Compare commits

..

4 Commits

Author SHA1 Message Date
Tom Pytleski
3e7beb0b9e Merge branch 'develop' into jason/review
* develop:
  Update send-http-requests.md
  Update overview.md
  Update run-test-terminal.md
2018-02-20 09:39:25 -06:00
Robert Wallach
38675fab3c Create authorization.md 2018-02-19 16:11:21 -06:00
Robert Wallach
dfd07af634 Create assertions.md 2018-02-19 15:11:01 -06:00
Tom Pytleski
c95c8b1011 - Update example, and comments 2018-02-13 14:51:06 -06:00
17 changed files with 130 additions and 391 deletions

View File

@@ -21,3 +21,13 @@ When you start the Stoplight desktop app, it will start an instance of Prism on
* The Stoplight desktop app can read/write specification files on your local file system. This is perfect for generating specification outside of Stoplight (like from code), want to use version control systems like Git, or want to use your favorite IDE to work on a spec.
* This feature is **NOT** available in the web app
the web app
<!--stackedit_data:
eyJoaXN0b3J5IjpbMTU3NDc5MjY0XX0=
-->

View File

@@ -1,32 +1 @@
# Publishing
## What
Publishing your documentation has never been easier. Stoplight has added:
- New **Integrations** for Segment, Intercom, and Google Analytics to allow for traffic monitoring and additional analytics
- New **Authorizations** to make sure your documentation is secure at all times. This includes basic user/password authentication and SSO provider integration, powered by Auth0, SAML, and more.
- New **Builds** section for tracking published Hubs, including when a Hub was published, who published it, and under what domain.
Take that Gutenberg!
## Who
- **Organization Owners**, **Account Owners**, and **Administrators** can publish Hubs
## How
1. From the Stoplight editor, click on **Publish** in the far left-hand toolbar
2. Input a **subdomain** under Stoplight's tech-docs.io domain
- Or input a **custom domain** (optional)
3. Once completed, click **Next ->**
4. Select the Hub you wish to publish under **Hub File**
5. Add Integrations to **Segment**, **Intercom**, and **Google Analytics** under Integrations (optional)
6. Add security via **Username/Passwords Login**, **Auth0**, or **SAML** (optional)
7. Once completed, click **Publish** in the top left
8. A confirmation window will ask you to confirm your selection, click **Okay**
9. Once confirmed, **Build Logs** will display your current progress
- The process usually takes 2-5 minutes
10. Once the process has completed, a **green success message** will display at the bottom of the screen, letting you know that the Hub was published succesfully
11. Once a Hub is published, it will appear under **Builds**
12. To unpublish a Hub, select **Unpublish** in the **Danger Zone** underneath **Builds**
- If you wish to delete all builds and release the domain you are currently using, select **Remove Domain**

