@lumifai/harness-tool-pack-docling 0.1.0

package/README.md

@@ -0,0 +1,81 @@
+# @lumifai/harness-tool-pack-docling
+
+Convert supported workspace files through a user-configured Docling Serve endpoint.
+
+```ts
+import { createDoclingToolPack } from '@lumifai/harness-tool-pack-docling';
+
+const doclingPack = createDoclingToolPack({
+  serviceUrl: process.env.DOCLING_SERVICE_URL!,
+  apiKey: process.env.DOCLING_API_KEY,
+});
+```
+
+Register the pack with the preset tools:
+
+```ts
+createPresetTools({ packs: [doclingPack] });
+```
+
+The pack exposes `process_file`. It reads a file from the workspace, sends it to
+`POST /v1/convert/file`, and writes the selected result to `processed/`.
+Markdown is the default output format; JSON, HTML, text, and doctags are also
+available. Docling automatically determines the input type from the uploaded file.
+
+The pack is disabled by default and requires `ask` permission. Configure the
+service URL and credentials when constructing the pack. Credentials are never
+part of the model-visible tool input. Use `headers` for services that require
+authentication other than Docling's `X-Api-Key` header.
+
+For broker or parser integrations that already have file bytes, use the shared
+converter API. It sends one Docling request and can return every requested
+representation without writing through the workspace:
+
+```ts
+import { createDoclingConverter } from '@lumifai/harness-tool-pack-docling';
+
+const converter = createDoclingConverter({
+  serviceUrl: process.env.DOCLING_SERVICE_URL!,
+  apiKey: process.env.DOCLING_API_KEY,
+});
+
+const result = await converter.convertFile({
+  bytes,
+  filename: 'report.pdf',
+  mimeType: 'application/pdf',
+  outputFormats: ['md', 'json', 'html', 'text', 'doctags'],
+});
+```
+
+The converter returns structured JSON content, status, processing time, and
+errors. `process_file` remains the agent-facing wrapper for workspace paths and
+writes one selected representation under `processed/`.
+
+For reusable application integrations that need shared defaults, normalization,
+or workspace persistence, use the high-level processor:
+
+```ts
+import { createDoclingDocumentProcessor } from '@lumifai/harness-tool-pack-docling';
+
+const processor = createDoclingDocumentProcessor({
+  serviceUrl: process.env.DOCLING_SERVICE_URL!,
+  apiKey: process.env.DOCLING_API_KEY,
+});
+
+const parsed = await processor.parse({
+  filename: 'report.pdf',
+  bytes,
+  mimeType: 'application/pdf',
+});
+
+const workspaceParsed = await processor.parseWorkspaceFile('/uploads/report.pdf', {
+  filesystem,
+  persistArtifacts: true,
+});
+```
+
+Programmatic parsing returns `processorId: 'docling'` and defaults to Markdown, JSON, HTML, text, and doctags with
+OCR enabled, accurate tables, the standard pipeline, and placeholder images.
+Callers can override these options per request. Workspace filesystems are
+provided per call, so local, mounted, and object-backed filesystem providers can
+use the same processor.

package/dist/converter.d.ts

@@ -0,0 +1,33 @@
+export type DoclingOutputFormat = 'md' | 'json' | 'html' | 'text' | 'doctags';
+export type DoclingConversionStatus = 'success' | 'partial_success' | 'skipped' | 'failure';
+export type DoclingFormValue = string | number | boolean | readonly (string | number | boolean)[];
+export interface DoclingToolPackOptions {
+    serviceUrl: string;
+    apiKey?: string;
+    headers?: Record<string, string>;
+    timeoutMs?: number;
+    fetchFn?: typeof fetch;
+}
+export interface DoclingConvertFileInput {
+    bytes: Uint8Array;
+    filename: string;
+    mimeType?: string;
+    outputFormats: readonly DoclingOutputFormat[];
+    doOcr?: boolean;
+    forceOcr?: boolean;
+    tableMode?: 'fast' | 'accurate';
+    imageExportMode?: 'placeholder' | 'embedded' | 'referenced';
+    pipeline?: 'standard' | 'vlm';
+    /** Additional Docling Serve form fields not covered by convenience options. */
+    formOptions?: Readonly<Record<string, DoclingFormValue | null | undefined>>;
+}
+export interface DoclingConversionResult {
+    outputs: Partial<Record<DoclingOutputFormat, unknown>>;
+    status: DoclingConversionStatus;
+    processingTimeSec?: number;
+    errors: unknown[];
+}
+export interface DoclingConverter {
+    convertFile(input: DoclingConvertFileInput): Promise<DoclingConversionResult>;
+}
+export declare function createDoclingConverter(options: DoclingToolPackOptions): DoclingConverter;

package/dist/converter.js

@@ -0,0 +1,282 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.createDoclingConverter = createDoclingConverter;
+const DEFAULT_DOCLING_TIMEOUT_MS = 120_000;
+const SUPPORTED_OUTPUT_FORMATS = new Set([
+    'md',
+    'json',
+    'html',
+    'text',
+    'doctags',
+]);
+function createDoclingConverter(options) {
+    const serviceUrl = normalizeServiceUrl(options.serviceUrl);
+    const timeoutMs = normalizeTimeout(options.timeoutMs);
+    const fetchFn = options.fetchFn ?? globalThis.fetch;
+    if (!fetchFn) {
+        throw new Error('Docling processing requires a fetch implementation.');
+    }
+    return {
+        convertFile: (input) => convertFile(serviceUrl, timeoutMs, fetchFn, options, normalizeInput(input)),
+    };
+}
+async function convertFile(serviceUrl, timeoutMs, fetchFn, options, input) {
+    const form = new FormData();
+    const fileBytes = new Uint8Array(input.bytes.byteLength);
+    fileBytes.set(input.bytes);
+    form.append('files', new Blob([fileBytes], {
+        type: input.mimeType ?? resolveMimeType(input.filename),
+    }), input.filename);
+    for (const outputFormat of input.outputFormats) {
+        form.append('to_formats', outputFormat);
+    }
+    const convenienceOptions = {
+        do_ocr: input.doOcr,
+        force_ocr: input.forceOcr,
+        table_mode: input.tableMode,
+        image_export_mode: input.imageExportMode,
+        pipeline: input.pipeline,
+    };
+    appendFormOptions(form, input.formOptions, convenienceOptions);
+    for (const [key, value] of Object.entries(convenienceOptions)) {
+        appendOptionalFormValue(form, key, value);
+    }
+    const controller = new AbortController();
+    const timeout = setTimeout(() => controller.abort(), timeoutMs);
+    try {
+        let response;
+        try {
+            response = await fetchFn(`${serviceUrl}/v1/convert/file`, {
+                method: 'POST',
+                headers: createRequestHeaders(options),
+                body: form,
+                signal: controller.signal,
+            });
+        }
+        catch (error) {
+            if (controller.signal.aborted) {
+                throw new Error(`Docling request timed out after ${timeoutMs} ms.`, { cause: error });
+            }
+            throw new Error(`Docling request failed: ${redactSecrets(errorMessage(error), options)}`, {
+                cause: error,
+            });
+        }
+        if (!response.ok) {
+            const responseText = await readResponseText(response, controller.signal);
+            throw new Error(redactSecrets(`Docling request failed with HTTP ${response.status}${responseText ? `: ${responseText}` : '.'}`, options));
+        }
+        const payload = await parseDoclingResponse(response, options, controller.signal);
+        const status = normalizeStatus(payload.status);
+        const outputs = extractOutputs(payload, input.outputFormats);
+        const outputValues = Object.values(outputs);
+        if ((status === 'failure' || status === 'skipped') && outputValues.length === 0) {
+            throw new Error(redactSecrets(formatConversionError(status, payload.errors), options));
+        }
+        const missingFormats = input.outputFormats.filter((format) => outputs[format] === undefined);
+        if (status === 'success' && missingFormats.length > 0) {
+            throw new Error(redactSecrets(`Docling returned no ${missingFormats.join(', ')} content for ${input.filename}.`, options));
+        }
+        return {
+            outputs,
+            status,
+            processingTimeSec: payload.processing_time,
+            errors: payload.errors ?? [],
+        };
+    }
+    catch (error) {
+        if (controller.signal.aborted &&
+            !(error instanceof Error && /timed out after/i.test(error.message))) {
+            throw new Error(`Docling request timed out after ${timeoutMs} ms.`, { cause: error });
+        }
+        throw error;
+    }
+    finally {
+        clearTimeout(timeout);
+    }
+}
+function normalizeInput(input) {
+    const filename = input.filename.trim().split(/[\\/]/).pop()?.trim();
+    if (!filename) {
+        throw new Error('Docling filename is required.');
+    }
+    if (input.bytes.byteLength === 0) {
+        throw new Error(`Docling file ${filename} is empty.`);
+    }
+    const outputFormats = [...new Set(input.outputFormats)];
+    if (outputFormats.length === 0) {
+        throw new Error('Docling outputFormats must contain at least one format.');
+    }
+    for (const format of outputFormats) {
+        if (!SUPPORTED_OUTPUT_FORMATS.has(format)) {
+            throw new Error(`Unsupported Docling output format: ${format}`);
+        }
+    }
+    return { ...input, filename, outputFormats };
+}
+function normalizeServiceUrl(serviceUrl) {
+    if (typeof serviceUrl !== 'string' || serviceUrl.trim().length === 0) {
+        throw new Error('Docling serviceUrl is required.');
+    }
+    let parsed;
+    try {
+        parsed = new URL(serviceUrl);
+    }
+    catch {
+        throw new Error('Docling serviceUrl must be a valid HTTP or HTTPS URL.');
+    }
+    if (parsed.protocol !== 'http:' && parsed.protocol !== 'https:') {
+        throw new Error('Docling serviceUrl must use HTTP or HTTPS.');
+    }
+    return serviceUrl.trim().replace(/\/+$/, '');
+}
+function normalizeTimeout(timeoutMs) {
+    if (timeoutMs === undefined) {
+        return DEFAULT_DOCLING_TIMEOUT_MS;
+    }
+    if (!Number.isFinite(timeoutMs) || timeoutMs <= 0) {
+        throw new Error('Docling timeoutMs must be a positive finite number.');
+    }
+    return timeoutMs;
+}
+function createRequestHeaders(options) {
+    const headers = new Headers(options.headers);
+    if (options.apiKey && !headers.has('X-Api-Key')) {
+        headers.set('X-Api-Key', options.apiKey);
+    }
+    // Fetch must generate the multipart boundary. A caller-supplied content type
+    // would replace it and make the upload invalid.
+    headers.delete('content-type');
+    return headers;
+}
+function appendOptionalFormValue(form, name, value) {
+    if (value !== undefined) {
+        appendFormValue(form, name, value);
+    }
+}
+function appendFormOptions(form, options, convenienceOptions) {
+    for (const [name, value] of Object.entries(options ?? {})) {
+        if (name === 'to_formats' || convenienceOptions[name] !== undefined || value == null) {
+            continue;
+        }
+        appendFormValue(form, name, value);
+    }
+}
+function appendFormValue(form, name, value) {
+    if (Array.isArray(value)) {
+        for (const item of value) {
+            form.append(name, String(item));
+        }
+        return;
+    }
+    form.append(name, String(value));
+}
+function extractOutputs(payload, outputFormats) {
+    return Object.fromEntries(outputFormats
+        .map((format) => {
+        const content = payload.document?.[`${format}_content`];
+        return content === undefined || content === null ? undefined : [format, content];
+    })
+        .filter((entry) => entry !== undefined));
+}
+function normalizeStatus(status) {
+    if (status === 'success' ||
+        status === 'partial_success' ||
+        status === 'skipped' ||
+        status === 'failure') {
+        return status;
+    }
+    // Fail closed: missing/unknown Docling statuses must not be treated as success.
+    return 'failure';
+}
+const MAX_DOCLING_RESPONSE_BYTES = 32 * 1024 * 1024;
+async function parseDoclingResponse(response, options, signal) {
+    let payload;
+    try {
+        const text = await readResponseBody(response, MAX_DOCLING_RESPONSE_BYTES, signal);
+        payload = JSON.parse(text);
+    }
+    catch (error) {
+        if (signal?.aborted)
+            throw error;
+        throw new Error(redactSecrets(`Docling returned invalid JSON: ${errorMessage(error)}`, options), {
+            cause: error,
+        });
+    }
+    if (!isDoclingResponse(payload)) {
+        throw new Error('Docling returned an unexpected response shape.');
+    }
+    return payload;
+}
+function isDoclingResponse(value) {
+    return value !== null && typeof value === 'object' && !Array.isArray(value);
+}
+async function readResponseText(response, signal) {
+    try {
+        return (await readResponseBody(response, 2_000, signal)).slice(0, 2_000);
+    }
+    catch {
+        return '';
+    }
+}
+async function readResponseBody(response, maxBytes, signal) {
+    if (!response.body) {
+        const text = await response.text();
+        if (text.length > maxBytes) {
+            throw new Error(`Docling response exceeded the ${maxBytes} byte safety limit.`);
+        }
+        return text;
+    }
+    const reader = response.body.getReader();
+    const chunks = [];
+    let total = 0;
+    try {
+        while (true) {
+            if (signal?.aborted) {
+                throw new Error('Docling response read was aborted.');
+            }
+            const { done, value } = await reader.read();
+            if (done)
+                break;
+            if (!value)
+                continue;
+            total += value.byteLength;
+            if (total > maxBytes) {
+                await reader.cancel();
+                throw new Error(`Docling response exceeded the ${maxBytes} byte safety limit.`);
+            }
+            chunks.push(value);
+        }
+    }
+    finally {
+        reader.releaseLock();
+    }
+    return Buffer.concat(chunks.map((chunk) => Buffer.from(chunk))).toString('utf8');
+}
+function formatConversionError(status, errors) {
+    const details = errors && errors.length > 0 ? `: ${JSON.stringify(errors)}` : '.';
+    return `Docling conversion ${status}${details}`;
+}
+function errorMessage(error) {
+    return error instanceof Error ? error.message : String(error);
+}
+function redactSecrets(message, options) {
+    const secrets = [options.apiKey, ...Object.values(options.headers ?? {})].filter((value) => Boolean(value));
+    return secrets.reduce((redacted, secret) => redacted.split(secret).join('[REDACTED]'), message);
+}
+function resolveMimeType(filename) {
+    const extension = filename.slice(filename.lastIndexOf('.') + 1).toLowerCase();
+    const mimeTypes = {
+        csv: 'text/csv',
+        doc: 'application/msword',
+        docx: 'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
+        html: 'text/html',
+        md: 'text/markdown',
+        pdf: 'application/pdf',
+        ppt: 'application/vnd.ms-powerpoint',
+        pptx: 'application/vnd.openxmlformats-officedocument.presentationml.presentation',
+        xls: 'application/vnd.ms-excel',
+        xlsx: 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
+    };
+    return mimeTypes[extension] ?? 'application/octet-stream';
+}
+//# sourceMappingURL=converter.js.map

