|
| 1 | +#standardSQL |
| 2 | +WITH base AS ( |
| 3 | + SELECT |
| 4 | + date, |
| 5 | + page, |
| 6 | + NET.HOST(page) AS host |
| 7 | + FROM |
| 8 | + `httparchive.crawl.pages` |
| 9 | + WHERE |
| 10 | + client = 'desktop' |
| 11 | + AND is_root_page |
| 12 | + AND date IN ( |
| 13 | + '2020-01-01','2020-04-01','2020-07-01','2020-10-01', |
| 14 | + '2021-01-01','2021-04-01','2021-07-01','2021-10-01', |
| 15 | + '2022-01-01','2022-04-01','2022-07-01','2022-10-01', |
| 16 | + '2023-01-01','2023-04-01','2023-07-01','2023-10-01', |
| 17 | + '2024-01-01','2024-04-01','2024-07-01','2024-10-01', |
| 18 | + '2025-01-01','2025-04-01','2025-07-01','2025-10-01' |
| 19 | + ) |
| 20 | +), |
| 21 | + |
| 22 | +classified AS ( |
| 23 | + SELECT |
| 24 | + date, |
| 25 | + page, |
| 26 | + CASE |
| 27 | + WHEN ENDS_WITH(host, 'vercel.app') THEN 'vercel' |
| 28 | + WHEN ENDS_WITH(host, 'pages.dev') THEN 'cloudflare_pages' |
| 29 | + WHEN ENDS_WITH(host, 'workers.dev') THEN 'cloudflare_workers' |
| 30 | + WHEN ENDS_WITH(host, 'lovable.app') OR ENDS_WITH(host, 'lovable.dev') THEN 'lovable' |
| 31 | + WHEN ENDS_WITH(host, 'bolt.new') OR ENDS_WITH(host, 'stackblitz.io') THEN 'bolt' |
| 32 | + WHEN ENDS_WITH(host, 'v0.dev') THEN 'v0' |
| 33 | + WHEN ENDS_WITH(host, 'replit.app') THEN 'replit' |
| 34 | + ELSE NULL |
| 35 | + END AS platform |
| 36 | + FROM |
| 37 | + base |
| 38 | +), |
| 39 | + |
| 40 | +totals AS ( |
| 41 | + SELECT |
| 42 | + date, |
| 43 | + COUNT(0) AS total_pages |
| 44 | + FROM |
| 45 | + base |
| 46 | + GROUP BY |
| 47 | + date |
| 48 | +) |
| 49 | + |
| 50 | +SELECT |
| 51 | + c.date, |
| 52 | + c.platform, |
| 53 | + COUNT(0) AS pages, |
| 54 | + SAFE_DIVIDE(COUNT(0), t.total_pages) AS pct_pages |
| 55 | +FROM |
| 56 | + classified c |
| 57 | +JOIN |
| 58 | + totals t |
| 59 | +USING (date) |
| 60 | +WHERE |
| 61 | + c.platform IS NOT NULL |
| 62 | +GROUP BY |
| 63 | + c.date, |
| 64 | + c.platform, |
| 65 | + t.total_pages |
| 66 | +ORDER BY |
| 67 | + c.date, |
| 68 | + pct_pages DESC; |
0 commit comments