india2actual 0.1.0

package/dist/env-file.js

@@ -0,0 +1,40 @@
+import { env, loadEnvFile } from 'node:process';
+/**
+ * A missing file is not an error: the CSV path needs no configuration, so most
+ * runs legitimately have no `.env` at all. Anything else — unreadable, a
+ * directory, malformed contents — is reported, because it means the user
+ * believes they have configured something that is in fact being ignored. That
+ * would otherwise surface much later as a confusing "ACTUAL_SERVER_URL is not
+ * set".
+ */
+export function loadEnvironmentFile(path) {
+    try {
+        loadEnvFile(path);
+        return { loaded: true, path };
+    }
+    catch (error) {
+        if (isMissingFile(error)) {
+            return { loaded: false, path, reason: 'missing' };
+        }
+        throw new Error(`Could not read ${path}: ${error instanceof Error ? error.message : String(error)}`);
+    }
+}
+/**
+ * Read a setting, treating blank as absent.
+ *
+ * `.env.example` ships its keys with empty values for the user to fill in, so
+ * a copied-but-unedited file leaves `ACTUAL_DATA_DIR=` set to the empty
+ * string. The empty string is not nullish, so a `??` default would not fire
+ * and the Actual API would be handed `''` as its data directory. Trimming also
+ * catches the trailing space left by `ACTUAL_SYNC_ID= ` .
+ */
+export function setting(name) {
+    return env[name]?.trim() || undefined;
+}
+function isMissingFile(error) {
+    return (typeof error === 'object' &&
+        error !== null &&
+        'code' in error &&
+        error.code === 'ENOENT');
+}
+//# sourceMappingURL=env-file.js.map

package/dist/env-file.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"env-file.js","sourceRoot":"","sources":["../src/env-file.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,GAAG,EAAE,WAAW,EAAE,MAAM,cAAc,CAAC;AAkBhD;;;;;;;GAOG;AACH,MAAM,UAAU,mBAAmB,CAAC,IAAY;IAC9C,IAAI,CAAC;QACH,WAAW,CAAC,IAAI,CAAC,CAAC;QAClB,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC;IAChC,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,IAAI,aAAa,CAAC,KAAK,CAAC,EAAE,CAAC;YACzB,OAAO,EAAE,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,MAAM,EAAE,SAAS,EAAE,CAAC;QACpD,CAAC;QAED,MAAM,IAAI,KAAK,CACb,kBAAkB,IAAI,KAAK,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,EAAE,CACpF,CAAC;IACJ,CAAC;AACH,CAAC;AAED;;;;;;;;GAQG;AACH,MAAM,UAAU,OAAO,CAAC,IAAY;IAClC,OAAO,GAAG,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,SAAS,CAAC;AACxC,CAAC;AAED,SAAS,aAAa,CAAC,KAAc;IACnC,OAAO,CACL,OAAO,KAAK,KAAK,QAAQ;QACzB,KAAK,KAAK,IAAI;QACd,MAAM,IAAI,KAAK;QACd,KAA4B,CAAC,IAAI,KAAK,QAAQ,CAChD,CAAC;AACJ,CAAC"}

package/dist/extract/csv.d.ts

@@ -0,0 +1,13 @@
+import type { Table } from './types.js';
+/**
+ * Guess the delimiter by counting candidates across the first few lines.
+ *
+ * Counting per-line and taking the most *consistent* candidate rather than the
+ * most frequent one matters here: Indian statements carry preamble lines
+ * (branch address, account holder name) whose punctuation would otherwise
+ * outvote the real delimiter.
+ */
+export declare function sniffDelimiter(contents: string): string;
+export declare function extractCsv(path: string, options?: {
+    delimiter?: string;
+}): Promise<Table>;

