@pranavraut033/ats-checker 1.1.0 → 1.2.0

package/dist/pdf/index.d.mts

@@ -0,0 +1,12 @@
+/**
+ * Extract plain text from a PDF buffer.
+ *
+ * Requires `pdfjs-dist` to be installed (optional peerDependency):
+ *   npm install pdfjs-dist
+ *
+ * @param data - Raw PDF bytes as Uint8Array or ArrayBuffer
+ * @returns Extracted text, ready to pass as `resumeText` to analyzeResume
+ */
+declare function extractTextFromPDF(data: Uint8Array | ArrayBuffer): Promise<string>;
+
+export { extractTextFromPDF };

package/dist/pdf/index.d.ts

@@ -0,0 +1,12 @@
+/**
+ * Extract plain text from a PDF buffer.
+ *
+ * Requires `pdfjs-dist` to be installed (optional peerDependency):
+ *   npm install pdfjs-dist
+ *
+ * @param data - Raw PDF bytes as Uint8Array or ArrayBuffer
+ * @returns Extracted text, ready to pass as `resumeText` to analyzeResume
+ */
+declare function extractTextFromPDF(data: Uint8Array | ArrayBuffer): Promise<string>;
+
+export { extractTextFromPDF };

package/dist/pdf/index.js

@@ -0,0 +1,81 @@
+'use strict';
+
+// src/pdf/index.ts
+async function extractTextFromPDF(data) {
+  let pdfjsLib;
+  try {
+    pdfjsLib = await import('pdfjs-dist');
+  } catch {
+    throw new Error(
+      "pdfjs-dist is required for PDF extraction. Install it: npm install pdfjs-dist"
+    );
+  }
+  const bytes = data instanceof ArrayBuffer ? new Uint8Array(data) : data;
+  const doc = await pdfjsLib.getDocument({ data: bytes }).promise;
+  const pages = [];
+  for (let i = 1; i <= doc.numPages; i++) {
+    const page = await doc.getPage(i);
+    const content = await page.getTextContent();
+    const items = [];
+    for (const item of content.items) {
+      if (!("str" in item) || !item.str.trim()) continue;
+      const transform = Array.isArray(
+        item.transform
+      ) ? item.transform : void 0;
+      if (!transform) {
+        items.push({ x: 0, y: 0, str: item.str });
+      } else {
+        items.push({ x: transform[4], y: transform[5], str: item.str });
+      }
+    }
+    const COLUMN_GAP_THRESHOLD = 80;
+    const xPositions = [...new Set(items.map((it) => Math.round(it.x)))].sort(
+      (a, b) => a - b
+    );
+    let columnBoundary = null;
+    let maxGap = 0;
+    for (let j = 1; j < xPositions.length; j++) {
+      const gap = xPositions[j] - xPositions[j - 1];
+      if (gap > maxGap) {
+        maxGap = gap;
+        columnBoundary = (xPositions[j - 1] + xPositions[j]) / 2;
+      }
+    }
+    if (maxGap < COLUMN_GAP_THRESHOLD) columnBoundary = null;
+    const columns = columnBoundary !== null ? [
+      items.filter((it) => it.x < columnBoundary),
+      items.filter((it) => it.x >= columnBoundary)
+    ] : [items];
+    const columnTexts = columns.map((col) => renderColumn(col));
+    pages.push(columnTexts.filter(Boolean).join("\n"));
+  }
+  return pages.join("\n");
+}
+function renderColumn(items) {
+  const Y_TOLERANCE = 2;
+  const lineMap = /* @__PURE__ */ new Map();
