microsoftgraph / microsoft-graph-explorer-v4 Goto Github PK

I think we should remove admin permissions in UI if the user does have those permissions.

I am using browser private mode to log in my account, you can see it has admin permissions by default, although I actually don't have admin permissions.

And even after I removing these permissions, when I log in again, they are back.

2121657

Store requests made in IndexedDb.

Based on this issue
AB#7513

Please update the Chinese version https://developer.microsoft.com/zh-cn/graph/graph-explorer-china.

to latest Graph Explorer.

(I heard it is because the AzureAD was blocked, but please work on it to get this work)

2121651
AB#7539

The Key and Value should each occupy a width of 45% with respect to the container and the delete button 10%.

Include

• Fabric React
• Monaco
• Graph SDK

Others?

Investigate adding appInsights instrumentation to report errors observed by users.

Customers are asking for this.
https://developer.microsoft.com/en-us/graph/docs/api-reference/v1.0/resources/schemaextension
AB#7534

Prettier Recommended Configuration

Use the issues in this repository to open issues about the Graph Explorer application. Questions about the Microsoft Graph should be directed to StackOverflow. Use the [microsoftgraph] tag on StackOverflow.

Expected behavior

Let's say I have a user "[email protected]" and I have an administrator "[email protected]" who has User.ReadBasicAll permissions to the domain.

If I visit the https://graph.microsoft.com/v1.0/users/[email protected] endpoint while logged in as '[email protected]', I expect:

{
    "@odata.context": "https://graph.microsoft.com/v1.0/$metadata#users/$entity",
    "businessPhones": [],
    "displayName": "Joe Smith",
    "givenName": "Joe",
    "jobTitle": "Senior Analyst, Analysis Services",
    "mail": "[email protected]",
    "mobilePhone": "123-456-7890",
    "officeLocation": "The Moon",
    "preferredLanguage": "en-US",
    "surname": "Smith",
    "userPrincipalName": "[email protected]",
    "id": "1532477f-026d-4742-ab6e-6c1796521530"
}

I do not believe we've set any special permissions on those fields that are returning null, nor would I even know how to do such a thing.

Actual behavior

When I run the query, I get back:

{
    "@odata.context": "https://graph.microsoft.com/v1.0/$metadata#users/$entity",
    "businessPhones": [],
    "displayName": "Joe Smith",
    "givenName": "Joe",
    "jobTitle": null,
    "mail": "[email protected]",
    "mobilePhone": null,
    "officeLocation": null,
    "preferredLanguage": null,
    "surname": "Smith",
    "userPrincipalName": "[email protected]",
    "id": "1532477f-026d-4742-ab6e-6c1796521530"
}

Steps to reproduce the behavior

1. Log in as a user with User.ReadBasicAll permissions on the domain
2. Look up another user

HTTP traffic capture

n/a
AB#7533

We need to be able to parse method, URL, headers and body out of example requests like this:

https://docs.microsoft.com/en-us/graph/api/user-list-calendars?view=graph-rest-1.0#http-request
https://docs.microsoft.com/en-us/graph/api/user-post-calendars?view=graph-rest-1.0#example

Where there are multiple requests in a single example, we need to determine which request to use. If an example has no parameter, it should be preferred.

Firing the my photo request does not show the related profile image response.

##Expected:
Firing a 'my photo' request results in an image being shown in the response component

##Actual results:

• the response contains an empty object
• The photo is not displayed

When a user changes the selected permissions those changes don't seem to be persisted. I believe this is due to the re-authentication not forcing the user to also re-consent to the new set of scopes.

Today the code is passing prompt=select_account in the query string. From the documentation, it doesn't look like select_account is supported in the v2 Endpoint. Possibly just a carried over option that predate switching to v2?

Regardless, manually switching this to prompt=consent successfully results in the new permissions being persisted.

Would it make sense to switch to prompt=consent (at least within the context scopes-dialog.component.ts)?
AB#7540

This bug happens when the user has a profile picture.

The profile picture overlays the user's name and email.

Second part of the conversation I just had with @yinaa it'd be awesome to have above or under the response json preview a list of links of the entities in the response.
When clicking on the link, it would pre-fill the url box with the proper url.
This would save me time during the demonstrations when I copy/past id's to build requests. Avoid typos as well.

AB#7535

We need to come up with a notification process that shows the status codes / status messages for the sent requests

Putting a part of the conversation I just had with @yinaa a couple of minutes ago.
It'd be awesome to save time and to avoid typos during demos to have a button "pre-fill payload" when building any POST/PUT request.
Clicking that button would look at the url template I currently have and get the standard JSON payload to put it in the payload box.

AB#7537

The Query button is clicked and the request has been created and awaiting the response.
We should reuse the submit button component and pass in a boolean value when running the query

Some queries can become exceptionally long making it quite challenging to read or edit the query.

The query editor should optionally support multiple lines, similar in functionality to Excel's formula editor.

2121659
AB#7538

when you run npm i you get this error

cc @jobala

the suggested fix would be running npm audit fix

The ClientId is hard coded. It should be accessed through env variables.

When the Customer authenticates, the Profile Component populates the QueryResponse Component.

The expected behaviour is that the QueryResponse is not populated on authentication.

The QueryResponse Component should only be populated on running a query.

There's a lot of space between the value and the button making the UI feel unbalanced.