@js-ak/excel-toolbox 1.2.6 → 1.2.7

package/build/esm/lib/zip/utils.js

@@ -1,152 +0,0 @@
-import { Buffer } from "node:buffer";
-/**
- * Precomputed CRC-32 lookup table for optimized checksum calculation.
- * The table is generated using the standard IEEE 802.3 (Ethernet) polynomial:
- * 0xEDB88320 (reversed representation of 0x04C11DB7).
- *
- * The table is immediately invoked and cached as a constant for performance,
- * following the common implementation pattern for CRC algorithms.
- */
-const crcTable = (() => {
-    // Create a typed array for better performance with 256 32-bit unsigned integers
-    const table = new Uint32Array(256);
-    // Generate table entries for all possible byte values (0-255)
-    for (let i = 0; i < 256; i++) {
-        let crc = i; // Initialize with current byte value
-        // Process each bit (8 times)
-        for (let j = 0; j < 8; j++) {
-            /*
-             * CRC division algorithm:
-             * 1. If LSB is set (crc & 1), XOR with polynomial
-             * 2. Right-shift by 1 (unsigned)
-             *
-             * The polynomial 0xEDB88320 is:
-             * - Bit-reversed version of 0x04C11DB7
-             * - Uses reflected input/output algorithm
-             */
-            crc = crc & 1
-                ? 0xedb88320 ^ (crc >>> 1) // XOR with polynomial if LSB is set
-                : crc >>> 1; // Just shift right if LSB is not set
-        }
-        // Store final 32-bit value (>>> 0 ensures unsigned 32-bit representation)
-        table[i] = crc >>> 0;
-    }
-    return table;
-})();
-/**
- * Computes a CRC-32 checksum for the given Buffer using the standard IEEE 802.3 polynomial.
- * This implementation uses a precomputed lookup table for optimal performance.
- *
- * The algorithm follows these characteristics:
- * - Polynomial: 0xEDB88320 (reversed representation of 0x04C11DB7)
- * - Initial value: 0xFFFFFFFF (inverted by ~0)
- * - Final XOR value: 0xFFFFFFFF (achieved by inverting the result)
- * - Input and output reflection: Yes
- *
- * @param {Buffer} buf - The input buffer to calculate checksum for
- * @returns {number} - The 32-bit unsigned CRC-32 checksum (0x00000000 to 0xFFFFFFFF)
- */
-export function crc32(buf) {
-    // Initialize CRC with all 1's (0xFFFFFFFF) using bitwise NOT
-    let crc = ~0;
-    // Process each byte in the buffer
-    for (let i = 0; i < buf.length; i++) {
-        /*
-         * CRC update algorithm steps:
-         * 1. XOR current CRC with next byte (lowest 8 bits)
-         * 2. Use result as index in precomputed table (0-255)
-         * 3. XOR the table value with right-shifted CRC (8 bits)
-         *
-         * The operation breakdown:
-         * - (crc ^ buf[i]) - XOR with next byte
-         * - & 0xff - Isolate lowest 8 bits
-         * - crc >>> 8 - Shift CRC right by 8 bits (unsigned)
-         * - ^ crcTable[...] - XOR with precomputed table value
-         */
-        crc = (crc >>> 8) ^ crcTable[(crc ^ buf[i]) & 0xff];
-    }
-    /*
-     * Final processing:
-     * 1. Invert all bits (~crc) to match standard CRC-32 output
-     * 2. Convert to unsigned 32-bit integer (>>> 0)
-     */
-    return ~crc >>> 0;
-}
-/**
- * Converts a JavaScript Date object to a 4-byte Buffer in MS-DOS date/time format
- * as specified in the ZIP file format specification (PKZIP APPNOTE.TXT).
- *
- * The MS-DOS date/time format packs both date and time into 4 bytes (32 bits) with
- * the following bit layout:
- *
- * Time portion (2 bytes/16 bits):
- * - Bits 00-04: Seconds divided by 2 (0-29, representing 0-58 seconds)
- * - Bits 05-10: Minutes (0-59)
- * - Bits 11-15: Hours (0-23)
- *
- * Date portion (2 bytes/16 bits):
- * - Bits 00-04: Day (1-31)
- * - Bits 05-08: Month (1-12)
- * - Bits 09-15: Year offset from 1980 (0-127, representing 1980-2107)
- *
- * @param {Date} date - The JavaScript Date object to convert
- * @returns {Buffer} - 4-byte Buffer containing:
- *                    - Bytes 0-1: DOS time (hours, minutes, seconds/2)
- *                    - Bytes 2-3: DOS date (year-1980, month, day)
- * @throws {RangeError} - If the date is before 1980 or after 2107
- */
-export function dosTime(date) {
-    // Pack time components into 2 bytes (16 bits):
-    // - Hours (5 bits) shifted left 11 positions (bits 11-15)
-    // - Minutes (6 bits) shifted left 5 positions (bits 5-10)
-    // - Seconds/2 (5 bits) in least significant bits (bits 0-4)
-    const time = (date.getHours() << 11) | // Hours occupy bits 11-15
-        (date.getMinutes() << 5) | // Minutes occupy bits 5-10
-        (Math.floor(date.getSeconds() / 2)); // Seconds/2 occupy bits 0-4
-    // Pack date components into 2 bytes (16 bits):
-    // - (Year-1980) (7 bits) shifted left 9 positions (bits 9-15)
-    // - Month (4 bits) shifted left 5 positions (bits 5-8)
-    // - Day (5 bits) in least significant bits (bits 0-4)
-    const day = ((date.getFullYear() - 1980) << 9) | // Years since 1980 (bits 9-15)
-        ((date.getMonth() + 1) << 5) | // Month 1-12 (bits 5-8)
-        date.getDate(); // Day 1-31 (bits 0-4)
-    // Combine both 2-byte values into a single 4-byte Buffer
-    // Note: Using little-endian byte order for each 2-byte segment
-    return Buffer.from([
-        ...toBytes(time, 2), // Convert time to 2 bytes (LSB first)
-        ...toBytes(day, 2), // Convert date to 2 bytes (LSB first)
-    ]);
-}
-/**
- * Converts a numeric value into a fixed-length Buffer representation,
- * storing the value in little-endian format with right-padding of zeros.
- *
- * This is particularly useful for binary protocols or file formats that
- * require fixed-width numeric fields.
- *
- * @param {number} value - The numeric value to convert to bytes.
- *        Note: JavaScript numbers are IEEE 754 doubles, but only the
- *        integer portion will be used (up to 53-bit precision).
- * @param {number} len - The desired length of the output Buffer in bytes.
- *        Must be a positive integer.
- * @returns {Buffer} - A new Buffer of exactly `len` bytes containing:
- *        1. The value's bytes in little-endian order (least significant byte first)
- *        2. Zero padding in any remaining higher-order bytes
- * @throws {RangeError} - If the value requires more bytes than `len` to represent
- *        (though this is currently not explicitly checked)
- */
-export function toBytes(value, len) {
-    // Allocate a new Buffer of the requested length, automatically zero-filled
-    const buf = Buffer.alloc(len);
-    // Process each byte position from least significant to most significant
-    for (let i = 0; i < len; i++) {
-        // Store the least significant byte of the current value
-        buf[i] = value & 0xff; // Mask to get bottom 8 bits
-        // Right-shift the value by 8 bits to process the next byte
-        // Note: This uses unsigned right shift (>>> would be signed)
-        value >>= 8;
-        // If the loop completes with value != 0, we've overflowed the buffer length,
-        // but this isn't currently checked/handled
-    }
-    return buf;
-}