package/dist/extract/csv.js

@@ -0,0 +1,60 @@
+import { readFile } from 'node:fs/promises';
+import { parse } from 'csv-parse/sync';
+const CANDIDATE_DELIMITERS = [',', '\t', ';', '|'];
+/**
+ * Guess the delimiter by counting candidates across the first few lines.
+ *
+ * Counting per-line and taking the most *consistent* candidate rather than the
+ * most frequent one matters here: Indian statements carry preamble lines
+ * (branch address, account holder name) whose punctuation would otherwise
+ * outvote the real delimiter.
+ */
+export function sniffDelimiter(contents) {
+    const lines = contents
+        .split(/\r?\n/)
+        .filter(line => line.trim())
+        .slice(0, 30);
+    let best = ',';
+    let bestScore = -1;
+    for (const delimiter of CANDIDATE_DELIMITERS) {
+        const counts = lines
+            .map(line => line.split(delimiter).length - 1)
+            .filter(count => count > 0);
+        if (counts.length < 2) {
+            continue;
+        }
+        // Reward the delimiter that appears the same number of times on the most
+        // lines — that is the signature of a real column structure.
+        const tally = new Map();
+        for (const count of counts) {
+            tally.set(count, (tally.get(count) ?? 0) + 1);
+        }
+        for (const [count, lineCount] of tally) {
+            const score = count * lineCount;
+            if (score > bestScore) {
+                bestScore = score;
+                best = delimiter;
+            }
+        }
+    }
+    return best;
+}
+export async function extractCsv(path, options = {}) {
+    const contents = await readFile(path, 'utf8');
+    const delimiter = options.delimiter ?? sniffDelimiter(contents);
+    const rows = parse(contents, {
+        delimiter,
+        bom: true,
+        // Statement preambles and footers have different column counts than the
+        // transaction table, so ragged rows are expected, not an error.
+        relax_column_count: true,
+        relax_quotes: true,
+        skip_empty_lines: true,
+        trim: true,
+    });
+    return {
+        rows: rows.map(row => row.map(cell => (cell ?? '').toString().trim())),
+        source: { path, format: 'csv' },
+    };
+}
+//# sourceMappingURL=csv.js.map

package/dist/extract/csv.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"csv.js","sourceRoot":"","sources":["../../src/extract/csv.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAE5C,OAAO,EAAE,KAAK,EAAE,MAAM,gBAAgB,CAAC;AAIvC,MAAM,oBAAoB,GAAG,CAAC,GAAG,EAAE,IAAI,EAAE,GAAG,EAAE,GAAG,CAAU,CAAC;AAE5D;;;;;;;GAOG;AACH,MAAM,UAAU,cAAc,CAAC,QAAgB;IAC7C,MAAM,KAAK,GAAG,QAAQ;SACnB,KAAK,CAAC,OAAO,CAAC;SACd,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC;SAC3B,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;IAEhB,IAAI,IAAI,GAAG,GAAG,CAAC;IACf,IAAI,SAAS,GAAG,CAAC,CAAC,CAAC;IAEnB,KAAK,MAAM,SAAS,IAAI,oBAAoB,EAAE,CAAC;QAC7C,MAAM,MAAM,GAAG,KAAK;aACjB,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC;aAC7C,MAAM,CAAC,KAAK,CAAC,EAAE,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC;QAE9B,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YACtB,SAAS;QACX,CAAC;QAED,yEAAyE;QACzE,4DAA4D;QAC5D,MAAM,KAAK,GAAG,IAAI,GAAG,EAAkB,CAAC;QACxC,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;YAC3B,KAAK,CAAC,GAAG,CAAC,KAAK,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QAChD,CAAC;QAED,KAAK,MAAM,CAAC,KAAK,EAAE,SAAS,CAAC,IAAI,KAAK,EAAE,CAAC;YACvC,MAAM,KAAK,GAAG,KAAK,GAAG,SAAS,CAAC;YAChC,IAAI,KAAK,GAAG,SAAS,EAAE,CAAC;gBACtB,SAAS,GAAG,KAAK,CAAC;gBAClB,IAAI,GAAG,SAAS,CAAC;YACnB,CAAC;QACH,CAAC;IACH,CAAC;IAED,OAAO,IAAI,CAAC;AACd,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,UAAU,CAC9B,IAAY,EACZ,UAAkC,EAAE;IAEpC,MAAM,QAAQ,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC;IAC9C,MAAM,SAAS,GAAG,OAAO,CAAC,SAAS,IAAI,cAAc,CAAC,QAAQ,CAAC,CAAC;IAEhE,MAAM,IAAI,GAAG,KAAK,CAAC,QAAQ,EAAE;QAC3B,SAAS;QACT,GAAG,EAAE,IAAI;QACT,wEAAwE;QACxE,gEAAgE;QAChE,kBAAkB,EAAE,IAAI;QACxB,YAAY,EAAE,IAAI;QAClB,gBAAgB,EAAE,IAAI;QACtB,IAAI,EAAE,IAAI;KACX,CAAe,CAAC;IAEjB,OAAO;QACL,IAAI,EAAE,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,IAAI,EAAE,CAAC,CAAC;QACtE,MAAM,EAAE,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE;KAChC,CAAC;AACJ,CAAC"}

