@dragon708/docmind-node 0.1.0-alpha.1 → 1.0.0

package/dist/index.d.ts

@@ -1,30 +1,34 @@
-import { DetectFileKindInput, AnalyzeOptions, AnalysisResult } from '@dragon708/docmind-core';
-export { AnalysisAnalyzer, AnalysisResult, AnalyzeOptions, DetectFileKindInput, FileKind, FileKindMetadata, GenericAnalysisResult, TextAnalysisResult } from '@dragon708/docmind-core';
-import { NamedInput } from '@dragon708/docmind-shared';
+import { DetectFileKindInput, NamedInput, AnalysisResult } from '@dragon708/docmind-shared';
+export { AnalysisAnalyzer, AnalysisResult, DetectFileKindInput, DocxAnalysisCoreResult, FileKind, FileKindMetadata, GenericAnalysisResult, ImageAnalysisCoreResult, PdfAnalysisCoreResult, TextAnalysisResult } from '@dragon708/docmind-shared';
+import { OcrOptions } from '@dragon708/docmind-ocr';
+import { PdfAnalyzeOptions } from '@dragon708/docmind-pdf';
+
+/** Options for {@link analyzeFile} in the Node entry (`pdf` / `ocr` forwarded to format packages). */
+interface NodeAnalyzeOptions {
+    readonly signal?: AbortSignal;
+    readonly pdf?: PdfAnalyzeOptions;
+    readonly ocr?: OcrOptions;
+}
 
 /**
  * Inputs accepted by {@link analyzeFile} in this package.
- * Paths and `file:` URLs are read with `fs`; other values pass through as core’s
- * {@link DetectFileKindInput}.
+ * Paths and `file:` URLs are read with `fs`; other values pass through as {@link DetectFileKindInput}.
  */
 type NodeAnalyzeInput = string | URL | DetectFileKindInput;
 /**
- * Reads a file from disk into a {@link NamedInput} suitable for `@dragon708/docmind-core`
- * (binary `Buffer`, basename as `name` for extension / MIME hints).
+ * Reads a file from disk into a {@link NamedInput} (binary `Buffer`, basename as `name` for hints).
  */
 declare function readFileToInput(path: string | URL): Promise<NamedInput<Buffer>>;
 /** Wraps a `Buffer` as a named payload when you already know the filename. */
 declare function bufferToInput(buffer: Buffer, name?: string): NamedInput<Buffer>;
 /**
- * Resolves paths / `file:` URLs to core’s internal shape; leaves other
- * {@link DetectFileKindInput} values untouched.
+ * Resolves paths / `file:` URLs to a {@link DetectFileKindInput}; leaves other values untouched.
  */
 declare function resolveNodeAnalyzeInput(input: NodeAnalyzeInput): Promise<DetectFileKindInput>;
 
 /**
- * Like `@dragon708/docmind-core`’s `analyzeFile`, but accepts filesystem paths and `file:` URLs
- * in addition to buffers and browser-style inputs.
+ * Node router: PDF, DOCX, images (OCR), and text. Paths and `file:` URLs are read via `fs`.
  */
-declare function analyzeFile(input: NodeAnalyzeInput, options?: AnalyzeOptions): Promise<AnalysisResult>;
+declare function analyzeFile(input: NodeAnalyzeInput, options?: NodeAnalyzeOptions): Promise<AnalysisResult>;
 
-export { type NodeAnalyzeInput, analyzeFile, bufferToInput, readFileToInput, resolveNodeAnalyzeInput };
+export { type NodeAnalyzeInput, type NodeAnalyzeOptions, analyzeFile, bufferToInput, readFileToInput, resolveNodeAnalyzeInput };

package/dist/index.js

@@ -1,9 +1,129 @@
-import { analyzeFile as analyzeFile$1 } from '@dragon708/docmind-core';
+import { assertValidAnalyzeFileInput, detectFileKind, notImplementedResult, UNKNOWN_FORMAT_WARNING, analyzeText, toUint8Array, isNamedInput, isBinaryInput, isBlob, isFile } from '@dragon708/docmind-shared';
+import { analyzeDocx } from '@dragon708/docmind-docx';
+import { ocr } from '@dragon708/docmind-ocr';
+import { analyzePdf } from '@dragon708/docmind-pdf';
 import { readFile } from 'fs/promises';
 import { basename } from 'path';
 import { fileURLToPath } from 'url';
 
 // src/analyze.ts
+function isByteBackedInput(input) {
+  return isNamedInput(input) || isBinaryInput(input) || isBlob(input) || isFile(input);
+}
+async function bytesFromDetectInput(input) {
+  if (!isByteBackedInput(input)) {
+    return new Uint8Array(0);
+  }
+  return toUint8Array(input);
+}
+
+// src/analyzers/docx.ts
+async function analyzeDocxForNode(input, signal) {
+  if (signal?.aborted) {
+    const err = new Error("The operation was aborted");
+    err.name = "AbortError";
+    throw err;
+  }
+  const data = await bytesFromDetectInput(input);
+  if (data.byteLength === 0) {
+    return {
+      fileKind: "docx",
+      analyzer: "docx",
+      status: "ok",
+      kind: "docx",
+      text: "",
+      html: "",
+      warnings: ["No document bytes were provided for analysis."]
