@query-doctor/core 0.7.2-rc.1 → 0.8.0-rc.1

package/dist/index.cjs

@@ -1817,14 +1817,18 @@ const ExportedStatsStatistics = zod.z.object({
 });
 const ExportedStatsColumns = zod.z.object({
 	columnName: zod.z.string(),
+	attlen: zod.z.number().nullable(),
 	stats: ExportedStatsStatistics.nullable()
 });
 const ExportedStatsIndex = zod.z.object({
 	indexName: zod.z.string(),
+	amname: zod.z.string().default("btree"),
 	relpages: zod.z.number(),
 	reltuples: zod.z.number(),
 	relallvisible: zod.z.number(),
-	relallfrozen: zod.z.number().optional()
+	relallfrozen: zod.z.number().optional(),
+	fillfactor: zod.z.number().default(90),
+	columns: zod.z.array(zod.z.object({ attlen: zod.z.number().nullable() })).default([])
 });
 const ExportedStatsV1 = zod.z.object({
 	tableName: zod.z.string(),
@@ -1833,14 +1837,13 @@ const ExportedStatsV1 = zod.z.object({
 	reltuples: zod.z.number(),
 	relallvisible: zod.z.number(),
 	relallfrozen: zod.z.number().optional(),
-	columns: zod.z.array(ExportedStatsColumns).nullable(),
+	columns: zod.z.array(ExportedStatsColumns).default([]),
 	indexes: zod.z.array(ExportedStatsIndex)
 });
 const ExportedStats = zod.z.union([ExportedStatsV1]);
 const StatisticsMode = zod.z.discriminatedUnion("kind", [zod.z.object({
 	kind: zod.z.literal("fromAssumption"),
-	reltuples: zod.z.number().min(0),
-	relpages: zod.z.number().min(0)
+	reltuples: zod.z.number().min(0)
 }), zod.z.object({
 	kind: zod.z.literal("fromStatisticsExport"),
 	stats: zod.z.array(ExportedStats),
@@ -1848,6 +1851,19 @@ const StatisticsMode = zod.z.discriminatedUnion("kind", [zod.z.object({
 })]);
 const DEFAULT_RELTUPLES = 1e4;
 const DEFAULT_RELPAGES = 1;
+const DEFAULT_PAGE_SIZE = 2 ** 13;
+function estimateStawidth(col) {
+	return col.attlen ?? 32;
+}
+function estimateRelpages(reltuples, columns) {
+	const rowWidth = columns.reduce((sum, col) => sum + estimateStawidth(col), 0) + 27;
+	return Math.ceil(reltuples * rowWidth / DEFAULT_PAGE_SIZE);
+}
+function estimateIndexRelpages(reltuples, columns, fillfactor, amname, tableRelpages) {
+	if (amname === "gin") return Math.ceil(tableRelpages * .3);
+	const keyWidth = columns.reduce((sum, col) => sum + estimateStawidth(col) + 16, 0);
+	return Math.ceil(reltuples * keyWidth / DEFAULT_PAGE_SIZE / fillfactor);
+}
 var Statistics = class Statistics {
 	constructor(db, postgresVersion, ownMetadata, statsMode) {
 		this.db = db;
@@ -1860,11 +1876,10 @@ var Statistics = class Statistics {
 			if (statsMode.kind === "fromStatisticsExport") this.exportedMetadata = statsMode.stats;
 		} else this.mode = Statistics.defaultStatsMode;
 	}
-	static statsModeFromAssumption({ reltuples, relpages }) {
+	static statsModeFromAssumption({ reltuples }) {
 		return {
 			kind: "fromAssumption",
-			reltuples,
-			relpages
+			reltuples
 		};
 	}
 	/**
@@ -1928,17 +1943,17 @@ var Statistics = class Statistics {
 		const columnStatsValues = [];
 		for (const table of this.ownMetadata) {
 			const target = (this.exportedMetadata?.find((m) => m.tableName === table.tableName && m.schemaName === table.schemaName))?.columns ?? table.columns;
-			if (!target) continue;
 			for (const column of target) {
 				const { stats } = column;
 				if (!stats || this.mode.kind === "fromAssumption") {
+					const stawidth = stats?.stawidth || estimateStawidth(column);
 					columnStatsValues.push({
 						schema_name: table.schemaName,