package/dist/converter.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"converter.js","sourceRoot":"","sources":["../src/converter.ts"],"names":[],"mappings":";;AAwDA,wDAaC;AArED,MAAM,0BAA0B,GAAG,OAAO,CAAC;AAE3C,MAAM,wBAAwB,GAAG,IAAI,GAAG,CAAsB;IAC5D,IAAI;IACJ,MAAM;IACN,MAAM;IACN,MAAM;IACN,SAAS;CACV,CAAC,CAAC;AAgDH,SAAgB,sBAAsB,CAAC,OAA+B;IACpE,MAAM,UAAU,GAAG,mBAAmB,CAAC,OAAO,CAAC,UAAU,CAAC,CAAC;IAC3D,MAAM,SAAS,GAAG,gBAAgB,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC;IACtD,MAAM,OAAO,GAAG,OAAO,CAAC,OAAO,IAAI,UAAU,CAAC,KAAK,CAAC;IAEpD,IAAI,CAAC,OAAO,EAAE,CAAC;QACb,MAAM,IAAI,KAAK,CAAC,qDAAqD,CAAC,CAAC;IACzE,CAAC;IAED,OAAO;QACL,WAAW,EAAE,CAAC,KAAK,EAAE,EAAE,CACrB,WAAW,CAAC,UAAU,EAAE,SAAS,EAAE,OAAO,EAAE,OAAO,EAAE,cAAc,CAAC,KAAK,CAAC,CAAC;KAC9E,CAAC;AACJ,CAAC;AAED,KAAK,UAAU,WAAW,CACxB,UAAkB,EAClB,SAAiB,EACjB,OAAqB,EACrB,OAA+B,EAC/B,KAA8B;IAE9B,MAAM,IAAI,GAAG,IAAI,QAAQ,EAAE,CAAC;IAC5B,MAAM,SAAS,GAAG,IAAI,UAAU,CAAC,KAAK,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC;IACzD,SAAS,CAAC,GAAG,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;IAC3B,IAAI,CAAC,MAAM,CACT,OAAO,EACP,IAAI,IAAI,CAAC,CAAC,SAAmC,CAAC,EAAE;QAC9C,IAAI,EAAE,KAAK,CAAC,QAAQ,IAAI,eAAe,CAAC,KAAK,CAAC,QAAQ,CAAC;KACxD,CAAC,EACF,KAAK,CAAC,QAAQ,CACf,CAAC;IAEF,KAAK,MAAM,YAAY,IAAI,KAAK,CAAC,aAAa,EAAE,CAAC;QAC/C,IAAI,CAAC,MAAM,CAAC,YAAY,EAAE,YAAY,CAAC,CAAC;IAC1C,CAAC;IAED,MAAM,kBAAkB,GAAG;QACzB,MAAM,EAAE,KAAK,CAAC,KAAK;QACnB,SAAS,EAAE,KAAK,CAAC,QAAQ;QACzB,UAAU,EAAE,KAAK,CAAC,SAAS;QAC3B,iBAAiB,EAAE,KAAK,CAAC,eAAe;QACxC,QAAQ,EAAE,KAAK,CAAC,QAAQ;KACzB,CAAC;IACF,iBAAiB,CAAC,IAAI,EAAE,KAAK,CAAC,WAAW,EAAE,kBAAkB,CAAC,CAAC;IAC/D,KAAK,MAAM,CAAC,GAAG,EAAE,KAAK,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,kBAAkB,CAAC,EAAE,CAAC;QAC9D,uBAAuB,CAAC,IAAI,EAAE,GAAG,EAAE,KAAK,CAAC,CAAC;IAC5C,CAAC;IAED,MAAM,UAAU,GAAG,IAAI,eAAe,EAAE,CAAC;IACzC,MAAM,OAAO,GAAG,UAAU,CAAC,GAAG,EAAE,CAAC,UAAU,CAAC,KAAK,EAAE,EAAE,SAAS,CAAC,CAAC;IAEhE,IAAI,CAAC;QACH,IAAI,QAAkB,CAAC;QACvB,IAAI,CAAC;YACH,QAAQ,GAAG,MAAM,OAAO,CAAC,GAAG,UAAU,kBAAkB,EAAE;gBACxD,MAAM,EAAE,MAAM;gBACd,OAAO,EAAE,oBAAoB,CAAC,OAAO,CAAC;gBACtC,IAAI,EAAE,IAAI;gBACV,MAAM,EAAE,UAAU,CAAC,MAAM;aAC1B,CAAC,CAAC;QACL,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACf,IAAI,UAAU,CAAC,MAAM,CAAC,OAAO,EAAE,CAAC;gBAC9B,MAAM,IAAI,KAAK,CAAC,mCAAmC,SAAS,MAAM,EAAE,EAAE,KAAK,EAAE,KAAK,EAAE,CAAC,CAAC;YACxF,CAAC;YAED,MAAM,IAAI,KAAK,CAAC,2BAA2B,aAAa,CAAC,YAAY,CAAC,KAAK,CAAC,EAAE,OAAO,CAAC,EAAE,EAAE;gBACxF,KAAK,EAAE,KAAK;aACb,CAAC,CAAC;QACL,CAAC;QAED,IAAI,CAAC,QAAQ,CAAC,EAAE,EAAE,CAAC;YACjB,MAAM,YAAY,GAAG,MAAM,gBAAgB,CAAC,QAAQ,EAAE,UAAU,CAAC,MAAM,CAAC,CAAC;YACzE,MAAM,IAAI,KAAK,CACb,aAAa,CACX,oCAAoC,QAAQ,CAAC,MAAM,GAAG,YAAY,CAAC,CAAC,CAAC,KAAK,YAAY,EAAE,CAAC,CAAC,CAAC,GAAG,EAAE,EAChG,OAAO,CACR,CACF,CAAC;QACJ,CAAC;QAED,MAAM,OAAO,GAAG,MAAM,oBAAoB,CAAC,QAAQ,EAAE,OAAO,EAAE,UAAU,CAAC,MAAM,CAAC,CAAC;QACjF,MAAM,MAAM,GAAG,eAAe,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;QAC/C,MAAM,OAAO,GAAG,cAAc,CAAC,OAAO,EAAE,KAAK,CAAC,aAAa,CAAC,CAAC;QAC7D,MAAM,YAAY,GAAG,MAAM,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;QAE5C,IAAI,CAAC,MAAM,KAAK,SAAS,IAAI,MAAM,KAAK,SAAS,CAAC,IAAI,YAAY,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YAChF,MAAM,IAAI,KAAK,CAAC,aAAa,CAAC,qBAAqB,CAAC,MAAM,EAAE,OAAO,CAAC,MAAM,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC;QACzF,CAAC;QAED,MAAM,cAAc,GAAG,KAAK,CAAC,aAAa,CAAC,MAAM,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,OAAO,CAAC,MAAM,CAAC,KAAK,SAAS,CAAC,CAAC;QAC7F,IAAI,MAAM,KAAK,SAAS,IAAI,cAAc,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YACtD,MAAM,IAAI,KAAK,CACb,aAAa,CACX,uBAAuB,cAAc,CAAC,IAAI,CAAC,IAAI,CAAC,gBAAgB,KAAK,CAAC,QAAQ,GAAG,EACjF,OAAO,CACR,CACF,CAAC;QACJ,CAAC;QAED,OAAO;YACL,OAAO;YACP,MAAM;YACN,iBAAiB,EAAE,OAAO,CAAC,eAAe;YAC1C,MAAM,EAAE,OAAO,CAAC,MAAM,IAAI,EAAE;SAC7B,CAAC;IACJ,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,IACE,UAAU,CAAC,MAAM,CAAC,OAAO;YACzB,CAAC,CAAC,KAAK,YAAY,KAAK,IAAI,kBAAkB,CAAC,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,EACnE,CAAC;YACD,MAAM,IAAI,KAAK,CAAC,mCAAmC,SAAS,MAAM,EAAE,EAAE,KAAK,EAAE,KAAK,EAAE,CAAC,CAAC;QACxF,CAAC;QACD,MAAM,KAAK,CAAC;IACd,CAAC;YAAS,CAAC;QACT,YAAY,CAAC,OAAO,CAAC,CAAC;IACxB,CAAC;AACH,CAAC;AAED,SAAS,cAAc,CAAC,KAA8B;IACpD,MAAM,QAAQ,GAAG,KAAK,CAAC,QAAQ,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,GAAG,EAAE,EAAE,IAAI,EAAE,CAAC;IACpE,IAAI,CAAC,QAAQ,EAAE,CAAC;QACd,MAAM,IAAI,KAAK,CAAC,+BAA+B,CAAC,CAAC;IACnD,CAAC;IAED,IAAI,KAAK,CAAC,KAAK,CAAC,UAAU,KAAK,CAAC,EAAE,CAAC;QACjC,MAAM,IAAI,KAAK,CAAC,gBAAgB,QAAQ,YAAY,CAAC,CAAC;IACxD,CAAC;IAED,MAAM,aAAa,GAAG,CAAC,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,aAAa,CAAC,CAAC,CAAC;IACxD,IAAI,aAAa,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QAC/B,MAAM,IAAI,KAAK,CAAC,yDAAyD,CAAC,CAAC;IAC7E,CAAC;IAED,KAAK,MAAM,MAAM,IAAI,aAAa,EAAE,CAAC;QACnC,IAAI,CAAC,wBAAwB,CAAC,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC;YAC1C,MAAM,IAAI,KAAK,CAAC,sCAAsC,MAAM,EAAE,CAAC,CAAC;QAClE,CAAC;IACH,CAAC;IAED,OAAO,EAAE,GAAG,KAAK,EAAE,QAAQ,EAAE,aAAa,EAAE,CAAC;AAC/C,CAAC;AAED,SAAS,mBAAmB,CAAC,UAAkB;IAC7C,IAAI,OAAO,UAAU,KAAK,QAAQ,IAAI,UAAU,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACrE,MAAM,IAAI,KAAK,CAAC,iCAAiC,CAAC,CAAC;IACrD,CAAC;IAED,IAAI,MAAW,CAAC;IAChB,IAAI,CAAC;QACH,MAAM,GAAG,IAAI,GAAG,CAAC,UAAU,CAAC,CAAC;IAC/B,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,KAAK,CAAC,uDAAuD,CAAC,CAAC;IAC3E,CAAC;IAED,IAAI,MAAM,CAAC,QAAQ,KAAK,OAAO,IAAI,MAAM,CAAC,QAAQ,KAAK,QAAQ,EAAE,CAAC;QAChE,MAAM,IAAI,KAAK,CAAC,4CAA4C,CAAC,CAAC;IAChE,CAAC;IAED,OAAO,UAAU,CAAC,IAAI,EAAE,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC;AAC/C,CAAC;AAED,SAAS,gBAAgB,CAAC,SAA6B;IACrD,IAAI,SAAS,KAAK,SAAS,