+    };
+  }
+  const r = await analyzeDocx(data);
+  return {
+    fileKind: "docx",
+    analyzer: "docx",
+    status: "ok",
+    kind: "docx",
+    text: r.text,
+    html: r.html,
+    warnings: [...r.warnings]
+  };
+}
+async function analyzeImageForNode(input, options) {
+  if (options?.signal?.aborted) {
+    const err = new Error("The operation was aborted");
+    err.name = "AbortError";
+    throw err;
+  }
+  const data = await bytesFromDetectInput(input);
+  if (data.byteLength === 0) {
+    return {
+      fileKind: "image",
+      analyzer: "image",
+      status: "ok",
+      kind: "image",
+      text: "",
+      confidence: 0,
+      ocrUsed: true,
+      warnings: ["No image bytes were provided for analysis."]
+    };
+  }
+  const ocrOpts = {
+    ...options?.ocr ?? {},
+    signal: options?.ocr?.signal ?? options?.signal
+  };
+  const r = await ocr(data, ocrOpts);
+  return {
+    fileKind: "image",
+    analyzer: "image",
+    status: "ok",
+    kind: "image",
+    text: r.text,
+    confidence: r.confidence,
+    ocrUsed: r.ocrUsed,
+    warnings: []
+  };
+}
+async function analyzePdfForNode(input, options) {
+  if (options?.signal?.aborted) {
+    const err = new Error("The operation was aborted");
+    err.name = "AbortError";
+    throw err;
+  }
+  const data = await bytesFromDetectInput(input);
+  if (data.byteLength === 0) {
+    return {
+      fileKind: "pdf",
+      analyzer: "pdf",
+      status: "ok",
+      kind: "pdf",
+      text: "",
+      pages: 0,
+      metadata: { info: {} },
+      warnings: ["No document bytes were provided for analysis."],
+      needsOCR: false,
+      ocrUsed: false
+    };
+  }
+  const userPdf = options?.pdf;
+  const pdfOpts = {
+    ...userPdf,
+    ocr: userPdf?.ocr ?? "auto",
+    ocrLangs: userPdf?.ocrLangs ?? options?.ocr?.langs,
+    signal: userPdf?.signal ?? options?.signal
+  };
+  const r = await analyzePdf(data, pdfOpts);
+  return {
+    fileKind: "pdf",
+    analyzer: "pdf",
+    status: "ok",
+    kind: "pdf",
+    text: r.text,
+    pages: r.pages,
+    metadata: r.metadata,
+    warnings: [...r.warnings],
+    needsOCR: r.needsOCR,
+    ocrUsed: r.ocrUsed
+  };
+}
 function toPathString(pathOrUrl) {
   return pathOrUrl instanceof URL ? fileURLToPath(pathOrUrl) : pathOrUrl;
 }
@@ -27,8 +147,26 @@ async function resolveNodeAnalyzeInput(input) {
 
 // src/analyze.ts
 async function analyzeFile(input, options) {
+  if (options?.signal?.aborted) {
+    const err = new Error("The operation was aborted");
+    err.name = "AbortError";
+    throw err;
+  }
   const resolved = await resolveNodeAnalyzeInput(input);
-  return analyzeFile$1(resolved, options);
+  assertValidAnalyzeFileInput(resolved);
+  const fileKind = detectFileKind(resolved);
+  switch (fileKind) {
+    case "pdf":
+      return analyzePdfForNode(resolved, options);
+    case "docx":
+      return analyzeDocxForNode(resolved, options?.signal);