+  const lineOrder = [];
+  for (const { x, y, str } of items) {
+    let bucketKey;
+    for (const key of lineOrder) {
+      if (Math.abs(key - y) <= Y_TOLERANCE) {
+        bucketKey = key;
+        break;
+      }
+    }
+    if (bucketKey === void 0) {
+      bucketKey = y;
+      lineOrder.push(y);
+      lineMap.set(y, []);
+    }
+    lineMap.get(bucketKey).push({ x, str });
+  }
+  lineOrder.sort((a, b) => b - a);
+  return lineOrder.map(
+    (key) => (lineMap.get(key) ?? []).sort((a, b) => a.x - b.x).map((it) => it.str).join(" ").replace(/[^\S\n]+/g, " ").trim()
+  ).filter(Boolean).join("\n");
+}
+
+exports.extractTextFromPDF = extractTextFromPDF;
+//# sourceMappingURL=index.js.map
+//# sourceMappingURL=index.js.map

package/dist/pdf/index.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../../src/pdf/index.ts"],"names":[],"mappings":";;;AASA,eAAsB,mBACpB,IAAA,EACiB;AAEjB,EAAA,IAAI,QAAA;AACJ,EAAA,IAAI;AACF,IAAA,QAAA,GAAW,MAAM,OAAO,YAAY,CAAA;AAAA,EACtC,CAAA,CAAA,MAAQ;AACN,IAAA,MAAM,IAAI,KAAA;AAAA,MACR;AAAA,KACF;AAAA,EACF;AAEA,EAAA,MAAM,QACJ,IAAA,YAAgB,WAAA,GAAc,IAAI,UAAA,CAAW,IAAI,CAAA,GAAI,IAAA;AAEvD,EAAA,MAAM,GAAA,GAAM,MAAM,QAAA,CAAS,WAAA,CAAY,EAAE,IAAA,EAAM,KAAA,EAAO,CAAA,CAAE,OAAA;AACxD,EAAA,MAAM,QAAkB,EAAC;AAEzB,EAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,IAAK,GAAA,CAAI,UAAU,CAAA,EAAA,EAAK;AACtC,IAAA,MAAM,IAAA,GAAO,MAAM,GAAA,CAAI,OAAA,CAAQ,CAAC,CAAA;AAChC,IAAA,MAAM,OAAA,GAAU,MAAM,IAAA,CAAK,cAAA,EAAe;AAG1C,IAAA,MAAM,QAAmB,EAAC;AAE1B,IAAA,KAAA,MAAW,IAAA,IAAQ,QAAQ,KAAA,EAAO;AAChC,MAAA,IAAI,EAAE,KAAA,IAAS,IAAA,CAAA,IAAS,CAAC,IAAA,CAAK,GAAA,CAAI,MAAK,EAAG;AAC1C,MAAA,MAAM,YAAkC,KAAA,CAAM,OAAA;AAAA,QAC3C,IAAA,CAAkC;AAAA,OACrC,GACK,KAAiC,SAAA,GAClC,MAAA;AAEJ,MAAA,IAAI,CAAC,SAAA,EAAW;AAEd,QAAA,KAAA,CAAM,IAAA,CAAK,EAAE,CAAA,EAAG,CAAA,EAAG,GAAG,CAAA,EAAG,GAAA,EAAK,IAAA,CAAK,GAAA,EAAK,CAAA;AAAA,MAC1C,CAAA,MAAO;AACL,QAAA,KAAA,CAAM,IAAA,CAAK,EAAE,CAAA,EAAG,SAAA,CAAU,CAAC,CAAA,EAAG,CAAA,EAAG,SAAA,CAAU,CAAC,CAAA,EAAG,GAAA,EAAK,IAAA,CAAK,KAAK,CAAA;AAAA,MAChE;AAAA,IACF;AAWA,IAAA,MAAM,oBAAA,GAAuB,EAAA;AAC7B,IAAA,MAAM,aAAa,CAAC,GAAG,IAAI,GAAA,CAAI,MAAM,GAAA,CAAI,CAAC,EAAA,KAAO,IAAA,CAAK,MAAM,EAAA,CAAG,CAAC,CAAC,CAAC,CAAC,CAAA,CAAE,IAAA;AAAA,MACnE,CAAC,CAAA,EAAG,CAAA,KAAM,CAAA,GAAI;AAAA,KAChB;AAEA,IAAA,IAAI,cAAA,GAAgC,IAAA;AACpC,IAAA,IAAI,MAAA,GAAS,CAAA;AACb,IAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,GAAI,UAAA,CAAW,QAAQ,CAAA,EAAA,EAAK;AAC1C,MAAA,MAAM,MAAM,UAAA,CAAW,CAAC,CAAA,GAAI,UAAA,CAAW,IAAI,CAAC,CAAA;AAC5C,MAAA,IAAI,MAAM,MAAA,EAAQ;AAChB,QAAA,MAAA,GAAS,GAAA;AACT,QAAA,cAAA,GAAA,CAAkB,WAAW,CAAA,GAAI,CAAC,CAAA,GAAI,UAAA,CAAW,CAAC,CAAA,IAAK,CAAA;AAAA,MACzD;AAAA,IACF;AACA,IAAA,IAAI,MAAA,GAAS,sBAAsB,cAAA,GAAiB,IAAA;AAEpD,IAAA,MAAM,OAAA,GACJ,mBAAmB,IAAA,GACf;AAAA,MACE,MAAM,MAAA,CAAO,CAAC,EAAA,KAAO,EAAA,CAAG,IAAI,cAAe,CAAA;AAAA,MAC3C,MAAM,MAAA,CAAO,CAAC,EAAA,KAAO,EAAA,CAAG,KAAK,cAAe;AAAA,KAC9C,GACA,CAAC,KAAK,CAAA;AAEZ,IAAA,MAAM,cAAc,OAAA,CAAQ,GAAA,CAAI,CAAC,GAAA,KAAQ,YAAA,CAAa,GAAG,CAAC,CAAA;AAC1D,IAAA,KAAA,CAAM,KAAK,WAAA,CAAY,MAAA,CAAO,OAAO,CAAA,CAAE,IAAA,CAAK,IAAI,CAAC,CAAA;AAAA,EACnD;AAEA,EAAA,OAAO,KAAA,CAAM,KAAK,IAAI,CAAA;AACxB;AAEA,SAAS,aAAa,KAAA,EAA6D;AACjF,EAAA,MAAM,WAAA,GAAc,CAAA;AACpB,EAAA,MAAM,OAAA,uBAA8D,GAAA,EAAI;AACxE,EAAA,MAAM,YAAsB,EAAC;AAE7B,EAAA,KAAA,MAAW,EAAE,CAAA,EAAG,CAAA,EAAG,GAAA,MAAS,KAAA,EAAO;AACjC,IAAA,IAAI,SAAA;AACJ,IAAA,KAAA,MAAW,OAAO,SAAA,EAAW;AAC3B,MAAA,IAAI,IAAA,CAAK,GAAA,CAAI,GAAA,GAAM,CAAC,KAAK,WAAA,EAAa;AACpC,QAAA,SAAA,GAAY,GAAA;AACZ,QAAA;AAAA,MACF;AAAA,IACF;AACA,IAAA,IAAI,cAAc,MAAA,EAAW;AAC3B,MAAA,SAAA,GAAY,CAAA;AACZ,MAAA,SAAA,CAAU,KAAK,CAAC,CAAA;AAChB,MAAA,OAAA,CAAQ,GAAA,CAAI,CAAA,EAAG,EAAE,CAAA;AAAA,IACnB;AACA,IAAA,OAAA,CAAQ,IAAI,SAAS,CAAA,CAAG,KAAK,EAAE,CAAA,EAAG,KAAK,CAAA;AAAA,EACzC;AAGA,EAAA,SAAA,CAAU,IAAA,CAAK,CAAC,CAAA,EAAG,CAAA,KAAM,IAAI,CAAC,CAAA;AAE9B,EAAA,OAAO,SAAA,CACJ,GAAA;AAAA,IAAI,CAAC,GAAA,KAAA,CACH,OAAA,CAAQ,GAAA,CAAI,GAAG,CAAA,IAAK,EAAC,EACnB,IAAA,CAAK,CAAC,CAAA,EAAG,CAAA,KAAM,EAAE,CAAA,GAAI,CAAA,CAAE,CAAC,CAAA,CACxB,GAAA,CAAI,CAAC,EAAA,KAAO,EAAA,CAAG,GAAG,CAAA,CAClB,IAAA,CAAK,GAAG,CAAA,CACR,OAAA,CAAQ,WAAA,EAAa,GAAG,EACxB,IAAA;AAAK,GACV,CACC,MAAA,CAAO,OAAO,CAAA,CACd,KAAK,IAAI,CAAA;AACd","file":"index.js","sourcesContent":["/**\n * Extract plain text from a PDF buffer.\n *\n * Requires `pdfjs-dist` to be installed (optional peerDependency):\n *   npm install pdfjs-dist\n *\n * @param data - Raw PDF bytes as Uint8Array or ArrayBuffer\n * @returns Extracted text, ready to pass as `resumeText` to analyzeResume\n */\nexport async function extractTextFromPDF(\n  data: Uint8Array | ArrayBuffer\n): Promise<string> {\n  // ponytail: lazy import keeps core zero-dep; missing peer throws with clear message\n  let pdfjsLib: typeof import(\"pdfjs-dist\");\n  try {\n    pdfjsLib = await import(\"pdfjs-dist\");\n  } catch {\n    throw new Error(\n      \"pdfjs-dist is required for PDF extraction. Install it: npm install pdfjs-dist\"\n    );\n  }\n\n  const bytes =\n    data instanceof ArrayBuffer ? new Uint8Array(data) : data;\n\n  const doc = await pdfjsLib.getDocument({ data: bytes }).promise;\n  const pages: string[] = [];\n\n  for (let i = 1; i <= doc.numPages; i++) {\n    const page = await doc.getPage(i);\n    const content = await page.getTextContent();\n\n    type RawItem = { x: number; y: number; str: string };\n    const items: RawItem[] = [];\n\n    for (const item of content.items) {\n      if (!(\"str\" in item) || !item.str.trim()) continue;\n      const transform: number[] | undefined = Array.isArray(\n        (item as { transform?: number[] }).transform\n      )\n        ? (item as { transform: number[] }).transform\n        : undefined;\n\n      if (!transform) {\n        // No positional info (unit-test mocks) — treat as single-column item\n        items.push({ x: 0, y: 0, str: item.str });\n      } else {\n        items.push({ x: transform[4], y: transform[5], str: item.str });\n      }\n    }\n\n    // Detect column boundary: find the largest x-gap among item start positions.