@nshiab/simple-data-analysis-core 2.1.2 → 2.1.3

package/script/helpers/updateColumnsWithJS.js

@@ -8,6 +8,8 @@ const quoteIdentifier_js_1 = __importDefault(require("./quoteIdentifier.js"));
 const readMutationRows_js_1 = __importDefault(require("./readMutationRows.js"));
 const tableRegistry_js_1 = require("./tableRegistry.js");
 const loadArray_js_1 = require("../methods/loadArray.js");
+const appendColumnBatches_js_1 = __importDefault(require("./appendColumnBatches.js"));
+const parseDuckDBType_js_1 = __importDefault(require("./parseDuckDBType.js"));
 /**
  * Generates columns from selected input columns, keeping other SQL values in
  * DuckDB. Call from an asynchronous extension barrier. The callback must return
@@ -74,7 +76,6 @@ async function updateColumnsWithJS(table, inputColumns, outputColumns, generate,
     const id = `__sda_id_${suffix}`;
     const snapshot = `__sda_source_${suffix}`;
     const staged = table.sdb.newTable(`__sda_generated_${suffix}`);
-    const batch = table.sdb.newTable(`__sda_batch_${suffix}`);
     const q = quoteIdentifier_js_1.default;
     try {
         // A SQL snapshot gives every row a stable identity, including duplicate rows
@@ -98,17 +99,17 @@ async function updateColumnsWithJS(table, inputColumns, outputColumns, generate,
                 throw new Error("Column generation must return one row per input row.");
             }
             // Generated vectors retain this helper's existing inference contract.
-            await (0, loadArray_js_1.executePreparedArray)(batch, (0, loadArray_js_1.prepareArray)(generated.map((row, i) => ({
+            const prepared = (0, loadArray_js_1.prepareArray)(generated.map((row, i) => ({
                 ...Object.fromEntries(outputColumns.map((name) => [name, row[name]])),
                 [id]: ids[i],
-            }))));
-            const batchTypes = await batch.getTypes();
+            })));
+            const batchTypes = new Map(prepared.keys.map((name, i) => [name, prepared.types[i]]));
             for (const name of outputColumns) {
                 const nonNull = generated.some((row) => row[name] !== null && row[name] !== undefined);
                 if (!nonNull)
                     continue;
                 const previous = knownTypes.get(name);
-                const current = batchTypes[name];
+                const current = batchTypes.get(name);
                 if (previous !== undefined && previous !== current) {
                     throw new Error(`Generated column ${JSON.stringify(name)} changed type from ${previous} to ${current}.`);
                 }
@@ -120,12 +121,13 @@ async function updateColumnsWithJS(table, inputColumns, outputColumns, generate,
                 }
             }
             if (!hasRows) {
-                await table.sdb.customQuery(`CREATE TEMP TABLE ${q(staged.name)} AS SELECT * FROM ${q(batch.name)}`);
+                await table.sdb.customQuery(`CREATE TEMP TABLE ${q(staged.name)} (${prepared.keys.map((name, i) => `${q(name)} ${prepared.types[i]}`)
+                    .join(", ")})`);
                 hasRows = true;
             }
-            else {
-                await table.sdb.customQuery(`INSERT INTO ${q(staged.name)} SELECT * FROM ${q(batch.name)}`);
-            }
+            // All-null batches use the column's established type, avoiding a separate
+            // batch table and SQL copy just to cast nulls into the staged schema.
+            await (0, appendColumnBatches_js_1.default)(table.connection, staged.name, prepared.keys.map((name, i) => (0, parseDuckDBType_js_1.default)(knownTypes.get(name) ?? prepared.types[i])), prepared.rowCount, (column, start, end) => prepared.columnsData[column].slice(start, end));
         }
         if (!hasRows)
             return;
@@ -145,7 +147,7 @@ async function updateColumnsWithJS(table, inputColumns, outputColumns, generate,
         await table.sdb.customQuery(`CREATE OR REPLACE TABLE ${q(table.name)} AS SELECT ${columns.join(", ")} FROM ${q(snapshot)} original JOIN ${q(staged.name)} generated ON original.${q(id)} = CAST(generated.${q(id)} AS BIGINT) ORDER BY original.${q(id)}`);
     }
     finally {
-        await table.sdb.customQuery(`DROP TABLE IF EXISTS ${q(snapshot)}; DROP TABLE IF EXISTS ${q(staged.name)}; DROP TABLE IF EXISTS ${q(batch.name)};`);
-        (0, tableRegistry_js_1.retainRegisteredTables)(table.sdb, (item) => item !== staged && item !== batch);
+        await table.sdb.customQuery(`DROP TABLE IF EXISTS ${q(snapshot)}; DROP TABLE IF EXISTS ${q(staged.name)};`);
+        (0, tableRegistry_js_1.retainRegisteredTables)(table.sdb, (item) => item !== staged);
     }
 }

