@opendataloader/pdf 2.5.10 → 2.5.12
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +74 -6
- package/dist/cli.cjs +7 -0
- package/dist/cli.cjs.map +1 -1
- package/dist/cli.js +7 -0
- package/dist/cli.js.map +1 -1
- package/dist/index.cjs +3 -0
- package/dist/index.cjs.map +1 -1
- package/dist/index.d.cts +2 -0
- package/dist/index.d.ts +2 -0
- package/dist/index.js +3 -0
- package/dist/index.js.map +1 -1
- package/lib/opendataloader-pdf-cli.jar +0 -0
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -130,7 +130,7 @@ opendataloader_pdf.convert(
|
|
|
130
130
|
| Standard digital PDF | Fast (default) | `pip install opendataloader-pdf` | None needed | `opendataloader-pdf file1.pdf file2.pdf folder/` |
|
|
131
131
|
| Complex or nested tables | **Hybrid** | `pip install "opendataloader-pdf[hybrid]"` | `opendataloader-pdf-hybrid --port 5002` | `opendataloader-pdf --hybrid docling-fast file1.pdf file2.pdf folder/` |
|
|
132
132
|
| Scanned / image-based PDF | Hybrid + OCR | `pip install "opendataloader-pdf[hybrid]"` | `opendataloader-pdf-hybrid --port 5002 --force-ocr` | `opendataloader-pdf --hybrid docling-fast file1.pdf file2.pdf folder/` |
|
|
133
|
-
| Non-English scanned PDF | Hybrid + OCR | `pip install "opendataloader-pdf[hybrid]"` | `opendataloader-pdf-hybrid --port 5002 --force-ocr --ocr-lang
|
|
133
|
+
| Non-English scanned PDF | Hybrid + OCR | `pip install "opendataloader-pdf[hybrid]"` | `opendataloader-pdf-hybrid --port 5002 --force-ocr --ocr-engine <engine> --ocr-lang <codes>` — see [Choosing an OCR Engine](#choosing-an-ocr-engine) | `opendataloader-pdf --hybrid docling-fast file1.pdf file2.pdf folder/` |
|
|
134
134
|
| Mathematical formulas | Hybrid + formula | `pip install "opendataloader-pdf[hybrid]"` | `opendataloader-pdf-hybrid --enrich-formula` | `opendataloader-pdf --hybrid docling-fast --hybrid-mode full file1.pdf file2.pdf folder/` |
|
|
135
135
|
| Charts needing description | Hybrid + picture | `pip install "opendataloader-pdf[hybrid]"` | `opendataloader-pdf-hybrid --enrich-picture-description` | `opendataloader-pdf --hybrid docling-fast --hybrid-mode full file1.pdf file2.pdf folder/` |
|
|
136
136
|
| Nested sections for RAG chunking | Hybrid + heading levels | `pip install "opendataloader-pdf[hybrid]"` | `opendataloader-pdf-hybrid --port 5002 --heading-hierarchy` | `opendataloader-pdf --hybrid docling-fast file1.pdf file2.pdf folder/` |
|
|
@@ -229,7 +229,44 @@ For non-English documents, specify the language:
|
|
|
229
229
|
opendataloader-pdf-hybrid --port 5002 --force-ocr --ocr-lang "ko,en"
|
|
230
230
|
```
|
|
231
231
|
|
|
232
|
-
|
|
232
|
+
The language codes above are for the default engine (EasyOCR): `en`, `ko`, `ja`, `ch_sim`, `ch_tra`, `de`, `fr`, `ar`, and more.
|
|
233
|
+
|
|
234
|
+
#### Choosing an OCR Engine
|
|
235
|
+
|
|
236
|
+
OCR quality varies with the engine, the language, and the document itself (scan resolution, fonts, layout). No engine is best for every document, so run a few representative pages through several engines and compare the output before processing a whole collection.
|
|
237
|
+
|
|
238
|
+
Select the engine with `--ocr-engine`. Each engine uses its own language code system for `--ocr-lang`:
|
|
239
|
+
|
|
240
|
+
| Engine | `--ocr-engine` | `--ocr-lang` example | Platform / extra install |
|
|
241
|
+
|--------|----------------|----------------------|--------------------------|
|
|
242
|
+
| EasyOCR (default) | `easyocr` | `ko,en` (ISO 639-1) | Included in `opendataloader-pdf[hybrid]` |
|
|
243
|
+
| RapidOCR | `rapidocr` | `korean` (RapidOCR codes: `en`, `ch`, `japan`, `korean`, ...) | `pip install rapidocr onnxruntime` |
|
|
244
|
+
| Tesseract (CLI) | `tesseract` | `kor,eng` (ISO 639-2) | Tesseract binary + language data (e.g. `tesseract-ocr-kor`) |
|
|
245
|
+
| Tesseract (Python binding) | `tesserocr` | `kor,eng` (ISO 639-2) | `pip install tesserocr` + libtesseract |
|
|
246
|
+
| Apple Vision | `ocrmac` | `ko-KR,en-US` (BCP-47) | macOS only, `pip install ocrmac` |
|
|
247
|
+
| NVIDIA Nemotron OCR | `nemotron-ocr` | `multilingual` or `english` | Linux x86_64 + CUDA, `pip install "docling[feat-ocr-nemotron]"` |
|
|
248
|
+
| Automatic | `auto` | Not configurable | Docling picks an available engine |
|
|
249
|
+
|
|
250
|
+
To compare engines on a sample, start one server per engine and send the same file to each.
|
|
251
|
+
|
|
252
|
+
**Terminal 1** — EasyOCR server:
|
|
253
|
+
|
|
254
|
+
```bash
|
|
255
|
+
opendataloader-pdf-hybrid --port 5002 --force-ocr --ocr-engine easyocr --ocr-lang "ko,en"
|
|
256
|
+
```
|
|
257
|
+
|
|
258
|
+
**Terminal 2** — Tesseract server:
|
|
259
|
+
|
|
260
|
+
```bash
|
|
261
|
+
opendataloader-pdf-hybrid --port 5003 --force-ocr --ocr-engine tesseract --ocr-lang "kor,eng"
|
|
262
|
+
```
|
|
263
|
+
|
|
264
|
+
**Terminal 3** — Send the same sample to both:
|
|
265
|
+
|
|
266
|
+
```bash
|
|
267
|
+
opendataloader-pdf sample.pdf --hybrid docling-fast --hybrid-mode full --hybrid-url http://localhost:5002 -o out-easyocr
|
|
268
|
+
opendataloader-pdf sample.pdf --hybrid docling-fast --hybrid-mode full --hybrid-url http://localhost:5003 -o out-tesseract
|
|
269
|
+
```
|
|
233
270
|
|
|
234
271
|
### Formula Extraction (LaTeX)
|
|
235
272
|
|
|
@@ -270,14 +307,45 @@ opendataloader-pdf --hybrid docling-fast --hybrid-mode full file1.pdf file2.pdf
|
|
|
270
307
|
Output in JSON:
|
|
271
308
|
```json
|
|
272
309
|
{
|
|
273
|
-
"type": "
|
|
310
|
+
"type": "image",
|
|
311
|
+
"pdfua_tag": "Figure",
|
|
312
|
+
"id": 285,
|
|
274
313
|
"page number": 1,
|
|
275
314
|
"bounding box": [72.0, 400.0, 540.0, 650.0],
|
|
276
|
-
"
|
|
315
|
+
"alt": "A bar chart showing waste generation by region from 2016 to 2030...",
|
|
316
|
+
"alt_source": "ai-generated"
|
|
277
317
|
}
|
|
278
318
|
```
|
|
279
319
|
|
|
320
|
+
`alt_source` says where the text came from: `original` (the PDF's own /Alt), `ai-generated`,
|
|
321
|
+
or `missing` — and when it is `missing`, there is no `alt` field at all.
|
|
322
|
+
|
|
323
|
+
With `--enrich-picture-description`, every picture goes to the model whatever its size. The
|
|
324
|
+
server reports the tally per request, so a picture that came back without a description shows
|
|
325
|
+
up in the log rather than only in the output:
|
|
326
|
+
|
|
327
|
+
```
|
|
328
|
+
picture_description pictures=4 described=4
|
|
329
|
+
```
|
|
330
|
+
|
|
331
|
+
Cost follows the number of pictures, not their size: a 40×20 pt logo costs about what a
|
|
332
|
+
full-page chart does, and on CPU one picture can take over a minute. A document full of icons
|
|
333
|
+
is therefore minutes of work per page, and conversions are serialized, so other requests wait.
|
|
334
|
+
`--picture-area-threshold 0.05` spends the model only on pictures covering at least 5% of
|
|
335
|
+
their page:
|
|
336
|
+
|
|
337
|
+
```
|
|
338
|
+
picture_description pictures=4 described=1 skipped=3 threshold=0.05
|
|
339
|
+
```
|
|
340
|
+
|
|
341
|
+
Skips are logged at INFO when you set a threshold, and at WARNING when you did not — in that
|
|
342
|
+
case nothing was asked to drop them, so the cause is worth a look.
|
|
343
|
+
|
|
280
344
|
> Uses SmolVLM (256M), a lightweight vision model. Custom prompts supported via `--picture-description-prompt`.
|
|
345
|
+
> Since the picture count comes from the uploaded file, a server reachable by untrusted callers
|
|
346
|
+
> should set `--picture-area-threshold`, and `--max-file-size` to cap the conversion work one
|
|
347
|
+
> file can ask for. That limit is applied once the request body has been received, so it bounds
|
|
348
|
+
> conversion cost rather than upload bandwidth or temporary disk.
|
|
281
349
|
|
|
282
350
|
### Heading Hierarchy
|
|
283
351
|
|
|
@@ -542,11 +610,11 @@ Yes. OpenDataLoader runs 100% locally. No API calls, no data transmission — yo
|
|
|
542
610
|
|
|
543
611
|
### Does it support OCR for scanned PDFs?
|
|
544
612
|
|
|
545
|
-
Yes, via hybrid mode. Install with `pip install "opendataloader-pdf[hybrid]"`, start the backend with `--force-ocr`, then process as usual. Supports multiple languages including Korean, Japanese, Chinese, Arabic, and more via `--ocr-lang`.
|
|
613
|
+
Yes, via hybrid mode. Install with `pip install "opendataloader-pdf[hybrid]"`, start the backend with `--force-ocr`, then process as usual. Supports multiple languages including Korean, Japanese, Chinese, Arabic, and more via `--ocr-lang`, with a choice of OCR engines via `--ocr-engine`. Quality varies by engine, language, and document — see [Choosing an OCR Engine](#choosing-an-ocr-engine).
|
|
546
614
|
|
|
547
615
|
### Does it work with Korean, Japanese, or Chinese documents?
|
|
548
616
|
|
|
549
|
-
Yes. For digital PDFs, text extraction works out of the box. For scanned PDFs, use hybrid mode with `--force-ocr --ocr-lang "ko,en"` (or `ja`, `ch_sim`, `ch_tra`). Coming soon: [Hancom Data Loader](https://sdk.hancom.com/en/services/1?utm_source=github&utm_medium=readme&utm_campaign=opendataloader-pdf) integration — enterprise-grade AI document analysis with built-in production-grade OCR and customer-customized models optimized for your specific document types and workflows.
|
|
617
|
+
Yes. For digital PDFs, text extraction works out of the box. For scanned PDFs, use hybrid mode with `--force-ocr --ocr-lang "ko,en"` (or `ja`, `ch_sim`, `ch_tra`), and compare OCR engines on a sample of your documents — see [Choosing an OCR Engine](#choosing-an-ocr-engine). Coming soon: [Hancom Data Loader](https://sdk.hancom.com/en/services/1?utm_source=github&utm_medium=readme&utm_campaign=opendataloader-pdf) integration — enterprise-grade AI document analysis with built-in production-grade OCR and customer-customized models optimized for your specific document types and workflows.