\n    // If it exceeds COLUMN_GAP_THRESHOLD, split into left / right columns and\n    // process each independently so headers in different columns don't merge.\n    // ponytail: single largest-gap heuristic handles the common 2-column resume;\n    // n-column needs k-means on x-distribution — upgrade if this proves insufficient.\n    // Column boundary heuristic: the largest gap in item x-positions.\n    // Real PDF column gutters show as a gap >>80px; normal word spacing is <50px.\n    // ponytail: magic number calibrated to PranavRaut2026.pdf (104px gap); raise\n    // if single-column PDFs with wide indentation start getting falsely split.\n    const COLUMN_GAP_THRESHOLD = 80;\n    const xPositions = [...new Set(items.map((it) => Math.round(it.x)))].sort(\n      (a, b) => a - b\n    );\n\n    let columnBoundary: number | null = null;\n    let maxGap = 0;\n    for (let j = 1; j < xPositions.length; j++) {\n      const gap = xPositions[j] - xPositions[j - 1];\n      if (gap > maxGap) {\n        maxGap = gap;\n        columnBoundary = (xPositions[j - 1] + xPositions[j]) / 2;\n      }\n    }\n    if (maxGap < COLUMN_GAP_THRESHOLD) columnBoundary = null;\n\n    const columns =\n      columnBoundary !== null\n        ? [\n            items.filter((it) => it.x < columnBoundary!),\n            items.filter((it) => it.x >= columnBoundary!),\n          ]\n        : [items];\n\n    const columnTexts = columns.map((col) => renderColumn(col));\n    pages.push(columnTexts.filter(Boolean).join(\"\\n\"));\n  }\n\n  return pages.join(\"\\n\");\n}\n\nfunction renderColumn(items: Array<{ x: number; y: number; str: string }>): string {\n  const Y_TOLERANCE = 2;\n  const lineMap: Map<number, Array<{ x: number; str: string }>> = new Map();\n  const lineOrder: number[] = [];\n\n  for (const { x, y, str } of items) {\n    let bucketKey: number | undefined;\n    for (const key of lineOrder) {\n      if (Math.abs(key - y) <= Y_TOLERANCE) {\n        bucketKey = key;\n        break;\n      }\n    }\n    if (bucketKey === undefined) {\n      bucketKey = y;\n      lineOrder.push(y);\n      lineMap.set(y, []);\n    }\n    lineMap.get(bucketKey)!.push({ x, str });\n  }\n\n  // pdfjs y=0 is bottom of page — sort descending so top comes first\n  lineOrder.sort((a, b) => b - a);\n\n  return lineOrder\n    .map((key) =>\n      (lineMap.get(key) ?? [])\n        .sort((a, b) => a.x - b.x)\n        .map((it) => it.str)\n        .join(\" \")\n        .replace(/[^\\S\\n]+/g, \" \")\n        .trim()\n    )\n    .filter(Boolean)\n    .join(\"\\n\");\n}\n"]}

package/dist/pdf/index.mjs

@@ -0,0 +1,79 @@
+// src/pdf/index.ts
+async function extractTextFromPDF(data) {
+  let pdfjsLib;
+  try {
+    pdfjsLib = await import('pdfjs-dist');
+  } catch {
+    throw new Error(
+      "pdfjs-dist is required for PDF extraction. Install it: npm install pdfjs-dist"