package/script/methods/similarityMahalanobis.d.ts

@@ -0,0 +1,5 @@
+import type SimpleTable from "../class/SimpleTable.js";
+type Options = NonNullable<Parameters<SimpleTable["similarityMahalanobis"]>[3]>;
+export default function similarityMahalanobis(table: SimpleTable, columns: string | string[], referencePoint: number[], newColumn: string, options?: Options): void;
+export {};
+//# sourceMappingURL=similarityMahalanobis.d.ts.map

package/script/methods/similarityMahalanobis.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"similarityMahalanobis.d.ts","sourceRoot":"","sources":["../../src/methods/similarityMahalanobis.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,WAAW,MAAM,yBAAyB,CAAC;AASvD,KAAK,OAAO,GAAG,WAAW,CAAC,UAAU,CAAC,WAAW,CAAC,uBAAuB,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AAEhF,MAAM,CAAC,OAAO,UAAU,qBAAqB,CAC3C,KAAK,EAAE,WAAW,EAClB,OAAO,EAAE,MAAM,GAAG,MAAM,EAAE,EAC1B,cAAc,EAAE,MAAM,EAAE,EACxB,SAAS,EAAE,MAAM,EACjB,OAAO,GAAE,OAAY,GACpB,IAAI,CAyBN"}

package/script/methods/{mahalanobis.js → similarityMahalanobis.js}

@@ -3,7 +3,7 @@ var __importDefault = (this && this.__importDefault) || function (mod) {
     return (mod && mod.__esModule) ? mod : { "default": mod };
 };
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.default = mahalanobis;
+exports.default = similarityMahalanobis;
 const computeCovariance_js_1 = __importDefault(require("../helpers/computeCovariance.js"));
 const computeMahalanobisDistances_js_1 = __importDefault(require("../helpers/computeMahalanobisDistances.js"));
 const foldIdentifier_js_1 = __importDefault(require("../helpers/foldIdentifier.js"));
@@ -11,17 +11,17 @@ const prepareNumericFeatures_js_1 = __importDefault(require("../helpers/prepareN
 const publishPreparedColumns_js_1 = __importDefault(require("../helpers/publishPreparedColumns.js"));
 const queueOp_js_1 = __importDefault(require("../helpers/queueOp.js"));
 const quoteIdentifier_js_1 = __importDefault(require("../helpers/quoteIdentifier.js"));
-function mahalanobis(table, columns, referencePoint, newColumn, options = {}) {
+function similarityMahalanobis(table, columns, referencePoint, newColumn, options = {}) {
     if (!Array.isArray(referencePoint) || referencePoint.length === 0 ||
         Array.from(referencePoint).some((value) => typeof value !== "number" || !Number.isFinite(value))) {
-        throw new Error("mahalanobis() referencePoint must be a nonempty array of finite numbers in feature-dimension order.");
+        throw new Error("similarityMahalanobis() referencePoint must be a nonempty array of finite numbers in feature-dimension order.");
     }
     const selected = typeof columns === "string" ? columns : [...columns];
     const reference = [...referencePoint];
     const settings = { ...options };
     (0, queueOp_js_1.default)(table, {
         kind: "barrier",
-        method: "mahalanobis()",
+        method: "similarityMahalanobis()",
         parameters: {
             columns: selected,
             referencePoint: reference,
@@ -39,23 +39,23 @@ async function execute(table, columns, referencePoint, newColumn, options) {
     const seen = new Set();
     for (const name of outputNames) {
         if (typeof name !== "string" || name.length === 0 || name.includes("\0")) {
-            throw new Error("mahalanobis() output column names must be nonempty strings without null characters.");
+            throw new Error("similarityMahalanobis() output column names must be nonempty strings without null characters.");
         }
         const folded = (0, foldIdentifier_js_1.default)(name);
         if (seen.has(folded)) {
-            throw new Error("mahalanobis() distance and similarity score columns must have different names (case-insensitive).");
+            throw new Error("similarityMahalanobis() distance and similarity score columns must have different names (case-insensitive).");
         }
         seen.add(folded);
         if (sourceColumns.some((column) => (0, foldIdentifier_js_1.default)(column) === folded)) {
-            throw new Error(`mahalanobis() cannot create ${(0, quoteIdentifier_js_1.default)(name)} because that column already exists. Remove it first or choose a different name.`);
+            throw new Error(`similarityMahalanobis() cannot create ${(0, quoteIdentifier_js_1.default)(name)} because that column already exists. Remove it first or choose a different name.`);
         }
     }
     const prepared = await (0, prepareNumericFeatures_js_1.default)(table, typeof columns === "string"
         ? { kind: "vector", column: columns }
-        : { kind: "scalars", columns }, { method: "mahalanobis()" });
+        : { kind: "scalars", columns }, { method: "similarityMahalanobis()" });
     try {
         if (referencePoint.length !== prepared.dimensions) {
-            throw new Error(`mahalanobis() referencePoint must contain ${prepared.dimensions} values in feature-dimension order; received ${referencePoint.length}.`);
+            throw new Error(`similarityMahalanobis() referencePoint must contain ${prepared.dimensions} values in feature-dimension order; received ${referencePoint.length}.`);
         }
         const model = await (0, computeCovariance_js_1.default)(table.connection, {
             relation: prepared.relation,
@@ -79,7 +79,7 @@ async function execute(table, columns, referencePoint, newColumn, options) {
                 });
             }
             await (0, publishPreparedColumns_js_1.default)(table, prepared, {
-                method: "mahalanobis()",
+                method: "similarityMahalanobis()",
                 parameters: { columns, referencePoint, newColumn, options },
                 result: {
                     relation: distances.relation,