|
|
550
618
|
|
|
551
619
|
### How fast is it?
|
|
552
620
|
|
package/dist/cli.cjs
CHANGED
|
@@ -115,6 +115,9 @@ function buildConvertOptions(cliOptions) {
|
|
|
115
115
|
if (cliOptions.hybridTimeout) {
|
|
116
116
|
convertOptions.hybridTimeout = cliOptions.hybridTimeout;
|
|
117
117
|
}
|
|
118
|
+
if (cliOptions.hybridChunkSize) {
|
|
119
|
+
convertOptions.hybridChunkSize = cliOptions.hybridChunkSize;
|
|
120
|
+
}
|
|
118
121
|
if (cliOptions.hybridFallback) {
|
|
119
122
|
convertOptions.hybridFallback = true;
|
|
120
123
|
}
|
|
@@ -221,6 +224,9 @@ function buildArgs(options) {
|
|
|
221
224
|
if (options.hybridTimeout) {
|
|
222
225
|
args.push("--hybrid-timeout", options.hybridTimeout);
|
|
223
226
|
}
|
|
227
|
+
if (options.hybridChunkSize) {
|
|
228
|
+
args.push("--hybrid-chunk-size", options.hybridChunkSize);
|
|
229
|
+
}
|
|
224
230
|
if (options.hybridFallback) {
|
|
225
231
|
args.push("--hybrid-fallback");
|
|
226
232
|
}
|
|
@@ -369,6 +375,7 @@ function registerCliOptions(program) {
|
|
|
369
375
|
program.option("--hybrid-mode <value>", "Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend)");
|
|
370
376
|
program.option("--hybrid-url <value>", "Hybrid backend server URL (overrides default)");
|
|
371
377
|
program.option("--hybrid-timeout <value>", "Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0");
|
|
378
|
+
program.option("--hybrid-chunk-size <value>", "Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50");
|
|
372
379
|
program.option("--hybrid-fallback", "Opt in to Java fallback on hybrid backend error (default: disabled)");
|
|
373
380
|
program.option("--to-stdout", "Write output to stdout instead of file (single format only)");
|
|
374
381
|
program.option("--threads <value>", "Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode");
|
package/dist/cli.cjs.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"sources":["../node_modules/.pnpm/tsup@8.5.1_postcss@8.5.28_typescript@6.0.3/node_modules/tsup/assets/cjs_shims.js","../src/cli.ts","../src/index.ts","../src/convert-options.generated.ts","../src/cli-options.generated.ts"],"sourcesContent":["// Shim globals in cjs bundle\n// There's a weird bug that esbuild will always inject importMetaUrl\n// if we export it as `const importMetaUrl = ... __filename ...`\n// But using a function will not cause this issue\n\nconst getImportMetaUrl = () => \n typeof document === \"undefined\" \n ? new URL(`file:${__filename}`).href \n : (document.currentScript && document.currentScript.tagName.toUpperCase() === 'SCRIPT') \n ? document.currentScript.src \n : new URL(\"main.js\", document.baseURI).href;\n\nexport const importMetaUrl = /* @__PURE__ */ getImportMetaUrl()\n","#!/usr/bin/env node\nimport { Command, CommanderError } from 'commander';\nimport { _runForCli } from './index.js';\nimport { CliOptions, buildConvertOptions } from './convert-options.generated.js';\nimport { registerCliOptions } from './cli-options.generated.js';\n\nfunction createProgram(): Command {\n const program = new Command();\n\n program\n .name('opendataloader-pdf')\n .usage('[options] <input...>')\n .description('Convert PDFs using the OpenDataLoader CLI.')\n .showHelpAfterError(\"Use '--help' to see available options.\")\n .showSuggestionAfterError(false)\n .argument('<input...>', 'Input files or directories to convert');\n\n // Register CLI options from auto-generated file\n registerCliOptions(program);\n\n program.configureOutput({\n writeErr: (str) => {\n console.error(str.trimEnd());\n },\n outputError: (str, write) => {\n write(str);\n },\n });\n\n return program;\n}\n\nasync function main(): Promise<number> {\n const program = createProgram();\n\n program.exitOverride();\n\n try {\n program.parse(process.argv);\n } catch (err) {\n if (err instanceof CommanderError) {\n if (err.code === 'commander.helpDisplayed') {\n return 0;\n }\n return err.exitCode ?? 1;\n }\n\n const message = err instanceof Error ? err.message : String(err);\n console.error(message);\n console.error(\"Use '--help' to see available options.\");\n return 1;\n }\n\n const cliOptions = program.opts<CliOptions>();\n const inputPaths = program.args;\n const convertOptions = buildConvertOptions(cliOptions);\n\n try {\n // _runForCli streams stdout/stderr to the parent process as they arrive;\n // we deliberately do not re-print anything here. (Issue #398.)\n await _runForCli(inputPaths, convertOptions);\n return 0;\n } catch (err) {\n // Subprocess-exit errors are already on the user's terminal via the live\n // stderr stream — re-printing would duplicate output and risk leaking\n // anything sensitive Java logged (e.g. a --password value echoed by an\n // underlying library). Wrapper-side failures (JAR not found, java not in\n // PATH, bad input path) still need to be surfaced.\n const isJavaExit =\n err instanceof Error && (err as Error & { isJavaExit?: boolean }).isJavaExit === true;\n if (!isJavaExit) {\n const message = err instanceof Error ? err.message : String(err);\n console.error(message);\n }\n return 1;\n }\n}\n\nmain().then((code) => {\n if (code !== 0) {\n process.exit(code);\n }\n});\n","import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { StringDecoder } from 'string_decoder';\nimport { fileURLToPath } from 'url';\n\n// Re-export types and utilities from auto-generated file\nexport type { ConvertOptions } from './convert-options.generated.js';\nexport { buildArgs } from './convert-options.generated.js';\nimport type { ConvertOptions } from './convert-options.generated.js';\nimport { buildArgs } from './convert-options.generated.js';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\ninterface JarExecutionOptions {\n /**\n * When true, forwards Java's stdout and stderr chunks to the parent\n * process in real time as well as accumulating them. Used by the bundled\n * CLI so long-running conversions show progress as it happens.\n */\n streamOutput?: boolean;\n}\n\nfunction executeJar(args: string[], executionOptions: JarExecutionOptions = {}): Promise<string> {\n const { streamOutput = false } = executionOptions;\n\n return new Promise((resolve, reject) => {\n const jarPath = path.join(__dirname, '..', 'lib', JAR_NAME);\n\n if (!fs.existsSync(jarPath)) {\n return reject(\n new Error(`JAR file not found at ${jarPath}. Please run the build script first.`),\n );\n }\n\n const command = 'java';\n // Force headless AWT so macOS doesn't surface a Dock icon (and steal focus)\n // every time the JVM touches ImageIO/PDFBox rendering. Safe on all OSes —\n // the CLI never opens a UI window, only manipulates BufferedImages.\n const commandArgs = [\n '-Djava.awt.headless=true',\n '-Dapple.awt.UIElement=true',\n '-jar',\n jarPath,\n ...args,\n ];\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n // StringDecoder buffers incomplete multi-byte UTF-8 sequences across\n // chunk boundaries — Buffer.toString() alone would emit replacement\n // characters when, e.g., a 3-byte Korean codepoint splits across two\n // 'data' events. One decoder per stream so they don't share state.\n const stdoutDecoder = new StringDecoder('utf8');\n const stderrDecoder = new StringDecoder('utf8');\n\n javaProcess.stdout.on('data', (data: Buffer) => {\n const chunk = stdoutDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n // Stream-only: don't also accumulate, or a multi-hour conversion\n // would buffer its entire (potentially gigabyte) stdout in memory\n // for no consumer.\n process.stdout.write(chunk);\n } else {\n stdout += chunk;\n }\n });\n\n javaProcess.stderr.on('data', (data: Buffer) => {\n const chunk = stderrDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n process.stderr.write(chunk);\n }\n // stderr is always accumulated (progress logs are small and we need\n // them for error messages on non-zero exit).\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n // Flush any trailing bytes the decoder is still holding (always emit\n // them — if we drop them on error paths, error messages with non-ASCII\n // characters lose their tail).\n const stdoutTail = stdoutDecoder.end();\n const stderrTail = stderrDecoder.end();\n if (stdoutTail.length > 0) {\n if (streamOutput) {\n process.stdout.write(stdoutTail);\n } else {\n stdout += stdoutTail;\n }\n }\n if (stderrTail.length > 0) {\n if (streamOutput) process.stderr.write(stderrTail);\n stderr += stderrTail;\n }\n\n if (code === 0) {\n resolve(stdout);\n } else {\n const errorOutput = stderr || stdout;\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${errorOutput}`,\n );\n // Tag so the CLI can suppress re-printing this message — Java's\n // stderr was already streamed live to the parent in CLI mode, and\n // re-printing risks leaking anything sensitive Java logged\n // (e.g. a --password value echoed by an underlying library).\n (error as Error & { isJavaExit?: boolean }).isJavaExit = true;\n reject(error);\n }\n });\n\n javaProcess.on('error', (err: Error) => {\n if (err.message.includes('ENOENT')) {\n reject(\n new Error(\n \"'java' command not found. Please ensure Java is installed and in your system's PATH.\",\n ),\n );\n } else {\n reject(err);\n }\n });\n });\n}\n\nfunction buildJarArgs(\n inputPaths: string | string[],\n options: ConvertOptions,\n): string[] | Error {\n const inputList = Array.isArray(inputPaths) ? inputPaths : [inputPaths];\n if (inputList.length === 0) {\n return new Error('At least one input path must be provided.');\n }\n\n for (const input of inputList) {\n if (!fs.existsSync(input)) {\n return new Error(`Input file or folder not found: ${input}`);\n }\n }\n\n return [...inputList, ...buildArgs(options)];\n}\n\nexport function convert(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<string> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n return Promise.reject(argsOrError);\n }\n // Library API: never streams to the parent process. Returns the full stdout\n // string so callers can do `const out = await convert(...)` without surprise\n // side-effects on process.stdout / process.stderr.\n return executeJar(argsOrError, { streamOutput: false });\n}\n\n/**\n * Internal entry point used by the bundled CLI. Streams Java's stdout and\n * stderr to the parent process in real time (so long-running conversions like\n * hybrid mode show progress as it happens) and resolves without a stdout\n * payload — preventing the caller from re-printing what was already streamed.\n *\n * Not part of the public API: do not import this from application code. Use\n * {@link convert} instead.\n *\n * @internal\n */\nexport async function _runForCli(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<void> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n throw argsOrError;\n }\n await executeJar(argsOrError, { streamOutput: true });\n}\n\n/**\n * @deprecated Use `convert()` and `ConvertOptions` instead. This function will be removed in a future version.\n */\nexport interface RunOptions {\n outputFolder?: string;\n password?: string;\n replaceInvalidChars?: string;\n generateMarkdown?: boolean;\n generateHtml?: boolean;\n generateAnnotatedPdf?: boolean;\n keepLineBreaks?: boolean;\n contentSafetyOff?: string;\n htmlInMarkdown?: boolean;\n addImageToMarkdown?: boolean;\n noJson?: boolean;\n debug?: boolean;\n useStructTree?: boolean;\n}\n\n/**\n * @deprecated Use `convert()` instead. This function will be removed in a future version.\n */\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n console.warn(\n 'Warning: run() is deprecated and will be removed in a future version. Use convert() instead.',\n );\n\n // Build format array based on legacy boolean options\n const formats: string[] = [];\n if (!options.noJson) {\n formats.push('json');\n }\n if (options.generateMarkdown) {\n if (options.addImageToMarkdown) {\n formats.push('markdown-with-images');\n } else if (options.htmlInMarkdown) {\n formats.push('markdown-with-html');\n } else {\n formats.push('markdown');\n }\n }\n if (options.generateHtml) {\n formats.push('html');\n }\n if (options.generateAnnotatedPdf) {\n formats.push('pdf');\n }\n\n return convert(inputPath, {\n outputDir: options.outputFolder,\n password: options.password,\n replaceInvalidChars: options.replaceInvalidChars,\n keepLineBreaks: options.keepLineBreaks,\n contentSafetyOff: options.contentSafetyOff,\n useStructTree: options.useStructTree,\n format: formats.length > 0 ? formats : undefined,\n quiet: !options.debug,\n });\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\n/**\n * Options for the convert function.\n */\nexport interface ConvertOptions {\n /** Directory where output files are written. Default: input file directory */\n outputDir?: string;\n /** Password for encrypted PDF files */\n password?: string;\n /** Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output. */\n format?: string | string[];\n /** Suppress console logging output */\n quiet?: boolean;\n /** Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background */\n contentSafetyOff?: string | string[];\n /** Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders */\n sanitize?: boolean;\n /** Preserve original line breaks in extracted text */\n keepLineBreaks?: boolean;\n /** Replacement character for invalid/unrecognized characters. Default: space */\n replaceInvalidChars?: string;\n /** Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called */\n useStructTree?: boolean;\n /** Table detection method. Values: default (border-based), cluster (border + cluster). Default: default */\n tableMethod?: string;\n /** Reading order algorithm. Values: off, xycut. Default: xycut */\n readingOrder?: string;\n /** Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none */\n markdownPageSeparator?: string;\n /** Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown. */\n markdownWithHtml?: boolean;\n /** Separator between pages in text output. Use %page-number% for page numbers. Default: none */\n textPageSeparator?: string;\n /** Separator between pages in HTML output. Use %page-number% for page numbers. Default: none */\n htmlPageSeparator?: string;\n /** Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external */\n imageOutput?: string;\n /** Output format for extracted images. Values: png, jpeg. Default: png */\n imageFormat?: string;\n /** Directory for extracted images (applies only with --image-output external) */\n imageDir?: string;\n /** Pages to extract (e.g., \"1,3,5-7\"). Default: all pages */\n pages?: string;\n /** Include page headers and footers in output */\n includeHeaderFooter?: boolean;\n /** Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental) */\n detectStrikethrough?: boolean;\n /** Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence) */\n hybrid?: string;\n /** Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend) */\n hybridMode?: string;\n /** Hybrid backend server URL (overrides default) */\n hybridUrl?: string;\n /** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */\n hybridTimeout?: string;\n /** Opt in to Java fallback on hybrid backend error (default: disabled) */\n hybridFallback?: boolean;\n /** Write output to stdout instead of file (single format only) */\n toStdout?: boolean;\n /** Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode */\n threads?: string;\n /** Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0. */\n imageResolution?: string;\n /** Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17 */\n spaceRatio?: string;\n}\n\n/**\n * Options as parsed from CLI (all values are strings from commander).