Files
family_budget/backend/src/services/vtbBrokerXlsx.ts

157 lines
8.6 KiB
TypeScript
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
import crypto from 'crypto';
import zlib from 'zlib';
import type { PortfolioFile, StatementFile } from '@family-budget/shared';
type Row = [number, Record<string, string>];
function xmlText(value: string): string {
return value.replace(/<[^>]+>/g, '').replace(/&amp;/g, '&').replace(/&lt;/g, '<').replace(/&gt;/g, '>').replace(/&#(d+);/g, (_, code) => String.fromCharCode(Number(code))).replace(/\s+/g, ' ').trim();
}
function zipFiles(buffer: Buffer): Map<string, Buffer> {
const end = buffer.lastIndexOf(Buffer.from([0x50, 0x4b, 0x05, 0x06]));
if (end < 0) throw new Error('Файл не является XLSX-архивом');
const files = new Map<string, Buffer>();
let offset = buffer.readUInt32LE(end + 16);
const count = buffer.readUInt16LE(end + 10);
for (let i = 0; i < count; i++) {
if (buffer.readUInt32LE(offset) !== 0x02014b50) throw new Error('Повреждён XLSX-архив');
const method = buffer.readUInt16LE(offset + 10);
const compressedSize = buffer.readUInt32LE(offset + 20);
const nameLength = buffer.readUInt16LE(offset + 28);
const extraLength = buffer.readUInt16LE(offset + 30);
const commentLength = buffer.readUInt16LE(offset + 32);
const localOffset = buffer.readUInt32LE(offset + 42);
const name = buffer.subarray(offset + 46, offset + 46 + nameLength).toString('utf8');
if (buffer.readUInt32LE(localOffset) !== 0x04034b50) throw new Error('Повреждён XLSX-архив');
const localNameLength = buffer.readUInt16LE(localOffset + 26);
const localExtraLength = buffer.readUInt16LE(localOffset + 28);
const data = buffer.subarray(localOffset + 30 + localNameLength + localExtraLength, localOffset + 30 + localNameLength + localExtraLength + compressedSize);
files.set(name, method === 0 ? data : method === 8 ? zlib.inflateRawSync(data) : (() => { throw new Error('Неподдерживаемое сжатие XLSX'); })());
offset += 46 + nameLength + extraLength + commentLength;
}
return files;
}
function readRows(buffer: Buffer): Row[] {
const files = zipFiles(buffer);
const sharedXml = files.get('xl/sharedStrings.xml')?.toString('utf8');
const sheetXml = files.get('xl/worksheets/sheet1.xml')?.toString('utf8');
if (!sharedXml || !sheetXml) throw new Error('В XLSX не найдены данные отчёта');
const shared = [...sharedXml.matchAll(/<si>([\s\S]*?)<\/si>/g)].map((match) => xmlText(match[1]));
return [...sheetXml.matchAll(/<row[^>]*\br="(\d+)"[^>]*>([\s\S]*?)<\/row>/g)].map((row) => {
const cells: Record<string, string> = {};
for (const cell of row[2].matchAll(/<c[^>]*\br="([A-Z]+)\d+"([^>]*)>([\s\S]*?)<\/c>/g)) {
const value = cell[3].match(/<v>([\s\S]*?)<\/v>/)?.[1] ?? '';
cells[cell[1]] = /\bt="s"/.test(cell[2]) && value ? shared[Number(value)] : xmlText(value);
}
return [Number(row[1]), cells];
});
}
function text(value: unknown): string {
return String(value ?? '').replace(/\s+/g, ' ').trim();
}
function rowText(cells: Record<string, string>): string {
return text(Object.values(cells).join(' '));
}
function findRow(rows: Row[], phrase: string, start = 0): number {
const index = rows.findIndex(([, cells], i) => i >= start && rowText(cells).toLowerCase().includes(phrase.toLowerCase()));
if (index < 0) throw new Error(`Не найден раздел: ${phrase}`);
return index;
}
function excelDate(value: string | undefined): string | null {
if (!value) return null;
const date = new Date(Date.UTC(1899, 11, 30) + Number(value) * 86_400_000);
if (Number.isNaN(date.valueOf())) return null;
return `${date.toISOString().slice(0, 19)}+03:00`;
}
function kopecks(value: string | undefined): number {
return value ? Math.round(Number(value.replace(',', '.')) * 100) : 0;
}
function metadata(rows: Row[]): { account: string; period: [string, string]; reportedAt: string | null } {
const period = rows.map(([, cells]) => rowText(cells)).join(' ').match(/период с (\d{2}\.\d{2}\.\d{4}) по (\d{2}\.\d{2}\.\d{4})/i);
let account: string | null = null;
let reportedAt: string | null = null;
for (const [, cells] of rows.slice(0, 35)) {