package/build/types/lib/zip/utils.d.ts

@@ -1,58 +0,0 @@
-import { Buffer } from "node:buffer";
-/**
- * Computes a CRC-32 checksum for the given Buffer using the standard IEEE 802.3 polynomial.
- * This implementation uses a precomputed lookup table for optimal performance.
- *
- * The algorithm follows these characteristics:
- * - Polynomial: 0xEDB88320 (reversed representation of 0x04C11DB7)
- * - Initial value: 0xFFFFFFFF (inverted by ~0)
- * - Final XOR value: 0xFFFFFFFF (achieved by inverting the result)
- * - Input and output reflection: Yes
- *
- * @param {Buffer} buf - The input buffer to calculate checksum for
- * @returns {number} - The 32-bit unsigned CRC-32 checksum (0x00000000 to 0xFFFFFFFF)
- */
-export declare function crc32(buf: Buffer): number;
-/**
- * Converts a JavaScript Date object to a 4-byte Buffer in MS-DOS date/time format
- * as specified in the ZIP file format specification (PKZIP APPNOTE.TXT).
- *
- * The MS-DOS date/time format packs both date and time into 4 bytes (32 bits) with
- * the following bit layout:
- *
- * Time portion (2 bytes/16 bits):
- * - Bits 00-04: Seconds divided by 2 (0-29, representing 0-58 seconds)
- * - Bits 05-10: Minutes (0-59)
- * - Bits 11-15: Hours (0-23)
- *
- * Date portion (2 bytes/16 bits):
- * - Bits 00-04: Day (1-31)
- * - Bits 05-08: Month (1-12)
- * - Bits 09-15: Year offset from 1980 (0-127, representing 1980-2107)
- *
- * @param {Date} date - The JavaScript Date object to convert
- * @returns {Buffer} - 4-byte Buffer containing:
- *                    - Bytes 0-1: DOS time (hours, minutes, seconds/2)
- *                    - Bytes 2-3: DOS date (year-1980, month, day)
- * @throws {RangeError} - If the date is before 1980 or after 2107
- */
-export declare function dosTime(date: Date): Buffer;
-/**
- * Converts a numeric value into a fixed-length Buffer representation,
- * storing the value in little-endian format with right-padding of zeros.
- *
- * This is particularly useful for binary protocols or file formats that
- * require fixed-width numeric fields.
- *
- * @param {number} value - The numeric value to convert to bytes.
- *        Note: JavaScript numbers are IEEE 754 doubles, but only the
- *        integer portion will be used (up to 53-bit precision).
- * @param {number} len - The desired length of the output Buffer in bytes.
- *        Must be a positive integer.
- * @returns {Buffer} - A new Buffer of exactly `len` bytes containing:
- *        1. The value's bytes in little-endian order (least significant byte first)
- *        2. Zero padding in any remaining higher-order bytes
- * @throws {RangeError} - If the value requires more bytes than `len` to represent
- *        (though this is currently not explicitly checked)
- */
-export declare function toBytes(value: number, len: number): Buffer;