@parseo/core 1.0.0

package/README.md

@@ -0,0 +1,32 @@
+# @parseo/core
+
+Universal document parser for underwriting PDFs. Auto-classifies the document and routes to the correct parser.
+
+## Installation
+
+```bash
+npm install @parseo/core
+```
+
+This installs all Parseo parsers (credit reports, background checks, appraisals, bank statements).
+
+## Usage
+
+```typescript
+import { parse } from "@parseo/core";
+
+const result = await parse(buffer);
+
+if (result) {
+  result.format;       // "chase", "xactus", "smartlinx", etc.
+  result.data;         // Parsed document (type depends on format)
+  result.confidence;   // Classifier confidence score
+  result.skippedPages; // Number of intro pages stripped
+}
+```
+
+`parse()` handles text extraction, classification, page skipping, and bounding box offset correction. Returns `null` if no known format is detected.
+
+## License
+
+MIT

package/dist/index.d.ts

@@ -0,0 +1,3 @@
+export { parse } from "./parse.js";
+export type { ParseResult } from "./parse.js";
+//# sourceMappingURL=index.d.ts.map

package/dist/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,YAAY,CAAC;AACnC,YAAY,EAAE,WAAW,EAAE,MAAM,YAAY,CAAC"}

package/dist/index.js

@@ -0,0 +1 @@
+export { parse } from "./parse.js";

package/dist/parse.d.ts

@@ -0,0 +1,19 @@
+import type { FormatName } from "@parseo/shared";
+export interface ParseResult {
+    /** Which format was detected */
+    format: FormatName;
+    /** Parsed data (type depends on format) */
+    data: unknown;
+    /** Number of intro pages that were skipped */
+    skippedPages: number;
+    /** Classifier confidence score */
+    confidence: number;
+}
+/**
+ * Universal parser. Extracts text from the PDF, classifies the document,
+ * and routes to the correct parser.
+ *
+ * Returns `null` if no known format is detected.
+ */
+export declare function parse(buffer: Buffer): Promise<ParseResult | null>;
+//# sourceMappingURL=parse.d.ts.map

package/dist/parse.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"parse.d.ts","sourceRoot":"","sources":["../src/parse.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAY,UAAU,EAAkB,MAAM,gBAAgB,CAAC;AA4B3E,MAAM,WAAW,WAAW;IAC1B,gCAAgC;IAChC,MAAM,EAAE,UAAU,CAAC;IACnB,2CAA2C;IAC3C,IAAI,EAAE,OAAO,CAAC;IACd,8CAA8C;IAC9C,YAAY,EAAE,MAAM,CAAC;IACrB,kCAAkC;IAClC,UAAU,EAAE,MAAM,CAAC;CACpB;AA4BD;;;;;GAKG;AACH,wBAAsB,KAAK,CAAC,MAAM,EAAE,MAAM,GAAG,OAAO,CAAC,WAAW,GAAG,IAAI,CAAC,CAgFvE"}