\n */\nexport interface CliOptions {\n outputDir?: string;\n password?: string;\n format?: string;\n quiet?: boolean;\n contentSafetyOff?: string;\n sanitize?: boolean;\n keepLineBreaks?: boolean;\n replaceInvalidChars?: string;\n useStructTree?: boolean;\n tableMethod?: string;\n readingOrder?: string;\n markdownPageSeparator?: string;\n markdownWithHtml?: boolean;\n textPageSeparator?: string;\n htmlPageSeparator?: string;\n imageOutput?: string;\n imageFormat?: string;\n imageDir?: string;\n pages?: string;\n includeHeaderFooter?: boolean;\n detectStrikethrough?: boolean;\n hybrid?: string;\n hybridMode?: string;\n hybridUrl?: string;\n hybridTimeout?: string;\n hybridFallback?: boolean;\n toStdout?: boolean;\n threads?: string;\n imageResolution?: string;\n spaceRatio?: string;\n}\n\n/**\n * Convert CLI options to ConvertOptions.\n */\nexport function buildConvertOptions(cliOptions: CliOptions): ConvertOptions {\n const convertOptions: ConvertOptions = {};\n\n if (cliOptions.outputDir) {\n convertOptions.outputDir = cliOptions.outputDir;\n }\n if (cliOptions.password) {\n convertOptions.password = cliOptions.password;\n }\n if (cliOptions.format) {\n convertOptions.format = cliOptions.format;\n }\n if (cliOptions.quiet) {\n convertOptions.quiet = true;\n }\n if (cliOptions.contentSafetyOff) {\n convertOptions.contentSafetyOff = cliOptions.contentSafetyOff;\n }\n if (cliOptions.sanitize) {\n convertOptions.sanitize = true;\n }\n if (cliOptions.keepLineBreaks) {\n convertOptions.keepLineBreaks = true;\n }\n if (cliOptions.replaceInvalidChars) {\n convertOptions.replaceInvalidChars = cliOptions.replaceInvalidChars;\n }\n if (cliOptions.useStructTree) {\n convertOptions.useStructTree = true;\n }\n if (cliOptions.tableMethod) {\n convertOptions.tableMethod = cliOptions.tableMethod;\n }\n if (cliOptions.readingOrder) {\n convertOptions.readingOrder = cliOptions.readingOrder;\n }\n if (cliOptions.markdownPageSeparator) {\n convertOptions.markdownPageSeparator = cliOptions.markdownPageSeparator;\n }\n if (cliOptions.markdownWithHtml) {\n convertOptions.markdownWithHtml = true;\n }\n if (cliOptions.textPageSeparator) {\n convertOptions.textPageSeparator = cliOptions.textPageSeparator;\n }\n if (cliOptions.htmlPageSeparator) {\n convertOptions.htmlPageSeparator = cliOptions.htmlPageSeparator;\n }\n if (cliOptions.imageOutput) {\n convertOptions.imageOutput = cliOptions.imageOutput;\n }\n if (cliOptions.imageFormat) {\n convertOptions.imageFormat = cliOptions.imageFormat;\n }\n if (cliOptions.imageDir) {\n convertOptions.imageDir = cliOptions.imageDir;\n }\n if (cliOptions.pages) {\n convertOptions.pages = cliOptions.pages;\n }\n if (cliOptions.includeHeaderFooter) {\n convertOptions.includeHeaderFooter = true;\n }\n if (cliOptions.detectStrikethrough) {\n convertOptions.detectStrikethrough = true;\n }\n if (cliOptions.hybrid) {\n convertOptions.hybrid = cliOptions.hybrid;\n }\n if (cliOptions.hybridMode) {\n convertOptions.hybridMode = cliOptions.hybridMode;\n }\n if (cliOptions.hybridUrl) {\n convertOptions.hybridUrl = cliOptions.hybridUrl;\n }\n if (cliOptions.hybridTimeout) {\n convertOptions.hybridTimeout = cliOptions.hybridTimeout;\n }\n if (cliOptions.hybridFallback) {\n convertOptions.hybridFallback = true;\n }\n if (cliOptions.toStdout) {\n convertOptions.toStdout = true;\n }\n if (cliOptions.threads) {\n convertOptions.threads = cliOptions.threads;\n }\n if (cliOptions.imageResolution) {\n convertOptions.imageResolution = cliOptions.imageResolution;\n }\n if (cliOptions.spaceRatio) {\n convertOptions.spaceRatio = cliOptions.spaceRatio;\n }\n\n return convertOptions;\n}\n\n/**\n * Build CLI arguments array from ConvertOptions.\n */\nexport function buildArgs(options: ConvertOptions): string[] {\n const args: string[] = [];\n\n if (options.outputDir) {\n args.push('--output-dir', options.outputDir);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.format) {\n if (Array.isArray(options.format)) {\n if (options.format.length > 0) {\n args.push('--format', options.format.join(','));\n }\n } else {\n args.push('--format', options.format);\n }\n }\n if (options.quiet) {\n args.push('--quiet');\n }\n if (options.contentSafetyOff) {\n if (Array.isArray(options.contentSafetyOff)) {\n if (options.contentSafetyOff.length > 0) {\n args.push('--content-safety-off', options.contentSafetyOff.join(','));\n }\n } else {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n }\n if (options.sanitize) {\n args.push('--sanitize');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.useStructTree) {\n args.push('--use-struct-tree');\n }\n if (options.tableMethod) {\n args.push('--table-method', options.tableMethod);\n }\n if (options.readingOrder) {\n args.push('--reading-order', options.readingOrder);\n }\n if (options.markdownPageSeparator) {\n args.push('--markdown-page-separator', options.markdownPageSeparator);\n }\n if (options.markdownWithHtml) {\n args.push('--markdown-with-html');\n }\n if (options.textPageSeparator) {\n args.push('--text-page-separator', options.textPageSeparator);\n }\n if (options.htmlPageSeparator) {\n args.push('--html-page-separator', options.htmlPageSeparator);\n }\n if (options.imageOutput) {\n args.push('--image-output', options.imageOutput);\n }\n if (options.imageFormat) {\n args.push('--image-format', options.imageFormat);\n }\n if (options.imageDir) {\n args.push('--image-dir', options.imageDir);\n }\n if (options.pages) {\n args.push('--pages', options.pages);\n }\n if (options.includeHeaderFooter) {\n args.push('--include-header-footer');\n }\n if (options.detectStrikethrough) {\n args.push('--detect-strikethrough');\n }\n if (options.hybrid) {\n args.push('--hybrid', options.hybrid);\n }\n if (options.hybridMode) {\n args.push('--hybrid-mode', options.hybridMode);\n }\n if (options.hybridUrl) {\n args.push('--hybrid-url', options.hybridUrl);\n }\n if (options.hybridTimeout) {\n args.push('--hybrid-timeout', options.hybridTimeout);\n }\n if (options.hybridFallback) {\n args.push('--hybrid-fallback');\n }\n if (options.toStdout) {\n args.push('--to-stdout');\n }\n if (options.threads) {\n args.push('--threads', options.threads);\n }\n if (options.imageResolution) {\n args.push('--image-resolution', options.imageResolution);\n }\n if (options.spaceRatio) {\n args.push('--space-ratio', options.spaceRatio);\n }\n\n return args;\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\nimport { Command } from 'commander';\n\n/**\n * Register all CLI options on the given Commander program.\n */\nexport function registerCliOptions(program: Command): void {\n program.option('-o, --output-dir <value>', 'Directory where output files are written. Default: input file directory');\n program.option('-p, --password <value>', 'Password for encrypted PDF files');\n program.option('-f, --format <value>', 'Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output.');\n program.option('-q, --quiet', 'Suppress console logging output');\n program.option('--content-safety-off <value>', 'Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background');\n program.option('--sanitize', 'Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders');\n program.option('--keep-line-breaks', 'Preserve original line breaks in extracted text');\n program.option('--replace-invalid-chars <value>', 'Replacement character for invalid/unrecognized characters. Default: space');\n program.option('--use-struct-tree', 'Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called');\n program.option('--table-method <value>', 'Table detection method. Values: default (border-based), cluster (border + cluster). Default: default');\n program.option('--reading-order <value>', 'Reading order algorithm. Values: off, xycut. Default: xycut');\n program.option('--markdown-page-separator <value>', 'Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none');\n program.option('--markdown-with-html', 'Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown.');\n program.option('--text-page-separator <value>', 'Separator between pages in text output. Use %page-number% for page numbers. Default: none');\n program.option('--html-page-separator <value>', 'Separator between pages in HTML output. Use %page-number% for page numbers. Default: none');\n program.option('--image-output <value>', 'Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external');\n program.option('--image-format <value>', 'Output format for extracted images. Values: png, jpeg. Default: png');\n program.option('--image-dir <value>', 'Directory for extracted images (applies only with --image-output external)');\n program.option('--pages <value>', 'Pages to extract (e.g., \"1,3,5-7\"). Default: all pages');\n program.option('--include-header-footer', 'Include page headers and footers in output');\n program.option('--detect-strikethrough', 'Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental)');\n program.option('--hybrid <value>', 'Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence)');\n program.option('--hybrid-mode <value>', 'Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend)');\n program.option('--hybrid-url <value>', 'Hybrid backend server URL (overrides default)');\n program.option('--hybrid-timeout <value>', 'Hybrid backend request timeout in milliseconds (0 = use the backend\\'s own default). Default: 0');\n program.option('--hybrid-fallback', 'Opt in to Java fallback on hybrid backend error (default: disabled)');\n program.option('--to-stdout', 'Write output to stdout instead of file (single format only)');\n program.option('--threads <value>', 'Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode');\n program.option('--image-resolution <value>', 'Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0.');\n program.option('--space-ratio <value>', 'Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17');\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;AAKA,IAAM,mBAAmB,MACvB,OAAO,aAAa,cAChB,IAAI,IAAI,QAAQ,UAAU,EAAE,EAAE,OAC7B,SAAS,iBAAiB,SAAS,cAAc,QAAQ,YAAY,MAAM,WAC1E,SAAS,cAAc,MACvB,IAAI,IAAI,WAAW,SAAS,OAAO,EAAE;AAEtC,IAAM,gBAAgC,iCAAiB;;;ACX9D,uBAAwC;;;ACDxC,2BAAsB;AACtB,WAAsB;AACtB,SAAoB;AACpB,4BAA8B;AAC9B,iBAA8B;;;ACwGvB,SAAS,oBAAoB,YAAwC;AAC1E,QAAM,iBAAiC,CAAC;AAExC,MAAI,WAAW,WAAW;AACxB,mBAAe,YAAY,WAAW;AAAA,EACxC;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW,WAAW;AAAA,EACvC;AACA,MAAI,WAAW,QAAQ;AACrB,mBAAe,SAAS,WAAW;AAAA,EACrC;AACA,MAAI,WAAW,OAAO;AACpB,mBAAe,QAAQ;AAAA,EACzB;AACA,MAAI,WAAW,kBAAkB;AAC/B,mBAAe,mBAAmB,WAAW;AAAA,EAC/C;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW;AAAA,EAC5B;AACA,MAAI,WAAW,gBAAgB;AAC7B,mBAAe,iBAAiB;AAAA,EAClC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB,WAAW;AAAA,EAClD;AACA,MAAI,WAAW,eAAe;AAC5B,mBAAe,gBAAgB;AAAA,EACjC;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,cAAc;AAC3B,mBAAe,eAAe,WAAW;AAAA,EAC3C;AACA,MAAI,WAAW,uBAAuB;AACpC,mBAAe,wBAAwB,WAAW;AAAA,EACpD;AACA,MAAI,WAAW,kBAAkB;AAC/B,mBAAe,mBAAmB;AAAA,EACpC;AACA,MAAI,WAAW,mBAAmB;AAChC,mBAAe,oBAAoB,WAAW;AAAA,EAChD;AACA,MAAI,WAAW,mBAAmB;AAChC,mBAAe,oBAAoB,WAAW;AAAA,EAChD;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW,WAAW;AAAA,EACvC;AACA,MAAI,WAAW,OAAO;AACpB,mBAAe,QAAQ,WAAW;AAAA,EACpC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB;AAAA,EACvC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB;AAAA,EACvC;AACA,MAAI,WAAW,QAAQ;AACrB,mBAAe,SAAS,WAAW;AAAA,EACrC;AACA,MAAI,WAAW,YAAY;AACzB,mBAAe,aAAa,WAAW;AAAA,EACzC;AACA,MAAI,WAAW,WAAW;AACxB,mBAAe,YAAY,WAAW;AAAA,EACxC;AACA,MAAI,WAAW,eAAe;AAC5B,mBAAe,gBAAgB,WAAW;AAAA,EAC5C;AACA,MAAI,WAAW,gBAAgB;AAC7B,mBAAe,iBAAiB;AAAA,EAClC;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW;AAAA,EAC5B;AACA,MAAI,WAAW,SAAS;AACtB,mBAAe,UAAU,WAAW;AAAA,EACtC;AACA,MAAI,WAAW,iBAAiB;AAC9B,mBAAe,kBAAkB,WAAW;AAAA,EAC9C;AACA,MAAI,WAAW,YAAY;AACzB,mBAAe,aAAa,WAAW;AAAA,EACzC;AAEA,SAAO;AACT;AAKO,SAAS,UAAU,SAAmC;AAC3D,QAAM,OAAiB,CAAC;AAExB,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,EAC1C;AACA,MAAI,QAAQ,QAAQ;AAClB,QAAI,MAAM,QAAQ,QAAQ,MAAM,GAAG;AACjC,UAAI,QAAQ,OAAO,SAAS,GAAG;AAC7B,aAAK,KAAK,YAAY,QAAQ,OAAO,KAAK,GAAG,CAAC;AAAA,MAChD;AAAA,IACF,OAAO;AACL,WAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,IACtC;AAAA,EACF;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,SAAS;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,MAAM,QAAQ,QAAQ,gBAAgB,GAAG;AAC3C,UAAI,QAAQ,iBAAiB,SAAS,GAAG;AACvC,aAAK,KAAK,wBAAwB,QAAQ,iBAAiB,KAAK,GAAG,CAAC;AAAA,MACtE;AAAA,IACF,OAAO;AACL,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AAAA,EACF;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,YAAY;AAAA,EACxB;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,oBAAoB;AAAA,EAChC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,EAClE;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,cAAc;AACxB,SAAK,KAAK,mBAAmB,QAAQ,YAAY;AAAA,EACnD;AACA,MAAI,QAAQ,uBAAuB;AACjC,SAAK,KAAK,6BAA6B,QAAQ,qBAAqB;AAAA,EACtE;AACA,MAAI,QAAQ,kBAAkB;AAC5B,SAAK,KAAK,sBAAsB;AAAA,EAClC;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,eAAe,QAAQ,QAAQ;AAAA,EAC3C;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,WAAW,QAAQ,KAAK;AAAA,EACpC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,yBAAyB;AAAA,EACrC