@nshiab/simple-data-analysis-core 2.0.3 → 2.0.5

package/script/methods/columnProportions.js

@@ -9,6 +9,10 @@ const assertNewColumns_js_1 = __importDefault(require("../helpers/assertNewColum
 const queueOp_js_1 = __importDefault(require("../helpers/queueOp.js"));
 const stringToArray_js_1 = __importDefault(require("../helpers/stringToArray.js"));
 function columnProportions(simpleTable, column, newColumn, options = {}) {
+    if (options.base !== undefined &&
+        (!Number.isFinite(options.base) || options.base <= 0)) {
+        throw new Error("columnProportions() options.base must be a finite number greater than 0.");
+    }
     options = structuredClone(options);
     (0, queueOp_js_1.default)(simpleTable, {
         kind: "fusable",
@@ -21,9 +25,10 @@ function columnProportions(simpleTable, column, newColumn, options = {}) {
             const partition = by.length === 0
                 ? ""
                 : `PARTITION BY ${by.map((d) => `${(0, quoteIdentifier_js_1.default)(d)}`).join(",")}`;
+            const proportion = `${(0, quoteIdentifier_js_1.default)(column)} / sum(${(0, quoteIdentifier_js_1.default)(column)}) OVER(${partition}) * ${options.base ?? 1}`;
             return typeof options.decimals === "number"
-                ? `SELECT *, ROUND(${(0, quoteIdentifier_js_1.default)(column)} / sum(${(0, quoteIdentifier_js_1.default)(column)}) OVER(${partition}), ${options.decimals}) AS ${(0, quoteIdentifier_js_1.default)(newColumn)} FROM ${input}`
-                : `SELECT *, ${(0, quoteIdentifier_js_1.default)(column)} / sum(${(0, quoteIdentifier_js_1.default)(column)}) OVER(${partition}) AS ${(0, quoteIdentifier_js_1.default)(newColumn)} FROM ${input}`;
+                ? `SELECT *, ROUND(${proportion}, ${options.decimals}) AS ${(0, quoteIdentifier_js_1.default)(newColumn)} FROM ${input}`
+                : `SELECT *, ${proportion} AS ${(0, quoteIdentifier_js_1.default)(newColumn)} FROM ${input}`;
         },
     });
 }

package/script/methods/loadArray.d.ts

@@ -2,7 +2,7 @@ import { type DuckDBValue } from "@duckdb/node-api";
 import type SimpleTable from "../class/SimpleTable.js";
 export default function loadArray(simpleTable: SimpleTable, rows: {
     [key: string]: unknown;
-}[]): void;
+}[], options?: NonNullable<Parameters<SimpleTable["loadArray"]>[1]>): void;
 type PreparedArray = {
     keys: string[];
     types: string[];
@@ -11,7 +11,9 @@ type PreparedArray = {
 };
 export declare function prepareArray(rows: {
     [key: string]: unknown;
-}[]): PreparedArray;
+}[], columnTypes?: {
+    [key: string]: string;
+}): PreparedArray;
 export declare function executePreparedArray(simpleTable: SimpleTable, prepared: PreparedArray): Promise<void>;
 export {};
 //# sourceMappingURL=loadArray.d.ts.map

