OSDatabase

Definition

class OSDatabase {
	// Setup
	constructor(cache = new Set()) {}
	configure(main, secondary = null, lang = "en", completion=()=>{}) {}

        // Interaction
	select(contains = null, key="keywords", range=null) {}
	add(records, main, secondary = null, lang="en") {}
	setPlugin(select, add, keywords) {}
}

Usage

OSDatabase is an object that enables OrionSearch to interact with any databases. It also contains a "keywords cache" that is used for rewriting the query. The cache will be auto-configured with the configure method, but if you already ran this function, you can just restore the cache by initiating the class with it:

const db1 = new OSDatabase()
db1.configure("title", "content")

const db2 = new OSDatabase(db1.keywordsCache) // same as db1. Use this when using the frontend package so that your visitors don't have to recompute the keywords cache every single time.

⚠️ Note about database structure:
Every database should have a column named keywords that will be used by OrionSearch to filter records.

Set up bindings for your DB

In this section, we'll show you a detailed example with the SQLite 3 database.

const db = new OSDatabase() // Creates the object

/* Implementing methods */
db.setPlugin(
	(key, contains, range) => { // Select method
		const r = range == null ? [0, Number.MAX_SAFE_INTEGER] : range  // Creates a range
		let rows = []
		if (contains == null) { // If OrionSearch needs to query everything. Used for the configure method.
			const query = data.prepare(`SELECT rowid, * FROM main WHERE rowid BETWEEN ${r[0]} AND ${r[1]}`)
			rows = query.all()
		} else { // Otherwise
			const query = data.prepare(`SELECT rowid, * FROM main WHERE ${key} LIKE "%${contains}%" AND rowid BETWEEN ${r[0]} AND ${r[1]}`)
			rows = query.all()
		}
		let out = []
		rows.forEach(row => { // Setting the main key.
			const record = new OSRecord(row) // Creating a record based on the retrieved data.
			record.main("headlines") // "headlines" is now the main column for this record.
			out.push(record)
		})
		return out // Returning all the records
	},
	add => { // Add a record
		const query = data.prepare(`INSERT INTO main VALUES (${add.values})`)
		query.run()
	},
	(keywords, record) => { // Manage "keywords" column. It will insert an array of records in the desired row.
		const keyQuery = data.prepare(`UPDATE main SET keywords = "${[...keywords].join(' ')}" WHERE rowid = ${record.data.rowid};`)
		keyQuery.run()
	}
)