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,wBAAwB;AAAA,EACpC;AACA,MAAI,QAAQ,QAAQ;AAClB,SAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,EACtC;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AACA,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,oBAAoB,QAAQ,aAAa;AAAA,EACrD;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,aAAa;AAAA,EACzB;AACA,MAAI,QAAQ,SAAS;AACnB,SAAK,KAAK,aAAa,QAAQ,OAAO;AAAA,EACxC;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,sBAAsB,QAAQ,eAAe;AAAA,EACzD;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AAEA,SAAO;AACT;;;AD/SA,IAAMA,kBAAa,0BAAc,aAAe;AAChD,IAAM,YAAiB,aAAQA,WAAU;AAEzC,IAAM,WAAW;AAWjB,SAAS,WAAW,MAAgB,mBAAwC,CAAC,GAAoB;AAC/F,QAAM,EAAE,eAAe,MAAM,IAAI;AAEjC,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,UAAe,UAAK,WAAW,MAAM,OAAO,QAAQ;AAE1D,QAAI,CAAI,cAAW,OAAO,GAAG;AAC3B,aAAO;AAAA,QACL,IAAI,MAAM,yBAAyB,OAAO,sCAAsC;AAAA,MAClF;AAAA,IACF;AAEA,UAAM,UAAU;AAIhB,UAAM,cAAc;AAAA,MAClB;AAAA,MACA;AAAA,MACA;AAAA,MACA;AAAA,MACA,GAAG;AAAA,IACL;AAEA,UAAM,kBAAc,4BAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAKb,UAAM,gBAAgB,IAAI,oCAAc,MAAM;AAC9C,UAAM,gBAAgB,IAAI,oCAAc,MAAM;AAE9C,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAIhB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B,OAAO;AACL,kBAAU;AAAA,MACZ;AAAA,IACF,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAChB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AAGA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAIhC,YAAM,aAAa,cAAc,IAAI;AACrC,YAAM,aAAa,cAAc,IAAI;AACrC,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,cAAc;AAChB,kBAAQ,OAAO,MAAM,UAAU;AAAA,QACjC,OAAO;AACL,oBAAU;AAAA,QACZ;AAAA,MACF;AACA,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,aAAc,SAAQ,OAAO,MAAM,UAAU;AACjD,kBAAU;AAAA,MACZ;AAEA,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,cAAc,UAAU;AAC9B,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,WAAW;AAAA,QACxE;AAKA,QAAC,MAA2C,aAAa;AACzD,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAe;AACtC,UAAI,IAAI,QAAQ,SAAS,QAAQ,GAAG;AAClC;AAAA,UACE,IAAI;AAAA,YACF;AAAA,UACF;AAAA,QACF;AAAA,MACF,OAAO;AACL,eAAO,GAAG;AAAA,MACZ;AAAA,IACF,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,aACP,YACA,SACkB;AAClB,QAAM,YAAY,MAAM,QAAQ,UAAU,IAAI,aAAa,CAAC,UAAU;AACtE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO,IAAI,MAAM,2CAA2C;AAAA,EAC9D;AAEA,aAAW,SAAS,WAAW;AAC7B,QAAI,CAAI,cAAW,KAAK,GAAG;AACzB,aAAO,IAAI,MAAM,mCAAmC,KAAK,EAAE;AAAA,IAC7D;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,WAAW,GAAG,UAAU,OAAO,CAAC;AAC7C;AA2BA,eAAsB,WACpB,YACA,UAA0B,CAAC,GACZ;AACf,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,UAAM;AAAA,EACR;AACA,QAAM,WAAW,aAAa,EAAE,cAAc,KAAK,CAAC;AACtD;;;AEjLO,SAAS,mBAAmB,SAAwB;AACzD,UAAQ,OAAO,4BAA4B,yEAAyE;AACpH,UAAQ,OAAO,0BAA0B,kCAAkC;AAC3E,UAAQ,OAAO,wBAAwB,2MAA2M;AAClP,UAAQ,OAAO,eAAe,iCAAiC;AAC/D,UAAQ,OAAO,gCAAgC,kGAAkG;AACjJ,UAAQ,OAAO,cAAc,mHAAmH;AAChJ,UAAQ,OAAO,sBAAsB,iDAAiD;AACtF,UAAQ,OAAO,mCAAmC,2EAA2E;AAC7H,UAAQ,OAAO,qBAAqB,yPAAyP;AAC7R,UAAQ,OAAO,0BAA0B,sGAAsG;AAC/I,UAAQ,OAAO,2BAA2B,6DAA6D;AACvG,UAAQ,OAAO,qCAAqC,+FAA+F;AACnJ,UAAQ,OAAO,wBAAwB,yHAAyH;AAChK,UAAQ,OAAO,iCAAiC,2FAA2F;AAC3I,UAAQ,OAAO,iCAAiC,2FAA2F;AAC3I,UAAQ,OAAO,0BAA0B,wHAAwH;AACjK,UAAQ,OAAO,0BAA0B,qEAAqE;AAC9G,UAAQ,OAAO,uBAAuB,4EAA4E;AAClH,UAAQ,OAAO,mBAAmB,wDAAwD;AAC1F,UAAQ,OAAO,2BAA2B,4CAA4C;AACtF,UAAQ,OAAO,0BAA0B,gHAAgH;AACzJ,UAAQ,OAAO,oBAAoB,+SAA+S;AAClV,UAAQ,OAAO,yBAAyB,sGAAsG;AAC9I,UAAQ,OAAO,wBAAwB,+CAA+C;AACtF,UAAQ,OAAO,4BAA4B,gGAAiG;AAC5I,UAAQ,OAAO,qBAAqB,qEAAqE;AACzG,UAAQ,OAAO,eAAe,6DAA6D;AAC3F,UAAQ,OAAO,qBAAqB,iTAAiT;AACrV,UAAQ,OAAO,8BAA8B,iPAAiP;AAC9R,UAAQ,OAAO,yBAAyB,kRAAkR;AAC5T;;;AHjCA,SAAS,gBAAyB;AAChC,QAAM,UAAU,IAAI,yBAAQ;AAE5B,UACG,KAAK,oBAAoB,EACzB,MAAM,sBAAsB,EAC5B,YAAY,4CAA4C,EACxD,mBAAmB,wCAAwC,EAC3D,yBAAyB,KAAK,EAC9B,SAAS,cAAc,uCAAuC;AAGjE,qBAAmB,OAAO;AAE1B,UAAQ,gBAAgB;AAAA,IACtB,UAAU,CAAC,QAAQ;AACjB,cAAQ,MAAM,IAAI,QAAQ,CAAC;AAAA,IAC7B;AAAA,IACA,aAAa,CAAC,KAAK,UAAU;AAC3B,YAAM,GAAG;AAAA,IACX;AAAA,EACF,CAAC;AAED,SAAO;AACT;AAEA,eAAe,OAAwB;AACrC,QAAM,UAAU,cAAc;AAE9B,UAAQ,aAAa;AAErB,MAAI;AACF,YAAQ,MAAM,QAAQ,IAAI;AAAA,EAC5B,SAAS,KAAK;AACZ,QAAI,eAAe,iCAAgB;AACjC,UAAI,IAAI,SAAS,2BAA2B;AAC1C,eAAO;AAAA,MACT;AACA,aAAO,IAAI,YAAY;AAAA,IACzB;AAEA,UAAM,UAAU,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC/D,YAAQ,MAAM,OAAO;AACrB,YAAQ,MAAM,wCAAwC;AACtD,WAAO;AAAA,EACT;AAEA,QAAM,aAAa,QAAQ,KAAiB;AAC5C,QAAM,aAAa,QAAQ;AAC3B,QAAM,iBAAiB,oBAAoB,UAAU;AAErD,MAAI;AAGF,UAAM,WAAW,YAAY,cAAc;AAC3C,WAAO;AAAA,EACT,SAAS,KAAK;AAMZ,UAAM,aACJ,eAAe,SAAU,IAAyC,eAAe;AACnF,QAAI,CAAC,YAAY;AACf,YAAM,UAAU,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC/D,cAAQ,MAAM,OAAO;AAAA,IACvB;AACA,WAAO;AAAA,EACT;AACF;AAEA,KAAK,EAAE,KAAK,CAAC,SAAS;AACpB,MAAI,SAAS,GAAG;AACd,YAAQ,KAAK,IAAI;AAAA,EACnB;AACF,CAAC;","names":["__filename"]}
|
|
1
|
+
{"version":3,"sources":["../node_modules/.pnpm/tsup@8.5.1_postcss@8.5.28_typescript@6.0.3/node_modules/tsup/assets/cjs_shims.js","../src/cli.ts","../src/index.ts","../src/convert-options.generated.ts","../src/cli-options.generated.ts"],"sourcesContent":["// Shim globals in cjs bundle\n// There's a weird bug that esbuild will always inject importMetaUrl\n// if we export it as `const importMetaUrl = ... __filename ...`\n// But using a function will not cause this issue\n\nconst getImportMetaUrl = () => \n typeof document === \"undefined\" \n ? new URL(`file:${__filename}`).href \n : (document.currentScript && document.currentScript.tagName.toUpperCase() === 'SCRIPT') \n ? document.currentScript.src \n : new URL(\"main.js\", document.baseURI).href;\n\nexport const importMetaUrl = /* @__PURE__ */ getImportMetaUrl()\n","#!/usr/bin/env node\nimport { Command, CommanderError } from 'commander';\nimport { _runForCli } from './index.js';\nimport { CliOptions, buildConvertOptions } from './convert-options.generated.js';\nimport { registerCliOptions } from './cli-options.generated.js';\n\nfunction createProgram(): Command {\n const program = new Command();\n\n program\n .name('opendataloader-pdf')\n .usage('[options] <input...>')\n .description('Convert PDFs using the OpenDataLoader CLI.')\n .showHelpAfterError(\"Use '--help' to see available options.\")\n .showSuggestionAfterError(false)\n .argument('<input...>', 'Input files or directories to convert');\n\n // Register CLI options from auto-generated file\n registerCliOptions(program);\n\n program.configureOutput({\n writeErr: (str) => {\n console.error(str.trimEnd());\n },\n outputError: (str, write) => {\n write(str);\n },\n });\n\n return program;\n}\n\nasync function main(): Promise<number> {\n const program = createProgram();\n\n program.exitOverride();\n\n try {\n program.parse(process.argv);\n } catch (err) {\n if (err instanceof CommanderError) {\n if (err.code === 'commander.helpDisplayed') {\n return 0;\n }\n return err.exitCode ?? 1;\n }\n\n const message = err instanceof Error ? err.message : String(err);\n console.error(message);\n console.error(\"Use '--help' to see available options.\");\n return 1;\n }\n\n const cliOptions = program.opts<CliOptions>();\n const inputPaths = program.args;\n const convertOptions = buildConvertOptions(cliOptions);\n\n try {\n // _runForCli streams stdout/stderr to the parent process as they arrive;\n // we deliberately do not re-print anything here. (Issue #398.)\n await _runForCli(inputPaths, convertOptions);\n return 0;\n } catch (err) {\n // Subprocess-exit errors are already on the user's terminal via the live\n // stderr stream — re-printing would duplicate output and risk leaking\n // anything sensitive Java logged (e.g. a --password value echoed by an\n // underlying library). Wrapper-side failures (JAR not found, java not in\n // PATH, bad input path) still need to be surfaced.\n const isJavaExit =\n err instanceof Error && (err as Error & { isJavaExit?: boolean }).isJavaExit === true;\n if (!isJavaExit) {\n const message = err instanceof Error ? err.message : String(err);\n console.error(message);\n }\n return 1;\n }\n}\n\nmain().then((code) => {\n if (code !== 0) {\n process.exit(code);\n }\n});\n","import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { StringDecoder } from 'string_decoder';\nimport { fileURLToPath } from 'url';\n\n// Re-export types and utilities from auto-generated file\nexport type { ConvertOptions } from './convert-options.generated.js';\nexport { buildArgs } from './convert-options.generated.js';\nimport type { ConvertOptions } from './convert-options.generated.js';\nimport { buildArgs } from './convert-options.generated.js';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\ninterface JarExecutionOptions {\n /**\n * When true, forwards Java's stdout and stderr chunks to the parent\n * process in real time as well as accumulating them. Used by the bundled\n * CLI so long-running conversions show progress as it happens.\n */\n streamOutput?: boolean;\n}\n\nfunction executeJar(args: string[], executionOptions: JarExecutionOptions = {}): Promise<string> {\n const { streamOutput = false } = executionOptions;\n\n return new Promise((resolve, reject) => {\n const jarPath = path.join(__dirname, '..', 'lib', JAR_NAME);\n\n if (!fs.existsSync(jarPath)) {\n return reject(\n new Error(`JAR file not found at ${jarPath}. Please run the build script first.`),\n );\n }\n\n const command = 'java';\n // Force headless AWT so macOS doesn't surface a Dock icon (and steal focus)\n // every time the JVM touches ImageIO/PDFBox rendering. Safe on all OSes —\n // the CLI never opens a UI window, only manipulates BufferedImages.\n const commandArgs = [\n '-Djava.awt.headless=true',\n '-Dapple.awt.UIElement=true',\n '-jar',\n jarPath,\n ...args,\n ];\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n // StringDecoder buffers incomplete multi-byte UTF-8 sequences across\n // chunk boundaries — Buffer.toString() alone would emit replacement\n // characters when, e.g., a 3-byte Korean codepoint splits across two\n // 'data' events. One decoder per stream so they don't share state.\n const stdoutDecoder = new StringDecoder('utf8');\n const stderrDecoder = new StringDecoder('utf8');\n\n javaProcess.stdout.on('data', (data: Buffer) => {\n const chunk = stdoutDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n // Stream-only: don't also accumulate, or a multi-hour conversion\n // would buffer its entire (potentially gigabyte) stdout in memory\n // for no consumer.\n process.stdout.write(chunk);\n } else {\n stdout += chunk;\n }\n });\n\n javaProcess.stderr.on('data', (data: Buffer) => {\n const chunk = stderrDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n process.stderr.write(chunk);\n }\n // stderr is always accumulated (progress logs are small and we need\n // them for error messages on non-zero exit).\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n // Flush any trailing bytes the decoder is still holding (always emit\n // them — if we drop them on error paths, error messages with non-ASCII\n // characters lose their tail).\n const stdoutTail = stdoutDecoder.end();\n const stderrTail = stderrDecoder.end();\n if (stdoutTail.length > 0) {\n if (streamOutput) {\n process.stdout.write(stdoutTail);\n } else {\n stdout += stdoutTail;\n }\n }\n if (stderrTail.length > 0) {\n if (streamOutput) process.stderr.write(stderrTail);\n stderr += stderrTail;\n }\n\n if (code === 0) {\n resolve(stdout);\n } else {\n const errorOutput = stderr || stdout;\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${errorOutput}`,\n );\n // Tag so the CLI can suppress re-printing this message — Java's\n // stderr was already streamed live to the parent in CLI mode, and\n // re-printing risks leaking anything sensitive Java logged\n // (e.g. a --password value echoed by an underlying library).\n (error as Error & { isJavaExit?: boolean }).isJavaExit = true;\n reject(error);\n }\n });\n\n javaProcess.on('error', (err: Error) => {\n if (err.message.includes('ENOENT')) {\n reject(\n new Error(\n \"'java' command not found. Please ensure Java is installed and in your system's PATH.\",\n ),\n );\n } else {\n reject(err);\n }\n });\n });\n}\n\nfunction buildJarArgs(\n inputPaths: string | string[],\n options: ConvertOptions,\n): string[] | Error {\n const inputList = Array.isArray(inputPaths) ? inputPaths : [inputPaths];\n if (inputList.length === 0) {\n return new Error('At least one input path must be provided.');\n }\n\n for (const input of inputList) {\n if (!fs.existsSync(input)) {\n return new Error(`Input file or folder not found: ${input}`);\n }\n }\n\n return [...inputList, ...buildArgs(options)];\n}\n\nexport function convert(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<string> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n return Promise.reject(argsOrError);\n }\n // Library API: never streams to the parent process. Returns the full stdout\n // string so callers can do `const out = await convert(...)` without surprise\n // side-effects on process.stdout / process.stderr.\n return executeJar(argsOrError, { streamOutput: false });\n}\n\n/**\n * Internal entry point used by the bundled CLI. Streams Java's stdout and\n * stderr to the parent process in real time (so long-running conversions like\n * hybrid mode show progress as it happens) and resolves without a stdout\n * payload — preventing the caller from re-printing what was already streamed.\n *\n * Not part of the public API: do not import this from application code. Use\n * {@link convert} instead.\n *\n * @internal\n */\nexport async function _runForCli(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<void> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n throw argsOrError;\n }\n await executeJar(argsOrError, { streamOutput: true });\n}\n\n/**\n * @deprecated Use `convert()` and `ConvertOptions` instead. This function will be removed in a future version.\n */\nexport interface RunOptions {\n outputFolder?: string;\n password?: string;\n replaceInvalidChars?: string;\n generateMarkdown?: boolean;\n generateHtml?: boolean;\n generateAnnotatedPdf?: boolean;\n keepLineBreaks?: boolean;\n contentSafetyOff?: string;\n htmlInMarkdown?: boolean;\n addImageToMarkdown?: boolean;\n noJson?: boolean;\n debug?: boolean;\n useStructTree?: boolean;\n}\n\n/**\n * @deprecated Use `convert()` instead. This function will be removed in a future version.