package/script/methods/updateWithJS.js

@@ -66,6 +66,7 @@ async function executeUpdateWithJS(simpleTable, dataModifier, options) {
     const accumulator = `updateWithJS_accumulator_${suffix}`;
     try {
         let first = true;
+        let accumulatorSchema;
         let lastRowid = null;
         let sawRows = false;
         let outputOffset = 0;
@@ -137,9 +138,17 @@ async function executeUpdateWithJS(simpleTable, dataModifier, options) {
                 ? (0, parseDuckDBType_js_1.default)("VARCHAR")
                 : source.types.get(key) ??
                     (0, parseDuckDBType_js_1.default)(added.types[newKeys.indexOf(key)]));
-            await simpleTable.sdb.customQuery(`CREATE OR REPLACE TABLE ${(0, quoteIdentifier_js_1.default)(scratch.name)} (${keys.map((key, i) => `${(0, quoteIdentifier_js_1.default)(key)} ${columnTypes[i]}`)
-                .join(", ")})`);
-            await (0, appendColumnBatches_js_1.default)(simpleTable.connection, scratch.name, columnTypes, modified.length, (column, start, end) => {
+            const schema = JSON.stringify(keys.map((key, i) => [key, columnTypes[i].toString()]));
+            // Append matching native batches directly. SQL still handles conversion
+            // and BY NAME alignment when a callback changes the batch schema.
+            const direct = outputGeometry.length === 0 && outputJSON.length === 0 &&
+                (first || schema === accumulatorSchema);
+            const destination = direct ? accumulator : scratch.name;
+            if (!direct || first) {
+                await simpleTable.sdb.customQuery(`CREATE OR REPLACE TABLE ${(0, quoteIdentifier_js_1.default)(destination)} (${keys.map((key, i) => `${(0, quoteIdentifier_js_1.default)(key)} ${columnTypes[i]}`)
+                    .join(", ")})`);
+            }
+            await (0, appendColumnBatches_js_1.default)(simpleTable.connection, destination, columnTypes, modified.length, (column, start, end) => {
                 const key = keys[column];
                 return outputGeometry.includes(key)
                     ? geometryData.get(key).slice(start, end)
@@ -152,7 +161,11 @@ async function executeUpdateWithJS(simpleTable, dataModifier, options) {
                 : outputJSON.includes(key)
                     ? `${(0, quoteIdentifier_js_1.default)(key)}::JSON AS ${(0, quoteIdentifier_js_1.default)(key)}`
                     : (0, quoteIdentifier_js_1.default)(key)).join(", ");
-            if (first) {
+            if (direct) {
+                accumulatorSchema = schema;
+                first = false;
+            }
+            else if (first) {
                 await simpleTable.sdb.customQuery(`CREATE OR REPLACE TABLE ${(0, quoteIdentifier_js_1.default)(accumulator)} AS SELECT ${projection} FROM ${(0, quoteIdentifier_js_1.default)(scratch.name)}`);
                 first = false;
             }
@@ -165,7 +178,13 @@ async function executeUpdateWithJS(simpleTable, dataModifier, options) {
         if (first) {
             throw new Error("The dataModifier returned no rows. updateWithJS can't infer the table schema from zero rows.");
         }
-        await simpleTable.sdb.customQuery(`CREATE OR REPLACE TABLE ${(0, quoteIdentifier_js_1.default)(simpleTable.name)} AS SELECT * FROM ${(0, quoteIdentifier_js_1.default)(accumulator)}`);
+        // Publish the completed ordinary table without copying its rows again.
+        // Qualify the creation namespace: a temporary table or a later search-path
+        // schema can shadow the name without being the destination of replacement.
+        const namespace = await simpleTable.connection.runAndReadAll("SELECT current_database(), current_schema()");
+        const prefix = namespace.getRows()[0].map((name) => (0, quoteIdentifier_js_1.default)(String(name))).join(".");
+        await simpleTable.sdb.customQuery(`DROP TABLE IF EXISTS ${prefix}.${(0, quoteIdentifier_js_1.default)(simpleTable.name)};
+ALTER TABLE ${prefix}.${(0, quoteIdentifier_js_1.default)(accumulator)} RENAME TO ${(0, quoteIdentifier_js_1.default)(simpleTable.name)};`);
     }
     finally {
         await simpleTable.sdb.customQuery(`DROP TABLE IF EXISTS ${(0, quoteIdentifier_js_1.default)(accumulator)};

