@dankolz/abstract-data-service
v1.0.3
Published
A fairly thin layer over a document storate mongodb collection(s), indexeddb stores, objects as a json file. This serves as an abstract class for concrete implementations.
Downloads
30
Readme
abstract-data-service
A fairly thin layer over a document storate mongodb collection(s), indexeddb stores, objects as a json file. This serves as an abstract class for concrete implementations.
A couple reasons to use it / a couple reasons I wrote it.
- Emits events for add, change, and delete events. Handy for caching or kicking off long running processes.
- Has hooks for processing records loaded which allows additional data to be added, data to be removed, encryption/decryption, or recreating the objects as a certain class before returning them.
- A little help in querying by id.
- A nice starting point to create complex services objects.
- Just enough of an abstraction so that I'll be able to use the same interface for something like mongodb, indexeddb, remote rest calls over http, or a file based store.
- By default adds a unique id so that it's more consistent and the objects more referenceable in data stores which don't have mongo's strong unique keys.
Install
npm i @dankolz/abstract-data-service
Usage
You should probably check out one of the extending classes like @dankolz/mongodb-data-service, but just as a feel for how the interface works...
Data persistence usage
The save
call returns a promise which resolves to an array of objects. The first is the
saved object (including any id attributes added). The second is the native result that
mongodb returns. Save works for both insertion and document update, deciding which to
use on the basis of the existance of the _id attribute.
let [r] = await dataService.save({msg: 'hello world'})
Loading a record by id is straightfoward. The method accepts a BSON ObjectId, a string which it will try to turn into an ObjectId, the unique identifier string it creates and assigns to the id attribute, or a normal mongo query. If an object is passed as a query it won't even attempt to do any of the id processing stuff.
let result = await dataService.fetchOne(r.id)
result = await dataService.fetchOne(r._id)
Getting multiple results is almost exactly like mongo. Pass an object as a query, get an array of matches back, or an empty array of there are no matches.
result = await dataService.fetch({name: 'Kolz'})
Loading multiple objects by id with the sort of help given with the fetchOne function is like:
result = await dataService.fetch(dataService.createIdQuery(['abc', '123']))
Saving multiple objects has the save behavior as calling save
in a loop. It returns an
array of promises, each of the type returned by save
.
let promises = dataService.saveMany([
{msg: 'hello'}
, {msg: 'world'}
])
await Promise.all(promises)
Events usuage
Add a listenter to the emitter:
events.on('object-change', (one, two) => {
console.log(`object change: ${JSON.stringify(one)} ${two}`)
})