+    );
+  }
+  const bytes = data instanceof ArrayBuffer ? new Uint8Array(data) : data;
+  const doc = await pdfjsLib.getDocument({ data: bytes }).promise;
+  const pages = [];
+  for (let i = 1; i <= doc.numPages; i++) {
+    const page = await doc.getPage(i);
+    const content = await page.getTextContent();
+    const items = [];
+    for (const item of content.items) {
+      if (!("str" in item) || !item.str.trim()) continue;
+      const transform = Array.isArray(
+        item.transform
+      ) ? item.transform : void 0;
+      if (!transform) {
+        items.push({ x: 0, y: 0, str: item.str });
+      } else {
+        items.push({ x: transform[4], y: transform[5], str: item.str });
+      }
+    }
+    const COLUMN_GAP_THRESHOLD = 80;
+    const xPositions = [...new Set(items.map((it) => Math.round(it.x)))].sort(
+      (a, b) => a - b
+    );
+    let columnBoundary = null;
+    let maxGap = 0;
+    for (let j = 1; j < xPositions.length; j++) {
+      const gap = xPositions[j] - xPositions[j - 1];
+      if (gap > maxGap) {
+        maxGap = gap;
+        columnBoundary = (xPositions[j - 1] + xPositions[j]) / 2;
+      }
+    }
+    if (maxGap < COLUMN_GAP_THRESHOLD) columnBoundary = null;
+    const columns = columnBoundary !== null ? [
+      items.filter((it) => it.x < columnBoundary),
+      items.filter((it) => it.x >= columnBoundary)
+    ] : [items];
+    const columnTexts = columns.map((col) => renderColumn(col));
+    pages.push(columnTexts.filter(Boolean).join("\n"));
+  }
+  return pages.join("\n");
+}
+function renderColumn(items) {
+  const Y_TOLERANCE = 2;
+  const lineMap = /* @__PURE__ */ new Map();
+  const lineOrder = [];
+  for (const { x, y, str } of items) {
+    let bucketKey;
+    for (const key of lineOrder) {
+      if (Math.abs(key - y) <= Y_TOLERANCE) {
+        bucketKey = key;
+        break;
+      }
+    }
+    if (bucketKey === void 0) {
+      bucketKey = y;
+      lineOrder.push(y);
+      lineMap.set(y, []);
+    }
+    lineMap.get(bucketKey).push({ x, str });
+  }
+  lineOrder.sort((a, b) => b - a);
+  return lineOrder.map(
+    (key) => (lineMap.get(key) ?? []).sort((a, b) => a.x - b.x).map((it) => it.str).join(" ").replace(/[^\S\n]+/g, " ").trim()
+  ).filter(Boolean).join("\n");
+}
+
+export { extractTextFromPDF };
+//# sourceMappingURL=index.mjs.map
+//# sourceMappingURL=index.mjs.map

package/dist/pdf/index.mjs.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../../src/pdf/index.ts"],"names":[],"mappings":";AASA,eAAsB,mBACpB,IAAA,EACiB;AAEjB,EAAA,IAAI,QAAA;AACJ,EAAA,IAAI;AACF,IAAA,QAAA,GAAW,MAAM,OAAO,YAAY,CAAA;AAAA,EACtC,CAAA,CAAA,MAAQ;AACN,IAAA,MAAM,IAAI,KAAA;AAAA,MACR;AAAA,KACF;AAAA,EACF;AAEA,EAAA,MAAM,QACJ,IAAA,YAAgB,WAAA,GAAc,IAAI,UAAA,CAAW,IAAI,CAAA,GAAI,IAAA;AAEvD,EAAA,MAAM,GAAA,GAAM,MAAM,QAAA,CAAS,WAAA,CAAY,EAAE,IAAA,EAAM,KAAA,EAAO,CAAA,CAAE,OAAA;AACxD,EAAA,MAAM,QAAkB,EAAC;AAEzB,EAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