\n */\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n console.warn(\n 'Warning: run() is deprecated and will be removed in a future version. Use convert() instead.',\n );\n\n // Build format array based on legacy boolean options\n const formats: string[] = [];\n if (!options.noJson) {\n formats.push('json');\n }\n if (options.generateMarkdown) {\n if (options.addImageToMarkdown) {\n formats.push('markdown-with-images');\n } else if (options.htmlInMarkdown) {\n formats.push('markdown-with-html');\n } else {\n formats.push('markdown');\n }\n }\n if (options.generateHtml) {\n formats.push('html');\n }\n if (options.generateAnnotatedPdf) {\n formats.push('pdf');\n }\n\n return convert(inputPath, {\n outputDir: options.outputFolder,\n password: options.password,\n replaceInvalidChars: options.replaceInvalidChars,\n keepLineBreaks: options.keepLineBreaks,\n contentSafetyOff: options.contentSafetyOff,\n useStructTree: options.useStructTree,\n format: formats.length > 0 ? formats : undefined,\n quiet: !options.debug,\n });\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\n/**\n * Options for the convert function.\n */\nexport interface ConvertOptions {\n /** Directory where output files are written. Default: input file directory */\n outputDir?: string;\n /** Password for encrypted PDF files */\n password?: string;\n /** Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output. */\n format?: string | string[];\n /** Suppress console logging output */\n quiet?: boolean;\n /** Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background */\n contentSafetyOff?: string | string[];\n /** Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders */\n sanitize?: boolean;\n /** Preserve original line breaks in extracted text */\n keepLineBreaks?: boolean;\n /** Replacement character for invalid/unrecognized characters. Default: space */\n replaceInvalidChars?: string;\n /** Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called */\n useStructTree?: boolean;\n /** Table detection method. Values: default (border-based), cluster (border + cluster). Default: default */\n tableMethod?: string;\n /** Reading order algorithm. Values: off, xycut. Default: xycut */\n readingOrder?: string;\n /** Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none */\n markdownPageSeparator?: string;\n /** Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown. */\n markdownWithHtml?: boolean;\n /** Separator between pages in text output. Use %page-number% for page numbers. Default: none */\n textPageSeparator?: string;\n /** Separator between pages in HTML output. Use %page-number% for page numbers. Default: none */\n htmlPageSeparator?: string;\n /** Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external */\n imageOutput?: string;\n /** Output format for extracted images. Values: png, jpeg. Default: png */\n imageFormat?: string;\n /** Directory for extracted images (applies only with --image-output external) */\n imageDir?: string;\n /** Pages to extract (e.g., \"1,3,5-7\"). Default: all pages */\n pages?: string;\n /** Include page headers and footers in output */\n includeHeaderFooter?: boolean;\n /** Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental) */\n detectStrikethrough?: boolean;\n /** Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence) */\n hybrid?: string;\n /** Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend) */\n hybridMode?: string;\n /** Hybrid backend server URL (overrides default) */\n hybridUrl?: string;\n /** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */\n hybridTimeout?: string;\n /** Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50 */\n hybridChunkSize?: string;\n /** Opt in to Java fallback on hybrid backend error (default: disabled) */\n hybridFallback?: boolean;\n /** Write output to stdout instead of file (single format only) */\n toStdout?: boolean;\n /** Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode */\n threads?: string;\n /** Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0. */\n imageResolution?: string;\n /** Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17 */\n spaceRatio?: string;\n}\n\n/**\n * Options as parsed from CLI (all values are strings from commander).\n */\nexport interface CliOptions {\n outputDir?: string;\n password?: string;\n format?: string;\n quiet?: boolean;\n contentSafetyOff?: string;\n sanitize?: boolean;\n keepLineBreaks?: boolean;\n replaceInvalidChars?: string;\n useStructTree?: boolean;\n tableMethod?: string;\n readingOrder?: string;\n markdownPageSeparator?: string;\n markdownWithHtml?: boolean;\n textPageSeparator?: string;\n htmlPageSeparator?: string;\n imageOutput?: string;\n imageFormat?: string;\n imageDir?: string;\n pages?: string;\n includeHeaderFooter?: boolean;\n detectStrikethrough?: boolean;\n hybrid?: string;\n hybridMode?: string;\n hybridUrl?: string;\n hybridTimeout?: string;\n hybridChunkSize?: string;\n hybridFallback?: boolean;\n toStdout?: boolean;\n threads?: string;\n imageResolution?: string;\n spaceRatio?: string;\n}\n\n/**\n * Convert CLI options to ConvertOptions.\n */\nexport function buildConvertOptions(cliOptions: CliOptions): ConvertOptions {\n const convertOptions: ConvertOptions = {};\n\n if (cliOptions.outputDir) {\n convertOptions.outputDir = cliOptions.outputDir;\n }\n if (cliOptions.password) {\n convertOptions.password = cliOptions.password;\n }\n if (cliOptions.format) {\n convertOptions.format = cliOptions.format;\n }\n if (cliOptions.quiet) {\n convertOptions.quiet = true;\n }\n if (cliOptions.contentSafetyOff) {\n convertOptions.contentSafetyOff = cliOptions.contentSafetyOff;\n }\n if (cliOptions.sanitize) {\n convertOptions.sanitize = true;\n }\n if (cliOptions.keepLineBreaks) {\n convertOptions.keepLineBreaks = true;\n }\n if (cliOptions.replaceInvalidChars) {\n convertOptions.replaceInvalidChars = cliOptions.replaceInvalidChars;\n }\n if (cliOptions.useStructTree) {\n convertOptions.useStructTree = true;\n }\n if (cliOptions.tableMethod) {\n convertOptions.tableMethod = cliOptions.tableMethod;\n }\n if (cliOptions.readingOrder) {\n convertOptions.readingOrder = cliOptions.readingOrder;\n }\n if (cliOptions.markdownPageSeparator) {\n convertOptions.markdownPageSeparator = cliOptions.markdownPageSeparator;\n }\n if (cliOptions.markdownWithHtml) {\n convertOptions.markdownWithHtml = true;\n }\n if (cliOptions.textPageSeparator) {\n convertOptions.textPageSeparator = cliOptions.textPageSeparator;\n }\n if (cliOptions.htmlPageSeparator) {\n convertOptions.htmlPageSeparator = cliOptions.htmlPageSeparator;\n }\n if (cliOptions.imageOutput) {\n convertOptions.imageOutput = cliOptions.imageOutput;\n }\n if (cliOptions.imageFormat) {\n convertOptions.imageFormat = cliOptions.imageFormat;\n }\n if (cliOptions.imageDir) {\n convertOptions.imageDir = cliOptions.imageDir;\n }\n if (cliOptions.pages) {\n convertOptions.pages = cliOptions.pages;\n }\n if (cliOptions.includeHeaderFooter) {\n convertOptions.includeHeaderFooter = true;\n }\n if (cliOptions.detectStrikethrough) {\n convertOptions.detectStrikethrough = true;\n }\n if (cliOptions.hybrid) {\n convertOptions.hybrid = cliOptions.hybrid;\n }\n if (cliOptions.hybridMode) {\n convertOptions.hybridMode = cliOptions.hybridMode;\n }\n if (cliOptions.hybridUrl) {\n convertOptions.hybridUrl = cliOptions.hybridUrl;\n }\n if (cliOptions.hybridTimeout) {\n convertOptions.hybridTimeout = cliOptions.hybridTimeout;\n }\n if (cliOptions.hybridChunkSize) {\n convertOptions.hybridChunkSize = cliOptions.hybridChunkSize;\n }\n if (cliOptions.hybridFallback) {\n convertOptions.hybridFallback = true;\n }\n if (cliOptions.toStdout) {\n convertOptions.toStdout = true;\n }\n if (cliOptions.threads) {\n convertOptions.threads = cliOptions.threads;\n }\n if (cliOptions.imageResolution) {\n convertOptions.imageResolution = cliOptions.imageResolution;\n }\n if (cliOptions.spaceRatio) {\n convertOptions.spaceRatio = cliOptions.spaceRatio;\n }\n\n return convertOptions;\n}\n\n/**\n * Build CLI arguments array from ConvertOptions.\n */\nexport function buildArgs(options: ConvertOptions): string[] {\n const args: string[] = [];\n\n if (options.outputDir) {\n args.push('--output-dir', options.outputDir);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.format) {\n if (Array.isArray(options.format)) {\n if (options.format.length > 0) {\n args.push('--format', options.format.join(','));\n }\n } else {\n args.push('--format', options.format);\n }\n }\n if (options.quiet) {\n args.push('--quiet');\n }\n if (options.contentSafetyOff) {\n if (Array.isArray(options.contentSafetyOff)) {\n if (options.contentSafetyOff.length > 0) {\n args.push('--content-safety-off', options.contentSafetyOff.join(','));\n }\n } else {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n }\n if (options.sanitize) {\n args.push('--sanitize');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.useStructTree) {\n args.push('--use-struct-tree');\n }\n if (options.tableMethod) {\n args.push('--table-method', options.tableMethod);\n }\n if (options.readingOrder) {\n args.push('--reading-order', options.readingOrder);\n }\n if (options.markdownPageSeparator) {\n args.push('--markdown-page-separator', options.markdownPageSeparator);\n }\n if (options.markdownWithHtml) {\n args.push('--markdown-with-html');\n }\n if (options.textPageSeparator) {\n args.push('--text-page-separator', options.textPageSeparator);\n }\n if (options.htmlPageSeparator) {\n args.push('--html-page-separator', options.htmlPageSeparator);\n }\n if (options.imageOutput) {\n args.push('--image-output', options.imageOutput);\n }\n if (options.imageFormat) {\n args.push('--image-format', options.imageFormat);\n }\n if (options.imageDir) {\n args.push('--image-dir', options.imageDir);\n }\n if (options.pages) {\n args.push('--pages', options.pages);\n }\n if (options.includeHeaderFooter) {\n args.push('--include-header-footer');\n }\n if (options.detectStrikethrough) {\n args.push('--detect-strikethrough');\n }\n if (options.hybrid) {\n args.push('--hybrid', options.hybrid);\n }\n if (options.hybridMode) {\n args.push('--hybrid-mode', options.hybridMode);\n }\n if (options.hybridUrl) {\n args.push('--hybrid-url', options.hybridUrl);\n }\n if (options.hybridTimeout) {\n args.push('--hybrid-timeout', options.hybridTimeout);\n }\n if (options.hybridChunkSize) {\n args.push('--hybrid-chunk-size', options.hybridChunkSize);\n }\n if (options.hybridFallback) {\n args.push('--hybrid-fallback');\n }\n if (options.toStdout) {\n args.push('--to-stdout');\n }\n if (options.threads) {\n args.push('--threads', options.threads);\n }\n if (options.imageResolution) {\n args.push('--image-resolution', options.imageResolution);\n }\n if (options.spaceRatio) {\n args.push('--space-ratio', options.spaceRatio);\n }\n\n return args;\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\nimport { Command } from 'commander';\n\n/**\n * Register all CLI options on the given Commander program.\n */\nexport function registerCliOptions(program: Command): void {\n program.option('-o, --output-dir <value>', 'Directory where output files are written. Default: input file directory');\n program.option('-p, --password <value>', 'Password for encrypted PDF files');\n program.option('-f, --format <value>', 'Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output.');\n program.option('-q, --quiet', 'Suppress console logging output');\n program.option('--content-safety-off <value>', 'Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background');\n program.option('--sanitize', 'Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders');\n program.option('--keep-line-breaks', 'Preserve original line breaks in extracted text');\n program.option('--replace-invalid-chars <value>', 'Replacement character for invalid/unrecognized characters. Default: space');\n program.option('--use-struct-tree', 'Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called');\n program.option('--table-method <value>', 'Table detection method. Values: default (border-based), cluster (border + cluster). Default: default');\n program.option('--reading-order <value>', 'Reading order algorithm. Values: off, xycut. Default: xycut');\n program.option('--markdown-page-separator <value>', 'Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none');\n program.option('--markdown-with-html', 'Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown.');\n program.option('--text-page-separator <value>', 'Separator between pages in text output. Use %page-number% for page numbers. Default: none');\n program.option('--html-page-separator <value>', 'Separator between pages in HTML output. Use %page-number% for page numbers. Default: none');\n program.option('--image-output <value>', 'Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external');\n program.option('--image-format <value>', 'Output format for extracted images. Values: png, jpeg. Default: png');\n program.option('--image-dir <value>', 'Directory for extracted images (applies only with --image-output external)');\n program.option('--pages <value>', 'Pages to extract (e.g., \"1,3,5-7\"). Default: all pages');\n program.option('--include-header-footer', 'Include page headers and footers in output');\n program.option('--detect-strikethrough', 'Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental)');\n program.option('--hybrid <value>', 'Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence)');\n program.option('--hybrid-mode <value>', 'Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend)');\n program.option('--hybrid-url <value>', 'Hybrid backend server URL (overrides default)');\n program.option('--hybrid-timeout <value>', 'Hybrid backend request timeout in milliseconds (0 = use the backend\\'s own default). Default: 0');\n program.option('--hybrid-chunk-size <value>', 'Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50');\n program.option('--hybrid-fallback', 'Opt in to Java fallback on hybrid backend error (default: disabled)');\n program.option('--to-stdout', 'Write output to stdout instead of file (single format only)');\n program.option('--threads <value>', 'Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode');\n program.option('--image-resolution <value>', 'Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0.');\n program.option('--space-ratio <value>', 'Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17');\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;AAKA,IAAM,mBAAmB,MACvB,OAAO,aAAa,cAChB,IAAI,IAAI,QAAQ,UAAU,EAAE,EAAE,OAC7B,SAAS,iBAAiB,SAAS,cAAc,QAAQ,YAAY,MAAM,WAC1E,SAAS,cAAc,MACvB,IAAI,IAAI,WAAW,SAAS,OAAO,EAAE;AAEtC,IAAM,gBAAgC,iCAAiB;;;ACX9D,uBAAwC;;;ACDxC,2BAAsB;AACtB,WAAsB;AACtB,SAAoB;AACpB,4BAA8B;AAC9B,iBAA8B;;;AC2GvB,SAAS,oBAAoB,YAAwC;AAC1E,QAAM,iBAAiC,CAAC;AAExC,MAAI,WAAW,WAAW;AACxB,mBAAe,YAAY,WAAW;AAAA,EACxC;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW,WAAW;AAAA,EACvC;AACA,MAAI,WAAW,QAAQ;AACrB,mBAAe,SAAS,WAAW;AAAA,EACrC;AACA,MAAI,WAAW,OAAO;AACpB,mBAAe,QAAQ;AAAA,EACzB;AACA,MAAI,WAAW,kBAAkB;AAC/B,mBAAe,mBAAmB,WAAW;AAAA,EAC/C;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW;AAAA,EAC5B;AACA,MAAI,WAAW,gBAAgB;AAC7B,mBAAe,iBAAiB;AAAA,EAClC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB,WAAW;AAAA,EAClD;AACA,MAAI,WAAW,eAAe;AAC5B,mBAAe,gBAAgB;AAAA,EACjC;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,cAAc;AAC3B,mBAAe,eAAe,WAAW;AAAA,EAC3C;AACA,MAAI,WAAW,uBAAuB;AACpC,mBAAe,wBAAwB,WAAW;AAAA,EACpD;AACA,MAAI,WAAW,kBAAkB;AAC/B,mBAAe,mBAAmB;AAAA,EACpC;AACA,MAAI,WAAW,mBAAmB;AAChC,mBAAe,oBAAoB,WAAW;AAAA,EAChD;AACA,MAAI,WAAW,mBAAmB;AAChC,mBAAe,oBAAoB,WAAW;AAAA,EAChD;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW,WAAW;AAAA,EACvC;AACA,MAAI,WAAW,OAAO;AACpB,mBAAe,QAAQ,WAAW;AAAA,EACpC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB;AAAA,EACvC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB;AAAA,EACvC;AACA,MAAI,WAAW,QAAQ;AACrB,mBAAe,SAAS,WAAW;AAAA,EACrC;AACA,MAAI,WAAW,YAAY;AACzB,mBAAe,aAAa,WAAW;AAAA,EACzC;AACA,MAAI,WAAW,WAAW;AACxB,mBAAe,YAAY,WAAW;AAAA,EACxC;AACA,MAAI,WAAW,eAAe;AAC5B,mBAAe,gBAAgB,WAAW;AAAA,EAC5C;AACA,MAAI,WAAW,iBAAiB;AAC9B,mBAAe,kBAAkB,WAAW;AAAA,EAC9C;AACA,MAAI,WAAW,gBAAgB;AAC7B,mBAAe,iBAAiB;AAAA,EAClC;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW;AAAA,EAC5B;AACA,MAAI,WAAW,SAAS;AACtB,mBAAe,UAAU,WAAW;AAAA,EACtC;AACA,MAAI,WAAW,iBAAiB;AAC9B,mBAAe,kBAAkB,WAAW;AAAA,EAC9C;AACA,MAAI,WAAW,YAAY;AACzB,mBAAe,aAAa,WAAW;AAAA,EACzC;AAEA,SAAO;AACT;AAKO,SAAS,UAAU,SAAmC;AAC3D,QAAM,OAAiB,CAAC;AAExB,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,EAC1C;AACA,MAAI,QAAQ,QAAQ;AAClB,QAAI,MAAM,QAAQ,QAAQ,MAAM,GAAG;AACjC,UAAI,QAAQ,OAAO,SAAS,GAAG;AAC7B,aAAK,KAAK,YAAY,QAAQ,OAAO,KAAK,GAAG,CAAC;AAAA,MAChD;AAAA,IACF,OAAO;AACL,WAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,IACtC;AAAA,EACF;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,SAAS;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,MAAM,QAAQ,QAAQ,gBAAgB,GAAG;AAC3C,UAAI,QAAQ,iBAAiB,SAAS,GAAG;AACvC,aAAK,KAAK,wBAAwB,QAAQ,iBAAiB,KAAK,GAAG,CAAC;AAAA,MACtE;AAAA,IACF,OAAO;AACL,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AAAA,EACF;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,YAAY;AAAA,EACxB;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,oBAAoB;AAAA,EAChC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,EAClE;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,cAAc;AACxB,SAAK,KAAK,mBAAmB,QAAQ,YAAY;AAAA,EACnD;AACA,MAAI,QAAQ,uBAAuB;AACjC,SAAK,KAAK,6BAA6B,QAAQ,qBAAqB;AAAA,EACtE;AACA,MAAI,QAAQ,kBAAkB;AAC5B,SAAK,KAAK,sBAAsB;AAAA,EAClC;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,eAAe,QAAQ,QAAQ;AAAA,EAC3C;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,WAAW,QAAQ,KAAK;AAAA,EACpC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,yBAAyB;AAAA,EACrC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,wBAAwB;AAAA,EACpC;AACA,MAAI,QAAQ,QAAQ;AAClB,SAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,EACtC;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AACA,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,oBAAoB,QAAQ,aAAa;AAAA,EACrD;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,uBAAuB,QAAQ,eAAe;AAAA,EAC1D;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,aAAa;AAAA,EACzB;AACA,MAAI,QAAQ,SAAS;AACnB,SAAK,KAAK,aAAa,QAAQ,OAAO;AAAA,EACxC;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,sBAAsB,QAAQ,eAAe;AAAA,EACzD;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AAEA,SAAO;AACT;;;ADxTA,IAAMA,kBAAa,0BAAc,aAAe;AAChD,IAAM,YAAiB,aAAQA,WAAU;AAEzC,IAAM,WAAW;AAWjB,SAAS,WAAW,MAAgB,mBAAwC,CAAC,GAAoB;AAC/F,QAAM,EAAE,eAAe,MAAM,IAAI;AAEjC,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,UAAe,UAAK,WAAW,MAAM,OAAO,QAAQ;AAE1D,QAAI,CAAI,cAAW,OAAO,GAAG;AAC3B,aAAO;AAAA,QACL,IAAI,MAAM,yBAAyB,OAAO,sCAAsC;AAAA,MAClF;AAAA,IACF;AAEA,UAAM,UAAU;AAIhB,UAAM,cAAc;AAAA,MAClB;AAAA,MACA;AAAA,MACA;AAAA,MACA;AAAA,MACA,GAAG;AAAA,IACL;AAEA,UAAM,kBAAc,4BAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAKb,UAAM,gBAAgB,IAAI,oCAAc,MAAM;AAC9C,UAAM,gBAAgB,IAAI,oCAAc,MAAM;AAE9C,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAIhB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B,OAAO;AACL,kBAAU;AAAA,MACZ;AAAA,IACF,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAChB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AAGA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAIhC,YAAM,aAAa,cAAc,IAAI;AACrC,YAAM,aAAa,cAAc,IAAI;AACrC,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,cAAc;AAChB,kBAAQ,OAAO,MAAM,UAAU;AAAA,QACjC,OAAO;AACL,oBAAU;AAAA,QACZ;AAAA,MACF;AACA,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,aAAc,SAAQ,OAAO,MAAM,UAAU;AACjD,kBAAU;AAAA,MACZ;AAEA,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,cAAc,UAAU;AAC9B,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,WAAW;AAAA,QACxE;AAKA,QAAC,MAA2C,aAAa;AACzD,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAe;AACtC,UAAI,IAAI,QAAQ,SAAS,QAAQ,GAAG;AAClC;AAAA,UACE,IAAI;AAAA,YACF;AAAA,UACF;AAAA,QACF;AAAA,MACF,OAAO;AACL,eAAO,GAAG;AAAA,MACZ;AAAA,IACF,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,aACP,YACA,SACkB;AAClB,QAAM,YAAY,MAAM,QAAQ,UAAU,IAAI,aAAa,CAAC,UAAU;AACtE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO,IAAI,MAAM,2CAA2C;AAAA,EAC9D;AAEA,aAAW,SAAS,WAAW;AAC7B,QAAI,CAAI,cAAW,KAAK,GAAG;AACzB,aAAO,IAAI,MAAM,mCAAmC,KAAK,EAAE;AAAA,IAC7D;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,WAAW,GAAG,UAAU,OAAO,CAAC;AAC7C;AA2BA,eAAsB,WACpB,YACA,UAA0B,CAAC,GACZ;AACf,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,UAAM;AAAA,EACR;AACA,QAAM,WAAW,aAAa,EAAE,cAAc,KAAK,CAAC;AACtD;;;AEjLO,SAAS,mBAAmB,SAAwB;AACzD,UAAQ,OAAO,4BAA4B,yEAAyE;AACpH,UAAQ,OAAO,0BAA0B,kCAAkC;AAC3E,UAAQ,OAAO,wBAAwB,2MAA2M;AAClP,UAAQ,OAAO,eAAe,iCAAiC;AAC/D,UAAQ,OAAO,gCAAgC,kGAAkG;AACjJ,UAAQ,OAAO,cAAc,mHAAmH;AAChJ,UAAQ,OAAO,sBAAsB,iDAAiD;AACtF,UAAQ,OAAO,mCAAmC,2EAA2E;AAC7H,UAAQ,OAAO,qBAAqB,yPAAyP;AAC7R,UAAQ,OAAO,0BAA0B,sGAAsG;AAC/I,UAAQ,OAAO,2BAA2B,6DAA6D;AACvG,UAAQ,OAAO,qCAAqC,+FAA+F;AACnJ,UAAQ,OAAO,wBAAwB,yHAAyH;AAChK,UAAQ,OAAO,iCAAiC,2FAA2F;AAC3I,UAAQ,OAAO,iCAAiC,2FAA2F;AAC3I,UAAQ,OAAO,0BAA0B,wHAAwH;AACjK,UAAQ,OAAO,0BAA0B,qEAAqE;AAC9G,UAAQ,OAAO,uBAAuB,4EAA4E;AAClH,UAAQ,OAAO,mBAAmB,wDAAwD;AAC1F,UAAQ,OAAO,2BAA2B,4CAA4C;AACtF,UAAQ,OAAO,0BAA0B,gHAAgH;AACzJ,UAAQ,OAAO,oBAAoB,+SAA+S;AAClV,UAAQ,OAAO,yBAAyB,sGAAsG;AAC9I,UAAQ,OAAO,wBAAwB,+CAA+C;AACtF,UAAQ,OAAO,4BAA4B,gGAAiG;AAC5I,UAAQ,OAAO,+BAA+B,yOAAyO;AACvR,UAAQ,OAAO,qBAAqB,qEAAqE;AACzG,UAAQ,OAAO,eAAe,6DAA6D;AAC3F,UAAQ,OAAO,qBAAqB,iTAAiT;AACrV,UAAQ,OAAO,8BAA8B,iPAAiP;AAC9R,UAAQ,OAAO,yBAAyB,kRAAkR;AAC5T;;;AHlCA,SAAS,gBAAyB;AAChC,QAAM,UAAU,IAAI,yBAAQ;AAE5B,UACG,KAAK,oBAAoB,EACzB,MAAM,sBAAsB,EAC5B,YAAY,4CAA4C,EACxD,mBAAmB,wCAAwC,EAC3D,yBAAyB,KAAK,EAC9B,SAAS,cAAc,uCAAuC;AAGjE,qBAAmB,OAAO;AAE1B,UAAQ,gBAAgB;AAAA,IACtB,UAAU,CAAC,QAAQ;AACjB,cAAQ,MAAM,IAAI,QAAQ,CAAC;AAAA,IAC7B;AAAA,IACA,aAAa,CAAC,KAAK,UAAU;AAC3B,YAAM,GAAG;AAAA,IACX;AAAA,EACF,CAAC;AAED,SAAO;AACT;AAEA,eAAe,OAAwB;AACrC,QAAM,UAAU,cAAc;AAE9B,UAAQ,aAAa;AAErB,MAAI;AACF,YAAQ,MAAM,QAAQ,IAAI;AAAA,EAC5B,SAAS,KAAK;AACZ,QAAI,eAAe,iCAAgB;AACjC,UAAI,IAAI,SAAS,2BAA2B;AAC1C,eAAO;AAAA,MACT;AACA,aAAO,IAAI,YAAY;AAAA,IACzB;AAEA,UAAM,UAAU,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC/D,YAAQ,MAAM,OAAO;AACrB,YAAQ,MAAM,wCAAwC;AACtD,WAAO;AAAA,EACT;AAEA,QAAM,aAAa,QAAQ,KAAiB;AAC5C,QAAM,aAAa,QAAQ;AAC3B,QAAM,iBAAiB,oBAAoB,UAAU;AAErD,MAAI;AAGF,UAAM,WAAW,YAAY,cAAc;AAC3C,WAAO;AAAA,EACT,SAAS,KAAK;AAMZ,UAAM,aACJ,eAAe,SAAU,IAAyC,eAAe;AACnF,QAAI,CAAC,YAAY;AACf,YAAM,UAAU,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC/D,cAAQ,MAAM,OAAO;AAAA,IACvB;AACA,WAAO;AAAA,EACT;AACF;AAEA,KAAK,EAAE,KAAK,CAAC,SAAS;AACpB,MAAI,SAAS,GAAG;AACd,YAAQ,KAAK,IAAI;AAAA,EACnB;AACF,CAAC;","names":["__filename"]}
|
package/dist/cli.js
CHANGED
|
@@ -88,6 +88,9 @@ function buildConvertOptions(cliOptions) {
|
|
|
88
88
|
if (cliOptions.hybridTimeout) {
|
|
89
89
|
convertOptions.hybridTimeout = cliOptions.hybridTimeout;
|
|
90
90
|
}
|
|
91
|
+
if (cliOptions.hybridChunkSize) {
|
|
92
|
+
convertOptions.hybridChunkSize = cliOptions.hybridChunkSize;
|
|
93
|
+
}
|
|
91
94
|
if (cliOptions.hybridFallback) {
|
|
92
95
|
convertOptions.hybridFallback = true;
|
|
93
96
|
}
|
|
@@ -194,6 +197,9 @@ function buildArgs(options) {
|
|
|
194
197
|
if (options.hybridTimeout) {
|
|
195
198
|
args.push("--hybrid-timeout", options.hybridTimeout);
|
|
196
199
|
}
|
|
200
|
+
if (options.hybridChunkSize) {
|
|
201
|
+
args.push("--hybrid-chunk-size", options.hybridChunkSize);
|
|
202
|
+
}
|
|
197
203
|
if (options.hybridFallback) {
|
|
198
204
|
args.push("--hybrid-fallback");
|
|
199
205
|
}
|
|
@@ -342,6 +348,7 @@ function registerCliOptions(program) {
|
|
|
342
348
|
program.option("--hybrid-mode <value>", "Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend)");
|
|
343
349
|
program.option("--hybrid-url <value>", "Hybrid backend server URL (overrides default)");
|
|
344
350
|
program.option("--hybrid-timeout <value>", "Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0");
|
|
351
|
+
program.option("--hybrid-chunk-size <value>", "Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50");
|
|
345
352
|
program.option("--hybrid-fallback", "Opt in to Java fallback on hybrid backend error (default: disabled)");
|
|
346
353
|
program.option("--to-stdout", "Write output to stdout instead of file (single format only)");
|
|
347
354
|
program.option("--threads <value>", "Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode");
|
package/dist/cli.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"sources":["../src/cli.ts","../src/index.ts","../src/convert-options.generated.ts","../src/cli-options.generated.ts"],"sourcesContent":["#!/usr/bin/env node\nimport { Command, CommanderError } from 'commander';\nimport { _runForCli } from './index.js';\nimport { CliOptions, buildConvertOptions } from './convert-options.generated.js';\nimport { registerCliOptions } from './cli-options.generated.js';\n\nfunction createProgram(): Command {\n const program = new Command();\n\n program\n .name('opendataloader-pdf')\n .usage('[options] <input...>')\n .description('Convert PDFs using the OpenDataLoader CLI.')\n .showHelpAfterError(\"Use '--help' to see available options.\")\n .showSuggestionAfterError(false)\n .argument('<input...>', 'Input files or directories to convert');\n\n // Register CLI options from auto-generated file\n registerCliOptions(program);\n\n program.configureOutput({\n writeErr: (str) => {\n console.error(str.trimEnd());\n },\n outputError: (str, write) => {\n write(str);\n },\n });\n\n return program;\n}\n\nasync function main(): Promise<number> {\n const program = createProgram();\n\n program.exitOverride();\n\n try {\n program.parse(process.argv);\n } catch (err) {\n if (err instanceof CommanderError) {\n if (err.code === 'commander.helpDisplayed') {\n return 0;\n }\n return err.exitCode ?? 1;\n }\n\n const message = err instanceof Error ? err.message : String(err);\n console.error(message);\n console.error(\"Use '--help' to see available options.\");\n return 1;\n }\n\n const cliOptions = program.opts<CliOptions>();\n const inputPaths = program.args;\n const convertOptions = buildConvertOptions(cliOptions);\n\n try {\n // _runForCli streams stdout/stderr to the parent process as they arrive;\n // we deliberately do not re-print anything here. (Issue #398.)\n await _runForCli(inputPaths, convertOptions);\n return 0;\n } catch (err) {\n // Subprocess-exit errors are already on the user's terminal via the live\n // stderr stream — re-printing would duplicate output and risk leaking\n // anything sensitive Java logged (e.g. a --password value echoed by an\n // underlying library). Wrapper-side failures (JAR not found, java not in\n // PATH, bad input path) still need to be surfaced.\n const isJavaExit =\n err instanceof Error && (err as Error & { isJavaExit?: boolean }).isJavaExit === true;\n if (!isJavaExit) {\n const message = err instanceof Error ? err.message : String(err);\n console.error(message);\n }\n return 1;\n }\n}\n\nmain().then((code) => {\n if (code !== 0) {\n process.exit(code);\n }\n});\n","import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { StringDecoder } from 'string_decoder';\nimport { fileURLToPath } from 'url';\n\n// Re-export types and utilities from auto-generated file\nexport type { ConvertOptions } from './convert-options.generated.js';\nexport { buildArgs } from './convert-options.generated.js';\nimport type { ConvertOptions } from './convert-options.generated.js';\nimport { buildArgs } from './convert-options.generated.js';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\ninterface JarExecutionOptions {\n /**\n * When true, forwards Java's stdout and stderr chunks to the parent\n * process in real time as well as accumulating them. Used by the bundled\n * CLI so long-running conversions show progress as it happens.\n */\n streamOutput?: boolean;\n}\n\nfunction executeJar(args: string[], executionOptions: JarExecutionOptions = {}): Promise<string> {\n const { streamOutput = false } = executionOptions;\n\n return new Promise((resolve, reject) => {\n const jarPath = path.join(__dirname, '..', 'lib', JAR_NAME);\n\n if (!fs.existsSync(jarPath)) {\n return reject(\n new Error(`JAR file not found at ${jarPath}. Please run the build script first.`),\n );\n }\n\n const command = 'java';\n // Force headless AWT so macOS doesn't surface a Dock icon (and steal focus)\n // every time the JVM touches ImageIO/PDFBox rendering. Safe on all OSes —\n // the CLI never opens a UI window, only manipulates BufferedImages.\n const commandArgs = [\n '-Djava.awt.headless=true',\n '-Dapple.awt.UIElement=true',\n '-jar',\n jarPath,\n ...args,\n ];\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n // StringDecoder buffers incomplete multi-byte UTF-8 sequences across\n // chunk boundaries — Buffer.toString() alone would emit replacement\n // characters when, e.g., a 3-byte Korean codepoint splits across two\n // 'data' events. One decoder per stream so they don't share state.