package/dist/extract/html-table.d.ts

@@ -0,0 +1,3 @@
+import type { Table } from './types.js';
+export declare function tableFromHtml(html: string, path?: string): Table;
+export declare function extractHtmlTable(path: string): Promise<Table>;

package/dist/extract/html-table.js

@@ -0,0 +1,77 @@
+import { readFile } from 'node:fs/promises';
+import { parse } from 'node-html-parser';
+const ENTITIES = {
+    ' ': ' ',
+    '&': '&',
+    '&lt;': '<',
+    '&gt;': '>',
+    '&quot;': '"',
+    '&apos;': "'",
+    '&#39;': "'",
+};
+/**
+ * Bank exports lean on `&nbsp;` for empty cells and `&amp;` in merchant names,
+ * and an undecoded `&nbsp;` is not whitespace — it would survive a `.trim()`
+ * and make an empty cell look populated.
+ */
+function decodeEntities(value) {
+    return value
+        .replace(/&nbsp;|&amp;|&lt;|&gt;|&quot;|&apos;|&#39;/g, match => ENTITIES[match] ?? match)
+        .replace(/&#(\d+);/g, (_, code) => String.fromCodePoint(Number(code)))
+        .replace(/&#x([0-9a-f]+);/gi, (_, code) => String.fromCodePoint(Number.parseInt(code, 16)));
+}
+function cellsOf(row) {
+    const cells = [];
+    for (const node of row.childNodes) {
+        const element = node;
+        const tag = element.tagName?.toUpperCase();
+        if (tag !== 'TD' && tag !== 'TH') {
+            continue;
+        }
+        const text = decodeEntities(element.textContent ?? '')
+            .replace(/\s+/g, ' ')
+            .trim();
+        cells.push(text);
+        // Expand colspan so later columns keep their index. Without this a
+        // merged header cell shifts every column after it by one, and amounts get
+        // read out of the wrong column.
+        const span = Number.parseInt(element.getAttribute('colspan') ?? '1', 10);
+        for (let extra = 1; extra < (Number.isFinite(span) ? span : 1); extra += 1) {
+            cells.push('');
+        }
+    }
+    return cells;
+}
+function rowsOf(table) {
+    return table.querySelectorAll('tr').map(cellsOf);
+}
+/** Rows with enough populated cells to plausibly be transaction data. */
+function density(rows) {
+    return rows.filter(row => row.filter(cell => cell.trim()).length >= 3).length;
+}
+export function tableFromHtml(html, path = 'inline') {
+    const root = parse(html);
+    // Bank exports nest the data table inside layout tables. No special handling
+    // is needed for that: `querySelectorAll('tr')` walks descendants, so an
+    // outer table's row list already contains the inner table's rows, each
+    // parsed from its own cells. Picking the table with the most dense rows
+    // therefore lands on a row set containing the real data, and the surrounding
+    // layout rows are dropped later by requiring a parseable date and amount.
+    const candidates = root.querySelectorAll('table');
+    let best = [];
+    let bestDensity = -1;
+    for (const candidate of candidates) {
+        const rows = rowsOf(candidate);
+        const score = density(rows);
+        if (score > bestDensity) {
+            bestDensity = score;
+            best = rows;
+        }
+    }
+    return { rows: best, source: { path, format: 'xlsx', part: 'html-table' } };
+}
+export async function extractHtmlTable(path) {
+    const contents = await readFile(path, 'utf8');
+    return tableFromHtml(contents, path);
+}
+//# sourceMappingURL=html-table.js.map