EAAE,CAAC;QAC5B,OAAO,0BAA0B,CAAC;IACpC,CAAC;IAED,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,SAAS,CAAC,IAAI,SAAS,IAAI,CAAC,EAAE,CAAC;QAClD,MAAM,IAAI,KAAK,CAAC,qDAAqD,CAAC,CAAC;IACzE,CAAC;IAED,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,SAAS,oBAAoB,CAAC,OAA+B;IAC3D,MAAM,OAAO,GAAG,IAAI,OAAO,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC;IAC7C,IAAI,OAAO,CAAC,MAAM,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,WAAW,CAAC,EAAE,CAAC;QAChD,OAAO,CAAC,GAAG,CAAC,WAAW,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC;IAC3C,CAAC;IAED,6EAA6E;IAC7E,gDAAgD;IAChD,OAAO,CAAC,MAAM,CAAC,cAAc,CAAC,CAAC;IAC/B,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,SAAS,uBAAuB,CAC9B,IAAc,EACd,IAAY,EACZ,KAAmC;IAEnC,IAAI,KAAK,KAAK,SAAS,EAAE,CAAC;QACxB,eAAe,CAAC,IAAI,EAAE,IAAI,EAAE,KAAK,CAAC,CAAC;IACrC,CAAC;AACH,CAAC;AAED,SAAS,iBAAiB,CACxB,IAAc,EACd,OAA+C,EAC/C,kBAAgE;IAEhE,KAAK,MAAM,CAAC,IAAI,EAAE,KAAK,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,OAAO,IAAI,EAAE,CAAC,EAAE,CAAC;QAC1D,IAAI,IAAI,KAAK,YAAY,IAAI,kBAAkB,CAAC,IAAI,CAAC,KAAK,SAAS,IAAI,KAAK,IAAI,IAAI,EAAE,CAAC;YACrF,SAAS;QACX,CAAC;QAED,eAAe,CAAC,IAAI,EAAE,IAAI,EAAE,KAAK,CAAC,CAAC;IACrC,CAAC;AACH,CAAC;AAED,SAAS,eAAe,CAAC,IAAc,EAAE,IAAY,EAAE,KAAuB;IAC5E,IAAI,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC;QACzB,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,IAAI,CAAC,MAAM,CAAC,IAAI,EAAE,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC;QAClC,CAAC;QACD,OAAO;IACT,CAAC;IAED,IAAI,CAAC,MAAM,CAAC,IAAI,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC;AACnC,CAAC;AAED,SAAS,cAAc,CACrB,OAAwB,EACxB,aAA6C;IAE7C,OAAO,MAAM,CAAC,WAAW,CACvB,aAAa;SACV,GAAG,CAAC,CAAC,MAAM,EAA8C,EAAE;QAC1D,MAAM,OAAO,GAAG,OAAO,CAAC,QAAQ,EAAE,CAAC,GAAG,MAAM,UAAU,CAAC,CAAC;QACxD,OAAO,OAAO,KAAK,SAAS,IAAI,OAAO,KAAK,IAAI,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;IACnF,CAAC,CAAC;SACD,MAAM,CAAC,CAAC,KAAK,EAA2C,EAAE,CAAC,KAAK,KAAK,SAAS,CAAC,CACnF,CAAC;AACJ,CAAC;AAED,SAAS,eAAe,CAAC,MAA0B;IACjD,IACE,MAAM,KAAK,SAAS;QACpB,MAAM,KAAK,iBAAiB;QAC5B,MAAM,KAAK,SAAS;QACpB,MAAM,KAAK,SAAS,EACpB,CAAC;QACD,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,gFAAgF;IAChF,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,MAAM,0BAA0B,GAAG,EAAE,GAAG,IAAI,GAAG,IAAI,CAAC;AAEpD,KAAK,UAAU,oBAAoB,CACjC,QAAkB,EAClB,OAA+B,EAC/B,MAAoB;IAEpB,IAAI,OAAgB,CAAC;IACrB,IAAI,CAAC;QACH,MAAM,IAAI,GAAG,MAAM,gBAAgB,CAAC,QAAQ,EAAE,0BAA0B,EAAE,MAAM,CAAC,CAAC;QAClF,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAY,CAAC;IACxC,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,IAAI,MAAM,EAAE,OAAO;YAAE,MAAM,KAAK,CAAC;QACjC,MAAM,IAAI,KAAK,CACb,aAAa,CAAC,kCAAkC,YAAY,CAAC,KAAK,CAAC,EAAE,EAAE,OAAO,CAAC,EAC/E;YACE,KAAK,EAAE,KAAK;SACb,CACF,CAAC;IACJ,CAAC;IAED,IAAI,CAAC,iBAAiB,CAAC,OAAO,CAAC,EAAE,CAAC;QAChC,MAAM,IAAI,KAAK,CAAC,gDAAgD,CAAC,CAAC;IACpE,CAAC;IAED,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,SAAS,iBAAiB,CAAC,KAAc;IACvC,OAAO,KAAK,KAAK,IAAI,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;AAC9E,CAAC;AAED,KAAK,UAAU,gBAAgB,CAAC,QAAkB,EAAE,MAAoB;IACtE,IAAI,CAAC;QACH,OAAO,CAAC,MAAM,gBAAgB,CAAC,QAAQ,EAAE,KAAK,EAAE,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC;IAC3E,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,EAAE,CAAC;IACZ,CAAC;AACH,CAAC;AAED,KAAK,UAAU,gBAAgB,CAC7B,QAAkB,EAClB,QAAgB,EAChB,MAAoB;IAEpB,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,CAAC;QACnB,MAAM,IAAI,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC;QACnC,IAAI,IAAI,CAAC,MAAM,GAAG,QAAQ,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,iCAAiC,QAAQ,qBAAqB,CAAC,CAAC;QAClF,CAAC;QACD,OAAO,IAAI,CAAC;IACd,CAAC;IAED,MAAM,MAAM,GAAG,QAAQ,CAAC,IAAI,CAAC,SAAS,EAAE,CAAC;IACzC,MAAM,MAAM,GAAiB,EAAE,CAAC;IAChC,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,CAAC;QACH,OAAO,IAAI,EAAE,CAAC;YACZ,IAAI,MAAM,EAAE,OAAO,EAAE,CAAC;gBACpB,MAAM,IAAI,KAAK,CAAC,oCAAoC,CAAC,CAAC;YACxD,CAAC;YACD,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,MAAM,MAAM,CAAC,IAAI,EAAE,CAAC;YAC5C,IAAI,IAAI;gBAAE,MAAM;YAChB,IAAI,CAAC,KAAK;gBAAE,SAAS;YACrB,KAAK,IAAI,KAAK,CAAC,UAAU,CAAC;YAC1B,IAAI,KAAK,GAAG,QAAQ,EAAE,CAAC;gBACrB,MAAM,MAAM,CAAC,MAAM,EAAE,CAAC;gBACtB,MAAM,IAAI,KAAK,CAAC,iCAAiC,QAAQ,qBAAqB,CAAC,CAAC;YAClF,CAAC;YACD,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACrB,CAAC;IACH,CAAC;YAAS,CAAC;QACT,MAAM,CAAC,WAAW,EAAE,CAAC;IACvB,CAAC;IAED,OAAO,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;AACnF,CAAC;AAED,SAAS,qBAAqB,CAC5B,MAA+B,EAC/B,MAA6B;IAE7B,MAAM,OAAO,GAAG,MAAM,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,KAAK,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC;IAClF,OAAO,sBAAsB,MAAM,GAAG,OAAO,EAAE,CAAC;AAClD,CAAC;AAED,SAAS,YAAY,CAAC,KAAc;IAClC,OAAO,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;AAChE,CAAC;AAED,SAAS,aAAa,CAAC,OAAe,EAAE,OAA+B;IACrE,MAAM,OAAO,GAAG,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,MAAM,CAAC,MAAM,CAAC,OAAO,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC,CAAC,MAAM,CAC9E,CAAC,KAAK,EAAmB,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,CAC3C,CAAC;IACF,OAAO,OAAO,CAAC,MAAM,CAAC,CAAC,QAAQ,EAAE,MAAM,EAAE,EAAE,CAAC,QAAQ,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,YAAY,CAAC,EAAE,OAAO,CAAC,CAAC;AAClG,CAAC;AAED,SAAS,eAAe,CAAC,QAAgB;IACvC,MAAM,SAAS,GAAG,QAAQ,CAAC,KAAK,CAAC,QAAQ,CAAC,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC;IAC9E,MAAM,SAAS,GAA2B;QACxC,GAAG,EAAE,UAAU;QACf,GAAG,EAAE,oBAAoB;QACzB,IAAI,EAAE,yEAAyE;QAC/E,IAAI,EAAE,WAAW;QACjB,EAAE,EAAE,eAAe;QACnB,GAAG,EAAE,iBAAiB;QACtB,GAAG,EAAE,+BAA+B;QACpC,IAAI,EAAE,2EAA2E;QACjF,GAAG,EAAE,0BAA0B;QAC/B,IAAI,EAAE,mEAAmE;KAC1E,CAAC;IAEF,OAAO,SAAS,CAAC,SAAS,CAAC,IAAI,0BAA0B,CAAC;AAC5D,CAAC"}