\n const stdoutDecoder = new StringDecoder('utf8');\n const stderrDecoder = new StringDecoder('utf8');\n\n javaProcess.stdout.on('data', (data: Buffer) => {\n const chunk = stdoutDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n // Stream-only: don't also accumulate, or a multi-hour conversion\n // would buffer its entire (potentially gigabyte) stdout in memory\n // for no consumer.\n process.stdout.write(chunk);\n } else {\n stdout += chunk;\n }\n });\n\n javaProcess.stderr.on('data', (data: Buffer) => {\n const chunk = stderrDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n process.stderr.write(chunk);\n }\n // stderr is always accumulated (progress logs are small and we need\n // them for error messages on non-zero exit).\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n // Flush any trailing bytes the decoder is still holding (always emit\n // them — if we drop them on error paths, error messages with non-ASCII\n // characters lose their tail).\n const stdoutTail = stdoutDecoder.end();\n const stderrTail = stderrDecoder.end();\n if (stdoutTail.length > 0) {\n if (streamOutput) {\n process.stdout.write(stdoutTail);\n } else {\n stdout += stdoutTail;\n }\n }\n if (stderrTail.length > 0) {\n if (streamOutput) process.stderr.write(stderrTail);\n stderr += stderrTail;\n }\n\n if (code === 0) {\n resolve(stdout);\n } else {\n const errorOutput = stderr || stdout;\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${errorOutput}`,\n );\n // Tag so the CLI can suppress re-printing this message — Java's\n // stderr was already streamed live to the parent in CLI mode, and\n // re-printing risks leaking anything sensitive Java logged\n // (e.g. a --password value echoed by an underlying library).\n (error as Error & { isJavaExit?: boolean }).isJavaExit = true;\n reject(error);\n }\n });\n\n javaProcess.on('error', (err: Error) => {\n if (err.message.includes('ENOENT')) {\n reject(\n new Error(\n \"'java' command not found. Please ensure Java is installed and in your system's PATH.\",\n ),\n );\n } else {\n reject(err);\n }\n });\n });\n}\n\nfunction buildJarArgs(\n inputPaths: string | string[],\n options: ConvertOptions,\n): string[] | Error {\n const inputList = Array.isArray(inputPaths) ? inputPaths : [inputPaths];\n if (inputList.length === 0) {\n return new Error('At least one input path must be provided.');\n }\n\n for (const input of inputList) {\n if (!fs.existsSync(input)) {\n return new Error(`Input file or folder not found: ${input}`);\n }\n }\n\n return [...inputList, ...buildArgs(options)];\n}\n\nexport function convert(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<string> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n return Promise.reject(argsOrError);\n }\n // Library API: never streams to the parent process. Returns the full stdout\n // string so callers can do `const out = await convert(...)` without surprise\n // side-effects on process.stdout / process.stderr.\n return executeJar(argsOrError, { streamOutput: false });\n}\n\n/**\n * Internal entry point used by the bundled CLI. Streams Java's stdout and\n * stderr to the parent process in real time (so long-running conversions like\n * hybrid mode show progress as it happens) and resolves without a stdout\n * payload — preventing the caller from re-printing what was already streamed.\n *\n * Not part of the public API: do not import this from application code. Use\n * {@link convert} instead.\n *\n * @internal\n */\nexport async function _runForCli(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<void> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n throw argsOrError;\n }\n await executeJar(argsOrError, { streamOutput: true });\n}\n\n/**\n * @deprecated Use `convert()` and `ConvertOptions` instead. This function will be removed in a future version.\n */\nexport interface RunOptions {\n outputFolder?: string;\n password?: string;\n replaceInvalidChars?: string;\n generateMarkdown?: boolean;\n generateHtml?: boolean;\n generateAnnotatedPdf?: boolean;\n keepLineBreaks?: boolean;\n contentSafetyOff?: string;\n htmlInMarkdown?: boolean;\n addImageToMarkdown?: boolean;\n noJson?: boolean;\n debug?: boolean;\n useStructTree?: boolean;\n}\n\n/**\n * @deprecated Use `convert()` instead. This function will be removed in a future version.\n */\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n console.warn(\n 'Warning: run() is deprecated and will be removed in a future version. Use convert() instead.',\n );\n\n // Build format array based on legacy boolean options\n const formats: string[] = [];\n if (!options.noJson) {\n formats.push('json');\n }\n if (options.generateMarkdown) {\n if (options.addImageToMarkdown) {\n formats.push('markdown-with-images');\n } else if (options.htmlInMarkdown) {\n formats.push('markdown-with-html');\n } else {\n formats.push('markdown');\n }\n }\n if (options.generateHtml) {\n formats.push('html');\n }\n if (options.generateAnnotatedPdf) {\n formats.push('pdf');\n }\n\n return convert(inputPath, {\n outputDir: options.outputFolder,\n password: options.password,\n replaceInvalidChars: options.replaceInvalidChars,\n keepLineBreaks: options.keepLineBreaks,\n contentSafetyOff: options.contentSafetyOff,\n useStructTree: options.useStructTree,\n format: formats.length > 0 ? formats : undefined,\n quiet: !options.debug,\n });\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\n/**\n * Options for the convert function.\n */\nexport interface ConvertOptions {\n /** Directory where output files are written. Default: input file directory */\n outputDir?: string;\n /** Password for encrypted PDF files */\n password?: string;\n /** Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output. */\n format?: string | string[];\n /** Suppress console logging output */\n quiet?: boolean;\n /** Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background */\n contentSafetyOff?: string | string[];\n /** Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders */\n sanitize?: boolean;\n /** Preserve original line breaks in extracted text */\n keepLineBreaks?: boolean;\n /** Replacement character for invalid/unrecognized characters. Default: space */\n replaceInvalidChars?: string;\n /** Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called */\n useStructTree?: boolean;\n /** Table detection method. Values: default (border-based), cluster (border + cluster). Default: default */\n tableMethod?: string;\n /** Reading order algorithm. Values: off, xycut. Default: xycut */\n readingOrder?: string;\n /** Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none */\n markdownPageSeparator?: string;\n /** Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown. */\n markdownWithHtml?: boolean;\n /** Separator between pages in text output. Use %page-number% for page numbers. Default: none */\n textPageSeparator?: string;\n /** Separator between pages in HTML output. Use %page-number% for page numbers. Default: none */\n htmlPageSeparator?: string;\n /** Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external */\n imageOutput?: string;\n /** Output format for extracted images. Values: png, jpeg. Default: png */\n imageFormat?: string;\n /** Directory for extracted images (applies only with --image-output external) */\n imageDir?: string;\n /** Pages to extract (e.g., \"1,3,5-7\"). Default: all pages */\n pages?: string;\n /** Include page headers and footers in output */\n includeHeaderFooter?: boolean;\n /** Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental) */\n detectStrikethrough?: boolean;\n /** Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence) */\n hybrid?: string;\n /** Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend) */\n hybridMode?: string;\n /** Hybrid backend server URL (overrides default) */\n hybridUrl?: string;\n /** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */\n hybridTimeout?: string;\n /** Opt in to Java fallback on hybrid backend error (default: disabled) */\n hybridFallback?: boolean;\n /** Write output to stdout instead of file (single format only) */\n toStdout?: boolean;\n /** Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode */\n threads?: string;\n /** Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0. */\n imageResolution?: string;\n /** Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17 */\n spaceRatio?: string;\n}\n\n/**\n * Options as parsed from CLI (all values are strings from commander).\n */\nexport interface CliOptions {\n outputDir?: string;\n password?: string;\n format?: string;\n quiet?: boolean;\n contentSafetyOff?: string;\n sanitize?: boolean;\n keepLineBreaks?: boolean;\n replaceInvalidChars?: string;\n useStructTree?: boolean;\n tableMethod?: string;\n readingOrder?: string;\n markdownPageSeparator?: string;\n markdownWithHtml?: boolean;\n textPageSeparator?: string;\n htmlPageSeparator?: string;\n imageOutput?: string;\n imageFormat?: string;\n imageDir?: string;\n pages?: string;\n includeHeaderFooter?: boolean;\n detectStrikethrough?: boolean;\n hybrid?: string;\n hybridMode?: string;\n hybridUrl?: string;\n hybridTimeout?: string;\n hybridFallback?: boolean;\n toStdout?: boolean;\n threads?: string;\n imageResolution?: string;\n spaceRatio?: string;\n}\n\n/**\n * Convert CLI options to ConvertOptions.\n */\nexport function buildConvertOptions(cliOptions: CliOptions): ConvertOptions {\n const convertOptions: ConvertOptions = {};\n\n if (cliOptions.outputDir) {\n convertOptions.outputDir = cliOptions.outputDir;\n }\n if (cliOptions.password) {\n convertOptions.password = cliOptions.password;\n }\n if (cliOptions.format) {\n convertOptions.format = cliOptions.format;\n }\n if (cliOptions.quiet) {\n convertOptions.quiet = true;\n }\n if (cliOptions.contentSafetyOff) {\n convertOptions.contentSafetyOff = cliOptions.contentSafetyOff;\n }\n if (cliOptions.sanitize) {\n convertOptions.sanitize = true;\n }\n if (cliOptions.keepLineBreaks) {\n convertOptions.keepLineBreaks = true;\n }\n if (cliOptions.replaceInvalidChars) {\n convertOptions.replaceInvalidChars = cliOptions.replaceInvalidChars;\n }\n if (cliOptions.useStructTree) {\n convertOptions.useStructTree = true;\n }\n if (cliOptions.tableMethod) {\n convertOptions.tableMethod = cliOptions.tableMethod;\n }\n if (cliOptions.readingOrder) {\n convertOptions.readingOrder = cliOptions.readingOrder;\n }\n if (cliOptions.markdownPageSeparator) {\n convertOptions.markdownPageSeparator = cliOptions.markdownPageSeparator;\n }\n if (cliOptions.markdownWithHtml) {\n convertOptions.markdownWithHtml = true;\n }\n if (cliOptions.textPageSeparator) {\n convertOptions.textPageSeparator = cliOptions.textPageSeparator;\n }\n if (cliOptions.htmlPageSeparator) {\n convertOptions.htmlPageSeparator = cliOptions.htmlPageSeparator;\n }\n if (cliOptions.imageOutput) {\n convertOptions.imageOutput = cliOptions.imageOutput;\n }\n if (cliOptions.imageFormat) {\n convertOptions.imageFormat = cliOptions.imageFormat;\n }\n if (cliOptions.imageDir) {\n convertOptions.imageDir = cliOptions.imageDir;\n }\n if (cliOptions.pages) {\n convertOptions.pages = cliOptions.pages;\n }\n if (cliOptions.includeHeaderFooter) {\n convertOptions.includeHeaderFooter = true;\n }\n if (cliOptions.detectStrikethrough) {\n convertOptions.detectStrikethrough = true;\n }\n if (cliOptions.hybrid) {\n convertOptions.hybrid = cliOptions.hybrid;\n }\n if (cliOptions.hybridMode) {\n convertOptions.hybridMode = cliOptions.hybridMode;\n }\n if (cliOptions.hybridUrl) {\n convertOptions.hybridUrl = cliOptions.hybridUrl;\n }\n if (cliOptions.hybridTimeout) {\n convertOptions.hybridTimeout = cliOptions.hybridTimeout;\n }\n if (cliOptions.hybridFallback) {\n convertOptions.hybridFallback = true;\n }\n if (cliOptions.toStdout) {\n convertOptions.toStdout = true;\n }\n if (cliOptions.threads) {\n convertOptions.threads = cliOptions.threads;\n }\n if (cliOptions.imageResolution) {\n convertOptions.imageResolution = cliOptions.imageResolution;\n }\n if (cliOptions.spaceRatio) {\n convertOptions.spaceRatio = cliOptions.spaceRatio;\n }\n\n return convertOptions;\n}\n\n/**\n * Build CLI arguments array from ConvertOptions.\n */\nexport function buildArgs(options: ConvertOptions): string[] {\n const args: string[] = [];\n\n if (options.outputDir) {\n args.push('--output-dir', options.outputDir);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.format) {\n if (Array.isArray(options.format)) {\n if (options.format.length > 0) {\n args.push('--format', options.format.join(','));\n }\n } else {\n args.push('--format', options.format);\n }\n }\n if (options.quiet) {\n args.push('--quiet');\n }\n if (options.contentSafetyOff) {\n if (Array.isArray(options.contentSafetyOff)) {\n if (options.contentSafetyOff.length > 0) {\n args.push('--content-safety-off', options.contentSafetyOff.join(','));\n }\n } else {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n }\n if (options.sanitize) {\n args.push('--sanitize');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.useStructTree) {\n args.push('--use-struct-tree');\n }\n if (options.tableMethod) {\n args.push('--table-method', options.tableMethod);\n }\n if (options.readingOrder) {\n args.push('--reading-order', options.readingOrder);\n }\n if (options.markdownPageSeparator) {\n args.push('--markdown-page-separator', options.markdownPageSeparator);\n }\n if (options.markdownWithHtml) {\n args.push('--markdown-with-html');\n }\n if (options.textPageSeparator) {\n args.push('--text-page-separator', options.textPageSeparator);\n }\n if (options.htmlPageSeparator) {\n args.push('--html-page-separator', options.htmlPageSeparator);\n }\n if (options.imageOutput) {\n args.push('--image-output', options.imageOutput);\n }\n if (options.imageFormat) {\n args.push('--image-format', options.imageFormat);\n }\n if (options.imageDir) {\n args.push('--image-dir', options.imageDir);\n }\n if (options.pages) {\n args.push('--pages', options.pages);\n }\n if (options.includeHeaderFooter) {\n args.push('--include-header-footer');\n }\n if (options.detectStrikethrough) {\n args.push('--detect-strikethrough');\n }\n if (options.hybrid) {\n args.push('--hybrid', options.hybrid);\n }\n if (options.hybridMode) {\n args.push('--hybrid-mode', options.hybridMode);\n }\n if (options.hybridUrl) {\n args.push('--hybrid-url', options.hybridUrl);\n }\n if (options.hybridTimeout) {\n args.push('--hybrid-timeout', options.hybridTimeout);\n }\n if (options.hybridFallback) {\n args.push('--hybrid-fallback');\n }\n if (options.toStdout) {\n args.push('--to-stdout');\n }\n if (options.threads) {\n args.push('--threads', options.threads);\n }\n if (options.imageResolution) {\n args.push('--image-resolution', options.imageResolution);\n }\n if (options.spaceRatio) {\n args.push('--space-ratio', options.spaceRatio);\n }\n\n return args;\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\nimport { Command } from 'commander';\n\n/**\n * Register all CLI options on the given Commander program.