+    case "image":
+      return analyzeImageForNode(resolved, options);
+    case "text":
+      return analyzeText(resolved, { signal: options?.signal });
+    default:
+      return notImplementedResult(fileKind, "none", [UNKNOWN_FORMAT_WARNING]);
+  }
 }
 
 export { analyzeFile, bufferToInput, readFileToInput, resolveNodeAnalyzeInput };

package/dist/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["../src/toCoreInput.ts","../src/analyze.ts"],"names":["analyzeWithCore"],"mappings":";;;;;;AAaA,SAAS,aAAa,SAAA,EAAiC;AACrD,EAAA,OAAO,SAAA,YAAqB,GAAA,GAAM,aAAA,CAAc,SAAS,CAAA,GAAI,SAAA;AAC/D;AAMA,eAAsB,gBAAgB,IAAA,EAAiD;AACrF,EAAA,MAAM,MAAA,GAAS,aAAa,IAAI,CAAA;AAChC,EAAA,MAAM,IAAA,GAAO,MAAM,QAAA,CAAS,MAAM,CAAA;AAClC,EAAA,OAAO;AAAA,IACL,IAAA;AAAA,IACA,IAAA,EAAM,SAAS,MAAM;AAAA,GACvB;AACF;AAGO,SAAS,aAAA,CAAc,QAAgB,IAAA,EAAmC;AAC/E,EAAA,OAAO,IAAA,KAAS,SAAY,EAAE,IAAA,EAAM,QAAQ,IAAA,EAAK,GAAI,EAAE,IAAA,EAAM,MAAA,EAAO;AACtE;AAMA,eAAsB,wBAAwB,KAAA,EAAuD;AACnG,EAAA,IAAI,OAAO,KAAA,KAAU,QAAA,IAAY,KAAA,YAAiB,GAAA,EAAK;AACrD,IAAA,OAAO,gBAAgB,KAAK,CAAA;AAAA,EAC9B;AACA,EAAA,OAAO,KAAA;AACT;;;ACpCA,eAAsB,WAAA,CACpB,OACA,OAAA,EACyB;AACzB,EAAA,MAAM,QAAA,GAAW,MAAM,uBAAA,CAAwB,KAAK,CAAA;AACpD,EAAA,OAAOA,aAAA,CAAgB,UAAU,OAAO,CAAA;AAC1C","file":"index.js","sourcesContent":["import type { DetectFileKindInput } from \"@dragon708/docmind-core\";\nimport type { NamedInput } from \"@dragon708/docmind-shared\";\nimport { readFile } from \"node:fs/promises\";\nimport { basename } from \"node:path\";\nimport { fileURLToPath } from \"node:url\";\n\n/**\n * Inputs accepted by {@link analyzeFile} in this package.\n * Paths and `file:` URLs are read with `fs`; other values pass through as core’s\n * {@link DetectFileKindInput}.\n */\nexport type NodeAnalyzeInput = string | URL | DetectFileKindInput;\n\nfunction toPathString(pathOrUrl: string | URL): string {\n  return pathOrUrl instanceof URL ? fileURLToPath(pathOrUrl) : pathOrUrl;\n}\n\n/**\n * Reads a file from disk into a {@link NamedInput} suitable for `@dragon708/docmind-core`\n * (binary `Buffer`, basename as `name` for extension / MIME hints).\n */\nexport async function readFileToInput(path: string | URL): Promise<NamedInput<Buffer>> {\n  const fsPath = toPathString(path);\n  const data = await readFile(fsPath);\n  return {\n    data,\n    name: basename(fsPath),\n  };\n}\n\n/** Wraps a `Buffer` as a named payload when you already know the filename. */\nexport function bufferToInput(buffer: Buffer, name?: string): NamedInput<Buffer> {\n  return name !== undefined ? { data: buffer, name } : { data: buffer };\n}\n\n/**\n * Resolves paths / `file:` URLs to core’s internal shape; leaves other\n * {@link DetectFileKindInput} values untouched.