package/dist/extract/html-table.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"html-table.js","sourceRoot":"","sources":["../../src/extract/html-table.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAE5C,OAAO,EAAE,KAAK,EAAE,MAAM,kBAAkB,CAAC;AAKzC,MAAM,QAAQ,GAA2B;IACvC,QAAQ,EAAE,GAAG;IACb,OAAO,EAAE,GAAG;IACZ,MAAM,EAAE,GAAG;IACX,MAAM,EAAE,GAAG;IACX,QAAQ,EAAE,GAAG;IACb,QAAQ,EAAE,GAAG;IACb,OAAO,EAAE,GAAG;CACb,CAAC;AAEF;;;;GAIG;AACH,SAAS,cAAc,CAAC,KAAa;IACnC,OAAO,KAAK;SACT,OAAO,CACN,6CAA6C,EAC7C,KAAK,CAAC,EAAE,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,CAClC;SACA,OAAO,CAAC,WAAW,EAAE,CAAC,CAAC,EAAE,IAAY,EAAE,EAAE,CACxC,MAAM,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,CACnC;SACA,OAAO,CAAC,mBAAmB,EAAE,CAAC,CAAC,EAAE,IAAY,EAAE,EAAE,CAChD,MAAM,CAAC,aAAa,CAAC,MAAM,CAAC,QAAQ,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC,CAChD,CAAC;AACN,CAAC;AAED,SAAS,OAAO,CAAC,GAAgB;IAC/B,MAAM,KAAK,GAAa,EAAE,CAAC;IAE3B,KAAK,MAAM,IAAI,IAAI,GAAG,CAAC,UAAU,EAAE,CAAC;QAClC,MAAM,OAAO,GAAG,IAAmB,CAAC;QACpC,MAAM,GAAG,GAAG,OAAO,CAAC,OAAO,EAAE,WAAW,EAAE,CAAC;QAC3C,IAAI,GAAG,KAAK,IAAI,IAAI,GAAG,KAAK,IAAI,EAAE,CAAC;YACjC,SAAS;QACX,CAAC;QAED,MAAM,IAAI,GAAG,cAAc,CAAC,OAAO,CAAC,WAAW,IAAI,EAAE,CAAC;aACnD,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC;aACpB,IAAI,EAAE,CAAC;QACV,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAEjB,mEAAmE;QACnE,0EAA0E;QAC1E,gCAAgC;QAChC,MAAM,IAAI,GAAG,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,YAAY,CAAC,SAAS,CAAC,IAAI,GAAG,EAAE,EAAE,CAAC,CAAC;QACzE,KACE,IAAI,KAAK,GAAG,CAAC,EACb,KAAK,GAAG,CAAC,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,EAC1C,KAAK,IAAI,CAAC,EACV,CAAC;YACD,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACjB,CAAC;IACH,CAAC;IAED,OAAO,KAAK,CAAC;AACf,CAAC;AAED,SAAS,MAAM,CAAC,KAAkB;IAChC,OAAO,KAAK,CAAC,gBAAgB,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;AACnD,CAAC;AAED,yEAAyE;AACzE,SAAS,OAAO,CAAC,IAAgB;IAC/B,OAAO,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,CAAC,MAAM,CAAC;AAChF,CAAC;AAED,MAAM,UAAU,aAAa,CAAC,IAAY,EAAE,IAAI,GAAG,QAAQ;IACzD,MAAM,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC,CAAC;IAEzB,6EAA6E;IAC7E,wEAAwE;IACxE,uEAAuE;IACvE,wEAAwE;IACxE,6EAA6E;IAC7E,0EAA0E;IAC1E,MAAM,UAAU,GAAG,IAAI,CAAC,gBAAgB,CAAC,OAAO,CAAC,CAAC;IAElD,IAAI,IAAI,GAAe,EAAE,CAAC;IAC1B,IAAI,WAAW,GAAG,CAAC,CAAC,CAAC;IAErB,KAAK,MAAM,SAAS,IAAI,UAAU,EAAE,CAAC;QACnC,MAAM,IAAI,GAAG,MAAM,CAAC,SAAS,CAAC,CAAC;QAC/B,MAAM,KAAK,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC;QAC5B,IAAI,KAAK,GAAG,WAAW,EAAE,CAAC;YACxB,WAAW,GAAG,KAAK,CAAC;YACpB,IAAI,GAAG,IAAI,CAAC;QACd,CAAC;IACH,CAAC;IAED,OAAO,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,EAAE,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,IAAI,EAAE,YAAY,EAAE,EAAE,CAAC;AAC9E,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,gBAAgB,CAAC,IAAY;IACjD,MAAM,QAAQ,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC;IAC9C,OAAO,aAAa,CAAC,QAAQ,EAAE,IAAI,CAAC,CAAC;AACvC,CAAC"}

