Postman Automated Collection Generation - api

has anyone done an enterprise integration of their public API with Postman? Checking Postman pages it seems like everything is straightforward, however, I have some concerns:
I don't see the way to automatically install pre-request scripts. Pre-request scripts allow an easy and straightforward way to call the endpoints without going through authentication step manually.
If you use sync with Github you'll need to give Postman full access. Not sure how people work around that.
You need to convert swagger to postman definition. Default Postman has limited nested levels of API schema, which means your API documentation will need additional processing step
So I don't know if it's worth integrating API release to the Postman with the internal API management system, or rather have a simple script on a virtual machine.

Related

How to automatically generate a web UI from a REST API

Is there any solution to automatically generate a web UI from a REST API?
I found Swagger codegen but it generates a client for the API, not a UI.
I need a basic UI, allowing directly from the browser to use the different endpoints and display the response prettily. Something like a basic Postman that would be directly integrated into my website.
I don't have constraint about how the generation is done. Can be done once at build time, or at runtime on server side or on client side.
I've heard good things about retool.com, it seems to do what you need.

Does Openshift Origin 1.1 REST API allow to create new applications based on templates?

We are developing a custom console to manage development environments. We have several application templates preloaded in openshift, and whenever a developer wants to create a new environment, we would need to tell openshift (via REST API) to create a new application based on one of those templates (oc new-app template).
I can't find anything in the REST API specification. Is there any alternative way to do this?
Thanks
There is no single API that today creates all of that in one go. The reason is that the create flow is intended to span multiple disjoint API servers (today, Kube and OpenShift resources can be created at once, and in the future, individual Kube extensions). We wanted to preserve the possibility that a client was authenticated to each individual API group. However, it makes it harder to write easy clients like this, so it is something we plan on adding.
Today the flow from the CLI and WebUI is:
Fetch the template
Invoke the POST /processedtemplates endpoint
For each "object" returned invoke the right create call.

What online REST API workbenches are available?

When creating a site/script to be on the client end of a RESTful API, what tools are available to create a "workbench" to explore the API, examining headers and responses while working through the design? Preferably one(s) that allow you to enter a custom endpoint, and create sample requests to see the responses. I recall seeing one nice workbench before, but its name has escaped me.
Re-found the one I remembered: The Apigee Console is a great interface for playing around with an existing API or building your own.
Mashery's I/O Docs is an open source workbench that you can deploy yourself on a Node.js server with Redis for storage.
If you have the wadl file of the ReST Services, you can load it in SOAP UI and use it.
EditedAnother much simpler tool Rest Client

Proper auth method against keystone ( openstack )

Should I directly query the API or rely on importing methods from the keystone client?
Thoughts?
It kind of depends on your requirements as to how you want to interact with keystone.
If you're querying the API directly then you're probably using curl in a bash script or from the command line. This can be particularly useful if you're working with bleeding edge keystone API code from trunk that doesn't even have methods in the keystone client yet.
If you're importing methods from the keystone client you're probably writing a python script or application. This is the better option if you're working with stable keystone code from a stable branch or package. The keystone client is just easier to work with rather than raw HTTP requests.
HTH
It's easier to use the keystone client methods. The python-keystoneclient package has documentation on how to do this in the doc directory. For example, see The client API.
I ended up using the python-keystoneclient bindings for initial authentication but for any query outside the scope of the keystoneclient API I simply referenced the auth_token and called to requests for direct API queries against the keystone ec2 url with the token in the header.
This worked well enough. Keystone needs work.

Auto-generate an API Explorer for WCF services

If you have ever used the Flickr API, you'll be familiar with their API Explorer. It is an awesome tool, that allows you to view the documentation for each API method, and the killer feature, being the execution of that API method (with a form to populate any request parameters). It even picks up when you are logged in, and completes the authentication part on your behalf. Gowalla has a similar API Explorer that is also really good.
Are there are tools for WCF that will auto-generate such an API Explorer, free or commercial?
Currently, we use Fiddler to build the JSON requests, but I would like to publish these service contracts, and allow potential developers to play around with them via a web based API explorer.
I am aware of the WCF Web HTTP Service Help Page, which I am using (and is awesome), but it is the API Explorer part that I am interested in.
You may want to look at I/O Docs - an open-sourced interactive documentation system for RESTful web APIs that any API owner can use to deploy for their own documentation. It runs on Node.js and uses Redis as a data store.
https://github.com/mashery/iodocs
Example: developer.klout.com/iodocs, developer.rottentomatoes.com/iodocs
It uses JSON schema based files to define API endpoints, method and parameters. Based on these JSON files, it generates a client interface that developers can use to learn and explore your API. API calls can be executed directly from the documentation interface, producing formatted responses.
It's Open-sourced, so you can be assured of regular updates and improvements. In fact this past weekend, Brandon West from SendGrid (who use I/O docs to power their documentation), created and open sourced the UI to create/edit the JSON schema files for I/O Docs. So you don't have to manually create the JSON files anymore.
https://github.com/brandonmwest/iodoctor
Not exactly what you were looking for, but....
WCF provides something called the WCF Test Client, for this purpose.
If you install Visual Studio, you get it. For example, for VS2008, installed in the usual place, you can find the WCF Test Client (WcfTestClient.exe) in the following location:
C:\Program Files\Microsoft Visual Studio 9.0\Common7\IDE\
Take a look at Apigee: http://apigee.com/