aaaa3ba797
The main changes are: - Updating the pipeline to allow for doing a background refresh of the cache - Remove the use of the EventAwareCacheWrapper for the posts public cache ### Background refresh This is just an initial implementation, and tbh it doesn't sit right with me that the logic for this is in the pipeline - I think this should sit in the cache implementation itself, and then we call out to it with something like: `cache.get(key, fetchData)` and then the updates can happen internally. The `cache-manager` project actually has a method like this called `wrap` - but every time I've used it it hangs, and debugging was a pain, so I don't really trust it. ### EventAwareCacheWrapper This is such a small amount of logic, I don't think it's worth creating an entire wrapper for it, at least not a class based one. I would be happy to refactor this to use a `Proxy` too, so that we don't have to add methods to it each time we wanna change the underlying cache implementation. |
||
---|---|---|
.. | ||
lib | ||
test | ||
.eslintrc.js | ||
index.js | ||
package.json | ||
README.md |
API Framework
API framework used by Ghost
Usage
Stages
Each request goes through the following stages:
- input validation
- input serialisation
- permissions
- query
- output serialisation
The framework we are building pipes a request through these stages in respect of the API controller configuration.
Frame
Is a class, which holds all the information for request processing. We pass this instance by reference. Each function can modify the original instance. No need to return the class instance.
Structure
{
original: Object,
options: Object,
data: Object,
user: Object,
file: Object,
files: Array
}
Example
{
original: {
include: 'tags'
},
options: {
withRelated: ['tags']
},
data: {
posts: []
}
}
API Controller
A controller is no longer just a function, it's a set of configurations.
Structure
edit: function || object
edit: {
headers: object,
options: Array,
data: Array,
validation: object | function,
permissions: boolean | object | function,
query: function
}
Examples
edit: {
headers: {
cacheInvalidate: true
},
// Allowed url/query params
options: ['include']
// Url/query param validation configuration
validation: {
options: {
include: {
required: true,
values: ['tags']
}
}
},
permissions: true,
// Returns a model response!
query(frame) {
return models.Post.edit(frame.data, frame.options);
}
}
read: {
// Allowed url/query params, which will be remembered inside `frame.data`
// This is helpful for READ requests e.g. `model.findOne(frame.data, frame.options)`.
// Our model layer requires sending the where clauses as first parameter.
data: ['slug']
validation: {
data: {
slug: {
values: ['eins']
}
}
},
permissions: true,
query(frame) {
return models.Post.findOne(frame.data, frame.options);
}
}
edit: {
validation() {
// custom validation, skip framework
},
permissions: {
unsafeAttrs: ['author']
},
query(frame) {
return models.Post.edit(frame.data, frame.options);
}
}
Develop
This is a monorepo package.
Follow the instructions for the top-level repo.
git clone
this repo &cd
into it as usual- Run
yarn
to install top-level dependencies.
Test
yarn lint
run just eslintyarn test
run lint and tests