Lovelacejs is a modern JavaScript Library to create objects that easily mutate data through relationships, filtering, and tranforming the shape of data.. Designed without dependencies and with modern JavaScript Syntax. Lovelacejs is a tool for robust data manipulation with ease.
Functionality is rooted in data being constructed as an instance of the Table class
import { Table } from 'lovelacejs`
const table = new Table({
id: 'abc',
label: 'Test Label',
rows: [
{ id: '2345676', type: 'row', lat: 54, long: 31 },
{ id: '2345676', type: 'lh', lat: 31, long: -71.34 }
]
})
Your data will be passed in as the rows property as an object. It is besed to pass it as an array of objects, yet if only a single object is passed it will be autocast as an array.
getProperties() // all props of the class
export() // array of rows
setRows([
{ id: '2345676', type: 'row', lat: 54, long: 31 },
{ id: '2345676', type: 'lh', lat: 31, long: -71.34 }
])
Nodules are classes used to transform Tables in various ways. There is a Nodule base class, but it has no purpose in the wild by itself. Create an instance of one of its child classes depending on the desired use case.
import { Nodule } from 'lovelacejs`
const nodule = new Nodule({
id: 'ABC',
label: 'Test Node',
tables: [
new Table({
id: 'XYZ',
label: 'Test Table',
rows: [{ id: 'abc', data: 'row' }]
})
]
})
getProperties() // all props of the class and the props of the tables
asTable() // return the Nodule as an instance of a new Table
setTables(new Table({
id: 'XYZ',
label: 'Test Table',
rows: [{ id: 'abc', data: 'row' }]
}))
For the use case of filtering out only the rows of data that fits the desired credentials. The two distinct properties of the FilterNodule are the filterParams and filterType.
import { Table } from 'lovelacejs`
import { FilterNodule } from 'lovelacejs`
const table = new Table({
id: 'XYZ',
label: 'Test Table',
rows: [
{ id: 'abc', data: 'row', contractor: 'AshBritt' },
{ id: 'qwe', data: 'lh', contractor: 'AshBritt' },
{ id: 'XYZ', data: 'row', contractor: 'AshBritt' },
{ id: 'XYZ', data: 'row', contractor: 'HeyDay' },
]
})
const filterNodule = new FilterNodule({
id: 'ABC',
label: 'Test Nodule',
tables: [table],
filterParams: {
data: 'row',
contractor: 'AshBritt'
},
filterType: 'EQUAL'
})
const output = filterNodule.export()
[
{ id: 'abc', data: 'row', contractor: 'AshBritt' },
{ id: 'XYZ', data: 'row', contractor: 'AshBritt' }
]
The filterParams is an object where the key is the name of the column that is desired to be evaluated and the value to that key is used in its evaluation.
A FilterNodule can take these different filterTypes (where the latter four only regard to numerical values):
'EQUAL', 'GREATER', 'GREATEREQUAL', 'LESSER, 'LESSEREQUAL'
addFilter({ contractor: 'AshBritt' })
filterNodule.setFilterType('EQUAL')
export() // get the filtered data from the Nodule
This class' usecase is akin to that of an SQL join. It takes in Table instances like any other Nodule. With it, there are some extra properties that dictate how the Nodule should relate them to one another.
import { Table } from 'lovelacejs`
import { JoinNodule } from 'lovelacejs`
const pickupTable = new Table({
id: 'abc',
label: 'receipts',
rows: [
{ id: '2345676', contractor: 'AshBritt', type: 'row', lat: 54, long: 31 },
{ id: '2345676', contractor: 'Jefferson',type: 'lh', lat: 31, long: -71.34 },
{ id: '2345676', contractor: 'AshBritt', type: 'lh', lat: 80, long: -41 },
]
})
const contractorTable = new Table({
id: 'XYZ',
label: 'contractors',
rows: [
{ id: '1WE3V6', name: 'AshBritt', employeeCount: 43, homeState: 'CA' },
{ id: 'FG4S67', name: 'Jefferson', employeeCount: 91, homeState: 'AL' }
]
})
const joinNodule = new JoinNodule({
id: 'QWE',
label: 'Receipts with Contractors',
tables: [pickupTable, contractorTable],
joinBy: {
baseTableLabel: 'receipts',
joinParams: [
{ foreignTable: 'contractors', primaryTableKey: 'contractor', matchingKey: 'name' }
]
}
})
const output = joinNodule.export()
[
{ id: '2345676', contractor: 'AshBritt', type: 'row', lat: 54, long: 31, 'contractors::id': '1WE3V6', 'contractors::name': 'AshBritt', 'contractors::employeeCount': 43, 'contractors::homeState': 'CA' },
{ id: '2345676', contractor: 'Jefferson',type: 'lh', lat: 31, long: -71.34, 'contractors::id': 'FG4S67', 'contractors::name': 'Jefferson', 'contractors::employeeCount': 91, 'contractors::homeState': 'AL' },
{ id: '2345676', contractor: 'AshBritt', type: 'lh', lat: 80, long: -41, 'contractors::id': '1WE3V6', 'contractors::name': 'AshBritt', 'contractors::employeeCount': 43, 'contractors::homeState': 'CA' }
]
The joinBy property is an object that tells the Nodule how the rows of the Table instances relate to one another.