package/dist/extract/index.d.ts

@@ -0,0 +1,21 @@
+import { describeFormat, detectFormat } from './sniff.js';
+import type { DetectedFormat } from './sniff.js';
+import type { Table } from './types.js';
+export type ExtractOptions = {
+    /** Only used for delimited text. */
+    delimiter?: string;
+    /** Only used for encrypted PDFs. */
+    password?: string;
+};
+export type Extraction = {
+    table: Table;
+    /** What the file turned out to be, which is often not what it is named. */
+    format: DetectedFormat;
+};
+/**
+ * Read a statement file into a table, choosing the parser by inspecting the
+ * contents rather than the extension.
+ */
+export declare function extractTable(path: string, options?: ExtractOptions): Promise<Extraction>;
+export { describeFormat, detectFormat };
+export type { DetectedFormat, Table };

package/dist/extract/index.js

@@ -0,0 +1,44 @@
+import { readFile } from 'node:fs/promises';
+import { extractCsv } from './csv.js';
+import { extractHtmlTable } from './html-table.js';
+import { extractPdf } from './pdf.js';
+import { describeFormat, detectFormat } from './sniff.js';
+import { extractSpreadsheetMl } from './spreadsheetml.js';
+import { extractXlsx } from './xlsx.js';
+/**
+ * Read a statement file into a table, choosing the parser by inspecting the
+ * contents rather than the extension.
+ */
+export async function extractTable(path, options = {}) {
+    const buffer = await readFile(path);
+    const format = detectFormat(buffer);
+    switch (format) {
+        case 'xlsx':
+            return { table: await extractXlsx(path), format };
+        case 'html':
+            return { table: await extractHtmlTable(path), format };
+        case 'spreadsheetml':
+            return { table: await extractSpreadsheetMl(path), format };
+        case 'pdf':
+            return {
+                table: await extractPdf(path, options.password ? { password: options.password } : {}),
+                format,
+            };
+        case 'biff':
+            // Legacy BIFF has no maintained, permissively licensed reader for Node,
+            // and re-saving is a one-step fix, so this stays an explicit refusal
+            // rather than a half-working parser.
+            throw new Error(`${path} is a ${describeFormat(format)}, which is not supported.\n` +
+                'Open it in Excel or LibreOffice and "Save As" .xlsx or .csv, then ' +
+                'run this again. If your bank offers CSV through internet banking, ' +
+                'that is the most reliable option.');
+        case 'text':
+        default:
+            return {
+                table: await extractCsv(path, options.delimiter ? { delimiter: options.delimiter } : {}),
+                format: 'text',
+            };
+    }
+}
+export { describeFormat, detectFormat };
+//# sourceMappingURL=index.js.map

