sineQL
sineQL is a web API query language that mimics graphQL, designed solely for fun.
sineQL consists of two languages - the schema language, and the query language. sineQL assumes that the records are related in a non-looping tree-structure, defined by the schema language. Also, each non-scalar type queried is returned as an array.
The handler's definition is left up to the user.
Feature List
- Easy to use schema language
- Easy to use query language
- Simple to set-up a server
- Each top-level keyword (and queries) is optional
Example Server
A simple express server using sineQL.
//express for testing
const express = require('express');
const app = express();
//uses text input
app.use(express.text());
//test the library
const sineQL = require('sineql');
const schema = require('./schema.js');
const queryHandler = require('./query-handler.js');
//omit 'queryHandler', 'createHandler', 'updateHandler' or 'deleteHandler' to disable those methods
const sine = sineQL(schema, { queryHandler }, { debug: true });
//open the endpoint
app.post('/sineql', async (req, res) => {
const [code, result] = await sine(req.body);
res.status(code).send(result);
});
//startup
const port = process.env.WEB_PORT || 4000;
app.listen(port, err => {
console.log(`listening to *:${port}`);
});
const schema = `
scalar Date
type Book {
String title
Date published
}
type Author {
String name
Book books
}
`;
module.exports = schema;
//there's a different handler object for query, create, update and delete
const queryHandler = {
Author: (query, graph) => {
//TODO: implement this
},
Book: (query, graph) => {
//TODO: implement this
},
};
module.exports = queryHandler;
Create a matching client-side function pointing to the server.
//create the wave function, wrapping a fetch to the server
const wave = body => fetch('http://example.com/sineql', {
method: 'POST',
headers: {
'Content-Type': 'text/plain'
},
body: body
});
//get a list of content
wave('Author { name books { title } }')
.then(blob => blob.text())
.then(text => console.log(text))
.catch(e => console.error(e))
;
The Schema Language
The schema language is a layout of how queries should be made, as well as what can be made with them. There are two built-in keywords for the schema language:
- type
- scalar
type is used for defining new compound types. scalar is for defining new scalar types, such as Date.
The built-in types for the schema language are:
- String
- Integer
- Float
- Boolean
These can be combined into compound types as so:
scalar Date
type Book {
String title
Date published
}
type Author {
String name
Book books
}
The Query Language
The query langauge can be used to request data from a server, either in whole or in part by listing its type and its needed fields:
Author {
name
books {
title
published
}
}
The fields can be altered as well, using the query language's built-in keywords:
- create
- update
- delete
- match
- set
- typeName
create, update and delete are still to be defined properly, but they'll probably work as follows.
Create
When using create, match will find an existing record and associate that with the created values (multiple matches is an error):
Author {
match name "Kenneth Grahame"
create books {
create title "The Wind in the Willows"
}
}
You can create multiple records at once by surrounding them with []:
create Book [
{
set title "The Philosepher's Kidney Stone"
}
{
set title "The Chamber Pot of Secrets"
}
{
set title "The Prisoner of Aunt Kazban"
}
{
set title "The Goblet of the Fire Cocktail"
}
{
set title "The Order for Kleenex"
}
{
set title "The Half-Priced Pharmacy"
}
{
set title "Yeah, I Got Nothing"
}
]
Update
When using update, match will find all existing records and update those using the set keyword:
update Book {
match title "The Wind in the Willows"
set published "15 June 1908"
}
update Book {
match title "The Wind in the Willows"
set title "The Fart in the Fronds"
}
Delete
When using delete, only match is valid, and will delete all matching records:
delete Book {
match title "The Fart in the Fronds"
}
You can use as many instances of match and set as you like, as long as the result is valid.