/v1/me";
HttpGet apiRequest = new HttpGet(apiURL);
apiRequest.addHeader("X-IBM-Client-Id","client_id");
apiRequest.addHeader("X-IBM-Client-Secret", "client_secret");
apiRequest.addHeader("Authorization", "Bearer " + ACCESS_TOKENS);
try {
HttpClient httpClient = new DefaultHttpClient();
HttpResponse apiResponse = httpClient.execute( apiRequest );
ByteArrayOutputStream stream = new ByteArrayOutputStream();
apiResponse.getEntity().writeTo( stream );
response.getWriter().write(stream.toString());
stream.close();
} catch (Exception e) {
System.err.println(e.getLocalizedMessage());
}

Related

Perform OAuth authentication from WPF application

I've got an existing Web API backend that uses OAuth to authenticate a vue.js frontend call's. This is an existing one and I can't modify it.
I need to perform the authentication from a new WPF Application I wrote.
I've composed the query using the HttpClient in the form
http://backend/api/signin?grant_type=password&username=user&password=1234hola
but I receive an error regarding the grant_type. Is there a tutorial I can follow? I didn't think it was that difficult to perform the authentication, but I think I'm missing something really stupid
Thanks in advance
You should add the credentials to the header as suggested here:
var client = new HttpClient() { BaseAddress = new Uri("http://url.com") };
var request = new HttpRequestMessage(HttpMethod.Post, "/path");
var byteArray = new UTF8Encoding().GetBytes("<clientid>:<clientsecret>");
client.DefaultRequestHeaders.Authorization = new AuthenticationHeaderValue("Basic", Convert.ToBase64String(byteArray));
...
var response = await client.SendAsync(request);

Azure B2C authentication without a ClientSecret

I've successfully implemented this tutorial and have a client + server working locally.
However, the front-end application that I'm building is an Angular app - this means that it isn't possible to store a client secret in it..
Relevant code:
ConfidentialClientApplication cca = new ConfidentialClientApplication(Startup.ClientId, Startup.Authority, Startup.RedirectUri, new ClientCredential(Startup.ClientSecret), userTokenCache, null);
var user = cca.Users.FirstOrDefault();
AuthenticationResult result = await cca.AcquireTokenSilentAsync(scope, user, Startup.Authority, false);
HttpClient client = new HttpClient();
HttpRequestMessage request = new HttpRequestMessage(HttpMethod.Get, apiEndpoint);
request.Headers.Authorization = new AuthenticationHeaderValue("Bearer", result.AccessToken);
How can I set my frontend up securely to work without having a client secret, based on the tutorial mentioned?
I'm currently using this angular library .
You can achieve this by using implicit grant flow
You can seamlessly integrate into any SPA application.
Follow https://learn.microsoft.com/en-us/azure/active-directory-b2c/active-directory-b2c-reference-spa, this article helps you.
There is an SPA sample already available in GitHub, you can try https://github.com/Azure-Samples/active-directory-b2c-javascript-hellojs-singlepageapp
you can use oidc-client library instead hello.js in above sample, both are very similar and easy to implement.

Authenticate a Google PubSub POST request using OAuth2

I need to form a POST to publish a Google PubSub message. I can't use the client libraries because they use gRPC which is incompatible with Google App Engine. I can form the critical POST request, but I'm not sure how to authenticate it using OAuth2.
This link shows what I'm doing, but it obscures the authentication part.
https://cloud.google.com/pubsub/docs/reference/rest/v1/projects.topics/publish
(If GAE standard environment would support gRPC this would not matter.)
JSONObject obj = new JSONObject();
JSONArray attr = new JSONArray();
obj.put("script_name","foo_script.py");
obj.put("script_args","arg1");
attr.put(obj);
JSONObject jsontop = new JSONObject();
jsontop.put("messages",attr);
URL url = new URL("https://pubsub.googleapis.com/v1/projects/{my-URL}/topics/topic_run_script:publish");
HttpURLConnection conn = (HttpURLConnection) url.openConnection();
This code comes back "401 : UNAUTHENTICATED". How do I authenticate it?
App Engine has an API to fetch an access token that you can use to when calling Google services. For documentation and an example, see https://cloud.google.com/appengine/docs/standard/java/appidentity/#asserting_identity_to_google_apis
You might also be able to use the pubsub client library on GAE Std if you switch to the Java 8 environment. This doc implies that it should work.

Call Parse.com REST services from Google App Engine

I have created a database in the Parse.com cloud. I now need to write a Servlet in my Google App Engine application to call the REST services on Parse. The REST services require
user authentication which is the Parse app id and the Javascript key.
...
URL url = new URL("https://api.parse.com/1/classes/OBJECT");
HttpURLConnection connection = (HttpURLConnection) url.openConnection();
connection.setRequestProperty("Content-Type", "application/json");
Base64 enc = new Base64();
String userpassword = "{PARSE_APP_ID}" + ":" + "javascript-key={PARSE_JS_KEY}";
String encodedAuthorization = enc.encodeBase64String(userpassword.getBytes());
connection.setRequestProperty("Authorization", "Basic " + encodedAuthorization);
connection.setRequestMethod("GET");
...
I use org.apache.commons.codec.binary.Base64 for encoding to get REST call authenticated.
The Parse.com REST API recommends using the following request format to make a HTTP call:
https://myAppID:javascript-key=myJavaScriptKey#api.parse.com/1/classes/GameScore/Ed1nuqPvcm
The problem is that I kept getting
{"error":"unauthorized"}
Is there anyone having the experience with working with calling a authenticated REST service? Thanks!
EDIT.
URL url = new URL("https://api.parse.com/1/classes/OBJECT");
HttpURLConnection connection = (HttpURLConnection) url.openConnection();
connection.setRequestMethod("GET");
connection.setRequestProperty("X-Parse-Application-Id", "{APP_ID}");
connection.setRequestProperty("X-Parse-REST-API-Key", "REST_ID");
connection.setRequestProperty("Content-Type", "application/json");
I still got the same error while the responseCode was "200".
Cheers,
Use the header version of authentication instead:
connection.setRequestProperty("X-Parse-Application-Id", "app id here");
connection.setRequestProperty("X-Parse-REST-API-Key", "rest key here");

Google Drive Access using AuthSubUtil.exchangeForSessionToken

On Google AppEngine i have been through authentication for Google Docs ... access using AuthSub authshub.
We managed to AuthSubUtil.exchangeForSessionToken(..).
QUESTION: Is it possible to follow up and get Access to Google Drive using this token?
... new Drive.Builder(httpTransport, jsonFactory, credential);
I'm not sure if AuthSub works with the Google Drive API, but if it does, this code snippet should resolve your problem:
new Drive.Builder(httpTransport, jsonFactory, new HttpRequestInitializer() {
#Override
public void initialize(HttpRequest request) {
HttpHeaders headers = request.getHeaders();
// Not sure if this test is necessary as headers might never be null.
if (headers == null) {
headers = new HttpHeaders();
request.setHeaders(headers);
}
headers.setAuthorization("AuthSub token=\"<TOKEN>\"");
}
}).build();
Important: please be aware that AuthSub is being deprecated and you should try to migrate your application to use OAuth 2.0 as soon as you can.

Resources