\n */\nexport async function resolveNodeAnalyzeInput(input: NodeAnalyzeInput): Promise<DetectFileKindInput> {\n  if (typeof input === \"string\" || input instanceof URL) {\n    return readFileToInput(input);\n  }\n  return input;\n}\n","import { analyzeFile as analyzeWithCore } from \"@dragon708/docmind-core\";\nimport type { AnalyzeOptions, AnalysisResult } from \"@dragon708/docmind-core\";\nimport { resolveNodeAnalyzeInput, type NodeAnalyzeInput } from \"./toCoreInput.js\";\n\n/**\n * Like `@dragon708/docmind-core`’s `analyzeFile`, but accepts filesystem paths and `file:` URLs\n * in addition to buffers and browser-style inputs.\n */\nexport async function analyzeFile(\n  input: NodeAnalyzeInput,\n  options?: AnalyzeOptions,\n): Promise<AnalysisResult> {\n  const resolved = await resolveNodeAnalyzeInput(input);\n  return analyzeWithCore(resolved, options);\n}\n"]}
+{"version":3,"sources":["../src/inputBytes.ts","../src/analyzers/docx.ts","../src/analyzers/image.ts","../src/analyzers/pdf.ts","../src/resolveNodeInput.ts","../src/analyze.ts"],"names":["extractDocx","runPdf"],"mappings":";;;;;;;;;AASO,SAAS,kBAAkB,KAAA,EAAoD;AACpF,EAAA,OAAO,YAAA,CAAa,KAAK,CAAA,IAAK,aAAA,CAAc,KAAK,KAAK,MAAA,CAAO,KAAK,CAAA,IAAK,MAAA,CAAO,KAAK,CAAA;AACrF;AAGA,eAAsB,qBAAqB,KAAA,EAAiD;AAC1F,EAAA,IAAI,CAAC,iBAAA,CAAkB,KAAK,CAAA,EAAG;AAC7B,IAAA,OAAO,IAAI,WAAW,CAAC,CAAA;AAAA,EACzB;AACA,EAAA,OAAO,aAAa,KAAK,CAAA;AAC3B;;;ACZA,eAAsB,kBAAA,CACpB,OACA,MAAA,EACyB;AACzB,EAAA,IAAI,QAAQ,OAAA,EAAS;AACnB,IAAA,MAAM,GAAA,GAAM,IAAI,KAAA,CAAM,2BAA2B,CAAA;AACjD,IAAA,GAAA,CAAI,IAAA,GAAO,YAAA;AACX,IAAA,MAAM,GAAA;AAAA,EACR;AAEA,EAAA,MAAM,IAAA,GAAO,MAAM,oBAAA,CAAqB,KAAK,CAAA;AAC7C,EAAA,IAAI,IAAA,CAAK,eAAe,CAAA,EAAG;AACzB,IAAA,OAAO;AAAA,MACL,QAAA,EAAU,MAAA;AAAA,MACV,QAAA,EAAU,MAAA;AAAA,MACV,MAAA,EAAQ,IAAA;AAAA,MACR,IAAA,EAAM,MAAA;AAAA,MACN,IAAA,EAAM,EAAA;AAAA,MACN,IAAA,EAAM,EAAA;AAAA,MACN,QAAA,EAAU,CAAC,+CAA+C;AAAA,KAC5D;AAAA,EACF;AAEA,EAAA,MAAM,CAAA,GAAI,MAAMA,WAAA,CAAY,IAAI,CAAA;AAChC,EAAA,OAAO;AAAA,IACL,QAAA,EAAU,MAAA;AAAA,IACV,QAAA,EAAU,MAAA;AAAA,IACV,MAAA,EAAQ,IAAA;AAAA,IACR,IAAA,EAAM,MAAA;AAAA,IACN,MAAM,CAAA,CAAE,IAAA;AAAA,IACR,MAAM,CAAA,CAAE,IAAA;AAAA,IACR,QAAA,EAAU,CAAC,GAAG,CAAA,CAAE,QAAQ;AAAA,GAC1B;AACF;AChCA,eAAsB,mBAAA,CACpB,OACA,OAAA,EACyB;AACzB,EAAA,IAAI,OAAA,EAAS,QAAQ,OAAA,EAAS;AAC5B,IAAA,MAAM,GAAA,GAAM,IAAI,KAAA,CAAM,2BAA2B,CAAA;AACjD,IAAA,GAAA,CAAI,IAAA,GAAO,YAAA;AACX,IAAA,MAAM,GAAA;AAAA,EACR;AAEA,EAAA,MAAM,IAAA,GAAO,MAAM,oBAAA,CAAqB,KAAK,CAAA;AAC7C,EAAA,IAAI,IAAA,CAAK,eAAe,CAAA,EAAG;AACzB,IAAA,OAAO;AAAA,MACL,QAAA,EAAU,OAAA;AAAA,MACV,QAAA,EAAU,OAAA;AAAA,MACV,MAAA,EAAQ,IAAA;AAAA,MACR,IAAA,EAAM,OAAA;AAAA,MACN,IAAA,EAAM,EAAA;AAAA,MACN,UAAA,EAAY,CAAA;AAAA,MACZ,OAAA,EAAS,IAAA;AAAA,MACT,QAAA,EAAU,CAAC,4CAA4C;AAAA,KACzD;AAAA,EACF;AAEA,EAAA,MAAM,OAAA,GAAU;AAAA,IACd,GAAI,OAAA,EAAS,GAAA,IAAO,EAAC;AAAA,IACrB,MAAA,EAAQ,OAAA,EAAS,GAAA,EAAK,MAAA,IAAU,OAAA,EAAS;AAAA,GAC3C;AAEA,EAAA,MAAM,CAAA,GAAI,MAAM,GAAA,CAAI,IAAA,EAAM,OAAO,CAAA;AACjC,EAAA,OAAO;AAAA,IACL,QAAA,EAAU,OAAA;AAAA,IACV,QAAA,EAAU,OAAA;AAAA,IACV,MAAA,EAAQ,IAAA;AAAA,IACR,IAAA,EAAM,OAAA;AAAA,IACN,MAAM,CAAA,CAAE,IAAA;AAAA,IACR,YAAY,CAAA,CAAE,UAAA;AAAA,IACd,SAAS,CAAA,CAAE,OAAA;AAAA,IACX,UAAU;AAAC,GACb;AACF;AClCA,eAAsB,iBAAA,CACpB,OACA,OAAA,EACyB;AACzB,EAAA,IAAI,OAAA,EAAS,QAAQ,OAAA,EAAS;AAC5B,IAAA,MAAM,GAAA,GAAM,IAAI,KAAA,CAAM,2BAA2B,CAAA;AACjD,IAAA,GAAA,CAAI,IAAA,GAAO,YAAA;AACX,IAAA,MAAM,GAAA;AAAA,EACR;AAEA,EAAA,MAAM,IAAA,GAAO,MAAM,oBAAA,CAAqB,KAAK,CAAA;AAC7C,EAAA,IAAI,IAAA,CAAK,eAAe,CAAA,EAAG;AACzB,IAAA,OAAO;AAAA,MACL,QAAA,EAAU,KAAA;AAAA,MACV,QAAA,EAAU,KAAA;AAAA,MACV,MAAA,EAAQ,IAAA;AAAA,MACR,IAAA,EAAM,KAAA;AAAA,MACN,IAAA,EAAM,EAAA;AAAA,MACN,KAAA,EAAO,CAAA;AAAA,MACP,QAAA,EAAU,EAAE,IAAA,EAAM,EAAC,EAAE;AAAA,MACrB,QAAA,EAAU,CAAC,+CAA+C,CAAA;AAAA,MAC1D,QAAA,EAAU,KAAA;AAAA,MACV,OAAA,EAAS;AAAA,KACX;AAAA,EACF;AAEA,EAAA,MAAM,UAAU,OAAA,EAAS,GAAA;AACzB,EAAA,MAAM,OAAA,GAA6B;AAAA,IACjC,GAAG,OAAA;AAAA,IACH,GAAA,EAAK,SAAS,GAAA,IAAO,MAAA;AAAA,IACrB,QAAA,EAAU,OAAA,EAAS,QAAA,IAAY,OAAA,EAAS,GAAA,EAAK,KAAA;AAAA,IAC7C,MAAA,EAAQ,OAAA,EAAS,MAAA,IAAU,OAAA,EAAS;AAAA,GACtC;AAEA,EAAA,MAAM,CAAA,GAAI,MAAMC,UAAA,CAAO,IAAA,EAAM,OAAO,CAAA;AACpC,EAAA,OAAO;AAAA,IACL,QAAA,EAAU,KAAA;AAAA,IACV,QAAA,EAAU,KAAA;AAAA,IACV,MAAA,EAAQ,IAAA;AAAA,IACR,IAAA,EAAM,KAAA;AAAA,IACN,MAAM,CAAA,CAAE,IAAA;AAAA,IACR,OAAO,CAAA,CAAE,KAAA;AAAA,IACT,UAAU,CAAA,CAAE,QAAA;AAAA,IACZ,QAAA,EAAU,CAAC,GAAG,CAAA,CAAE,QAAQ,CAAA;AAAA,IACxB,UAAU,CAAA,CAAE,QAAA;AAAA,IACZ,SAAS,CAAA,CAAE;AAAA,GACb;AACF;AClDA,SAAS,aAAa,SAAA,EAAiC;AACrD,EAAA,OAAO,SAAA,YAAqB,GAAA,GAAM,aAAA,CAAc,SAAS,CAAA,GAAI,SAAA;AAC/D;AAKA,eAAsB,gBAAgB,IAAA,EAAiD;AACrF,EAAA,MAAM,MAAA,GAAS,aAAa,IAAI,CAAA;AAChC,EAAA,MAAM,IAAA,GAAO,MAAM,QAAA,CAAS,MAAM,CAAA;AAClC,EAAA,OAAO;AAAA,IACL,IAAA;AAAA,IACA,IAAA,EAAM,SAAS,MAAM;AAAA,GACvB;AACF;AAGO,SAAS,aAAA,CAAc,QAAgB,IAAA,EAAmC;AAC/E,EAAA,OAAO,IAAA,KAAS,SAAY,EAAE,IAAA,EAAM,QAAQ,IAAA,EAAK,GAAI,EAAE,IAAA,EAAM,MAAA,EAAO;AACtE;AAKA,eAAsB,wBAAwB,KAAA,EAAuD;AACnG,EAAA,IAAI,OAAO,KAAA,KAAU,QAAA,IAAY,KAAA,YAAiB,GAAA,EAAK;AACrD,IAAA,OAAO,gBAAgB,KAAK,CAAA;AAAA,EAC9B;AACA,EAAA,OAAO,KAAA;AACT;;;ACvBA,eAAsB,WAAA,CACpB,OACA,OAAA,EACyB;AACzB,EAAA,IAAI,OAAA,EAAS,QAAQ,OAAA,EAAS;AAC5B,IAAA,MAAM,GAAA,GAAM,IAAI,KAAA,CAAM,2BAA2B,CAAA;AACjD,IAAA,GAAA,CAAI,IAAA,GAAO,YAAA;AACX,IAAA,MAAM,GAAA;AAAA,EACR;AAEA,EAAA,MAAM,QAAA,GAAW,MAAM,uBAAA,CAAwB,KAAK,CAAA;AACpD,EAAA,2BAAA,CAA4B,QAAQ,CAAA;AAEpC,EAAA,MAAM,QAAA,GAAW,eAAe,QAAQ,CAAA;AAExC,EAAA,QAAQ,QAAA;AAAU,IAChB,KAAK,KAAA;AACH,MAAA,OAAO,iBAAA,CAAkB,UAAiC,OAAO,CAAA;AAAA,IACnE,KAAK,MAAA;AACH,MAAA,OAAO,kBAAA,CAAmB,QAAA,EAAiC,OAAA,EAAS,MAAM,CAAA;AAAA,IAC5E,KAAK,OAAA;AACH,MAAA,OAAO,mBAAA,CAAoB,UAAiC,OAAO,CAAA;AAAA,IACrE,KAAK,MAAA;AACH,MAAA,OAAO,YAAY,QAAA,EAAiC,EAAE,MAAA,EAAQ,OAAA,EAAS,QAAQ,CAAA;AAAA,IACjF;AACE,MAAA,OAAO,oBAAA,CAAqB,QAAA,EAAU,MAAA,EAAQ,CAAC,sBAAsB,CAAC,CAAA;AAAA;AAE5E","file":"index.js","sourcesContent":["import type { DetectFileKindInput, FileLikeInput } from \"@dragon708/docmind-shared\";\nimport {\n  isBinaryInput,\n  isBlob,\n  isFile,\n  isNamedInput,\n  toUint8Array,\n} from \"@dragon708/docmind-shared\";\n\nexport function isByteBackedInput(input: DetectFileKindInput): input is FileLikeInput {\n  return