package/dist/docling.d.ts

@@ -0,0 +1,14 @@
+import type { ToolsInput } from '@mastra/core/agent';
+import { type DoclingDocumentProcessorOptions, type DocumentParseOutput } from './document-processor';
+import type { DoclingOutputFormat } from './converter';
+export interface DoclingProcessResult {
+    source_path: string;
+    output_path: string;
+    output_format: DoclingOutputFormat;
+    status: DocumentParseOutput['status'];
+    processing_time?: number;
+    errors: unknown[];
+}
+export declare function resolveDoclingSkillPath(): string;
+export declare function createDoclingTools(options: DoclingDocumentProcessorOptions): ToolsInput;
+export declare function createDoclingToolPack(options: DoclingDocumentProcessorOptions): import("@lumifai/harness-tool-packs").ToolPack;

package/dist/docling.js

@@ -0,0 +1,97 @@
+"use strict";
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.resolveDoclingSkillPath = resolveDoclingSkillPath;
+exports.createDoclingTools = createDoclingTools;
+exports.createDoclingToolPack = createDoclingToolPack;
+const node_path_1 = __importDefault(require("node:path"));
+const tools_1 = require("@mastra/core/tools");
+const zod_1 = require("zod");
+const harness_tool_packs_1 = require("@lumifai/harness-tool-packs");
+const document_processor_1 = require("./document-processor");
+const DEFAULT_OUTPUT_FORMAT = 'md';
+const outputFormatSchema = zod_1.z.enum(['md', 'json', 'html', 'text', 'doctags']);
+const processFileInputSchema = zod_1.z
+    .object({
+    path: zod_1.z
+        .string()
+        .min(1)
+        .describe('Workspace-relative path to a supported file, e.g. documents/report.pdf'),
+    output_format: outputFormatSchema
+        .optional()
+        .describe('Output format to write. Defaults to Markdown.'),
+    do_ocr: zod_1.z.boolean().optional().describe('Enable OCR for scanned documents.'),
+    force_ocr: zod_1.z.boolean().optional().describe('Force OCR even when text is available.'),
+    table_mode: zod_1.z.enum(['fast', 'accurate']).optional().describe('Table extraction mode.'),
+    image_export_mode: zod_1.z
+        .enum(['placeholder', 'embedded', 'referenced'])
+        .optional()
+        .describe('How images are represented in the exported document.'),
+    pipeline: zod_1.z
+        .enum(['standard', 'vlm'])
+        .optional()
+        .describe('Docling processing pipeline to use.'),
+})
+    .strict();
+function resolveDoclingSkillPath() {
+    const packageRoot = node_path_1.default.dirname(require.resolve('@lumifai/harness-tool-pack-docling/package.json'));
+    return node_path_1.default.join(packageRoot, 'skills', 'docling');
+}
+function createDoclingTools(options) {
+    const processor = (0, document_processor_1.createDoclingDocumentProcessor)(options);
+    const processFileTool = (0, tools_1.createTool)({
+        id: 'process_file',
+        description: 'Convert a supported workspace file through Docling and write the selected result under processed/. Supports PDF, Office documents, spreadsheets, CSV, HTML, Markdown, images, and other formats supported by the configured Docling service.',
+        inputSchema: processFileInputSchema,
+        execute: async (input, context) => processFile(processor, input, context),
+    });
+    return { process_file: processFileTool };
+}
+function createDoclingToolPack(options) {
+    return (0, harness_tool_packs_1.defineToolPack)({
+        id: 'docling',
+        description: 'Docling document conversion for supported workspace files.',
+        enabledByDefault: false,
+        skills: [resolveDoclingSkillPath()],
+        tools: createDoclingTools(options),
+        toolConfig: {
+            process_file: { permissionPolicy: 'ask' },
+        },
+    });
+}
+async function processFile(processor, input, context) {
+    const filesystem = context.workspace?.filesystem;
+    if (!filesystem) {
+        throw new Error('Docling processing requires a workspace filesystem.');
+    }
+    const outputFormat = input.output_format ?? DEFAULT_OUTPUT_FORMAT;
+    const result = await processor.parseWorkspaceFile(input.path, {
+        filesystem,
+        persistArtifacts: true,
+        options: {
+            outputFormats: [outputFormat],
+            ...(input.do_ocr === undefined ? {} : { doOcr: input.do_ocr }),
+            ...(input.force_ocr === undefined ? {} : { forceOcr: input.force_ocr }),
+            ...(input.table_mode === undefined ? {} : { tableMode: input.table_mode }),
+            ...(input.image_export_mode === undefined
+                ? {}
+                : { imageExportMode: input.image_export_mode }),
+            ...(input.pipeline === undefined ? {} : { pipeline: input.pipeline }),
+        },
+    });
+    const outputPath = result.artifactPaths?.[outputFormat];
+    if (!outputPath) {
+        throw new Error(`Docling returned no ${outputFormat} content for ${result.sourcePath}.`);
+    }
+    return {
+        source_path: result.sourcePath,
+        output_path: outputPath,
+        output_format: outputFormat,
+        status: result.status,
+        processing_time: result.processingTimeSec,
+        errors: result.errors,
+    };
+}
+//# sourceMappingURL=docling.js.map