package/dist/extract/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/extract/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAE5C,OAAO,EAAE,UAAU,EAAE,MAAM,UAAU,CAAC;AACtC,OAAO,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AACnD,OAAO,EAAE,UAAU,EAAE,MAAM,UAAU,CAAC;AACtC,OAAO,EAAE,cAAc,EAAE,YAAY,EAAE,MAAM,YAAY,CAAC;AAE1D,OAAO,EAAE,oBAAoB,EAAE,MAAM,oBAAoB,CAAC;AAE1D,OAAO,EAAE,WAAW,EAAE,MAAM,WAAW,CAAC;AAexC;;;GAGG;AACH,MAAM,CAAC,KAAK,UAAU,YAAY,CAChC,IAAY,EACZ,UAA0B,EAAE;IAE5B,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,IAAI,CAAC,CAAC;IACpC,MAAM,MAAM,GAAG,YAAY,CAAC,MAAM,CAAC,CAAC;IAEpC,QAAQ,MAAM,EAAE,CAAC;QACf,KAAK,MAAM;YACT,OAAO,EAAE,KAAK,EAAE,MAAM,WAAW,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC;QACpD,KAAK,MAAM;YACT,OAAO,EAAE,KAAK,EAAE,MAAM,gBAAgB,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC;QACzD,KAAK,eAAe;YAClB,OAAO,EAAE,KAAK,EAAE,MAAM,oBAAoB,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC;QAC7D,KAAK,KAAK;YACR,OAAO;gBACL,KAAK,EAAE,MAAM,UAAU,CACrB,IAAI,EACJ,OAAO,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,QAAQ,EAAE,OAAO,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC,EAAE,CACvD;gBACD,MAAM;aACP,CAAC;QACJ,KAAK,MAAM;YACT,wEAAwE;YACxE,qEAAqE;YACrE,qCAAqC;YACrC,MAAM,IAAI,KAAK,CACb,GAAG,IAAI,SAAS,cAAc,CAAC,MAAM,CAAC,6BAA6B;gBACjE,oEAAoE;gBACpE,oEAAoE;gBACpE,mCAAmC,CACtC,CAAC;QACJ,KAAK,MAAM,CAAC;QACZ;YACE,OAAO;gBACL,KAAK,EAAE,MAAM,UAAU,CACrB,IAAI,EACJ,OAAO,CAAC,SAAS,CAAC,CAAC,CAAC,EAAE,SAAS,EAAE,OAAO,CAAC,SAAS,EAAE,CAAC,CAAC,CAAC,EAAE,CAC1D;gBACD,MAAM,EAAE,MAAM;aACf,CAAC;IACN,CAAC;AACH,CAAC;AAED,OAAO,EAAE,cAAc,EAAE,YAAY,EAAE,CAAC"}

package/dist/extract/pdf.d.ts

@@ -0,0 +1,47 @@
+import type { Table } from './types.js';
+export type PdfExtractOptions = {
+    password?: string;
+};
+export type Item = {
+    x: number;
+    right: number;
+    y: number;
+    text: string;
+};
+/**
+ * `page` matters for more than reporting: y coordinates restart on every page,
+ * so two lines' y values are only comparable within the same page.
+ */
+export type Line = {
+    y: number;
+    page: number;
+    items: Item[];
+};
+/** A column, as the horizontal span its text occupies. */
+export type Band = {
+    left: number;
+    right: number;
+};
+/**
+ * Infer columns from the vertical whitespace that separates them.
+ *
+ * Every column of a statement table is divided from its neighbour by a gutter
+ * that stays empty on every row, so accumulating the horizontal extent of all
+ * table text and splitting on runs of emptiness recovers the columns directly.
+ *
+ * This is markedly more robust than comparing where text starts or merging
+ * overlapping spans, both of which were tried first and failed on real files:
+ *
+ * - Start positions alone split a right-aligned amount from its own header,
+ *   which can begin 36 points to its left.
+ * - Merging overlapping spans bridges columns whenever a header label is wider
+ *   than the column spacing. Federal packs columns 45 points apart with labels
+ *   nearly that wide, collapsing `Withdrawals` and `Deposits` into one cell.
+ *
+ * Only header and dated rows take part. Wrapped narration lines and page
+ * furniture would otherwise bridge gutters — ICICI's footer URL spans the gap
+ * between the serial-number and date columns.
+ */
+export declare function inferBands(lines: Line[]): Band[];
+export declare function assembleRows(lines: Line[], bands: Band[]): string[][];
+export declare function extractPdf(path: string, options?: PdfExtractOptions): Promise<Table>;