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

@@ -1 +1 @@
-{"version":3,"file":"loadArray.d.ts","sourceRoot":"","sources":["../../src/methods/loadArray.ts"],"names":[],"mappings":"AAAA,OAAO,EAKL,KAAK,WAAW,EACjB,MAAM,kBAAkB,CAAC;AAE1B,OAAO,KAAK,WAAW,MAAM,yBAAyB,CAAC;AAKvD,MAAM,CAAC,OAAO,UAAU,SAAS,CAC/B,WAAW,EAAE,WAAW,EACxB,IAAI,EAAE;IAAE,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAA;CAAE,EAAE,QAiBnC;AAED,KAAK,aAAa,GAAG;IACnB,IAAI,EAAE,MAAM,EAAE,CAAC;IACf,KAAK,EAAE,MAAM,EAAE,CAAC;IAChB,WAAW,EAAE,WAAW,EAAE,EAAE,CAAC;IAC7B,QAAQ,EAAE,MAAM,CAAC;CAClB,CAAC;AAEF,wBAAgB,YAAY,CAC1B,IAAI,EAAE;IAAE,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAA;CAAE,EAAE,GACjC,aAAa,CA8Df;AAED,wBAAsB,oBAAoB,CACxC,WAAW,EAAE,WAAW,EACxB,QAAQ,EAAE,aAAa,iBA+BxB"}
+{"version":3,"file":"loadArray.d.ts","sourceRoot":"","sources":["../../src/methods/loadArray.ts"],"names":[],"mappings":"AAAA,OAAO,EAQL,KAAK,WAAW,EACjB,MAAM,kBAAkB,CAAC;AAE1B,OAAO,KAAK,WAAW,MAAM,yBAAyB,CAAC;AAKvD,MAAM,CAAC,OAAO,UAAU,SAAS,CAC/B,WAAW,EAAE,WAAW,EACxB,IAAI,EAAE;IAAE,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAA;CAAE,EAAE,EAClC,OAAO,GAAE,WAAW,CAAC,UAAU,CAAC,WAAW,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,CAAC,CAAM,QAiBnE;AAED,KAAK,aAAa,GAAG;IACnB,IAAI,EAAE,MAAM,EAAE,CAAC;IACf,KAAK,EAAE,MAAM,EAAE,CAAC;IAChB,WAAW,EAAE,WAAW,EAAE,EAAE,CAAC;IAC7B,QAAQ,EAAE,MAAM,CAAC;CAClB,CAAC;AAEF,wBAAgB,YAAY,CAC1B,IAAI,EAAE;IAAE,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAA;CAAE,EAAE,EAClC,WAAW,GAAE;IAAE,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAAA;CAAO,GAC1C,aAAa,CA2Ff;AAED,wBAAsB,oBAAoB,CACxC,WAAW,EAAE,WAAW,EACxB,QAAQ,EAAE,aAAa,iBA+BxB"}