,IAAK,GAAA,CAAI,UAAU,CAAA,EAAA,EAAK;AACtC,IAAA,MAAM,IAAA,GAAO,MAAM,GAAA,CAAI,OAAA,CAAQ,CAAC,CAAA;AAChC,IAAA,MAAM,OAAA,GAAU,MAAM,IAAA,CAAK,cAAA,EAAe;AAG1C,IAAA,MAAM,QAAmB,EAAC;AAE1B,IAAA,KAAA,MAAW,IAAA,IAAQ,QAAQ,KAAA,EAAO;AAChC,MAAA,IAAI,EAAE,KAAA,IAAS,IAAA,CAAA,IAAS,CAAC,IAAA,CAAK,GAAA,CAAI,MAAK,EAAG;AAC1C,MAAA,MAAM,YAAkC,KAAA,CAAM,OAAA;AAAA,QAC3C,IAAA,CAAkC;AAAA,OACrC,GACK,KAAiC,SAAA,GAClC,MAAA;AAEJ,MAAA,IAAI,CAAC,SAAA,EAAW;AAEd,QAAA,KAAA,CAAM,IAAA,CAAK,EAAE,CAAA,EAAG,CAAA,EAAG,GAAG,CAAA,EAAG,GAAA,EAAK,IAAA,CAAK,GAAA,EAAK,CAAA;AAAA,MAC1C,CAAA,MAAO;AACL,QAAA,KAAA,CAAM,IAAA,CAAK,EAAE,CAAA,EAAG,SAAA,CAAU,CAAC,CAAA,EAAG,CAAA,EAAG,SAAA,CAAU,CAAC,CAAA,EAAG,GAAA,EAAK,IAAA,CAAK,KAAK,CAAA;AAAA,MAChE;AAAA,IACF;AAWA,IAAA,MAAM,oBAAA,GAAuB,EAAA;AAC7B,IAAA,MAAM,aAAa,CAAC,GAAG,IAAI,GAAA,CAAI,MAAM,GAAA,CAAI,CAAC,EAAA,KAAO,IAAA,CAAK,MAAM,EAAA,CAAG,CAAC,CAAC,CAAC,CAAC,CAAA,CAAE,IAAA;AAAA,MACnE,CAAC,CAAA,EAAG,CAAA,KAAM,CAAA,GAAI;AAAA,KAChB;AAEA,IAAA,IAAI,cAAA,GAAgC,IAAA;AACpC,IAAA,IAAI,MAAA,GAAS,CAAA;AACb,IAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,GAAI,UAAA,CAAW,QAAQ,CAAA,EAAA,EAAK;AAC1C,MAAA,MAAM,MAAM,UAAA,CAAW,CAAC,CAAA,GAAI,UAAA,CAAW,IAAI,CAAC,CAAA;AAC5C,MAAA,IAAI,MAAM,MAAA,EAAQ;AAChB,QAAA,MAAA,GAAS,GAAA;AACT,QAAA,cAAA,GAAA,CAAkB,WAAW,CAAA,GAAI,CAAC,CAAA,GAAI,UAAA,CAAW,CAAC,CAAA,IAAK,CAAA;AAAA,MACzD;AAAA,IACF;AACA,IAAA,IAAI,MAAA,GAAS,sBAAsB,cAAA,GAAiB,IAAA;AAEpD,IAAA,MAAM,OAAA,GACJ,mBAAmB,IAAA,GACf;AAAA,MACE,MAAM,MAAA,CAAO,CAAC,EAAA,KAAO,EAAA,CAAG,IAAI,cAAe,CAAA;AAAA,MAC3C,MAAM,MAAA,CAAO,CAAC,EAAA,KAAO,EAAA,CAAG,KAAK,cAAe;AAAA,KAC9C,GACA,CAAC,KAAK,CAAA;AAEZ,IAAA,MAAM,cAAc,OAAA,CAAQ,GAAA,CAAI,CAAC,GAAA,KAAQ,YAAA,CAAa,GAAG,CAAC,CAAA;AAC1D,IAAA,KAAA,CAAM,KAAK,WAAA,CAAY,MAAA,CAAO,OAAO,CAAA,CAAE,IAAA,CAAK,IAAI,CAAC,CAAA;AAAA,EACnD;AAEA,EAAA,OAAO,KAAA,CAAM,KAAK,IAAI,CAAA;AACxB;AAEA,SAAS,aAAa,KAAA,EAA6D;AACjF,EAAA,MAAM,WAAA,GAAc,CAAA;AACpB,EAAA,MAAM,OAAA,uBAA8D,GAAA,EAAI;AACxE,EAAA,MAAM,YAAsB,EAAC;AAE7B,EAAA,KAAA,MAAW,EAAE,CAAA,EAAG,CAAA,EAAG,GAAA,MAAS,KAAA,EAAO;AACjC,IAAA,IAAI,SAAA;AACJ,IAAA,KAAA,MAAW,OAAO,SAAA,EAAW;AAC3B,MAAA,IAAI,IAAA,CAAK,GAAA,CAAI,GAAA,GAAM,CAAC,KAAK,WAAA,EAAa;AACpC,QAAA,SAAA,GAAY,GAAA;AACZ,QAAA;AAAA,MACF;AAAA,IACF;AACA,IAAA,IAAI,cAAc,MAAA,EAAW;AAC3B,MAAA,SAAA,GAAY,CAAA;AACZ,MAAA,SAAA,CAAU,KAAK,CAAC,CAAA;AAChB,MAAA,OAAA,CAAQ,GAAA,CAAI,CAAA,EAAG,EAAE,CAAA;AAAA,IACnB;AACA,IAAA,OAAA,CAAQ,IAAI,SAAS,CAAA,CAAG,KAAK,EAAE,CAAA,EAAG,KAAK,CAAA;AAAA,EACzC;AAGA,EAAA,SAAA,CAAU,IAAA,CAAK,CAAC,CAAA,EAAG,CAAA,KAAM,IAAI,CAAC,CAAA;AAE9B,EAAA,OAAO,SAAA,CACJ,GAAA;AAAA,IAAI,CAAC,GAAA,KAAA,CACH,OAAA,CAAQ,GAAA,CAAI,GAAG,CAAA,IAAK,EAAC,EACnB,IAAA,CAAK,CAAC,CAAA,EAAG,CAAA,KAAM,EAAE,CAAA,GAAI,CAAA,CAAE,CAAC,CAAA,CACxB,GAAA,CAAI,CAAC,EAAA,KAAO,EAAA,CAAG,GAAG,CAAA,CAClB,IAAA,CAAK,GAAG,CAAA,CACR,OAAA,CAAQ,WAAA,EAAa,GAAG,EACxB,IAAA;AAAK,GACV,CACC,MAAA,CAAO,OAAO,CAAA,CACd,KAAK,IAAI,CAAA;AACd","file":"index.mjs","sourcesContent":["/**\n * Extract plain text from a PDF buffer.