View File

@@ -1,143 +1,38 @@
# Object Inheritance
# Object Inheritance
## What
## What
- A **model** contains common resuable information that can be referenced in your endpoint definitions or other models in your API design.
- When a model derives its properties from another model, the event is called **inheritance**.
- The model which contains the common set of properties and fields becomes a parent to other models, and it is called the **base type**.
- The model which inherits the common set of properties and fields is known as the **derived type**.
- If a base type inherits its properties from another model, the derived type automatically inherits the same properties indicating that inheritance is **transitive**.
- OpenAPI Specification v2 uses the **allOf** syntax to declare inheritance.
- **allOf** obtains a collection of object definitions validated independently but, collectively make up a single object.
* A **model** contains properties that can be reused and referenced by endpoint
definitions, shared objects, and other models. For more information on what
models are and how they can be used, please see the API model overview
[here](./api-models.md).
## Why
- Inheritance makes your API design more compact. It helps avoid duplication of common properties and fields.
* **Inheritance** is when a model derives its properties from another model.
## Best Practices
* When a model inherits properties from another model, the model being inherited from is
known as a **base type** (or parent). A model that is inheriting
properties from a base type is known as a **derived type** (or child).
<!-- theme: info -->
> Avoid using contradictory declarations such as declaring properties with the samer name but dissimilar data type in your base model and derived model.
* When a base type inherits properties from another model, any derived types
will also automatically inherit the properties as well. For example, if model
C is a derived type of model B, and model B is a derived type of model A, then
model C is also a derived type of model A. In mathematics, this is known as
the [transitive property](https://en.wikipedia.org/wiki/Transitive_relation).
### Example
* To specify that a model should inherit from a base type, use the **allOf**
option (under "Combination Types") when building the model. By specifying
allOf and referencing the base type, the model will automatically inherit all
of the parent model's properties. A model can also inherit from multiple base
types as needed.
## Why
* Inheritance makes your API design more compact. It helps avoid duplication of
common properties and fields, reducing the complexity of the specification and the chance of errors.
## Best Practices
<!-- theme: info -->
> Avoid using contradictory declarations such as declaring properties with the
> same name but dissimilar data type in your base model and derived model.
### Example
As an example, imagine you are creating an API that stores and categorizes
different types of vehicles. To begin working on the API, you will need a base
"car" model with a few attributes that are common across all vehicles. This
might look similar to:
```javascript
// the car base type
```
{
"type": "object",
"properties": {
// number of wheels
"wheels": {
"type": "integer"
},
// number of doors
"doors": {
"type": "integer"
},
// brand of car
"make": {
"type": "string"
},
// model of car
"model": {
"type": "string"
}
}
Vehicle:
type: object
properties:
brand:
type: string
Sedan:
allOf: # (This keyword combines the Vehicle model and the Sedan model)
$ref: '#/definitions/Vehicle'
type: object
properties:
isNew:
type: boolean
}
```
<!--FIXME Insert image of creating model from UI -->
Now that we have a base type model, we now need a derived type that extends the
base type. Since we're dealing with cars, let's create a model that defines a
SUV type (or a Sport Utility Vehicle):
```javascript
// the SUV model
{
"allOf": [
// a reference to the car base type
{
"$ref": "#/definitions/car"
},
// properties that are only applied to the SUV model
{
"type": "object",
"properties": {
// whether the vehicle can go "off road"
"off-road": {
"type": "boolean"
},
// the amount of ground clearance
"ground-clearance": {
"type": "integer"
}
}
}
]
}
```
<!--FIXME Insert image of creating derived model in UI -->
As shown above, by wrapping our SUV model inside of an `allOf` block, we are
able to include all of the properties that are included in the car base model
above.
When fully de-referenced (the car reference is replaced with the car
properties), the derived SUV model will have the following JSON properties:
```javascript
{
"type": "object",
"properties": {
// number of wheels
"wheels": {
"type": "integer"
},
// number of doors
"doors": {
"type": "integer"
},
// brand of car
"make": {
"type": "string"
},
// model of car
"model": {
"type": "string"
},
// whether the vehicle can go "off road"
"off-road": {
"type": "boolean"
},
// the amount of ground clearance
"ground-clearance": {
"type": "integer"
}
}