package/script/methods/loadArray.js

@@ -11,12 +11,12 @@ const queueOp_js_1 = __importDefault(require("../helpers/queueOp.js"));
 const parseTypes_js_1 = __importDefault(require("../helpers/parseTypes.js"));
 const parseDuckDBType_js_1 = __importDefault(require("../helpers/parseDuckDBType.js"));
 const quoteIdentifier_js_1 = __importDefault(require("../helpers/quoteIdentifier.js"));
-function loadArray(simpleTable, rows) {
+function loadArray(simpleTable, rows, options = {}) {
     // This validation doesn't need the database, so it stays at call time.
     if (rows.length === 0) {
         throw new Error("The array is empty. loadArray needs at least one object to infer the column types.");
     }
-    const prepared = prepareArray(rows);
+    const prepared = prepareArray(rows, options.columnTypes);
     (0, queueOp_js_1.default)(simpleTable, {
         kind: "barrier",
         method: "loadArray()",
@@ -24,14 +24,38 @@ function loadArray(simpleTable, rows) {
         execute: () => executePreparedArray(simpleTable, prepared),
     });
 }
-function prepareArray(rows) {
+function prepareArray(rows, columnTypes = {}) {
     const keys = Object.keys(rows[0]);
+    const overrides = new Map();
+    for (const [key, type] of Object.entries(columnTypes)) {
+        if (!keys.includes(key)) {
+            throw new Error(`Unknown column ${JSON.stringify(key)} in loadArray columnTypes.`);
+        }
+        try {
+            const normalized = /^FLOAT\[[1-9]\d*\]$/i.test(type)
+                ? type.toUpperCase()
+                : (0, parseTypes_js_1.default)(type);
+            (0, parseDuckDBType_js_1.default)(normalized);
+            overrides.set(key, normalized);
+        }
+        catch {
+            throw new Error(`Unsupported type ${JSON.stringify(type)} for column ${JSON.stringify(key)}.`);
+        }
+    }
     const firstNonNullValue = keys.map((key) => rows.find((obj) => obj[key] !== null && obj[key] !== undefined)?.[key]);
     const types = [];
     const columnsData = keys.map(() => new Array(rows.length));
     for (let i = 0; i < keys.length; i++) {
         const key = keys[i];
         const columnData = columnsData[i];
+        const override = overrides.get(key);
+        if (override !== undefined) {
+            types[i] = override;
+            for (let j = 0; j < rows.length; j++) {
+                columnData[j] = prepareTypedValue(rows[j][key], override, key, j + 1);
+            }
+            continue;
+        }
         const value = firstNonNullValue[i];
         const type = typeof value;
         if (type === "symbol" || type === "function") {
@@ -104,3 +128,88 @@ async function executePreparedArray(simpleTable, prepared) {
     }
     appender.flushSync();
 }
+/** Validates and snapshots a value for an explicitly declared destination type. */
+function prepareTypedValue(value, type, column, row) {
+    if (value === null || value === undefined || Number.isNaN(value))
+        return null;
+    let compatible = false;
+    if (type === "VARCHAR" && typeof value === "string")
+        return value;
+    if (type === "BOOLEAN" && typeof value === "boolean")
+        return value;
+    if (type === "DOUBLE") {
+        if (typeof value === "number")
+            return value;
+        compatible = typeof value === "bigint";
+        if (typeof value === "bigint") {
+            const number = Number(value);
+            if (Number.isFinite(number) && BigInt(number) === value)
+                return number;
+        }
+    }
+    else if (type === "INTEGER") {
+        compatible = typeof value === "number" || typeof value === "bigint";
+        if (typeof value === "number") {
+            if (Number.isInteger(value) && value >= -2147483648 && value <= 2147483647) {
+                return value;
+            }
+        }
+        else if (typeof value === "bigint" && value >= -2147483648n && value <= 2147483647n) {
+            return Number(value);
+        }
+    }
+    else if (type === "BIGINT") {
+        compatible = typeof value === "number" || typeof value === "bigint";
+        if (typeof value === "bigint") {
+            if (value >= -9223372036854775808n && value <= 9223372036854775807n) {
+                return value;
+            }
+        }
+        else if (typeof value === "number" && Number.isInteger(value) &&
+            value >= -(2 ** 63) && value < 2 ** 63) {
+            // The upper bound is exclusive: 2 ** 63 - 1 rounds up as a JS number.
+            return BigInt(value);
+        }
+    }
+    else if (["DATE", "TIME", "TIMESTAMP", "TIMESTAMP WITH TIME ZONE"].includes(type)) {
+        if (type === "TIME" && value instanceof node_api_1.DuckDBTimeValue) {
+            if (value.micros >= 0n && value.micros <= 86400000000n) {
+                return new node_api_1.DuckDBTimeValue(value.micros);
+            }
+            compatible = true;
+        }
+        else if (value instanceof Date) {
+            const millis = value.getTime();
+            compatible = Number.isFinite(millis);
+            if (compatible) {
+                if (type === "DATE" && millis % 86400000 === 0) {
+                    return new node_api_1.DuckDBDateValue(millis / 86400000);
+                }
+                const micros = BigInt(millis) * 1000n;
+                if (type === "TIMESTAMP")
+                    return new node_api_1.DuckDBTimestampValue(micros);
+                if (type === "TIMESTAMP WITH TIME ZONE") {
+                    return new node_api_1.DuckDBTimestampTZValue(micros);
+                }
+            }
+        }
+    }
+    else if (type.startsWith("FLOAT[") && Array.isArray(value)) {
+        compatible = true;
+        const size = Number(type.slice(6, -1));
+        if (value.length === size &&
+            value.every((item) => item === null || item === undefined ||
+                (typeof item === "number" &&
+                    (Number.isNaN(item) || Math.fround(item) === item)))) {
+            return (0, node_api_1.arrayValue)(value.map((item) => item === undefined || Number.isNaN(item)
+                ? null
+                : item));
+        }
+    }
+    const displayed = typeof value === "string"
+        ? JSON.stringify(value)
+        : String(value);
+    throw new Error(`Column ${JSON.stringify(column)}, row ${row}: value ${displayed} ${compatible
+        ? `cannot be stored as ${type} without losing information`
+        : `is incompatible with ${type}`}. Select a type compatible with this value.`);
+}

package/script/methods/loadYahooFinanceData.d.ts

@@ -0,0 +1,10 @@
+import type SimpleTable from "../class/SimpleTable.js";
+declare const INTERVAL_SECONDS: {
+    readonly "1d": number;
+    readonly "1h": number;
+    readonly "1m": 60;
+};
+type YahooInterval = keyof typeof INTERVAL_SECONDS;
+export default function loadYahooFinanceData(table: SimpleTable, symbol: string, startDate: Date, endDate: Date, interval: YahooInterval): void;
+export {};
+//# sourceMappingURL=loadYahooFinanceData.d.ts.map