\n *\n * Requires `pdfjs-dist` to be installed (optional peerDependency):\n *   npm install pdfjs-dist\n *\n * @param data - Raw PDF bytes as Uint8Array or ArrayBuffer\n * @returns Extracted text, ready to pass as `resumeText` to analyzeResume\n */\nexport async function extractTextFromPDF(\n  data: Uint8Array | ArrayBuffer\n): Promise<string> {\n  // ponytail: lazy import keeps core zero-dep; missing peer throws with clear message\n  let pdfjsLib: typeof import(\"pdfjs-dist\");\n  try {\n    pdfjsLib = await import(\"pdfjs-dist\");\n  } catch {\n    throw new Error(\n      \"pdfjs-dist is required for PDF extraction. Install it: npm install pdfjs-dist\"\n    );\n  }\n\n  const bytes =\n    data instanceof ArrayBuffer ? new Uint8Array(data) : data;\n\n  const doc = await pdfjsLib.getDocument({ data: bytes }).promise;\n  const pages: string[] = [];\n\n  for (let i = 1; i <= doc.numPages; i++) {\n    const page = await doc.getPage(i);\n    const content = await page.getTextContent();\n\n    type RawItem = { x: number; y: number; str: string };\n    const items: RawItem[] = [];\n\n    for (const item of content.items) {\n      if (!(\"str\" in item) || !item.str.trim()) continue;\n      const transform: number[] | undefined = Array.isArray(\n        (item as { transform?: number[] }).transform\n      )\n        ? (item as { transform: number[] }).transform\n        : undefined;\n\n      if (!transform) {\n        // No positional info (unit-test mocks) — treat as single-column item\n        items.push({ x: 0, y: 0, str: item.str });\n      } else {\n        items.push({ x: transform[4], y: transform[5], str: item.str });\n      }\n    }\n\n    // Detect column boundary: find the largest x-gap among item start positions.\n    // If it exceeds COLUMN_GAP_THRESHOLD, split into left / right columns and\n    // process each independently so headers in different columns don't merge.\n    // ponytail: single largest-gap heuristic handles the common 2-column resume;\n    // n-column needs k-means on x-distribution — upgrade if this proves insufficient.\n    // Column boundary heuristic: the largest gap in item x-positions.\n    // Real PDF column gutters show as a gap >>80px; normal word spacing is <50px.\n    // ponytail: magic number calibrated to PranavRaut2026.pdf (104px gap); raise\n    // if single-column PDFs with wide indentation start getting falsely split.\n    const COLUMN_GAP_THRESHOLD = 80;\n    const xPositions = [...new Set(items.map((it) => Math.round(it.x)))].sort(\n      (a, b) => a - b\n    );\n\n    let columnBoundary: number | null = null;\n    let maxGap = 0;\n    for (let j = 1; j < xPositions.length; j++) {\n      const gap = xPositions[j] - xPositions[j - 1];\n      if (gap > maxGap) {\n        maxGap = gap;\n        columnBoundary = (xPositions[j - 1] + xPositions[j]) / 2;\n      }\n    }\n    if (maxGap < COLUMN_GAP_THRESHOLD) columnBoundary = null;\n\n    const columns =\n      columnBoundary !