package/dist/docling.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"docling.js","sourceRoot":"","sources":["../src/docling.ts"],"names":[],"mappings":";;;;;AAuDA,0DAKC;AAED,gDAYC;AAED,sDAWC;AAvFD,0DAA6B;AAI7B,8CAAgD;AAChD,6BAAwB;AAExB,oEAA6D;AAE7D,6DAK8B;AAG9B,MAAM,qBAAqB,GAAwB,IAAI,CAAC;AAExD,MAAM,kBAAkB,GAAG,OAAC,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,SAAS,CAAC,CAAC,CAAC;AAE7E,MAAM,sBAAsB,GAAG,OAAC;KAC7B,MAAM,CAAC;IACN,IAAI,EAAE,OAAC;SACJ,MAAM,EAAE;SACR,GAAG,CAAC,CAAC,CAAC;SACN,QAAQ,CAAC,wEAAwE,CAAC;IACrF,aAAa,EAAE,kBAAkB;SAC9B,QAAQ,EAAE;SACV,QAAQ,CAAC,+CAA+C,CAAC;IAC5D,MAAM,EAAE,OAAC,CAAC,OAAO,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAAC,mCAAmC,CAAC;IAC5E,SAAS,EAAE,OAAC,CAAC,OAAO,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAAC,wCAAwC,CAAC;IACpF,UAAU,EAAE,OAAC,CAAC,IAAI,CAAC,CAAC,MAAM,EAAE,UAAU,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAAC,wBAAwB,CAAC;IACtF,iBAAiB,EAAE,OAAC;SACjB,IAAI,CAAC,CAAC,aAAa,EAAE,UAAU,EAAE,YAAY,CAAC,CAAC;SAC/C,QAAQ,EAAE;SACV,QAAQ,CAAC,sDAAsD,CAAC;IACnE,QAAQ,EAAE,OAAC;SACR,IAAI,CAAC,CAAC,UAAU,EAAE,KAAK,CAAC,CAAC;SACzB,QAAQ,EAAE;SACV,QAAQ,CAAC,qCAAqC,CAAC;CACnD,CAAC;KACD,MAAM,EAAE,CAAC;AAaZ,SAAgB,uBAAuB;IACrC,MAAM,WAAW,GAAG,mBAAI,CAAC,OAAO,CAC9B,OAAO,CAAC,OAAO,CAAC,iDAAiD,CAAC,CACnE,CAAC;IACF,OAAO,mBAAI,CAAC,IAAI,CAAC,WAAW,EAAE,QAAQ,EAAE,SAAS,CAAC,CAAC;AACrD,CAAC;AAED,SAAgB,kBAAkB,CAAC,OAAwC;IACzE,MAAM,SAAS,GAAG,IAAA,mDAA8B,EAAC,OAAO,CAAC,CAAC;IAE1D,MAAM,eAAe,GAAG,IAAA,kBAAU,EAAC;QACjC,EAAE,EAAE,cAAc;QAClB,WAAW,EACT,8OAA8O;QAChP,WAAW,EAAE,sBAAsB;QACnC,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAE,EAAE,CAAC,WAAW,CAAC,SAAS,EAAE,KAAK,EAAE,OAAO,CAAC;KAC1E,CAAC,CAAC;IAEH,OAAO,EAAE,YAAY,EAAE,eAAe,EAAE,CAAC;AAC3C,CAAC;AAED,SAAgB,qBAAqB,CAAC,OAAwC;IAC5E,OAAO,IAAA,mCAAc,EAAC;QACpB,EAAE,EAAE,SAAS;QACb,WAAW,EAAE,4DAA4D;QACzE,gBAAgB,EAAE,KAAK;QACvB,MAAM,EAAE,CAAC,uBAAuB,EAAE,CAAC;QACnC,KAAK,EAAE,kBAAkB,CAAC,OAAO,CAAC;QAClC,UAAU,EAAE;YACV,YAAY,EAAE,EAAE,gBAAgB,EAAE,KAAK,EAAE;SAC1C;KACF,CAAC,CAAC;AACL,CAAC;AAED,KAAK,UAAU,WAAW,CACxB,SAA4D,EAC5D,KAAuB,EACvB,OAA6B;IAE7B,MAAM,UAAU,GAAG,OAAO,CAAC,SAAS,EAAE,UAAU,CAAC;IACjD,IAAI,CAAC,UAAU,EAAE,CAAC;QAChB,MAAM,IAAI,KAAK,CAAC,qDAAqD,CAAC,CAAC;IACzE,CAAC;IAED,MAAM,YAAY,GAAG,KAAK,CAAC,aAAa,IAAI,qBAAqB,CAAC;IAClE,MAAM,MAAM,GAAG,MAAM,SAAS,CAAC,kBAAkB,CAAC,KAAK,CAAC,IAAI,EAAE;QAC5D,UAAU;QACV,gBAAgB,EAAE,IAAI;QACtB,OAAO,EAAE;YACP,aAAa,EAAE,CAAC,YAAY,CAAC;YAC7B,GAAG,CAAC,KAAK,CAAC,MAAM,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,KAAK,CAAC,MAAM,EAAE,CAAC;YAC9D,GAAG,CAAC,KAAK,CAAC,SAAS,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,QAAQ,EAAE,KAAK,CAAC,SAAS,EAAE,CAAC;YACvE,GAAG,CAAC,KAAK,CAAC,UAAU,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,SAAS,EAAE,KAAK,CAAC,UAAU,EAAE,CAAC;YAC1E,GAAG,CAAC,KAAK,CAAC,iBAAiB,KAAK,SAAS;gBACvC,CAAC,CAAC,EAAE;gBACJ,CAAC,CAAC,EAAE,eAAe,EAAE,KAAK,CAAC,iBAAiB,EAAE,CAAC;YACjD,GAAG,CAAC,KAAK,CAAC,QAAQ,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,QAAQ,EAAE,KAAK,CAAC,QAAQ,EAAE,CAAC;SAC5B;KAC5C,CAAC,CAAC;IACH,MAAM,UAAU,GAAG,MAAM,CAAC,aAAa,EAAE,CAAC,YAAY,CAAC,CAAC;IAExD,IAAI,CAAC,UAAU,EAAE,CAAC;QAChB,MAAM,IAAI,KAAK,CAAC,uBAAuB,YAAY,gBAAgB,MAAM,CAAC,UAAU,GAAG,CAAC,CAAC;IAC3F,CAAC;IAED,OAAO;QACL,WAAW,EAAE,MAAM,CAAC,UAAU;QAC9B,WAAW,EAAE,UAAU;QACvB,aAAa,EAAE,YAAY;QAC3B,MAAM,EAAE,MAAM,CAAC,MAAM;QACrB,eAAe,EAAE,MAAM,CAAC,iBAAiB;QACzC,MAAM,EAAE,MAAM,CAAC,MAAM;KACtB,CAAC;AACJ,CAAC"}

package/dist/document-processor.d.ts

@@ -0,0 +1,69 @@
+import type { WorkspaceFilesystem } from '@mastra/core/workspace';
+import { type DoclingFormValue, type DoclingOutputFormat, type DoclingToolPackOptions } from './converter';
+export type DocumentParseStatus = 'success' | 'partial_success' | 'skipped' | 'failure';
+export interface DocumentParseInput {
+    filename: string;
+    bytes: Uint8Array;
+    mimeType?: string;
+    options?: Partial<DoclingDocumentOptions>;
+}
+export interface DocumentParseOutput {
+    processorId: string;
+    sourcePath: string;
+    markdown?: string;
+    json?: unknown;
+    extras: {
+        html?: string;
+        text?: string;
+        doctags?: string;
+    };
+    status: DocumentParseStatus;
+    processingTimeSec?: number;
+    errors: unknown[];
+    artifactPaths?: Partial<Record<DoclingOutputFormat, string>>;
+}
+export interface DocumentArtifactPathInput {
+    sourcePath: string;
+    filename: string;
+    format: DoclingOutputFormat;
+}
+export type DocumentArtifactPathResolver = (input: DocumentArtifactPathInput) => string;
+export interface DocumentWorkspaceParseOptions {
+    filesystem: WorkspaceFilesystem;
+    mimeType?: string;
+    options?: Partial<DoclingDocumentOptions>;
+    persistArtifacts?: boolean;
+    artifactPathResolver?: DocumentArtifactPathResolver;
+}
+export interface DocumentProcessor {
+    readonly processorId: string;
+    parse(input: DocumentParseInput): Promise<DocumentParseOutput>;
+    parseWorkspaceFile(path: string, options: DocumentWorkspaceParseOptions): Promise<DocumentParseOutput>;
+}
+export interface DoclingDocumentOptions {
+    outputFormats: readonly DoclingOutputFormat[];
+    doOcr: boolean;
+    forceOcr: boolean;
+    tableMode: 'fast' | 'accurate';
+    imageExportMode: 'placeholder' | 'embedded' | 'referenced';
+    pipeline: 'standard' | 'vlm';
+    ocrEngine: string;
+    pdfBackend: string;
+    abortOnError: boolean;
+    returnAsFile: boolean;
+    doCodeEnrichment: boolean;
+    doFormulaEnrichment: boolean;
+    doPictureClassification: boolean;
+    doPictureDescription: boolean;
+    normalizeEmbeddedImages: boolean;
+    imagePlaceholder: string;
+    formOptions: Readonly<Record<string, DoclingFormValue | null | undefined>>;
+}
+export interface DoclingDocumentProcessorOptions extends DoclingToolPackOptions {
+    defaults?: Partial<DoclingDocumentOptions>;
+}
+export declare const DOCLING_PROCESSOR_ID = "docling";
+export declare const DEFAULT_DOCLING_DOCUMENT_OPTIONS: Readonly<DoclingDocumentOptions>;
+export declare const IMAGE_PLACEHOLDER = "[Image]";
+export declare function createDoclingDocumentProcessor(options: DoclingDocumentProcessorOptions): DocumentProcessor;
+export declare function replaceEmbeddedImagesWithPlaceholder(markdown: string, placeholder?: string): string;