isNamedInput(input) || isBinaryInput(input) || isBlob(input) || isFile(input);\n}\n\n/** Resolves bytes when the input carries a body; otherwise an empty `Uint8Array`. */\nexport async function bytesFromDetectInput(input: DetectFileKindInput): Promise<Uint8Array> {\n  if (!isByteBackedInput(input)) {\n    return new Uint8Array(0);\n  }\n  return toUint8Array(input);\n}\n","import { analyzeDocx as extractDocx } from \"@dragon708/docmind-docx\";\nimport type { AnalysisResult, DetectFileKindInput } from \"@dragon708/docmind-shared\";\nimport { bytesFromDetectInput } from \"../inputBytes.js\";\n\n/**\n * DOCX → `@dragon708/docmind-docx`.\n */\nexport async function analyzeDocxForNode(\n  input: DetectFileKindInput,\n  signal?: AbortSignal,\n): Promise<AnalysisResult> {\n  if (signal?.aborted) {\n    const err = new Error(\"The operation was aborted\");\n    err.name = \"AbortError\";\n    throw err;\n  }\n\n  const data = await bytesFromDetectInput(input);\n  if (data.byteLength === 0) {\n    return {\n      fileKind: \"docx\",\n      analyzer: \"docx\",\n      status: \"ok\",\n      kind: \"docx\",\n      text: \"\",\n      html: \"\",\n      warnings: [\"No document bytes were provided for analysis.\"],\n    };\n  }\n\n  const r = await extractDocx(data);\n  return {\n    fileKind: \"docx\",\n    analyzer: \"docx\",\n    status: \"ok\",\n    kind: \"docx\",\n    text: r.text,\n    html: r.html,\n    warnings: [...r.warnings],\n  };\n}\n","import { ocr } from \"@dragon708/docmind-ocr\";\nimport type { AnalysisResult, DetectFileKindInput } from \"@dragon708/docmind-shared\";\nimport type { NodeAnalyzeOptions } from \"../nodeAnalyzeOptions.js\";\nimport { bytesFromDetectInput } from \"../inputBytes.js\";\n\n/**\n * Image → `@dragon708/docmind-ocr`.\n */\nexport async function analyzeImageForNode(\n  input: DetectFileKindInput,\n  options?: NodeAnalyzeOptions,\n): Promise<AnalysisResult> {\n  if (options?.signal?.aborted) {\n    const err = new Error(\"The operation was aborted\");\n    err.name = \"AbortError\";\n    throw err;\n  }\n\n  const data = await bytesFromDetectInput(input);\n  if (data.byteLength === 0) {\n    return {\n      fileKind: \"image\",\n      analyzer: \"image\",\n      status: \"ok\",\n      kind: \"image\",\n      text: \"\",\n      confidence: 0,\n      ocrUsed: true,\n      warnings: [\"No image bytes were provided for analysis.\"],\n    };\n  }\n\n  const ocrOpts = {\n    ...(options?.ocr ?? {}),\n    signal: options?.ocr?.signal ?? options?.signal,\n  };\n\n  const r = await ocr(data, ocrOpts);\n  return {\n    fileKind: \"image\",\n    analyzer: \"image\",\n    status: \"ok\",\n    kind: \"image\",\n    text: r.text,\n    confidence: r.confidence,\n    ocrUsed: r.ocrUsed,\n    warnings: [],\n  };\n}\n","import { analyzePdf as runPdf } from \"@dragon708/docmind-pdf\";\nimport type { PdfAnalyzeOptions } from \"@dragon708/docmind-pdf\";\nimport type { AnalysisResult, DetectFileKindInput } from \"@dragon708/docmind-shared\";\nimport type { NodeAnalyzeOptions } from \"../nodeAnalyzeOptions.js\";\nimport { bytesFromDetectInput } from \"../inputBytes.js\";\n\n/**\n * PDF → `@dragon708/docmind-pdf` (Node / pdf-parse + OCR).