}
```

View File

@@ -0,0 +1,42 @@
# Polymorphic Objects
## What
* Resources in your API are polymorphic. They can be returned as XML or JSON and can have a flexible amount of fields. You can also have requests and responses in your API design that can be depicted by a number of alternative schemas.
* **Polymorphism** is the capacity to present the same interface for differing underlying forms.
* The **discriminator** keyword is used to designate the name of the property that decides which schema definition validates the structure of the model.
## Why
* Polymorphism permits combining and extending model definitions.
## Best Practices
<!-- theme: warning -->
> The discriminator property **must** be a mandatory or required field. When it is used, the value **must** be the name of the schema or any schema that inherits it.
### Example
```yaml
definitions:
Vehicle:
type: object,
discriminator: model
properties:
model:
type: string
color:
type: string
required:
-model
Sedan: # If Vehicle.model is Sedan then use Sedan model for validation.
allOf:
- $ref: '#/definitions/Vehicle'
- type: object
properties:
dateManufactured:
type: date
required:
- dateManufactured
```

View File

@@ -1,194 +1 @@
# Shared Parameters and Responses
While designing API's in Stoplight, it is common to have multiple endpoints
share a set of query parameters and API responses. To help reduce extra
work (and the chance of introducing errors), it is important to:
* Identify endpoints with common parameters
* Use _shared components_ to reference the same property multiple times instead
of rewriting the properties for each individual endpoint.
Shared components in Stoplight come in two forms:
* __Parameters__ - These are shared parameters that can be applied to requests
across multiple endpoints.
* __Responses__ - These are shared response objects that can be applied to
multiple endpoints.
## Parameters
Shared parameters provide a way to use request properties across multiple API
endpoints without having to duplicate effort.
![](../../assets/gifs/shared-params-responses-param.gif)
Shared parameters are supported in the following request property locations:
* __path__ - The request URL path
* __query__ - The request URL query string
* __header__ - The request HTTP Header field object
* __body__ - The request HTTP message body
* __form-data__ - The request HTTP message body in the `multipart/form-data` format
<!-- theme: info -->
> For more information the above properties, see the [OpenAPI v2 Specification](https://github.com/OAI/OpenAPI-Specification/blob/master/versions/2.0.md#parameter-object)
Similar to generic request parameters, restrictions on the parameter values can
also be applied based on type, expected default value, minimum/maximum length,
and regular expression (regex).
![](../../assets/images/shared-params-responses.png)
To use a shared parameter, navigate to an API endpoint's _Request_ section and
create a reference to the shared parameter using the "chain" button as shown in
the image above. Once the parameter has been referenced, any updates to the
shared parameter will automatically be propagated to every endpoint using that
parameter.
![](../../assets/gifs/shared-params-responses-param2.gif)
Like other references in Stoplight, shared parameters can also be shared across
files, projects, and other external sources.
### Shared Parameters Example
Let's say you are creating an API to serve thousands of cooking recipes. When dealing with large volumes of
data, you typically want to avoid sending _all_ data in a request. To help avoid
sending more data than is necessary, most applications implement a "paging"
feature that allows clients to retrieve a small portion of results (i.e. a single
page).
There are multiple ways to approach a paging feature. For this example, we
want to add two query string parameters to every request:
* `limit` - The number of results to return when viewing a page. For example,
setting `limit` to `20` means that, at most, 20 results will be returned in the
request.
* `offset` - The number of results to skip before returning results. For
example, setting an `offset` of `20` means that the API will discard the first
20 results.
By using the two parameters above, a client can efficiently "page" through
results, only returning items that are within the requested bounds. To demonstrate, let's use the parameters to display the first page of our recipe
results:
```
GET /recipes?limit=20&offset=0
```
Since the `offset` is set to `0`, the API will not discard any results. Paired
with a `limit` of `20`, we will only see the first 20 results (1 through 20).
To view the second page of recipes, we would use:
```
GET /recipes?limit=20&offset=20
```
By setting an `offset` of `20`, the API will discard the first 20 results. Paired
again with a `limit` of `20`, we will see the second page of results (21 through
40).
### How
Now that we know how we want the components to behave, let's create them in
Stoplight. To get started, create a new shared parameter for an OpenAPI file