package/dist/document-processor.js

@@ -0,0 +1,208 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.IMAGE_PLACEHOLDER = exports.DEFAULT_DOCLING_DOCUMENT_OPTIONS = exports.DOCLING_PROCESSOR_ID = void 0;
+exports.createDoclingDocumentProcessor = createDoclingDocumentProcessor;
+exports.replaceEmbeddedImagesWithPlaceholder = replaceEmbeddedImagesWithPlaceholder;
+const harness_tool_packs_1 = require("@lumifai/harness-tool-packs");
+const converter_1 = require("./converter");
+exports.DOCLING_PROCESSOR_ID = 'docling';
+exports.DEFAULT_DOCLING_DOCUMENT_OPTIONS = {
+    outputFormats: ['md', 'json', 'html', 'text', 'doctags'],
+    imageExportMode: 'placeholder',
+    pipeline: 'standard',
+    doOcr: true,
+    forceOcr: false,
+    ocrEngine: 'auto',
+    pdfBackend: 'docling_parse',
+    tableMode: 'accurate',
+    abortOnError: false,
+    returnAsFile: false,
+    doCodeEnrichment: false,
+    doFormulaEnrichment: false,
+    doPictureClassification: false,
+    doPictureDescription: false,
+    normalizeEmbeddedImages: true,
+    imagePlaceholder: '[Image]',
+    formOptions: {},
+};
+const DEFAULT_ARTIFACT_PATH_RESOLVER = ({ filename, format }) => `processed/${filename}.${format}`;
+exports.IMAGE_PLACEHOLDER = '[Image]';
+function createDoclingDocumentProcessor(options) {
+    const converter = (0, converter_1.createDoclingConverter)(options);
+    const defaults = resolveDocumentOptions(options.defaults);
+    return {
+        processorId: exports.DOCLING_PROCESSOR_ID,
+        parse: (input) => parseBytes(converter, defaults, input, input.filename),
+        parseWorkspaceFile: (path, workspaceOptions) => parseWorkspaceFile(converter, defaults, path, workspaceOptions),
+    };
+}
+function replaceEmbeddedImagesWithPlaceholder(markdown, placeholder = exports.IMAGE_PLACEHOLDER) {
+    return markdown
+        .replace(/!\[[^\]]*\]\(data:image\/[^)]+\)/gi, placeholder)
+        .replace(/!\[[^\]]*\]\(data:image\/[A-Za-z0-9+/=\s]*$/gim, placeholder)
+        .replace(/<img\b[^>]*\bsrc=["']data:image\/[^"']+["'][^>]*>/gi, placeholder);
+}
+async function parseBytes(converter, defaults, input, sourcePath) {
+    const documentOptions = resolveDocumentOptions(defaults, input.options);
+    const filename = normalizeFilename(input.filename);
+    const conversion = await converter.convertFile({
+        bytes: input.bytes,
+        filename,
+        mimeType: input.mimeType,
+        outputFormats: documentOptions.outputFormats,
+        doOcr: documentOptions.doOcr,
+        forceOcr: documentOptions.forceOcr,
+        tableMode: documentOptions.tableMode,
+        imageExportMode: documentOptions.imageExportMode,
+        pipeline: documentOptions.pipeline,
+        formOptions: toFormOptions(documentOptions),
+    });
+    return mapConversionResult(sourcePath, conversion, documentOptions);
+}
+async function parseWorkspaceFile(converter, defaults, inputPath, workspaceOptions) {
+    const filesystem = workspaceOptions.filesystem;
+    if (!filesystem?.readFile) {
+        throw new Error(`Docling processing requires a workspace filesystem to read ${inputPath}.`);
+    }
+    const workspace = (0, harness_tool_packs_1.createWorkspaceFiles)();
+    const context = { workspace: { filesystem } };
+    const file = await workspace.read(context, inputPath);
+    const filename = getFilename(file.path);
+    const result = await parseBytes(converter, defaults, {
+        filename,
+        bytes: file.bytes,
+        mimeType: workspaceOptions.mimeType,
+        options: workspaceOptions.options,
+    }, file.path);
+    if (!workspaceOptions.persistArtifacts) {
+        return result;
+    }
+    return persistArtifacts(result, file.path, filename, workspace, context, workspaceOptions.artifactPathResolver ?? DEFAULT_ARTIFACT_PATH_RESOLVER);
+}
+function mapConversionResult(sourcePath, conversion, options) {
+    const markdown = asStringOutput(conversion.outputs.md);
+    const normalizedMarkdown = markdown !== undefined && options.normalizeEmbeddedImages
+        ? replaceEmbeddedImagesWithPlaceholder(markdown, options.imagePlaceholder)
+        : markdown;
+    return {
+        processorId: exports.DOCLING_PROCESSOR_ID,
+        sourcePath,
+        markdown: normalizedMarkdown,
+        json: conversion.outputs.json,
+        extras: {
+            html: asStringOutput(conversion.outputs.html),
+            text: asStringOutput(conversion.outputs.text),
+            doctags: asStringOutput(conversion.outputs.doctags),
+        },
+        status: conversion.status,
+        processingTimeSec: conversion.processingTimeSec,
+        errors: conversion.errors,
+    };
+}
+async function persistArtifacts(result, sourcePath, filename, workspace, context, pathResolver) {
+    const values = [
+        ['md', result.markdown],
+        ['json', result.json],
+        ['html', result.extras.html],
+        ['text', result.extras.text],
+        ['doctags', result.extras.doctags],
+    ];
+    const artifactPaths = {};
+    for (const [format, value] of values) {
+        if (value === undefined)
+            continue;
+        const artifactPath = (0, harness_tool_packs_1.normalizeWorkspacePath)(pathResolver({ sourcePath, filename, format }));
+        await workspace.write(context, artifactPath, serializeOutput(format, value), {
+            overwrite: true,
+        });
+        artifactPaths[format] = artifactPath;
+    }
+    return { ...result, artifactPaths };
+}
+function resolveDocumentOptions(...overrides) {
+    const merged = overrides.reduce((current, override) => ({
+        ...current,
+        ...(override ?? {}),
+        formOptions: {
+            ...current.formOptions,
+            ...(override?.formOptions ?? {}),
+        },
+    }), { ...exports.DEFAULT_DOCLING_DOCUMENT_OPTIONS, formOptions: {} });
+    const outputFormats = normalizeOutputFormats(merged.outputFormats);
+    if (typeof merged.doOcr !== 'boolean' || typeof merged.forceOcr !== 'boolean') {
+        throw new Error('Docling OCR options must be boolean values.');
+    }
+    if (merged.tableMode !== 'fast' && merged.tableMode !== 'accurate') {
+        throw new Error(`Unsupported Docling tableMode: ${String(merged.tableMode)}`);
+    }
+    if (!['placeholder', 'embedded', 'referenced'].includes(merged.imageExportMode)) {
+        throw new Error(`Unsupported Docling imageExportMode: ${String(merged.imageExportMode)}`);
+    }
+    if (!['standard', 'vlm'].includes(merged.pipeline)) {
+        throw new Error(`Unsupported Docling pipeline: ${String(merged.pipeline)}`);
+    }
+    for (const name of ['ocrEngine', 'pdfBackend', 'imagePlaceholder']) {
+        if (typeof merged[name] !== 'string' || merged[name].trim().length === 0) {
+            throw new Error(`Docling ${name} must be a non-empty string.`);
+        }
+    }
+    for (const name of [
+        'abortOnError',
+        'returnAsFile',
+        'doCodeEnrichment',
+        'doFormulaEnrichment',
+        'doPictureClassification',
+        'doPictureDescription',
+        'normalizeEmbeddedImages',
+    ]) {
+        if (typeof merged[name] !== 'boolean') {
+            throw new Error(`Docling ${name} must be a boolean value.`);
+        }
+    }
+    return { ...merged, outputFormats };
+}
+function normalizeOutputFormats(value) {
+    if (!Array.isArray(value) || value.length === 0) {
+        throw new Error('Docling outputFormats must contain at least one format.');
+    }
+    const formats = [...new Set(value)];
+    for (const format of formats) {
+        if (!['md', 'json', 'html', 'text', 'doctags'].includes(format)) {
+            throw new Error(`Unsupported Docling output format: ${String(format)}`);
+        }
+    }
+    return formats;
+}
+function toFormOptions(options) {
+    return {
+        ...options.formOptions,
+        ocr_engine: options.ocrEngine,
+        pdf_backend: options.pdfBackend,
+        abort_on_error: options.abortOnError,
+        return_as_file: options.returnAsFile,
+        do_code_enrichment: options.doCodeEnrichment,
+        do_formula_enrichment: options.doFormulaEnrichment,
+        do_picture_classification: options.doPictureClassification,
+        do_picture_description: options.doPictureDescription,
+    };
+}
+function asStringOutput(value) {
+    if (value === undefined || value === null)
+        return undefined;
+    return typeof value === 'string' ? value : JSON.stringify(value);
+}
+function serializeOutput(format, value) {
+    if (format !== 'json' && typeof value === 'string')
+        return value;
+    return JSON.stringify(value, null, 2);
+}
+function normalizeFilename(inputPath) {
+    const filename = inputPath.trim().split(/[\\/]/).pop()?.trim();
+    if (!filename)
+        throw new Error('Docling filename is required.');
+    return filename;
+}
+function getFilename(normalizedPath) {
+    return normalizeFilename(normalizedPath);
+}
+//# sourceMappingURL=document-processor.js.map