const joined = rowText(cells);
account ??= joined.match(/(\d{20})\s*\(RUR\)/)?.[1] ?? null;
if (joined.includes('Дата формирования отчета')) {
reportedAt ??= Object.values(cells).map(excelDate).find(Boolean) ?? null;
}
}
if (!account || !period) throw new Error('Не удалось определить счёт или период отчёта');
return { account, period: [period[1], period[2]], reportedAt };
}
function isoDate(value: string): string {
const [day, month, year] = value.split('.');
return `${year}-${month}-${day}`;
}
export function convertVtbBrokerXlsx(buffer: Buffer): { cash: StatementFile; portfolio: PortfolioFile } {
const rows = readRows(buffer);
const { account, period, reportedAt } = metadata(rows);
const holdingsStart = findRow(rows, 'Отчёт об остатках ценных бумаг');
const movementStart = findRow(rows, 'Движение ценных бумаг', holdingsStart);
const cashStart = findRow(rows, 'Движение денежных средств');
const tradesStart = findRow(rows, 'Заключенные в отчетном периоде сделки с ценными бумагами');
const tradesEnd = findRow(rows, 'Завершенные в отчетном периоде сделки с ценными бумагами', tradesStart + 1);
const occurrences = new Map<string, number>();
const transactions = rows.slice(cashStart + 1, holdingsStart).flatMap(([, cells]) => {
if (!cells.B || !cells.C || !cells.J) return [];
const operationAt = excelDate(cells.B);
if (!operationAt) return [];
const amountSigned = kopecks(cells.C);
const description = text(`${text(cells.J)}. ${text(cells.P)}`.replace(/\. $/, ''));
const digest = crypto.createHash('sha256').update(`${operationAt}|${amountSigned}|${description}`).digest('hex').slice(0, 16);
const occurrence = (occurrences.get(digest) ?? 0) + 1;
occurrences.set(digest, occurrence);
return [{ operationAt, amountSigned, commission: 0, description, sourceId: `vtb-broker-cash:${digest}:${occurrence}` }];
});
if (!transactions.length) throw new Error('Операции движения денежных средств не найдены');
const balanceStart = findRow(rows, 'Отчёт об остатках денежных средств');
const openingBalance = kopecks(rows[balanceStart + 3]?.[1].L);
const closingBalance = kopecks(rows[balanceStart + 3]?.[1].AF);
if (openingBalance + transactions.reduce((sum, tx) => sum + tx.amountSigned, 0) !== closingBalance) throw new Error('Баланс cash-операций не сходится с отчётом');
const positions = rows.slice(holdingsStart + 1, movementStart).flatMap(([, cells]) => {
const instrument = text(cells.B);
if (!instrument || instrument.toLowerCase().startsWith('итого') || !/RU[A-Z0-9]{10}/.test(instrument)) return [];
return [{ instrument, isin: instrument.split(', ').find((part) => /^RU[A-Z0-9]{10}$/.test(part)) ?? null, quantity: cells.L || cells.M || cells.I || cells.J, price: cells.P || null, valuation: cells.AF || cells.AJ || null }];
});
const trades = rows.slice(tradesStart + 1, tradesEnd).flatMap(([number, cells]) => {
if (!cells.B || !cells.C || !cells.F) return [];
const concludedAt = excelDate(cells.C);
if (!concludedAt) return [];
const tradeId = text(cells.Z);
return [{
sourceId: `vtb-broker-trade:${tradeId || number}`,
instrument: text(cells.B),
isin: text(cells.B).split(', ').find((part) => /^RU[A-Z0-9]{10}$/.test(part)) ?? null,
concludedAt,
side: text(cells.F),
quantity: text(cells.H),
priceCurrency: text(cells.I) || null,
price: text(cells.J) || null,
settlementCurrency: text(cells.L) || null,
settlementAmount: text(cells.M) || null,
nkd: text(cells.O) || null,
settlementCommission: text(cells.P) || null,
tradeCommission: text(cells.R) || null,
orderId: text(cells.W) || null,
tradeId: tradeId || null,
venue: text(cells.AK) || null,
comment: text(cells.AN) || null,
}];
});
return {
cash: { schemaVersion: '1.0', bank: 'VTB_BROKER', statement: { accountNumber: account, currency: 'RUB', openingBalance, closingBalance, exportedAt: reportedAt ?? transactions[transactions.length - 1].operationAt }, transactions },
portfolio: { schemaVersion: 'broker-portfolio-1.0', bank: 'VTB_BROKER', accountNumber: account, reportPeriod: { from: isoDate(period[0]), to: isoDate(period[1]) }, reportedAt, positions, trades },
};
}