package/dist/parse.js

@@ -0,0 +1,111 @@
+import { extractLines, classifyDocument } from "@parseo/shared";
+import { parseSmartLinxReportFromLines } from "@parseo/background-checks";
+import { parseCreditReportFromLines } from "@parseo/credit-reports";
+import { parseRicherValuesReportFromLines, parseForm1004MCFromLines, parseForm1073FromLines, } from "@parseo/appraisals";
+import { parseWellsFargoFromLines, parseTDBankFromLines, parseChaseFromLines, parseBankOfAmericaFromLines, parseNavyFederalFromLines, parseThirdFederalFromLines, parseCitibankFromLines, parseRelayFromLines, parseGroveBankFromLines, parseCapitalOneFromLines, parseTruistFromLines, parsePNCFromLines, parseDiscoverFromLines, parseSynovusFromLines, } from "@parseo/bank-statements";
+// ── Page helpers ─────────────────────────────────────────────
+function skipPages(lines, pagesToSkip) {
+    const minPage = lines.length > 0 ? lines[0].page : 1;
+    const firstKeptPage = minPage + pagesToSkip;
+    return lines
+        .filter((l) => l.page >= firstKeptPage)
+        .map((l) => ({ ...l, page: l.page - pagesToSkip }));
+}
+function offsetBoundingBoxPages(obj, offset) {
+    if (offset === 0 || obj == null || typeof obj !== "object")
+        return;
+    const record = obj;
+    if (typeof record.pageNumber === "number" && "x" in record && "y" in record) {
+        record.pageNumber = record.pageNumber + offset;
+        return;
+    }
+    for (const value of Object.values(record)) {
+        if (value != null && typeof value === "object") {
+            offsetBoundingBoxPages(value, offset);
+        }
+    }
+}
+// ── Main ─────────────────────────────────────────────────────
+/**
+ * Universal parser. Extracts text from the PDF, classifies the document,
+ * and routes to the correct parser.
+ *
+ * Returns `null` if no known format is detected.
+ */
+export async function parse(buffer) {
+    const allLines = await extractLines(buffer);
+    const classification = classifyDocument(allLines);
+    if (!classification)
+        return null;
+    const { format, skip, confidence } = classification;
+    const lines = skip > 0 ? skipPages(allLines, skip) : allLines;
+    let data;
+    switch (format) {
+        case "smartlinx":
+            data = parseSmartLinxReportFromLines(lines);
+            break;
+        case "credit-report": {
+            const cr = parseCreditReportFromLines(lines);
+            data = {
+                format: cr.format,
+                report: cr.report,
+                ...(cr.creditXpert ? { creditXpert: cr.creditXpert } : {}),
+            };
+            break;
+        }
+        case "richer-values":
+            data = parseRicherValuesReportFromLines(lines);
+            break;
+        case "form-1004mc":
+            data = await parseForm1004MCFromLines(lines, buffer, skip);
+            break;
+        case "form-1073":
+            data = await parseForm1073FromLines(lines, buffer, skip);
+            break;
+        case "wells-fargo":
+            data = parseWellsFargoFromLines(lines);
+            break;
+        case "td-bank":
+            data = parseTDBankFromLines(lines);
+            break;
+        case "chase":
+            data = parseChaseFromLines(lines);
+            break;
+        case "bank-of-america":
+            data = parseBankOfAmericaFromLines(lines);
+            break;
+        case "navy-federal":
+            data = parseNavyFederalFromLines(lines);
+            break;
+        case "third-federal":
+            data = parseThirdFederalFromLines(lines);
+            break;
+        case "citibank":
+            data = parseCitibankFromLines(lines);
+            break;
+        case "relay":
+            data = parseRelayFromLines(lines);
+            break;
+        case "grove-bank":
+            data = parseGroveBankFromLines(lines);
+            break;
+        case "capital-one":
+            data = parseCapitalOneFromLines(lines);
+            break;
+        case "truist":
+            data = parseTruistFromLines(lines);
+            break;
+        case "pnc":
+            data = parsePNCFromLines(lines);
+            break;
+        case "discover":
+            data = parseDiscoverFromLines(lines);
+            break;
+        case "synovus":
+            data = parseSynovusFromLines(lines);
+            break;
+    }
+    if (skip > 0)
+        offsetBoundingBoxPages(data, skip);
+    return { format, data, skippedPages: skip, confidence };
+}

package/package.json

@@ -0,0 +1,28 @@
+{
+  "name": "@parseo/core",
+  "version": "1.0.0",
+  "type": "module",
+  "main": "./dist/index.js",
+  "types": "./dist/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./dist/index.d.ts",
+      "import": "./dist/index.js"
+    }
+  },
+  "license": "MIT",
+  "publishConfig": {
+    "access": "public"
+  },
+  "files": ["dist"],
+  "scripts": {
+    "build": "tsc"
+  },
+  "dependencies": {
+    "@parseo/shared": "1.0.0",
+    "@parseo/credit-reports": "1.0.0",
+    "@parseo/background-checks": "1.0.0",
+    "@parseo/appraisals": "1.0.0",
+    "@parseo/bank-statements": "1.0.0"
+  }
+}