under the "Shared" section of the menu.
![](../../assets/images/shared-params-responses2.png)
As shown in the image above, set the properties for each parameter based on our
requirements:
1. Be sure to set the display names for both components properly. In our
example, we only have two parameters, however, if there are multiple shared
parameters with similar names, you may want to set a more descriptive name
(i.e. 'recipe-limit' instead of 'limit').
2. Since both components are query string parameters, set the property location
for each as 'query'.
3. Set the name of the parameter, which is how it will be set in HTTP requests.
4. Optional type restrictions can be applied to the values. In our case, since
both parameters are integer values, we can use the 'integer' format
restriction.
5. Setting a default value can be useful if you don't need the client to specify
each parameter for every request. For our example, it makes sense to set
defaults that will return the first page (limit of 20, offset of 0).
![](../../assets/images/shared-params-responses3.png)
Once the shared parameters are created, reference them in any API endpoint under the
__Query Parameters__ block of the request section in the editor.
## Shared Responses
Shared responses provide a practical way to re-use response objects across multiple API
endpoints without having to duplicate effort. Similar to the shared components
discussed above, shared responses allow you to reference a single response
multiple times without having to recreate each response manually. The added
benefit of this approach is that updates to the shared response object are
automatically propagated to any endpoint using that object, no extra changes
required.
![](../../assets/gifs/shared-params-responses-response.gif)
Shared responses allow you to configure the following properties:
* Headers - Customize the HTTP Headers returned in the response
* Response body - Customize the HTTP message body contents using the Stoplight
modeling tool (or reference a pre-existing model)
<!-- theme: info -->
> For more information on the above properties, see the OpenAPI v2 Specification
[here](https://github.com/OAI/OpenAPI-Specification/blob/master/versions/2.0.md#responseObject)
![](../../assets/gifs/shared-params-responses-response2.gif)
To use a shared response, navigate to an API endpoint's __Response__ section and
create a reference to the shared response by choosing the _Type_ of the response
as "Reference". Once the Response type is set to "Reference", you can then
choose the shared response to use for that endpoint. Shared responses can also
be shared across files, projects, and other external sources.
### Shared Responses Example
Let's say you have a set of
API endpoints that should return the same error response when a failure occurs.
Every time an error is returned from the API, you want to make sure the
following properties are returned:
* `error_message` - A descriptive error message about the failure in string format.
* `error_code` - A code representing the category of the failure in integer format.
* `tracking_id` - A tracking ID that can be used by the caller to follow-up with
an administrator for more information (ie, debug an issue with customer
support).
Now that we know what should be returned, let's create a shared response in
Stoplight. To get started, create a new shared response for an OpenAPI file
under the "Shared" section of the menu.
![](../../assets/images/shared-params-responses4.png)
As shown in the image above, set the properties for each portion of the response
based on our requirements:
1. Set the name of the shared response. In our example, we only have one error
type, however, if there are multiple error responses with similar names, you
may want to set a more descriptive name (ie, 'api-tracking-error' instead of
'error').
2. A short description of the error response, such as why it is needed and how
it is used.
3. The contents of the shared response object based on the three required
properties above.
![](../../assets/images/shared-params-responses5.png)
Once the shared response is created, it can be referenced in any API endpoint by
using a _Reference_ type under a response. A shared response can also be used
multiple times under the same endpoint.
***
**Related**
* [OpenAPI v2 Parameter Objects Reference](https://github.com/OAI/OpenAPI-Specification/blob/master/versions/2.0.md#parameter-object)
* [OpenAPI v2 Response Objects Reference](https://github.com/OAI/OpenAPI-Specification/blob/master/versions/2.0.md#responseObject)

View File

@@ -0,0 +1,30 @@
# Assertions
## What is an Assertion?
- An API test consists of a series of steps (these are sometimes HTTP requests) that can be executed collectively or individually.
- An **assertion** is a specification that indicates the expected outcome (response) to a request executed in a test.
- A test is unsuccesful if an assertion fails i.e. the actual outcome is not equal to the expected outcome
