Files
umami/src/queries/sql/reports/getRevenueChart.ts
T

147 lines
4.1 KiB
TypeScript

import clickhouse from '@/lib/clickhouse';
import { EVENT_TYPE } from '@/lib/constants';
import { CLICKHOUSE, PRISMA, runQuery } from '@/lib/db';
import prisma from '@/lib/prisma';
import type { QueryFilters } from '@/lib/types';
export interface RevenuParameters {
startDate: Date;
endDate: Date;
unit: string;
timezone: string;
currency: string;
compare?: string;
}
export interface RevenueChartResult {
chart: {
x: string;
t: string;
y: number;
count: number;
}[];
}
export async function getRevenueChart(
...args: [websiteId: string, parameters: RevenuParameters, filters: QueryFilters]
): Promise<RevenueChartResult> {
return runQuery({
[PRISMA]: () => relationalQuery(...args),
[CLICKHOUSE]: () => clickhouseQuery(...args),
});
}
async function relationalQuery(
websiteId: string,
parameters: RevenuParameters,
filters: QueryFilters,
) {
const { startDate, endDate, unit = 'day', timezone = 'utc', currency } = parameters;
const { getDateSQL, rawQuery, parseFilters } = prisma;
const { queryParams, filterQuery, cohortQuery, joinSessionQuery, dateQuery } = parseFilters({
...filters,
websiteId,
startDate,
endDate,
currency,
});
const chart = await rawQuery(
`
with
filtered_sessions as (
select distinct website_event.website_id, website_event.session_id
from website_event
${cohortQuery}
${joinSessionQuery}
where website_event.website_id = {{websiteId::uuid}}
and website_event.event_type != ${EVENT_TYPE.performance}
${dateQuery}
${filterQuery}
),
filtered_revenue as (
select
revenue.event_id,
revenue.event_name,
revenue.created_at,
revenue.revenue
from revenue
join filtered_sessions
on filtered_sessions.website_id = revenue.website_id
and filtered_sessions.session_id = revenue.session_id
where revenue.website_id = {{websiteId::uuid}}
and revenue.created_at between {{startDate}} and {{endDate}}
and upper(revenue.currency) = {{currency}}
)
select
filtered_revenue.event_name x,
${getDateSQL('filtered_revenue.created_at', unit, timezone)} t,
sum(filtered_revenue.revenue) y,
count(filtered_revenue.event_id) count
from filtered_revenue
group by x, t
order by t
`,
queryParams,
);
return { chart };
}
async function clickhouseQuery(
websiteId: string,
parameters: RevenuParameters,
filters: QueryFilters,
) {
const { startDate, endDate, unit = 'day', timezone = 'utc', currency } = parameters;
const { getDateSQL, rawQuery, parseFilters } = clickhouse;
const { filterQuery, cohortQuery, dateQuery, queryParams } = parseFilters({
...filters,
websiteId,
startDate,
endDate,
currency,
});
const chart = await rawQuery<{ x: string; t: string; y: number; count: number }[]>(
`
with
filtered_sessions as (
select website_id, session_id
from website_event
${cohortQuery}
where website_id = {websiteId:UUID}
and event_type != ${EVENT_TYPE.performance}
${dateQuery}
${filterQuery}
group by website_id, session_id
),
filtered_revenue as (
select
website_revenue.event_id,
website_revenue.event_name,
website_revenue.created_at,
website_revenue.revenue
from website_revenue
inner join filtered_sessions
on filtered_sessions.website_id = website_revenue.website_id
and filtered_sessions.session_id = website_revenue.session_id
where website_revenue.website_id = {websiteId:UUID}
and website_revenue.created_at between {startDate:DateTime64} and {endDate:DateTime64}
and upper(website_revenue.currency) = {currency:String}
)
select
filtered_revenue.event_name x,
${getDateSQL('filtered_revenue.created_at', unit, timezone)} t,
sum(filtered_revenue.revenue) y,
count(filtered_revenue.event_id) count
from filtered_revenue
group by x, t
order by t
`,
queryParams,
);
return { chart };
}