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

@@ -0,0 +1 @@
+{"version":3,"file":"loadYahooFinanceData.d.ts","sourceRoot":"","sources":["../../src/methods/loadYahooFinanceData.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,WAAW,MAAM,yBAAyB,CAAC;AAKvD,QAAA,MAAM,gBAAgB;;;;CAIZ,CAAC;AAeX,KAAK,aAAa,GAAG,MAAM,OAAO,gBAAgB,CAAC;AA6BnD,MAAM,CAAC,OAAO,UAAU,oBAAoB,CAC1C,KAAK,EAAE,WAAW,EAClB,MAAM,EAAE,MAAM,EACd,SAAS,EAAE,IAAI,EACf,OAAO,EAAE,IAAI,EACb,QAAQ,EAAE,aAAa,GACtB,IAAI,CAeN"}

package/script/methods/loadYahooFinanceData.js

@@ -0,0 +1,139 @@
+"use strict";
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.default = loadYahooFinanceData;
+const SDAError_js_1 = __importDefault(require("../class/SDAError.js"));
+const queueOp_js_1 = __importDefault(require("../helpers/queueOp.js"));
+const loadArray_js_1 = require("./loadArray.js");
+const INTERVAL_SECONDS = {
+    "1d": 24 * 60 * 60,
+    "1h": 60 * 60,
+    "1m": 60,
+};
+const COLUMN_TYPES = {
+    datetime: "TIMESTAMP",
+    open: "DOUBLE",
+    high: "DOUBLE",
+    low: "DOUBLE",
+    close: "DOUBLE",
+    adjustedClose: "DOUBLE",
+    volume: "DOUBLE",
+};
+const USER_AGENT = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36";
+function loadYahooFinanceData(table, symbol, startDate, endDate, interval) {
+    assertArguments(symbol, startDate, endDate, interval);
+    const request = {
+        symbol,
+        startDate: new Date(startDate),
+        endDate: new Date(endDate),
+        interval,
+    };
+    (0, queueOp_js_1.default)(table, {
+        kind: "barrier",
+        method: "loadYahooFinanceData()",
+        parameters: request,
+        execute: () => executeLoadYahooFinanceData(table, request),
+    });
+}
+async function executeLoadYahooFinanceData(table, request) {
+    try {
+        const rows = await getYahooFinanceData(request);
+        await (0, loadArray_js_1.executePreparedArray)(table, (0, loadArray_js_1.prepareArray)(rows, COLUMN_TYPES));
+    }
+    catch (error) {
+        if (error instanceof SDAError_js_1.default) {
+            throw error;
+        }
+        throw new SDAError_js_1.default({
+            method: "loadYahooFinanceData()",
+            parameters: request,
+            query: "",
+            cause: error,
+        });
+    }
+}
+async function getYahooFinanceData(request) {
+    const { symbol, startDate, endDate, interval } = request;
+    const startTime = startDate.getTime();
+    const exclusiveEndTime = getExclusiveEndTime(endDate, interval);
+    const url = new URL(`https://query1.finance.yahoo.com/v8/finance/chart/${encodeURIComponent(symbol)}`);
+    url.search = new URLSearchParams({
+        includeAdjustedClose: "true",
+        interval,
+        period1: String(Math.floor(startTime / 1000)),
+        period2: String(Math.ceil(exclusiveEndTime / 1000)),
+    }).toString();
+    const response = await fetch(url, {
+        headers: { "User-Agent": USER_AGENT },
+    });
+    if (!response.ok) {
+        const text = (await response.text()).slice(0, 500).trim();
+        throw new Error(`Failed to fetch Yahoo Finance data: ${response.status} ${response.statusText}${text.length > 0 ? `. ${text}` : ""}. Yahoo may have changed, rate-limited, or disabled this undocumented endpoint.`);
+    }
+    const data = await response.json();
+    if (data.chart?.error) {
+        throw new Error(data.chart.error.description ??
+            "Yahoo Finance returned an unknown error.");
+    }
+    const result = data.chart?.result?.[0];
+    const timestamps = result?.timestamp;
+    if (!result || !timestamps?.length) {
+        throw new Error("No Yahoo Finance data found.");
+    }
+    const quote = result.indicators?.quote?.[0];
+    const adjustedClose = result.indicators?.adjclose?.[0]?.adjclose;
+    const rows = [];
+    for (let index = 0; index < timestamps.length; index++) {
+        const timestamp = timestamps[index] * 1000;
+        if (Number.isFinite(timestamp) && timestamp >= startTime &&
+            timestamp < exclusiveEndTime) {
+            rows.push({
+                datetime: new Date(timestamp),
+                open: getFiniteValue(quote?.open, index),
+                high: getFiniteValue(quote?.high, index),
+                low: getFiniteValue(quote?.low, index),
+                close: getFiniteValue(quote?.close, index),
+                adjustedClose: getFiniteValue(adjustedClose, index),
+                volume: getFiniteValue(quote?.volume, index),
+            });
+        }
+    }
+    if (rows.length === 0) {
+        throw new Error(`No Yahoo Finance data found for ${symbol}.`);
+    }
+    return rows;
+}
+function getFiniteValue(values, index) {
+    const value = values?.[index];
+    return typeof value === "number" && Number.isFinite(value) ? value : null;
+}
+function getExclusiveEndTime(endDate, interval) {
+    const end = new Date(endDate);
+    if (interval === "1d") {
+        end.setUTCHours(0, 0, 0, 0);
+    }
+    else if (interval === "1h") {
+        end.setUTCMinutes(0, 0, 0);
+    }
+    else {
+        end.setUTCSeconds(0, 0);
+    }
+    return end.getTime() + INTERVAL_SECONDS[interval] * 1000;
+}
+function assertArguments(symbol, startDate, endDate, interval) {
+    if (typeof symbol !== "string" || symbol.trim().length === 0) {
+        throw new TypeError("loadYahooFinanceData() symbol must be a non-empty string.");
+    }
+    if (!(startDate instanceof Date) || !(endDate instanceof Date) ||
+        !Number.isFinite(startDate.getTime()) || !Number.isFinite(endDate.getTime())) {
+        throw new RangeError("loadYahooFinanceData() startDate and endDate must be valid dates.");
+    }
+    if (endDate.getTime() < startDate.getTime()) {
+        throw new RangeError("loadYahooFinanceData() endDate must be equal to or later than startDate.");
+    }
+    if (interval !== "1d" && interval !== "1h" && interval !== "1m") {
+        throw new Error('loadYahooFinanceData() interval must be "1d", "1h", or "1m".');
+    }
+}