 						table_name: table.tableName,
 						column_name: column.columnName,
 						stainherit: false,
 						stanullfrac: .04,
-						stawidth: stats?.stawidth ?? 32,
+						stawidth,
 						stadistinct: -.9,
 						stakind1: 0,
 						stakind2: 0,
@@ -2239,7 +2254,7 @@ var Statistics = class Statistics {
 			let relallfrozen;
 			if (this.mode.kind === "fromAssumption") {
 				reltuples = this.mode.reltuples;
-				relpages = this.mode.relpages;
+				relpages = estimateRelpages(reltuples, table.columns);
 			} else if (targetTable) {
 				reltuples = targetTable.reltuples;
 				relpages = targetTable.relpages;
@@ -2258,15 +2273,18 @@ var Statistics = class Statistics {
 				relallfrozen,
 				relallvisible
 			});
-			if (this.mode.kind === "fromAssumption") for (const index of table.indexes) reltuplesValues.push({
-				relname: index.indexName,
-				schema_name: table.schemaName,
-				reltuples: this.mode.reltuples,
-				relpages: Math.ceil(this.mode.relpages * .2),
-				relallfrozen: 0,
-				relallvisible: this.mode.relpages
-			});
-			if (targetTable) for (const index of targetTable.indexes) reltuplesValues.push({
+			if (this.mode.kind === "fromAssumption") for (const index of table.indexes) {
+				const indexRelpages = estimateIndexRelpages(this.mode.reltuples, index.columns, index.fillfactor / 100, index.amname, relpages);
+				reltuplesValues.push({
+					relname: index.indexName,
+					schema_name: table.schemaName,
+					reltuples: this.mode.reltuples,
+					relpages: indexRelpages,
+					relallfrozen: 0,
+					relallvisible: indexRelpages
+				});
+			}
+			else if (targetTable) for (const index of targetTable.indexes) reltuplesValues.push({
 				relname: index.indexName,
 				schema_name: targetTable.schemaName,
 				reltuples: index.reltuples,
@@ -2309,19 +2327,19 @@ var Statistics = class Statistics {
 	static async dumpStats(db, postgresVersion, kind) {
 		const fullDump = kind === "full";
 		console.log(`dumping stats for postgres ${(0, colorette.gray)(postgresVersion)}`);
-		return (await db.exec(`
+		const stats = await db.exec(`
       WITH table_columns AS (
         SELECT
-          c.table_name,
-          c.table_schema,
+          cl.relname,
+          n.nspname,
           cl.reltuples,
           cl.relpages,
           cl.relallvisible,
           -- cl.relallfrozen,
-          n.nspname AS schema_name,
           json_agg(
             json_build_object(
-              'columnName', c.column_name,
+              'columnName', a.attname,
+              'attlen', CASE WHEN a.attlen > 0 THEN a.attlen ELSE NULL END,
               'stats', (
                 SELECT json_build_object(
                   'starelid', s.starelid,
@@ -2344,21 +2362,15 @@ var Statistics = class Statistics {
                 WHERE s.starelid = a.attrelid AND s.staattnum = a.attnum
               )
             )
-            ORDER BY c.ordinal_position
+            ORDER BY a.attnum
           ) AS columns
-        FROM information_schema.columns c
-        JOIN pg_attribute a
-          ON a.attrelid = (quote_ident(c.table_schema) || '.' || quote_ident(c.table_name))::regclass
-          AND a.attname = c.column_name
-        JOIN pg_class cl
-          ON cl.oid = a.attrelid
-        JOIN pg_namespace n
-          ON n.oid = cl.relnamespace
-        WHERE c.table_name NOT LIKE 'pg_%'
-          AND n.nspname <> 'information_schema'
