0.0.50 • Published 1 month ago

transmutation v0.0.50

Weekly downloads
54
License
MIT
Repository
github
Last release
1 month ago

Transmutation

npm npm Travis Ask Me Anything ! Conventional Changelog GitHub license Known Vulnerabilities

Enhanced promises for immutable data processing, inspired by RxJS and Lodash.

NOTICE: Active Development

Transmutation is still in its early stages of development. Feel free to watch this project as it grows and use the issues section to put in any questions or ideas you may have.

Nothing in this project promises to be stable until we reach v0.1.0.


Work Log

Transmutation is developed with an agile methodology that I'm currently developing called Dex. This methodology revolves around doing 25 minute work sessions (inspired by the Pomodoro Technique) and then logging what was done in that session with a work log entry.

You can find all of the work log entries here: ./worklog.md


Inception

Transmutation originally came out of my regular usage of RxJS for API development. I loved the utilities given among map, filter, mergeMap, and all of the handling of async processing. Being able to declaritively manage dozens of different types of calls to external resources proved to be extremely valuable in development. As I did more and more creation of APIs, I noticed a steady pattern emerging. Instead of just pushing one input into a part of the stream and coming out with a different output, I realized that I would have to refactor less if I just extended the input with the output at each portion of the stream. Mainly because I constantly found myself having to go back and introduce new dependencies at certain parts, and having to do a bunch of refactoring to make sure it got to certain parts of the stream.

From this I started to form a theory of data snowballing. More than just immutability and avoiding state changes in my code, if I never got rid of the ability to access any data I have collected, I never have to create a coupling of a later portion of the process back up to an earlier portion. Every bit of data is instead collected on the way down and can NEVER be removed until you are done with the request. At first it felt like it would create a lot of data management, but ultimately it proved to improve my code to the point where making future modifications were actually very easy. As long as a request has a "single responsibility" and it does a limited amount, it should never run into problems. In theory... as time goes on I'll probably discover some falacies in this plan.

Along the way I learned more things that would prove useful in this thought process and formulated Transmutation to capture these capabilities. After using it in several projects now, I'm fairly convinced this is a solid approach to doing logic. Check out the #Usage section below for examples.

Install

Pick one - Listed in order of recommendation

yarn add transmutation
npm i -S transmutation
npm install --save transmutation

Usage

The documentation below covers basic concepts in Transmutation including its promise like interface, basic extending, deep manipulation, conditionals, and external service interaction.

For more detailed readable usage documentation: ./docs/usage.md

For all tested use cases: ./src/node_modules/tests/*


Transmutation Promise

Transmutation works like a stream or chained promises. Every chained function call builds on the previous. The output of the one goes in as input for the next. The stream begins by passing data into the default export transmute function, and the stream ends by ending with .then(). Like promises, then will accept the output of the stream and anything can be done with it.

import transmute from 'transmutation';

transmute({ first: 'data' })
    .then(value => console.log(value));

// Prints
{ first: 'data' }

Basic Extending

Once data has been transmuted, it can now be transformed with various operators. The core operator is extend. Data in keys in a transmutation cannot be modified or overwritten. New keys may be added, new items may be pushed to arrays, and new children can be added, but once it has been added it can never be removed in a later part of the pipeline. The final then() will always be a full "snowball" of all data gathered through the stream.

import transmute from 'transmutation';

transmute({ first: 'data' })
    // Next line adds a new 'second' key to the base object
    .extend({ second: { data: 'new' } })
    // Next line adds a new child to the 'second' key, since it wouldn't remove any data
    .extend({ second: { child: 'a new child' } })
    // Next line will not modify the 'first' key since it is a string
    .extend({ first: { child: 'a child' } })
    .then(value => console.log(value));

// Prints
{ first: 'data', second: { child: 'a new child', data: 'new' } }

Deep Manipulation

Because data can never be removed in a transmutation, there is great value in working with nested objects. Eventually nested objects get to be very difficult to work with, so Transmutation takes the JSON path based approach used in lodash and other libraries to make working with them easier. Most operators in Transmutation support path based manipulation.

import transmute from 'transmutation';

transmute({ first: 'data' })
    .extend({ second: { data: 'new' } })
    // Next line does the exact same thing as the previous line
    .extend('second.data', 'new')
    .then(value => console.log(value));

// Prints
{ first: 'data', second: { data: 'new' } }

External Data

Part of the value out of RxJS was doing async processing with external services both in the form of adding new data to work with and committing side effects like saving to a database. Basic extending goes beyond just basic data, it will take in a promise as well, allowing you to easily add external data to your transmutation stream.

import transmute from 'transmutation';
import request from 'request-promise-native';

transmute({})
    .extend('firstUserName', request('https://jsonplaceholder.typicode.com/users')
        .then(JSON.parse)
        .then(users => users[0])
        .then(user => user.name)
    )
    .then(value => console.log(value));

// Prints
{ firstUserName: 'Leanne Graham' }

And of course, if there is something you need from the "snowball" most operators allow for functions to be passed where the argument given is the snowball at that current state. So if you start with the user ID, then you can use that.

import transmute from 'transmutation';
import request from 'request-promise-native';

transmute({ userId: 1 })
    .extend('userName', data => request(`https://jsonplaceholder.typicode.com/users/${data.userId}`)
        .then(JSON.parse)
        .then(user => user.name)
    )
    .then(value => console.log(value));

// Prints
{ userName: 'Leanne Graham', userId: 1 }

Side Effects

And of course we are able to send data outwards. The key with side effects is that typically we just want to send a portion of the data out to a system, but have that part of the pipeline return the same data it was given without any modification. In Transmutation we do that with the do() operator. It will execute the promise without modifying the base "snowball" of data and when it is done the stream will continue.

import transmute from 'transmutation';
import request from 'request-promise-native';

transmute({ name: 'Ray Benefield' })
    .do(user => request.put({
        uri: 'https://jsonplaceholder.typicode.com/users/1',
        json: true,
        body: user,
    }))
    .then(value => console.log(value));

// Prints
{ name: 'Ray Benefield' }

Team

Ray Benefield
Ray Benefield
Chief Architect

License

MIT © Ray Benefield

0.0.50

1 month ago

0.0.48

1 month ago

0.0.49

1 month ago

0.0.47

9 months ago

0.0.46

3 years ago

0.0.45

3 years ago

0.0.44

3 years ago

0.0.43

3 years ago

0.0.42

3 years ago

0.0.41

3 years ago

0.0.40

3 years ago

0.0.39

3 years ago

0.0.38

3 years ago

0.0.37

3 years ago

0.0.36

3 years ago

0.0.35

3 years ago

0.0.34

3 years ago

0.0.33

3 years ago

0.0.32

3 years ago

0.0.31

3 years ago

0.0.30

3 years ago

0.0.29

3 years ago

0.0.28

3 years ago

0.0.27

3 years ago

0.0.26

3 years ago

0.0.25

3 years ago

0.0.24

3 years ago

0.0.23

3 years ago

0.0.22

3 years ago

0.0.21

3 years ago

0.0.20

3 years ago

0.0.19

3 years ago

0.0.18

3 years ago

0.0.17

3 years ago

0.0.16

3 years ago

0.0.15

3 years ago

0.0.14

3 years ago

0.0.13

3 years ago

0.0.12

3 years ago

0.0.11

3 years ago

0.0.10

3 years ago

0.0.9

3 years ago

0.0.8

3 years ago

0.0.7

3 years ago

0.0.6

3 years ago

0.0.5

3 years ago

0.0.4

3 years ago

0.0.3

3 years ago

0.0.2

3 years ago

0.0.1

3 years ago