== null\n        ? [\n            items.filter((it) => it.x < columnBoundary!),\n            items.filter((it) => it.x >= columnBoundary!),\n          ]\n        : [items];\n\n    const columnTexts = columns.map((col) => renderColumn(col));\n    pages.push(columnTexts.filter(Boolean).join(\"\\n\"));\n  }\n\n  return pages.join(\"\\n\");\n}\n\nfunction renderColumn(items: Array<{ x: number; y: number; str: string }>): string {\n  const Y_TOLERANCE = 2;\n  const lineMap: Map<number, Array<{ x: number; str: string }>> = new Map();\n  const lineOrder: number[] = [];\n\n  for (const { x, y, str } of items) {\n    let bucketKey: number | undefined;\n    for (const key of lineOrder) {\n      if (Math.abs(key - y) <= Y_TOLERANCE) {\n        bucketKey = key;\n        break;\n      }\n    }\n    if (bucketKey === undefined) {\n      bucketKey = y;\n      lineOrder.push(y);\n      lineMap.set(y, []);\n    }\n    lineMap.get(bucketKey)!.push({ x, str });\n  }\n\n  // pdfjs y=0 is bottom of page — sort descending so top comes first\n  lineOrder.sort((a, b) => b - a);\n\n  return lineOrder\n    .map((key) =>\n      (lineMap.get(key) ?? [])\n        .sort((a, b) => a.x - b.x)\n        .map((it) => it.str)\n        .join(\" \")\n        .replace(/[^\\S\\n]+/g, \" \")\n        .trim()\n    )\n    .filter(Boolean)\n    .join(\"\\n\");\n}\n"]}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@pranavraut033/ats-checker",
-  "version": "1.1.0",
+  "version": "1.2.0",
   "description": "Deterministic, configurable ATS (Applicant Tracking System) compatibility checker with no external dependencies. Analyze resumes, generate scores, and get actionable suggestions.",
   "license": "MIT",
   "author": {
@@ -34,6 +34,20 @@
       "import": "./dist/index.mjs",
       "require": "./dist/index.cjs",
       "default": "./dist/index.mjs"
+    },
+    "./pdf": {
+      "types": "./dist/pdf/index.d.ts",
+      "import": "./dist/pdf/index.mjs",
+      "require": "./dist/pdf/index.cjs",
+      "default": "./dist/pdf/index.mjs"
+    }
+  },
+  "peerDependencies": {
+    "pdfjs-dist": ">=4"
+  },
+  "peerDependenciesMeta": {
+    "pdfjs-dist": {
+      "optional": true
     }
   },
   "sideEffects": false,
@@ -56,8 +70,8 @@
   ],
   "scripts": {
     "build": "tsup && npm run build:ui",
-    "build:ui": "mkdir -p ui/public/dist && cp dist/index.mjs ui/public/dist/ && cp dist/index.d.ts ui/public/dist/",
-    "serve": "npx http-server ui/public -p 3005",
+    "build:ui": "rm -rf ui/public/dist && cp -r dist ui/public/dist",
+    "serve": "npx http-server ui/public -p ${PORT:-3005}",
     "dev": "npm run serve",
     "test": "vitest run",
     "test:watch": "vitest",
@@ -71,8 +85,9 @@
     "@types/node": "^25.0.3",
     "@vitest/ui": "^4.0.16",
     "express": "^4.18.2",
-    "tsx": "^4.7.0",
+    "pdfjs-dist": "^4.10.38",
     "tsup": "^8.3.0",
+    "tsx": "^4.7.0",
     "typescript": "^5.6.3",
     "vitest": "^4.0.16"
   }