-          AND n.nspname NOT IN ('tiger', 'tiger_data', 'topology')
-          AND c.table_name NOT IN ('pg_stat_statements', 'pg_stat_statements_info')
-        GROUP BY c.table_name, c.table_schema, cl.reltuples, cl.relpages, cl.relallvisible, n.nspname
+        FROM pg_class cl
+        JOIN pg_namespace n ON n.oid = cl.relnamespace
+        JOIN pg_attribute a ON a.attrelid = cl.oid AND a.attnum > 0 AND NOT a.attisdropped
+        WHERE cl.relkind = 'r'
+          AND n.nspname NOT IN ('pg_catalog', 'information_schema', 'tiger', 'tiger_data', 'topology')
+          AND cl.relname NOT IN ('pg_stat_statements', 'pg_stat_statements_info')
+        GROUP BY cl.relname, n.nspname, cl.reltuples, cl.relpages, cl.relallvisible
       ),
       table_indexes AS (
         SELECT
@@ -2366,15 +2378,37 @@ var Statistics = class Statistics {
           json_agg(
             json_build_object(
               'indexName', i.relname,
+              'amname', am.amname,
               'reltuples', i.reltuples,
               'relpages', i.relpages,
-              'relallvisible', i.relallvisible
-              -- 'relallfrozen', i.relallfrozen
+              'relallvisible', i.relallvisible,
+              -- 'relallfrozen', i.relallfrozen,
+              'fillfactor', COALESCE(
+                (
+                  SELECT (regexp_match(opt, 'fillfactor=(\\d+)'))[1]::integer
+                  FROM unnest(i.reloptions) AS opt
+                  WHERE opt LIKE 'fillfactor=%'
+                  LIMIT 1
+                ),
+                90
+              ),
+              'columns', COALESCE(
+                (
+                  SELECT json_agg(json_build_object(
+                    'attlen', CASE WHEN a.attlen > 0 THEN a.attlen ELSE NULL END
+                  ) ORDER BY col_pos.ord)
+                  FROM unnest(ix.indkey) WITH ORDINALITY AS col_pos(attnum, ord)
+                  JOIN pg_attribute a ON a.attrelid = t.oid AND a.attnum = col_pos.attnum
+                  WHERE col_pos.attnum > 0
+                ),
+                '[]'::json
+              )
             )
           ) AS indexes
         FROM pg_class t
         JOIN pg_index ix ON ix.indrelid = t.oid
         JOIN pg_class i ON i.oid = ix.indexrelid
+        JOIN pg_am am ON am.oid = i.relam
         JOIN pg_namespace n ON n.oid = t.relnamespace
         WHERE t.relname NOT LIKE 'pg_%'
           AND n.nspname <> 'information_schema'
@@ -2383,20 +2417,21 @@ var Statistics = class Statistics {
       )
       SELECT json_agg(
         json_build_object(
-          'tableName', tc.table_name,
-          'schemaName', tc.table_schema,
+          'tableName', tc.relname,
+          'schemaName', tc.nspname,
           'reltuples', tc.reltuples,
           'relpages', tc.relpages,
           'relallvisible', tc.relallvisible,
           -- 'relallfrozen', tc.relallfrozen,
-          'columns', tc.columns,
+          'columns', COALESCE(tc.columns, '[]'::json),
           'indexes', COALESCE(ti.indexes, '[]'::json)
         )
       )
       FROM table_columns tc
       LEFT JOIN table_indexes ti
-        ON ti.table_name = tc.table_name;
-      `, [fullDump]))[0].json_agg;
+        ON ti.table_name = tc.relname;
+      `, [fullDump]);
+		return zod.z.array(ExportedStats).parse(stats[0].json_agg);
 	}
 	/**
 	* Returns all indexes in the database.