\n *\n * Unlike `analyzePdf` from `@dragon708/docmind-pdf` (OCR off unless set), `analyzeFile` defaults\n * to `pdf.ocr: \"auto\"`: when the PDF has\n * pages but almost no extractable text (typical scan), the raster OCR pipeline runs. Pass\n * `pdf: { ocr: \"off\" }` to skip OCR for speed.\n */\nexport async function analyzePdfForNode(\n  input: DetectFileKindInput,\n  options?: NodeAnalyzeOptions,\n): Promise<AnalysisResult> {\n  if (options?.signal?.aborted) {\n    const err = new Error(\"The operation was aborted\");\n    err.name = \"AbortError\";\n    throw err;\n  }\n\n  const data = await bytesFromDetectInput(input);\n  if (data.byteLength === 0) {\n    return {\n      fileKind: \"pdf\",\n      analyzer: \"pdf\",\n      status: \"ok\",\n      kind: \"pdf\",\n      text: \"\",\n      pages: 0,\n      metadata: { info: {} },\n      warnings: [\"No document bytes were provided for analysis.\"],\n      needsOCR: false,\n      ocrUsed: false,\n    };\n  }\n\n  const userPdf = options?.pdf;\n  const pdfOpts: PdfAnalyzeOptions = {\n    ...userPdf,\n    ocr: userPdf?.ocr ?? \"auto\",\n    ocrLangs: userPdf?.ocrLangs ?? options?.ocr?.langs,\n    signal: userPdf?.signal ?? options?.signal,\n  };\n\n  const r = await runPdf(data, pdfOpts);\n  return {\n    fileKind: \"pdf\",\n    analyzer: \"pdf\",\n    status: \"ok\",\n    kind: \"pdf\",\n    text: r.text,\n    pages: r.pages,\n    metadata: r.metadata,\n    warnings: [...r.warnings],\n    needsOCR: r.needsOCR,\n    ocrUsed: r.ocrUsed,\n  };\n}\n","import type { DetectFileKindInput, NamedInput } from \"@dragon708/docmind-shared\";\nimport { readFile } from \"node:fs/promises\";\nimport { basename } from \"node:path\";\nimport { fileURLToPath } from \"node:url\";\n\n/**\n * Inputs accepted by {@link analyzeFile} in this package.\n * Paths and `file:` URLs are read with `fs`; other values pass through as {@link DetectFileKindInput}.\n */\nexport type NodeAnalyzeInput = string | URL | DetectFileKindInput;\n\nfunction toPathString(pathOrUrl: string | URL): string {\n  return pathOrUrl instanceof URL ? fileURLToPath(pathOrUrl) : pathOrUrl;\n}\n\n/**\n * Reads a file from disk into a {@link NamedInput} (binary `Buffer`, basename as `name` for hints).\n */\nexport async function readFileToInput(path: string | URL): Promise<NamedInput<Buffer>> {\n  const fsPath = toPathString(path);\n  const data = await readFile(fsPath);\n  return {\n    data,\n    name: basename(fsPath),\n  };\n}\n\n/** Wraps a `Buffer` as a named payload when you already know the filename. */\nexport function bufferToInput(buffer: Buffer, name?: string): NamedInput<Buffer> {\n  return name !