Filter 0-byte requests and paths not starting with an integer

PRX · May 8, 2024 · 6f284dc · 6f284dc
1 parent 00fdaf2
commit 6f284dc
Showing 1 changed file with 3 additions and 1 deletion.
diff --git a/lib/athena.js b/lib/athena.js
@@ -77,7 +77,9 @@ exports.queryUsage = async (day) => {
   const removeRegion = "REGEXP_REPLACE(uri, '\\A/(use1/|usw2/)?')";
   const uri = `REGEXP_REPLACE(${removeRegion}, '/[^/]+/[^/]+\\z')`;
   const tbl = process.env.ATHENA_TABLE;
-  const sql = `SELECT ${uri}, SUM(bytes) FROM ${tbl} WHERE date = DATE(?) GROUP BY ${uri}`;
+  const prefix = `REGEXP_LIKE(${removeRegion}, '\\A[1-9][0-9]*/')`;
+  const where = `date = DATE(?) AND bytes > 0 AND ${prefix}`;
+  const sql = `SELECT ${uri}, SUM(bytes) FROM ${tbl} WHERE ${where} GROUP BY ${uri}`;
 
   // NOTE: for some reason, date params must be quoted
   return exports.query(sql, [`'${dayStr}'`]);