package/script/methods/rowProportions.d.ts

@@ -1,6 +1,7 @@
 import type SimpleTable from "../class/SimpleTable.js";
 export default function rowProportions(simpleTable: SimpleTable, columns: string[], options?: {
     suffix?: string;
+    base?: number;
     decimals?: number;
 }): void;
 //# sourceMappingURL=rowProportions.d.ts.map

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

@@ -1 +1 @@
-{"version":3,"file":"rowProportions.d.ts","sourceRoot":"","sources":["../../src/methods/rowProportions.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,WAAW,MAAM,yBAAyB,CAAC;AAEvD,MAAM,CAAC,OAAO,UAAU,cAAc,CACpC,WAAW,EAAE,WAAW,EACxB,OAAO,EAAE,MAAM,EAAE,EACjB,OAAO,GAAE;IACP,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,QAAQ,CAAC,EAAE,MAAM,CAAC;CACd,QAoCP"}
+{"version":3,"file":"rowProportions.d.ts","sourceRoot":"","sources":["../../src/methods/rowProportions.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,WAAW,MAAM,yBAAyB,CAAC;AAEvD,MAAM,CAAC,OAAO,UAAU,cAAc,CACpC,WAAW,EAAE,WAAW,EACxB,OAAO,EAAE,MAAM,EAAE,EACjB,OAAO,GAAE;IACP,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,QAAQ,CAAC,EAAE,MAAM,CAAC;CACd,QA4CP"}

package/script/methods/rowProportions.js

@@ -8,6 +8,10 @@ const quoteIdentifier_js_1 = __importDefault(require("../helpers/quoteIdentifier
 const assertNewColumns_js_1 = __importDefault(require("../helpers/assertNewColumns.js"));
 const queueOp_js_1 = __importDefault(require("../helpers/queueOp.js"));
 function rowProportions(simpleTable, columns, options = {}) {
+    if (options.base !== undefined &&
+        (!Number.isFinite(options.base) || options.base <= 0)) {
+        throw new Error("rowProportions() options.base must be a finite number greater than 0.");
+    }
     columns = [...columns];
     options = structuredClone(options);
     (0, queueOp_js_1.default)(simpleTable, {
@@ -20,7 +24,7 @@ function rowProportions(simpleTable, columns, options = {}) {
             (0, assertNewColumns_js_1.default)(schema, columns.map((col) => `${col}${suffix}`), "rowProportions()");
             let query = `SELECT *,`;
             for (const col of columns) {
-                const tempQuery = `${(0, quoteIdentifier_js_1.default)(col)} / (${columns.map((d) => `${(0, quoteIdentifier_js_1.default)(d)}`).join(" + ")})`;
+                const tempQuery = `${(0, quoteIdentifier_js_1.default)(col)} / (${columns.map((d) => `${(0, quoteIdentifier_js_1.default)(d)}`).join(" + ")}) * ${options.base ?? 1}`;
                 if (typeof options.decimals === "number") {
                     query += ` ROUND(${tempQuery}, ${options.decimals})`;
                 }