== undefined ? { data: buffer, name } : { data: buffer };\n}\n\n/**\n * Resolves paths / `file:` URLs to a {@link DetectFileKindInput}; leaves other values untouched.\n */\nexport async function resolveNodeAnalyzeInput(input: NodeAnalyzeInput): Promise<DetectFileKindInput> {\n  if (typeof input === \"string\" || input instanceof URL) {\n    return readFileToInput(input);\n  }\n  return input;\n}\n","import type { AnalysisResult, DetectFileKindInput } from \"@dragon708/docmind-shared\";\nimport {\n  analyzeText,\n  assertValidAnalyzeFileInput,\n  detectFileKind,\n  notImplementedResult,\n  UNKNOWN_FORMAT_WARNING,\n} from \"@dragon708/docmind-shared\";\nimport { analyzeDocxForNode } from \"./analyzers/docx.js\";\nimport { analyzeImageForNode } from \"./analyzers/image.js\";\nimport { analyzePdfForNode } from \"./analyzers/pdf.js\";\nimport type { NodeAnalyzeOptions } from \"./nodeAnalyzeOptions.js\";\nimport { resolveNodeAnalyzeInput, type NodeAnalyzeInput } from \"./resolveNodeInput.js\";\n\n/**\n * Node router: PDF, DOCX, images (OCR), and text. Paths and `file:` URLs are read via `fs`.\n */\nexport async function analyzeFile(\n  input: NodeAnalyzeInput,\n  options?: NodeAnalyzeOptions,\n): Promise<AnalysisResult> {\n  if (options?.signal?.aborted) {\n    const err = new Error(\"The operation was aborted\");\n    err.name = \"AbortError\";\n    throw err;\n  }\n\n  const resolved = await resolveNodeAnalyzeInput(input);\n  assertValidAnalyzeFileInput(resolved);\n\n  const fileKind = detectFileKind(resolved);\n\n  switch (fileKind) {\n    case \"pdf\":\n      return analyzePdfForNode(resolved as DetectFileKindInput, options);\n    case \"docx\":\n      return analyzeDocxForNode(resolved as DetectFileKindInput, options?.signal);\n    case \"image\":\n      return analyzeImageForNode(resolved as DetectFileKindInput, options);\n    case \"text\":\n      return analyzeText(resolved as DetectFileKindInput, { signal: options?.signal });\n    default:\n      return notImplementedResult(fileKind, \"none\", [UNKNOWN_FORMAT_WARNING]);\n  }\n}\n"]}

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@dragon708/docmind-node",
-  "version": "0.1.0-alpha.1",
-  "description": "Node.js file I/O helpers and core integration for DocMind (does not run in browsers).",
+  "version": "1.0.0",
+  "description": "Node.js DocMind entry: PDF, DOCX, OCR, text, and fs helpers.",
   "type": "module",
   "main": "./dist/index.js",
   "module": "./dist/index.js",
@@ -31,13 +31,16 @@
   ],
   "license": "MIT",
   "dependencies": {
-    "@dragon708/docmind-core": "^0.1.0-alpha.1",
-    "@dragon708/docmind-shared": "^0.1.0-alpha.1"
+    "@dragon708/docmind-docx": "^1.0.0",
+    "@dragon708/docmind-ocr": "^1.0.0",
+    "@dragon708/docmind-pdf": "^1.0.0",
+    "@dragon708/docmind-shared": "^1.0.0"
   },
   "devDependencies": {
     "@types/node": "^20.19.37",
     "tsup": "^8.5.1",
-    "typescript": "^5.9.3"
+    "typescript": "^5.9.3",
+    "vitest": "^1.6.1"
   },
   "engines": {
     "node": ">=18"