mirror of
https://github.com/escalante29/WealthySmart.git
synced 2026-05-19 12:28:47 +02:00
All checks were successful
Deploy to VPS / deploy (push) Successful in 14s
The bank website puts field labels and amounts on separate lines. Parser now handles both inline and split formats. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
180 lines
5.2 KiB
TypeScript
180 lines
5.2 KiB
TypeScript
export interface PensionParsedEntry {
|
|
fund: string;
|
|
period_start: string; // YYYY-MM-DD
|
|
period_end: string;
|
|
saldo_anterior: number;
|
|
aportes: number;
|
|
rendimientos: number;
|
|
retiros: number;
|
|
traslados: number;
|
|
comision: number;
|
|
correccion: number;
|
|
bonificacion: number;
|
|
saldo_final: number;
|
|
}
|
|
|
|
function parseAmount(raw: string): number {
|
|
const cleaned = raw.replace(/[¢\s]/g, '').replace(/,/g, '');
|
|
const num = parseFloat(cleaned);
|
|
return isNaN(num) ? 0 : num;
|
|
}
|
|
|
|
function parseDateDMY(raw: string): string {
|
|
const m = raw.match(/(\d{2})\/(\d{2})\/(\d{4})/);
|
|
if (!m) return '';
|
|
return `${m[3]}-${m[2]}-${m[1]}`;
|
|
}
|
|
|
|
function extractAmounts(line: string): number[] {
|
|
const matches = line.match(/¢\s*-?[\d,.]+/g);
|
|
if (!matches) return [];
|
|
return matches.map(parseAmount);
|
|
}
|
|
|
|
// Field labels in the order they appear in the bank statement
|
|
const FIELD_LABELS: [RegExp, string][] = [
|
|
[/saldo\s*anterior/i, 'saldo_anterior'],
|
|
[/aportes/i, 'aportes'],
|
|
[/rendimientos/i, 'rendimientos'],
|
|
[/retiros/i, 'retiros'],
|
|
[/traslados/i, 'traslados'],
|
|
[/comisi[oó]n/i, 'comision'],
|
|
[/bonificaci[oó]n/i, 'bonificacion'],
|
|
];
|
|
|
|
interface BlockResult {
|
|
funds: string[];
|
|
fields: Record<string, number[]>;
|
|
period_start: string;
|
|
period_end: string;
|
|
}
|
|
|
|
function parseBlock(lines: string[]): BlockResult | null {
|
|
const result: BlockResult = {
|
|
funds: [],
|
|
fields: {},
|
|
period_start: '',
|
|
period_end: '',
|
|
};
|
|
|
|
// Detect fund columns from header
|
|
const headerLine = lines.find((l) => /resumen del per[ií]odo/i.test(l));
|
|
if (!headerLine) return null;
|
|
|
|
if (/\bROP\b/i.test(headerLine) && /\bFCL\b/i.test(headerLine)) {
|
|
result.funds = ['ROP', 'FCL'];
|
|
} else if (/voluntario/i.test(headerLine) || /\bVOL\b/i.test(headerLine)) {
|
|
result.funds = ['VOL'];
|
|
} else {
|
|
return null;
|
|
}
|
|
|
|
// Strategy 1: Try same-line parsing (label + amounts on same line)
|
|
// Strategy 2: Collect standalone amount lines for split-format parsing
|
|
const detectedFieldOrder: string[] = [];
|
|
const standaloneAmounts: number[] = [];
|
|
|
|
for (const line of lines) {
|
|
// Check for period
|
|
const periodMatch = line.match(/del\s+(\d{2}\/\d{2}\/\d{4})\s+al\s+(\d{2}\/\d{2}\/\d{4})/i);
|
|
if (periodMatch) {
|
|
result.period_start = parseDateDMY(periodMatch[1]);
|
|
result.period_end = parseDateDMY(periodMatch[2]);
|
|
continue;
|
|
}
|
|
|
|
// Check for "Saldo Actual" line (always has amounts inline)
|
|
if (/saldo\s*actual/i.test(line)) {
|
|
const amounts = extractAmounts(line);
|
|
if (amounts.length > 0) {
|
|
result.fields['saldo_final'] = amounts;
|
|
}
|
|
continue;
|
|
}
|
|
|
|
// Check if this line matches a field label
|
|
let matchedLabel = false;
|
|
for (const [regex, key] of FIELD_LABELS) {
|
|
if (regex.test(line)) {
|
|
matchedLabel = true;
|
|
const amounts = extractAmounts(line);
|
|
if (amounts.length > 0) {
|
|
// Strategy 1: amounts on same line as label
|
|
result.fields[key] = amounts;
|
|
} else {
|
|
// Strategy 2: label-only line, record the order
|
|
detectedFieldOrder.push(key);
|
|
}
|
|
break;
|
|
}
|
|
}
|
|
|
|
// If not a label line, check if it's a standalone amount line
|
|
if (!matchedLabel) {
|
|
const amounts = extractAmounts(line);
|
|
if (amounts.length === 1) {
|
|
standaloneAmounts.push(amounts[0]);
|
|
}
|
|
}
|
|
}
|
|
|
|
// If we have standalone amounts and field labels, map them
|
|
// Format: N labels, then N amounts for fund1, then N amounts for fund2, ...
|
|
if (detectedFieldOrder.length > 0 && standaloneAmounts.length > 0) {
|
|
const numFields = detectedFieldOrder.length;
|
|
const numFunds = result.funds.length;
|
|
|
|
if (standaloneAmounts.length >= numFields * numFunds) {
|
|
for (let f = 0; f < numFunds; f++) {
|
|
for (let i = 0; i < numFields; i++) {
|
|
const key = detectedFieldOrder[i];
|
|
if (!result.fields[key]) result.fields[key] = [];
|
|
result.fields[key].push(standaloneAmounts[f * numFields + i]);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
return result;
|
|
}
|
|
|
|
export function parsePensionPaste(text: string): PensionParsedEntry[] {
|
|
// Split into blocks by "---" or multiple blank lines
|
|
const blocks = text.split(/(?:^|\n)-{3,}(?:\n|$)|\n{3,}/);
|
|
const entries: PensionParsedEntry[] = [];
|
|
|
|
for (const block of blocks) {
|
|
const lines = block.split('\n').filter((l) => l.trim());
|
|
if (lines.length < 3) continue;
|
|
|
|
const parsed = parseBlock(lines);
|
|
if (!parsed || !parsed.period_start || !parsed.period_end) continue;
|
|
|
|
for (let i = 0; i < parsed.funds.length; i++) {
|
|
const fund = parsed.funds[i];
|
|
const get = (key: string): number => {
|
|
const vals = parsed.fields[key];
|
|
if (!vals) return 0;
|
|
return vals[i] ?? vals[0] ?? 0;
|
|
};
|
|
|
|
entries.push({
|
|
fund,
|
|
period_start: parsed.period_start,
|
|
period_end: parsed.period_end,
|
|
saldo_anterior: get('saldo_anterior'),
|
|
aportes: get('aportes'),
|
|
rendimientos: get('rendimientos'),
|
|
retiros: get('retiros'),
|
|
traslados: get('traslados'),
|
|
comision: get('comision'),
|
|
correccion: 0,
|
|
bonificacion: get('bonificacion'),
|
|
saldo_final: get('saldo_final'),
|
|
});
|
|
}
|
|
}
|
|
|
|
return entries;
|
|
}
|