Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/gajus/unnest
Creates a cartesian product of properties identified using value pointers.
https://github.com/gajus/unnest
array cartesian declarative template
Last synced: 9 days ago
JSON representation
Creates a cartesian product of properties identified using value pointers.
- Host: GitHub
- URL: https://github.com/gajus/unnest
- Owner: gajus
- License: other
- Created: 2018-11-03T07:12:05.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2019-01-17T20:02:48.000Z (almost 6 years ago)
- Last Synced: 2024-10-19T00:18:21.816Z (about 1 month ago)
- Topics: array, cartesian, declarative, template
- Language: JavaScript
- Size: 29.3 KB
- Stars: 3
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Unnest
[](https://travis-ci.org/gajus/unnest)
[](https://coveralls.io/github/gajus/unnest)
[](https://www.npmjs.org/package/unnest)
[](https://github.com/gajus/canonical)
[](https://twitter.com/kuizinas)
Creates a cartesian product of all properties identified using value pointers.
## API
```js
/**
* @param {Object} tree A hierarchical data structure.
* @returns {Array} A cartesian product of all values identified using value-pointers in the input object.
*/
unnest(tree);
```
### Value-pointer
In the context of Unnest, value-pointer refers to an object property whose key begins with `@`, e.g.
```js
{
'@foo': {
name: 'bar'
}
}
```
`@foo` is a value-pointer.
Data pointers are used to identify all members that are used to create the cartesian product.
## Use case
Unnest solves the problem of translating hierarchical dataset into a collection of atomic data records. This is a common requirement when extracting information from a hierarchical document (e.g. HTML) with intent to store it.
### Example
To illustrate an example use case, consider an HTML document that describes event locations, dates and times:
```html
-
foo0
-
bar0
- baz0
- baz1
-
bar2
- baz2
- baz3
-
```
We want to extract location, date and time information into a collection of objects that each describe all attributes of the event, i.e. The desired result is a cartesian product of all 3 variables (location, date and time):
```json
[
{
"date": "bar0",
"location": "foo",
"time": "baz0"
},
{
"date": "bar0",
"location": "foo",
"time": "baz1"
},
{
"date": "bar1",
"location": "foo",
"time": "baz2"
},
{
"date": "bar1",
"location": "foo",
"time": "baz3"
}
]
```
We can extract the subject data from the document using [Surgeon](https://github.com/gajus/surgeon). Surgeon uses declarative instructions to extract information out of a HTML document, e.g.
```yaml
- sm .location
- '@location': so h1 | rdtc
children:
- sm .date
- '@date': so h2 | rdtc
children:
- '@time': sm .time | rdtc
```
The result is a hierarchical object describing the relevant variables contained in the HTML document:
```json
[
{
"@location": "foo0",
"children": [
{
"@date": "bar0",
"children": [
{
"@time": "baz0"
},
{
"@time": "baz1"
}
]
},
{
"@date": "bar1",
"children": [
{
"@time": "baz2"
},
{
"@time": "baz3"
}
]
}
]
}
]
```
To get a cartesian product of all the variables, we need to iterate the tree data structure:
```js
const locations = input;
const result = {};
for (const locationDatum of locations) {
for (const dateDatum of locationDatum.children) {
for (const timeDatum of dateDatum.children) {
result.push({
date: dateDatum['@date'],
location: locationDatum['@location']
time: timeDatum['@time']
});
}
}
}
```
Unnest replaces the last step:
```js
import unnest from 'unnest';
unnest(input);
// [
// {
// "@date": "bar0",
// "@location": "foo",
// "@time": "baz0"
// },
// {
// "@date": "bar0",
// "@location": "foo",
// "@time": "baz1"
// },
// {
// "@date": "bar1",
// "@location": "foo",
// "@time": "baz2"
// },
// {
// "@date": "bar1",
// "@location": "foo",
// "@time": "baz3"
// }
// ]
```