package/dist/document-processor.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"document-processor.js","sourceRoot":"","sources":["../src/document-processor.ts"],"names":[],"mappings":";;;AAkHA,wEAYC;AAED,oFAQC;AAvID,oEAA2F;AAG3F,2CAOqB;AA4ER,QAAA,oBAAoB,GAAG,SAAS,CAAC;AAEjC,QAAA,gCAAgC,GAAqC;IAChF,aAAa,EAAE,CAAC,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,SAAS,CAAC;IACxD,eAAe,EAAE,aAAa;IAC9B,QAAQ,EAAE,UAAU;IACpB,KAAK,EAAE,IAAI;IACX,QAAQ,EAAE,KAAK;IACf,SAAS,EAAE,MAAM;IACjB,UAAU,EAAE,eAAe;IAC3B,SAAS,EAAE,UAAU;IACrB,YAAY,EAAE,KAAK;IACnB,YAAY,EAAE,KAAK;IACnB,gBAAgB,EAAE,KAAK;IACvB,mBAAmB,EAAE,KAAK;IAC1B,uBAAuB,EAAE,KAAK;IAC9B,oBAAoB,EAAE,KAAK;IAC3B,uBAAuB,EAAE,IAAI;IAC7B,gBAAgB,EAAE,SAAS;IAC3B,WAAW,EAAE,EAAE;CAChB,CAAC;AAEF,MAAM,8BAA8B,GAAiC,CAAC,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,CAC5F,aAAa,QAAQ,IAAI,MAAM,EAAE,CAAC;AAEvB,QAAA,iBAAiB,GAAG,SAAS,CAAC;AAE3C,SAAgB,8BAA8B,CAC5C,OAAwC;IAExC,MAAM,SAAS,GAAG,IAAA,kCAAsB,EAAC,OAAO,CAAC,CAAC;IAClD,MAAM,QAAQ,GAAG,sBAAsB,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC;IAE1D,OAAO;QACL,WAAW,EAAE,4BAAoB;QACjC,KAAK,EAAE,CAAC,KAAK,EAAE,EAAE,CAAC,UAAU,CAAC,SAAS,EAAE,QAAQ,EAAE,KAAK,EAAE,KAAK,CAAC,QAAQ,CAAC;QACxE,kBAAkB,EAAE,CAAC,IAAI,EAAE,gBAAgB,EAAE,EAAE,CAC7C,kBAAkB,CAAC,SAAS,EAAE,QAAQ,EAAE,IAAI,EAAE,gBAAgB,CAAC;KAClE,CAAC;AACJ,CAAC;AAED,SAAgB,oCAAoC,CAClD,QAAgB,EAChB,cAAsB,yBAAiB;IAEvC,OAAO,QAAQ;SACZ,OAAO,CAAC,oCAAoC,EAAE,WAAW,CAAC;SAC1D,OAAO,CAAC,gDAAgD,EAAE,WAAW,CAAC;SACtE,OAAO,CAAC,qDAAqD,EAAE,WAAW,CAAC,CAAC;AACjF,CAAC;AAED,KAAK,UAAU,UAAU,CACvB,SAA2B,EAC3B,QAAgC,EAChC,KAAyB,EACzB,UAAkB;IAElB,MAAM,eAAe,GAAG,sBAAsB,CAAC,QAAQ,EAAE,KAAK,CAAC,OAAO,CAAC,CAAC;IACxE,MAAM,QAAQ,GAAG,iBAAiB,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;IACnD,MAAM,UAAU,GAAG,MAAM,SAAS,CAAC,WAAW,CAAC;QAC7C,KAAK,EAAE,KAAK,CAAC,KAAK;QAClB,QAAQ;QACR,QAAQ,EAAE,KAAK,CAAC,QAAQ;QACxB,aAAa,EAAE,eAAe,CAAC,aAAa;QAC5C,KAAK,EAAE,eAAe,CAAC,KAAK;QAC5B,QAAQ,EAAE,eAAe,CAAC,QAAQ;QAClC,SAAS,EAAE,eAAe,CAAC,SAAS;QACpC,eAAe,EAAE,eAAe,CAAC,eAAe;QAChD,QAAQ,EAAE,eAAe,CAAC,QAAQ;QAClC,WAAW,EAAE,aAAa,CAAC,eAAe,CAAC;KAC5C,CAAC,CAAC;IAEH,OAAO,mBAAmB,CAAC,UAAU,EAAE,UAAU,EAAE,eAAe,CAAC,CAAC;AACtE,CAAC;AAED,KAAK,UAAU,kBAAkB,CAC/B,SAA2B,EAC3B,QAAgC,EAChC,SAAiB,EACjB,gBAA+C;IAE/C,MAAM,UAAU,GAAG,gBAAgB,CAAC,UAAU,CAAC;IAC/C,IAAI,CAAC,UAAU,EAAE,QAAQ,EAAE,CAAC;QAC1B,MAAM,IAAI,KAAK,CAAC,8DAA8D,SAAS,GAAG,CAAC,CAAC;IAC9F,CAAC;IAED,MAAM,SAAS,GAAG,IAAA,yCAAoB,GAAE,CAAC;IACzC,MAAM,OAAO,GAAG,EAAE,SAAS,EAAE,EAAE,UAAU,EAAE,EAA0B,CAAC;IACtE,MAAM,IAAI,GAAG,MAAM,SAAS,CAAC,IAAI,CAAC,OAAO,EAAE,SAAS,CAAC,CAAC;IACtD,MAAM,QAAQ,GAAG,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACxC,MAAM,MAAM,GAAG,MAAM,UAAU,CAC7B,SAAS,EACT,QAAQ,EACR;QACE,QAAQ;QACR,KAAK,EAAE,IAAI,CAAC,KAAK;QACjB,QAAQ,EAAE,gBAAgB,CAAC,QAAQ;QACnC,OAAO,EAAE,gBAAgB,CAAC,OAAO;KAClC,EACD,IAAI,CAAC,IAAI,CACV,CAAC;IAEF,IAAI,CAAC,gBAAgB,CAAC,gBAAgB,EAAE,CAAC;QACvC,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,OAAO,gBAAgB,CACrB,MAAM,EACN,IAAI,CAAC,IAAI,EACT,QAAQ,EACR,SAAS,EACT,OAAO,EACP,gBAAgB,CAAC,oBAAoB,IAAI,8BAA8B,CACxE,CAAC;AACJ,CAAC;AAED,SAAS,mBAAmB,CAC1B,UAAkB,EAClB,UAAmC,EACnC,OAA+B;IAE/B,MAAM,QAAQ,GAAG,cAAc,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC;IACvD,MAAM,kBAAkB,GACtB,QAAQ,KAAK,SAAS,IAAI,OAAO,CAAC,uBAAuB;QACvD,CAAC,CAAC,oCAAoC,CAAC,QAAQ,EAAE,OAAO,CAAC,gBAAgB,CAAC;QAC1E,CAAC,CAAC,QAAQ,CAAC;IAEf,OAAO;QACL,WAAW,EAAE,4BAAoB;QACjC,UAAU;QACV,QAAQ,EAAE,kBAAkB;QAC5B,IAAI,EAAE,UAAU,CAAC,OAAO,CAAC,IAAI;QAC7B,MAAM,EAAE;YACN,IAAI,EAAE,cAAc,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,CAAC;YAC7C,IAAI,EAAE,cAAc,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,CAAC;YAC7C,OAAO,EAAE,cAAc,CAAC,UAAU,CAAC,OAAO,CAAC,OAAO,CAAC;SACpD;QACD,MAAM,EAAE,UAAU,CAAC,MAAM;QACzB,iBAAiB,EAAE,UAAU,CAAC,iBAAiB;QAC/C,MAAM,EAAE,UAAU,CAAC,MAAM;KAC1B,CAAC;AACJ,CAAC;AAED,KAAK,UAAU,gBAAgB,CAC7B,MAA2B,EAC3B,UAAkB,EAClB,QAAgB,EAChB,SAAkD,EAClD,OAA6B,EAC7B,YAA0C;IAE1C,MAAM,MAAM,GAA0C;QACpD,CAAC,IAAI,EAAE,MAAM,CAAC,QAAQ,CAAC;QACvB,CAAC,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC;QACrB,CAAC,MAAM,EAAE,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC;QAC5B,CAAC,MAAM,EAAE,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC;QAC5B,CAAC,SAAS,EAAE,MAAM,CAAC,MAAM,CAAC,OAAO,CAAC;KACnC,CAAC;IACF,MAAM,aAAa,GAAiD,EAAE,CAAC;IAEvE,KAAK,MAAM,CAAC,MAAM,EAAE,KAAK,CAAC,IAAI,MAAM,EAAE,CAAC;QACrC,IAAI,KAAK,KAAK,SAAS;YAAE,SAAS;QAElC,MAAM,YAAY,GAAG,IAAA,2CAAsB,EAAC,YAAY,CAAC,EAAE,UAAU,EAAE,QAAQ,EAAE,MAAM,EAAE,CAAC,CAAC,CAAC;QAC5F,MAAM,SAAS,CAAC,KAAK,CAAC,OAAO,EAAE,YAAY,EAAE,eAAe,CAAC,MAAM,EAAE,KAAK,CAAC,EAAE;YAC3E,SAAS,EAAE,IAAI;SAChB,CAAC,CAAC;QACH,aAAa,CAAC,MAAM,CAAC,GAAG,YAAY,CAAC;IACvC,CAAC;IAED,OAAO,EAAE,GAAG,MAAM,EAAE,aAAa,EAAE,CAAC;AACtC,CAAC;AAED,SAAS,sBAAsB,CAC7B,GAAG,SAA6D;IAEhE,MAAM,MAAM,GAAG,SAAS,CAAC,MAAM,CAC7B,CAAC,OAAO,EAAE,QAAQ,EAAE,EAAE,CAAC,CAAC;QACtB,GAAG,OAAO;QACV,GAAG,CAAC,QAAQ,IAAI,EAAE,CAAC;QACnB,WAAW,EAAE;YACX,GAAG,OAAO,CAAC,WAAW;YACtB,GAAG,CAAC,QAAQ,EAAE,WAAW,IAAI,EAAE,CAAC;SACjC;KACF,CAAC,EACF,EAAE,GAAG,wCAAgC,EAAE,WAAW,EAAE,EAAE,EAAE,CACzD,CAAC;IAEF,MAAM,aAAa,GAAG,sBAAsB,CAAC,MAAM,CAAC,aAAa,CAAC,CAAC;IACnE,IAAI,OAAO,MAAM,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,MAAM,CAAC,QAAQ,KAAK,SAAS,EAAE,CAAC;QAC9E,MAAM,IAAI,KAAK,CAAC,6CAA6C,CAAC,CAAC;IACjE,CAAC;IACD,IAAI,MAAM,CAAC,SAAS,KAAK,MAAM,IAAI,MAAM,CAAC,SAAS,KAAK,UAAU,EAAE,CAAC;QACnE,MAAM,IAAI,KAAK,CAAC,kCAAkC,MAAM,CAAC,MAAM,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;IAChF,CAAC;IACD,IAAI,CAAC,CAAC,aAAa,EAAE,UAAU,EAAE,YAAY,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,eAAe,CAAC,EAAE,CAAC;QAChF,MAAM,IAAI,KAAK,CAAC,wCAAwC,MAAM,CAAC,MAAM,CAAC,eAAe,CAAC,EAAE,CAAC,CAAC;IAC5F,CAAC;IACD,IAAI,CAAC,CAAC,UAAU,EAAE,KAAK,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,QAAQ,CAAC,EAAE,CAAC;QACnD,MAAM,IAAI,KAAK,CAAC,iCAAiC,MAAM,CAAC,MAAM,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC;IAC9E,CAAC;IACD,KAAK,MAAM,IAAI,IAAI,CAAC,WAAW,EAAE,YAAY,EAAE,kBAAkB,CAAU,EAAE,CAAC;QAC5E,IAAI,OAAO,MAAM,CAAC,IAAI,CAAC,KAAK,QAAQ,IAAI,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACzE,MAAM,IAAI,KAAK,CAAC,WAAW,IAAI,8BAA8B,CAAC,CAAC;QACjE,CAAC;IACH,CAAC;IACD,KAAK,MAAM,IAAI,IAAI;QACjB,cAAc;QACd,cAAc;QACd,kBAAkB;QAClB,qBAAqB;QACrB,yBAAyB;QACzB,sBAAsB;QACtB,yBAAyB;KACjB,EAAE,CAAC;QACX,IAAI,OAAO,MAAM,CAAC,IAAI,CAAC,KAAK,SAAS,EAAE,CAAC;YACtC,MAAM,IAAI,KAAK,CAAC,WAAW,IAAI,2BAA2B,CAAC,CAAC;QAC9D,CAAC;IACH,CAAC;IAED,OAAO,EAAE,GAAG,MAAM,EAAE,aAAa,EAAE,CAAC;AACtC,CAAC;AAED,SAAS,sBAAsB,CAAC,KAAqC;IACnE,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QAChD,MAAM,IAAI,KAAK,CAAC,yDAAyD,CAAC,CAAC;IAC7E,CAAC;IAED,MAAM,OAAO,GAAG,CAAC,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC;IACpC,KAAK,MAAM,MAAM,IAAI,OAAO,EAAE,CAAC;QAC7B,IAAI,CAAC,CAAC,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,SAAS,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAChE,MAAM,IAAI,KAAK,CAAC,sCAAsC,MAAM,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC;QAC1E,CAAC;IACH,CAAC;IACD,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,SAAS,aAAa,CAAC,OAA+B;IACpD,OAAO;QACL,GAAG,OAAO,CAAC,WAAW;QACtB,UAAU,EAAE,OAAO,CAAC,SAAS;QAC7B,WAAW,EAAE,OAAO,CAAC,UAAU;QAC/B,cAAc,EAAE,OAAO,CAAC,YAAY;QACpC,cAAc,EAAE,OAAO,CAAC,YAAY;QACpC,kBAAkB,EAAE,OAAO,CAAC,gBAAgB;QAC5C,qBAAqB,EAAE,OAAO,CAAC,mBAAmB;QAClD,yBAAyB,EAAE,OAAO,CAAC,uBAAuB;QAC1D,sBAAsB,EAAE,OAAO,CAAC,oBAAoB;KACrD,CAAC;AACJ,CAAC;AAED,SAAS,cAAc,CAAC,KAAc;IACpC,IAAI,KAAK,KAAK,SAAS,IAAI,KAAK,KAAK,IAAI;QAAE,OAAO,SAAS,CAAC;IAC5D,OAAO,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC;AACnE,CAAC;AAED,SAAS,eAAe,CAAC,MAA2B,EAAE,KAAc;IAClE,IAAI,MAAM,KAAK,MAAM,IAAI,OAAO,KAAK,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAC;IACjE,OAAO,IAAI,CAAC,SAAS,CAAC,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC;AACxC,CAAC;AAED,SAAS,iBAAiB,CAAC,SAAiB;IAC1C,MAAM,QAAQ,GAAG,SAAS,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,GAAG,EAAE,EAAE,IAAI,EAAE,CAAC;IAC/D,IAAI,CAAC,QAAQ;QAAE,MAAM,IAAI,KAAK,CAAC,+BAA+B,CAAC,CAAC;IAChE,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED,SAAS,WAAW,CAAC,cAAsB;IACzC,OAAO,iBAAiB,CAAC,cAAc,CAAC,CAAC;AAC3C,CAAC"}