The baseTableLabel finds an imported Table by its label property. This is the Table that will be "joined into".
The joinParams property is an array of objects that state which Tables should relate to which rows and in what way.
The foreignTable is a string that will find, by the label, the Table that is desired to join into the base Table. The primaryKey is the column of the base Table. The matchingKey is the column of the foreign Table. The matchingKey and the primaryKey are used to match with one another to create the join.
After the join the column names coming from the foreign Table will be altered to make sure none of the original data is overwritten. The naming convention will appear as seen in the example above and as the following:
'foreignTableName::foreignValue'
setJoinBy({
baseTableLabel: 'receipts',
joinParams: [
{ foreignTable: 'contractors', primaryTableKey: 'contractor', matchingKey: 'name' }
]
})
export() // get the joined data from the Nodule
The TransformNodule class, just as the others, extends from the Nodule base class. This usecase however is to mold the shape of data from Tables into an entirly new one; renaming columns and/or extracting select data.
import { Table } from 'lovelacejs`
import { TransformNodule } from 'lovelacejs`
const table = new Table({
id: 'XYZ',
label: 'Test Table',
rows: [
{ id: 'abc', data: 'row', contractor: 'AshBritt' },
{ id: 'qwe', data: 'lh', contractor: 'AshBritt' },
{ id: 'XYZ', data: 'row', contractor: 'AshBritt' },
{ id: 'XYZ', data: 'row', contractor: 'HeyDay' },
]
})
const transformNodule = new TransformNodule({
id: 'ABC',
label: 'Transform Test',
tables: [table],
structure: {
'id': 'identifier',
'data': 'ticketType',
'contractor': 'contractor'
}
})
const output = transformNodule.export()
[
{ identifier: 'abc', ticketType: 'row', contractor: 'AshBritt' },
{ identifier: 'qwe', ticketType: 'lh', contractor: 'AshBritt' },
{ identifier: 'XYZ', ticketType: 'row', contractor: 'AshBritt' },
{ identifier: 'XYZ', ticketType: 'row', contractor: 'HeyDay' },
]
The difference in properties with this Nodule is the structure object. The keys on this object refer to a column that currently exists in the Tables, the value of that key is what the new column name should be in the exported data.
The value does not need to be used to rename the column, meaning that the value can be the same as the curent key.
Only the columns specified in the structure object will be exported from the Nodule.
setStructure({
'id': 'identifier',
'data': 'ticketType',
'contractor': 'contractor'
})
export() // get the joined data from the Nodule
The GroupByNodule class can take in Tables and a single value of groupByValue to group rows by a table header. On export() this will return an object with keys representing an array of objects (or rows)
The asTable() method is overloaded and actually returns an array of new Table instances, operating differently than the Nodule base class asTable()
A set of declared constant variables has been provided for safer typing. Although importing them is not essential, the values they represent are the only options for certain options.
Filter types for the FilterNodule can be one of only these five string values.
{
EQUAL: 'EQUAL',
GREATER: 'GREATER',
GREATEREQUAL: 'GREATEREQUAL',
LESSER: 'LESSER',
LESSEREQUAL: 'LESSEREQUAL',
}
This library is in its early stages, and improvements are to come. The main priority being speed and readability when using the options in these classes. Extra features are not the priority, but will be worked on. Adding features is agains the philosophy of this project however. Lovelacejs is to be a powerful building block to making dynamic and programatic data mutation. It is to be a tool to help craft software, not a wall that gets in the way of work.
If you have any questions or issues please open one up or reach out to me at joshua@jshoemaker.dev