package/esm/methods/mahalanobis.d.ts

@@ -1,5 +0,0 @@
-import type SimpleTable from "../class/SimpleTable.js";
-type Options = NonNullable<Parameters<SimpleTable["mahalanobis"]>[3]>;
-export default function mahalanobis(table: SimpleTable, columns: string | string[], referencePoint: number[], newColumn: string, options?: Options): void;
-export {};
-//# sourceMappingURL=mahalanobis.d.ts.map

package/esm/methods/mahalanobis.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"mahalanobis.d.ts","sourceRoot":"","sources":["../../src/methods/mahalanobis.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,WAAW,MAAM,yBAAyB,CAAC;AASvD,KAAK,OAAO,GAAG,WAAW,CAAC,UAAU,CAAC,WAAW,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AAEtE,MAAM,CAAC,OAAO,UAAU,WAAW,CACjC,KAAK,EAAE,WAAW,EAClB,OAAO,EAAE,MAAM,GAAG,MAAM,EAAE,EAC1B,cAAc,EAAE,MAAM,EAAE,EACxB,SAAS,EAAE,MAAM,EACjB,OAAO,GAAE,OAAY,GACpB,IAAI,CAyBN"}

package/script/methods/mahalanobis.d.ts

@@ -1,5 +0,0 @@
-import type SimpleTable from "../class/SimpleTable.js";
-type Options = NonNullable<Parameters<SimpleTable["mahalanobis"]>[3]>;
-export default function mahalanobis(table: SimpleTable, columns: string | string[], referencePoint: number[], newColumn: string, options?: Options): void;
-export {};
-//# sourceMappingURL=mahalanobis.d.ts.map

package/script/methods/mahalanobis.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"mahalanobis.d.ts","sourceRoot":"","sources":["../../src/methods/mahalanobis.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,WAAW,MAAM,yBAAyB,CAAC;AASvD,KAAK,OAAO,GAAG,WAAW,CAAC,UAAU,CAAC,WAAW,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AAEtE,MAAM,CAAC,OAAO,UAAU,WAAW,CACjC,KAAK,EAAE,WAAW,EAClB,OAAO,EAAE,MAAM,GAAG,MAAM,EAAE,EAC1B,cAAc,EAAE,MAAM,EAAE,EACxB,SAAS,EAAE,MAAM,EACjB,OAAO,GAAE,OAAY,GACpB,IAAI,CAyBN"}