package/dist/index.d.ts

@@ -0,0 +1,4 @@
+export { createDoclingTools, createDoclingToolPack, resolveDoclingSkillPath } from './docling';
+export type { DoclingProcessResult } from './docling';
+export { createDoclingDocumentProcessor, DEFAULT_DOCLING_DOCUMENT_OPTIONS, DOCLING_PROCESSOR_ID, IMAGE_PLACEHOLDER, replaceEmbeddedImagesWithPlaceholder, type DoclingDocumentOptions, type DoclingDocumentProcessorOptions, type DocumentArtifactPathInput, type DocumentArtifactPathResolver, type DocumentParseInput, type DocumentParseOutput, type DocumentParseStatus, type DocumentProcessor, type DocumentWorkspaceParseOptions, } from './document-processor';
+export { createDoclingConverter, type DoclingConvertFileInput, type DoclingConversionResult, type DoclingConversionStatus, type DoclingConverter, type DoclingFormValue, type DoclingOutputFormat, type DoclingToolPackOptions, } from './converter';

package/dist/index.js

@@ -0,0 +1,16 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.createDoclingConverter = exports.replaceEmbeddedImagesWithPlaceholder = exports.IMAGE_PLACEHOLDER = exports.DOCLING_PROCESSOR_ID = exports.DEFAULT_DOCLING_DOCUMENT_OPTIONS = exports.createDoclingDocumentProcessor = exports.resolveDoclingSkillPath = exports.createDoclingToolPack = exports.createDoclingTools = void 0;
+var docling_1 = require("./docling");
+Object.defineProperty(exports, "createDoclingTools", { enumerable: true, get: function () { return docling_1.createDoclingTools; } });
+Object.defineProperty(exports, "createDoclingToolPack", { enumerable: true, get: function () { return docling_1.createDoclingToolPack; } });
+Object.defineProperty(exports, "resolveDoclingSkillPath", { enumerable: true, get: function () { return docling_1.resolveDoclingSkillPath; } });
+var document_processor_1 = require("./document-processor");
+Object.defineProperty(exports, "createDoclingDocumentProcessor", { enumerable: true, get: function () { return document_processor_1.createDoclingDocumentProcessor; } });
+Object.defineProperty(exports, "DEFAULT_DOCLING_DOCUMENT_OPTIONS", { enumerable: true, get: function () { return document_processor_1.DEFAULT_DOCLING_DOCUMENT_OPTIONS; } });
+Object.defineProperty(exports, "DOCLING_PROCESSOR_ID", { enumerable: true, get: function () { return document_processor_1.DOCLING_PROCESSOR_ID; } });
+Object.defineProperty(exports, "IMAGE_PLACEHOLDER", { enumerable: true, get: function () { return document_processor_1.IMAGE_PLACEHOLDER; } });
+Object.defineProperty(exports, "replaceEmbeddedImagesWithPlaceholder", { enumerable: true, get: function () { return document_processor_1.replaceEmbeddedImagesWithPlaceholder; } });
+var converter_1 = require("./converter");
+Object.defineProperty(exports, "createDoclingConverter", { enumerable: true, get: function () { return converter_1.createDoclingConverter; } });
+//# sourceMappingURL=index.js.map

package/dist/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":";;;AAAA,qCAA+F;AAAtF,6GAAA,kBAAkB,OAAA;AAAE,gHAAA,qBAAqB,OAAA;AAAE,kHAAA,uBAAuB,OAAA;AAE3E,2DAe8B;AAd5B,oIAAA,8BAA8B,OAAA;AAC9B,sIAAA,gCAAgC,OAAA;AAChC,0HAAA,oBAAoB,OAAA;AACpB,uHAAA,iBAAiB,OAAA;AACjB,0IAAA,oCAAoC,OAAA;AAWtC,yCASqB;AARnB,mHAAA,sBAAsB,OAAA"}

package/package.json

@@ -0,0 +1,42 @@
+{
+  "name": "@lumifai/harness-tool-pack-docling",
+  "version": "0.1.0",
+  "license": "UNLICENSED",
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/Lumif-ai/lumif-harness.git",
+    "directory": "packages/tool-packs/docling"
+  },
+  "engines": {
+    "node": ">=22.13.0"
+  },
+  "exports": {
+    ".": {
+      "types": "./dist/index.d.ts",
+      "import": "./dist/index.js",
+      "require": "./dist/index.js",
+      "default": "./dist/index.js"
+    },
+    "./package.json": "./package.json"
+  },
+  "main": "dist/index.js",
+  "types": "dist/index.d.ts",
+  "files": [
+    "dist",
+    "!dist/**/*.test.*",
+    "skills"
+  ],
+  "dependencies": {
+    "@lumifai/harness-tool-packs": "0.2.0",
+    "@mastra/core": "1.67.0",
+    "zod": "^4.4.3"
+  },
+  "publishConfig": {
+    "access": "public",
+    "registry": "https://registry.npmjs.org/"
+  },
+  "scripts": {
+    "build": "tsc -p tsconfig.json",
+    "test": "vitest run --root ../../../ packages/tool-packs/docling/src"
+  }
+}

package/skills/docling/SKILL.md

@@ -0,0 +1,33 @@
+---
+name: docling
+description: Convert supported workspace files through the configured Docling service
+version: 1.0.0
+tags:
+  - docling
+  - file-processing
+  - pdf
+  - office
+  - spreadsheets
+---
+
+# Docling File Processing
+
+Use `process_file` to convert a workspace file through the configured Docling
+Serve endpoint. The tool supports the formats enabled by that Docling service,
+including PDF, DOC/DOCX, PPT/PPTX, XLS/XLSX, CSV, HTML, Markdown, images, and
+other Docling-supported formats.
+
+## Workflow
+
+1. Call `process_file` with an workspace-relative path.
+2. Select `output_format` when Markdown is not appropriate.
+3. Use the returned `output_path` to read the converted file.
+
+Results are written under `processed/` and the converted content is not
+returned inline. Service credentials are configured by the host when the
+tool-pack is constructed; never request credentials from the user or model.
+
+Application integrations that already have bytes or need structured outputs
+should use `createDoclingDocumentProcessor`. It shares the same Docling defaults
+as the tool, returns Markdown/JSON/extras with status and timing metadata, and
+can read and persist files through an explicitly supplied workspace filesystem.