- You can create assertions for status codes, response time, reponse content, header values, etc.
- When you execute an assertion, you can determine the type of operation you want to perform with your expected outcomes.
### Comparison Logic Available in Scenarios
- equals
- greater than
- greater than or equals
- less than
- less than or equals
- no equal
- exists
- length equals
- contains
- validate pass
- validate fail
## Why
- Assertions are checked any time a test is executed.
- Assertions are used to detemine the state of a test (pass or fail).
- Assertions are ideal for discovering if an API satisfies stipulated objectives.
## Assertions in Scenarios
- Scenarios in Stoplight are grouped into collections. To create an assertion for a step in Scenarios, you need to create a collection and add your Scenarios to it.

View File

@@ -0,0 +1,20 @@
# Authorization
## What is Authorization?
- **Authentication** is the process of verifying if a user or API consumer can have access to an API.
- **Authorization** defines the resources an authenticated (properly identified) user can access. For example, a user might be authorized to use an endpoint for retrieving results but denied access to the endpoint for updating a data store.
- Authentication strategies can be implemented using basic HTTP authentication or OAuth methods. Authorization can be implemented using roles and claims.
## Authentication Schemes
- **Basic Authentication** is easy to implement and utilizes HTTP headers to validate API consumers. While the credentials might be encoded, they are not encrypted. It is advisable to use this method over HTTPS/SSL.
- **OAuth 1.0** has its foundation in cryptography. Digital signatures are used to authenticate and ensure the data originates from an expected source. It can be used with or without SSL.
- While OAuth 1.0 works primarily with web clients, **OAuth 2.0** works with web and non-web clients. OAuth 2.0 is easy to implement and focuses on bearer tokens. It wokrs with HTTPS/SSL for its security requirement.
- **AWS Signature** is a secutiry protocol that defines authentication information added to AWS requests. It consists of an access key ID and a security access key. Users who generate manual HTTP requests to AWS are required to sign the requests using AWS Signature. [Click here to learn more about AWS Signature](https://docs.aws.amazon.com/AmazonS3/latest/API/sig-v4-authenticating-requests.html)
## Best Practices for Authentication and Authorization when Testing APIs
- Use the same authentication and authorization your users would use during testing. Doing so will help you effectively identify and resolve issues users might face in a live scenario.
- Avoid creating too many test accounts with administrative access to all endpoints and resources during your test phase. It can be exploited, putting your resources and data at risk when the API is avaialble to consumers.
- Use ecryption to store valid IDs and credentials. Ensure test procedures containing valid user IDs or tokens are not displayed as unmasked text in test results or test logs.
- Test your authentication and authorization procedures rigorously by attempting to access secured resources with invalid credentials or session tokens or by atttempting to access a resource denied to an authenticated user.

Binary file not shown.

Before

Width:  |  Height:  |  Size: 1.6 MiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 2.5 MiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 2.3 MiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 1.0 MiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 41 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 52 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 50 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 75 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 52 KiB

34
size.sh
View File

@@ -1,34 +0,0 @@
#!/bin/bash
#set -x
# Shows you the largest objects in your repo's pack file.
# Written for osx.
#
# @see https://stubbisms.wordpress.com/2009/07/10/git-script-to-show-largest-pack-objects-and-trim-your-waist-line/
# @author Antony Stubbs
# set the internal field spereator to line break, so that we can iterate easily over the verify-pack output
IFS=$'\n';
# list all objects including their size, sort by size, take top 10
objects=`git verify-pack -v .git/objects/pack/pack-*.idx | grep -v chain | sort -k3nr | head`
echo "All sizes are in kB's. The pack column is the size of the object, compressed, inside the pack file."
output="size,pack,SHA,location"
allObjects=`git rev-list --all --objects`
for y in $objects
do
# extract the size in bytes
size=$((`echo $y | cut -f 5 -d ' '`/1024))
# extract the compressed size in bytes
compressedSize=$((`echo $y | cut -f 6 -d ' '`/1024))
# extract the SHA
sha=`echo $y | cut -f 1 -d ' '`
# find the objects location in the repository tree
other=`echo "${allObjects}" | grep $sha`
#lineBreak=`echo -e "\n"`
output="${output}\n${size},${compressedSize},${other}"
done
echo -e $output | column -t -s ', '