Query structure

Every JQL query says what result you want and which JSON file contains the data.

Every query needs select and from:

select {fields} from 'file.json'

For example:

jql "select {name, email} from 'customers.json'"

Main clauses

Clause	What it does
select	Describes the fields and shape of the result.
from	Chooses the JSON file.
join	Combines related records from another file.
where	Keeps records that match a condition.
group by	Creates groups for summaries.
having	Keeps groups that match a condition.
sort	Orders the final array by output fields.

The remaining clauses are optional.

A complete query

select {region, revenue: sum(amount)}
from 'orders.json'
where status = 'paid'
group by region
having revenue >= 100
sort !revenue

JQL reads this query in a predictable order:

1. Read orders.json.
2. Keep paid orders.
3. Group them by region.
4. Calculate revenue for every region.
5. Keep regions whose revenue is at least 100.
6. Sort the resulting regions by revenue.

Object and summary results

Braces create one object for every matching record:

select {id, name} from 'customers.json'

A named function list without braces creates one summary object:

select customers: count(id) from 'customers.json'

Keywords

Keywords and function names are case-insensitive, but JQL documentation uses lowercase as the conventional style. Field names and string values keep their original case. Unicode field names are supported.

JQL also accepts where before from:

select {id} where active = true from 'customers.json'

The usual from ... where ... order is easier to read and is used throughout this documentation.