\n */\nexport function registerCliOptions(program: Command): void {\n program.option('-o, --output-dir <value>', 'Directory where output files are written. Default: input file directory');\n program.option('-p, --password <value>', 'Password for encrypted PDF files');\n program.option('-f, --format <value>', 'Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output.');\n program.option('-q, --quiet', 'Suppress console logging output');\n program.option('--content-safety-off <value>', 'Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background');\n program.option('--sanitize', 'Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders');\n program.option('--keep-line-breaks', 'Preserve original line breaks in extracted text');\n program.option('--replace-invalid-chars <value>', 'Replacement character for invalid/unrecognized characters. Default: space');\n program.option('--use-struct-tree', 'Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called');\n program.option('--table-method <value>', 'Table detection method. Values: default (border-based), cluster (border + cluster). Default: default');\n program.option('--reading-order <value>', 'Reading order algorithm. Values: off, xycut. Default: xycut');\n program.option('--markdown-page-separator <value>', 'Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none');\n program.option('--markdown-with-html', 'Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown.');\n program.option('--text-page-separator <value>', 'Separator between pages in text output. Use %page-number% for page numbers. Default: none');\n program.option('--html-page-separator <value>', 'Separator between pages in HTML output. Use %page-number% for page numbers. Default: none');\n program.option('--image-output <value>', 'Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external');\n program.option('--image-format <value>', 'Output format for extracted images. Values: png, jpeg. Default: png');\n program.option('--image-dir <value>', 'Directory for extracted images (applies only with --image-output external)');\n program.option('--pages <value>', 'Pages to extract (e.g., \"1,3,5-7\"). Default: all pages');\n program.option('--include-header-footer', 'Include page headers and footers in output');\n program.option('--detect-strikethrough', 'Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental)');\n program.option('--hybrid <value>', 'Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence)');\n program.option('--hybrid-mode <value>', 'Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend)');\n program.option('--hybrid-url <value>', 'Hybrid backend server URL (overrides default)');\n program.option('--hybrid-timeout <value>', 'Hybrid backend request timeout in milliseconds (0 = use the backend\\'s own default). Default: 0');\n program.option('--hybrid-fallback', 'Opt in to Java fallback on hybrid backend error (default: disabled)');\n program.option('--to-stdout', 'Write output to stdout instead of file (single format only)');\n program.option('--threads <value>', 'Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode');\n program.option('--image-resolution <value>', 'Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0.');\n program.option('--space-ratio <value>', 'Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17');\n}\n"],"mappings":";;;AACA,SAAS,SAAS,sBAAsB;;;ACDxC,SAAS,aAAa;AACtB,YAAY,UAAU;AACtB,YAAY,QAAQ;AACpB,SAAS,qBAAqB;AAC9B,SAAS,qBAAqB;;;ACwGvB,SAAS,oBAAoB,YAAwC;AAC1E,QAAM,iBAAiC,CAAC;AAExC,MAAI,WAAW,WAAW;AACxB,mBAAe,YAAY,WAAW;AAAA,EACxC;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW,WAAW;AAAA,EACvC;AACA,MAAI,WAAW,QAAQ;AACrB,mBAAe,SAAS,WAAW;AAAA,EACrC;AACA,MAAI,WAAW,OAAO;AACpB,mBAAe,QAAQ;AAAA,EACzB;AACA,MAAI,WAAW,kBAAkB;AAC/B,mBAAe,mBAAmB,WAAW;AAAA,EAC/C;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW;AAAA,EAC5B;AACA,MAAI,WAAW,gBAAgB;AAC7B,mBAAe,iBAAiB;AAAA,EAClC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB,WAAW;AAAA,EAClD;AACA,MAAI,WAAW,eAAe;AAC5B,mBAAe,gBAAgB;AAAA,EACjC;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,cAAc;AAC3B,mBAAe,eAAe,WAAW;AAAA,EAC3C;AACA,MAAI,WAAW,uBAAuB;AACpC,mBAAe,wBAAwB,WAAW;AAAA,EACpD;AACA,MAAI,WAAW,kBAAkB;AAC/B,mBAAe,mBAAmB;AAAA,EACpC;AACA,MAAI,WAAW,mBAAmB;AAChC,mBAAe,oBAAoB,WAAW;AAAA,EAChD;AACA,MAAI,WAAW,mBAAmB;AAChC,mBAAe,oBAAoB,WAAW;AAAA,EAChD;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW,WAAW;AAAA,EACvC;AACA,MAAI,WAAW,OAAO;AACpB,mBAAe,QAAQ,WAAW;AAAA,EACpC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB;AAAA,EACvC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB;AAAA,EACvC;AACA,MAAI,WAAW,QAAQ;AACrB,mBAAe,SAAS,WAAW;AAAA,EACrC;AACA,MAAI,WAAW,YAAY;AACzB,mBAAe,aAAa,WAAW;AAAA,EACzC;AACA,MAAI,WAAW,WAAW;AACxB,mBAAe,YAAY,WAAW;AAAA,EACxC;AACA,MAAI,WAAW,eAAe;AAC5B,mBAAe,gBAAgB,WAAW;AAAA,EAC5C;AACA,MAAI,WAAW,gBAAgB;AAC7B,mBAAe,iBAAiB;AAAA,EAClC;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW;AAAA,EAC5B;AACA,MAAI,WAAW,SAAS;AACtB,mBAAe,UAAU,WAAW;AAAA,EACtC;AACA,MAAI,WAAW,iBAAiB;AAC9B,mBAAe,kBAAkB,WAAW;AAAA,EAC9C;AACA,MAAI,WAAW,YAAY;AACzB,mBAAe,aAAa,WAAW;AAAA,EACzC;AAEA,SAAO;AACT;AAKO,SAAS,UAAU,SAAmC;AAC3D,QAAM,OAAiB,CAAC;AAExB,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,EAC1C;AACA,MAAI,QAAQ,QAAQ;AAClB,QAAI,MAAM,QAAQ,QAAQ,MAAM,GAAG;AACjC,UAAI,QAAQ,OAAO,SAAS,GAAG;AAC7B,aAAK,KAAK,YAAY,QAAQ,OAAO,KAAK,GAAG,CAAC;AAAA,MAChD;AAAA,IACF,OAAO;AACL,WAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,IACtC;AAAA,EACF;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,SAAS;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,MAAM,QAAQ,QAAQ,gBAAgB,GAAG;AAC3C,UAAI,QAAQ,iBAAiB,SAAS,GAAG;AACvC,aAAK,KAAK,wBAAwB,QAAQ,iBAAiB,KAAK,GAAG,CAAC;AAAA,MACtE;AAAA,IACF,OAAO;AACL,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AAAA,EACF;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,YAAY;AAAA,EACxB;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,oBAAoB;AAAA,EAChC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,EAClE;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,cAAc;AACxB,SAAK,KAAK,mBAAmB,QAAQ,YAAY;AAAA,EACnD;AACA,MAAI,QAAQ,uBAAuB;AACjC,SAAK,KAAK,6BAA6B,QAAQ,qBAAqB;AAAA,EACtE;AACA,MAAI,QAAQ,kBAAkB;AAC5B,SAAK,KAAK,sBAAsB;AAAA,EAClC;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,eAAe,QAAQ,QAAQ;AAAA,EAC3C;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,WAAW,QAAQ,KAAK;AAAA,EACpC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,yBAAyB;AAAA,EACrC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,wBAAwB;AAAA,EACpC;AACA,MAAI,QAAQ,QAAQ;AAClB,SAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,EACtC;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AACA,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,oBAAoB,QAAQ,aAAa;AAAA,EACrD;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,aAAa;AAAA,EACzB;AACA,MAAI,QAAQ,SAAS;AACnB,SAAK,KAAK,aAAa,QAAQ,OAAO;AAAA,EACxC;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,sBAAsB,QAAQ,eAAe;AAAA,EACzD;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AAEA,SAAO;AACT;;;AD/SA,IAAMA,cAAa,cAAc,YAAY,GAAG;AAChD,IAAMC,aAAiB,aAAQD,WAAU;AAEzC,IAAM,WAAW;AAWjB,SAAS,WAAW,MAAgB,mBAAwC,CAAC,GAAoB;AAC/F,QAAM,EAAE,eAAe,MAAM,IAAI;AAEjC,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,UAAe,UAAKC,YAAW,MAAM,OAAO,QAAQ;AAE1D,QAAI,CAAI,cAAW,OAAO,GAAG;AAC3B,aAAO;AAAA,QACL,IAAI,MAAM,yBAAyB,OAAO,sCAAsC;AAAA,MAClF;AAAA,IACF;AAEA,UAAM,UAAU;AAIhB,UAAM,cAAc;AAAA,MAClB;AAAA,MACA;AAAA,MACA;AAAA,MACA;AAAA,MACA,GAAG;AAAA,IACL;AAEA,UAAM,cAAc,MAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAKb,UAAM,gBAAgB,IAAI,cAAc,MAAM;AAC9C,UAAM,gBAAgB,IAAI,cAAc,MAAM;AAE9C,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAIhB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B,OAAO;AACL,kBAAU;AAAA,MACZ;AAAA,IACF,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAChB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AAGA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAIhC,YAAM,aAAa,cAAc,IAAI;AACrC,YAAM,aAAa,cAAc,IAAI;AACrC,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,cAAc;AAChB,kBAAQ,OAAO,MAAM,UAAU;AAAA,QACjC,OAAO;AACL,oBAAU;AAAA,QACZ;AAAA,MACF;AACA,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,aAAc,SAAQ,OAAO,MAAM,UAAU;AACjD,kBAAU;AAAA,MACZ;AAEA,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,cAAc,UAAU;AAC9B,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,WAAW;AAAA,QACxE;AAKA,QAAC,MAA2C,aAAa;AACzD,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAe;AACtC,UAAI,IAAI,QAAQ,SAAS,QAAQ,GAAG;AAClC;AAAA,UACE,IAAI;AAAA,YACF;AAAA,UACF;AAAA,QACF;AAAA,MACF,OAAO;AACL,eAAO,GAAG;AAAA,MACZ;AAAA,IACF,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,aACP,YACA,SACkB;AAClB,QAAM,YAAY,MAAM,QAAQ,UAAU,IAAI,aAAa,CAAC,UAAU;AACtE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO,IAAI,MAAM,2CAA2C;AAAA,EAC9D;AAEA,aAAW,SAAS,WAAW;AAC7B,QAAI,CAAI,cAAW,KAAK,GAAG;AACzB,aAAO,IAAI,MAAM,mCAAmC,KAAK,EAAE;AAAA,IAC7D;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,WAAW,GAAG,UAAU,OAAO,CAAC;AAC7C;AA2BA,eAAsB,WACpB,YACA,UAA0B,CAAC,GACZ;AACf,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,UAAM;AAAA,EACR;AACA,QAAM,WAAW,aAAa,EAAE,cAAc,KAAK,CAAC;AACtD;;;AEjLO,SAAS,mBAAmB,SAAwB;AACzD,UAAQ,OAAO,4BAA4B,yEAAyE;AACpH,UAAQ,OAAO,0BAA0B,kCAAkC;AAC3E,UAAQ,OAAO,wBAAwB,2MAA2M;AAClP,UAAQ,OAAO,eAAe,iCAAiC;AAC/D,UAAQ,OAAO,gCAAgC,kGAAkG;AACjJ,UAAQ,OAAO,cAAc,mHAAmH;AAChJ,UAAQ,OAAO,sBAAsB,iDAAiD;AACtF,UAAQ,OAAO,mCAAmC,2EAA2E;AAC7H,UAAQ,OAAO,qBAAqB,yPAAyP;AAC7R,UAAQ,OAAO,0BAA0B,sGAAsG;AAC/I,UAAQ,OAAO,2BAA2B,6DAA6D;AACvG,UAAQ,OAAO,qCAAqC,+FAA+F;AACnJ,UAAQ,OAAO,wBAAwB,yHAAyH;AAChK,UAAQ,OAAO,iCAAiC,2FAA2F;AAC3I,UAAQ,OAAO,iCAAiC,2FAA2F;AAC3I,UAAQ,OAAO,0BAA0B,wHAAwH;AACjK,UAAQ,OAAO,0BAA0B,qEAAqE;AAC9G,UAAQ,OAAO,uBAAuB,4EAA4E;AAClH,UAAQ,OAAO,mBAAmB,wDAAwD;AAC1F,UAAQ,OAAO,2BAA2B,4CAA4C;AACtF,UAAQ,OAAO,0BAA0B,gHAAgH;AACzJ,UAAQ,OAAO,oBAAoB,+SAA+S;AAClV,UAAQ,OAAO,yBAAyB,sGAAsG;AAC9I,UAAQ,OAAO,wBAAwB,+CAA+C;AACtF,UAAQ,OAAO,4BAA4B,gGAAiG;AAC5I,UAAQ,OAAO,qBAAqB,qEAAqE;AACzG,UAAQ,OAAO,eAAe,6DAA6D;AAC3F,UAAQ,OAAO,qBAAqB,iTAAiT;AACrV,UAAQ,OAAO,8BAA8B,iPAAiP;AAC9R,UAAQ,OAAO,yBAAyB,kRAAkR;AAC5T;;;AHjCA,SAAS,gBAAyB;AAChC,QAAM,UAAU,IAAI,QAAQ;AAE5B,UACG,KAAK,oBAAoB,EACzB,MAAM,sBAAsB,EAC5B,YAAY,4CAA4C,EACxD,mBAAmB,wCAAwC,EAC3D,yBAAyB,KAAK,EAC9B,SAAS,cAAc,uCAAuC;AAGjE,qBAAmB,OAAO;AAE1B,UAAQ,gBAAgB;AAAA,IACtB,UAAU,CAAC,QAAQ;AACjB,cAAQ,MAAM,IAAI,QAAQ,CAAC;AAAA,IAC7B;AAAA,IACA,aAAa,CAAC,KAAK,UAAU;AAC3B,YAAM,GAAG;AAAA,IACX;AAAA,EACF,CAAC;AAED,SAAO;AACT;AAEA,eAAe,OAAwB;AACrC,QAAM,UAAU,cAAc;AAE9B,UAAQ,aAAa;AAErB,MAAI;AACF,YAAQ,MAAM,QAAQ,IAAI;AAAA,EAC5B,SAAS,KAAK;AACZ,QAAI,eAAe,gBAAgB;AACjC,UAAI,IAAI,SAAS,2BAA2B;AAC1C,eAAO;AAAA,MACT;AACA,aAAO,IAAI,YAAY;AAAA,IACzB;AAEA,UAAM,UAAU,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC/D,YAAQ,MAAM,OAAO;AACrB,YAAQ,MAAM,wCAAwC;AACtD,WAAO;AAAA,EACT;AAEA,QAAM,aAAa,QAAQ,KAAiB;AAC5C,QAAM,aAAa,QAAQ;AAC3B,QAAM,iBAAiB,oBAAoB,UAAU;AAErD,MAAI;AAGF,UAAM,WAAW,YAAY,cAAc;AAC3C,WAAO;AAAA,EACT,SAAS,KAAK;AAMZ,UAAM,aACJ,eAAe,SAAU,IAAyC,eAAe;AACnF,QAAI,CAAC,YAAY;AACf,YAAM,UAAU,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC/D,cAAQ,MAAM,OAAO;AAAA,IACvB;AACA,WAAO;AAAA,EACT;AACF;AAEA,KAAK,EAAE,KAAK,CAAC,SAAS;AACpB,MAAI,SAAS,GAAG;AACd,YAAQ,KAAK,IAAI;AAAA,EACnB;AACF,CAAC;","names":["__filename","__dirname"]}
|
|
1
|
+
{"version":3,"sources":["../src/cli.ts","../src/index.ts","../src/convert-options.generated.ts","../src/cli-options.generated.ts"],"sourcesContent":["#!/usr/bin/env node\nimport { Command, CommanderError } from 'commander';\nimport { _runForCli } from './index.js';\nimport { CliOptions, buildConvertOptions } from './convert-options.generated.js';\nimport { registerCliOptions } from './cli-options.generated.js';\n\nfunction createProgram(): Command {\n const program = new Command();\n\n program\n .name('opendataloader-pdf')\n .usage('[options] <input...>')\n .description('Convert PDFs using the OpenDataLoader CLI.')\n .showHelpAfterError(\"Use '--help' to see available options.\")\n .showSuggestionAfterError(false)\n .argument('<input...>', 'Input files or directories to convert');\n\n // Register CLI options from auto-generated file\n registerCliOptions(program);\n\n program.configureOutput({\n writeErr: (str) => {\n console.error(str.trimEnd());\n },\n outputError: (str, write) => {\n write(str);\n },\n });\n\n return program;\n}\n\nasync function main(): Promise<number> {\n const program = createProgram();\n\n program.exitOverride();\n\n try {\n program.parse(process.argv);\n } catch (err) {\n if (err instanceof CommanderError) {\n if (err.code === 'commander.helpDisplayed') {\n return 0;\n }\n return err.exitCode ?? 1;\n }\n\n const message = err instanceof Error ? err.message : String(err);\n console.error(message);\n console.error(\"Use '--help' to see available options.\");\n return 1;\n }\n\n const cliOptions = program.opts<CliOptions>();\n const inputPaths = program.args;\n const convertOptions = buildConvertOptions(cliOptions);\n\n try {\n // _runForCli streams stdout/stderr to the parent process as they arrive;\n // we deliberately do not re-print anything here. (Issue #398.)\n await _runForCli(inputPaths, convertOptions);\n return 0;\n } catch (err) {\n // Subprocess-exit errors are already on the user's terminal via the live\n // stderr stream — re-printing would duplicate output and risk leaking\n // anything sensitive Java logged (e.g. a --password value echoed by an\n // underlying library). Wrapper-side failures (JAR not found, java not in\n // PATH, bad input path) still need to be surfaced.\n const isJavaExit =\n err instanceof Error && (err as Error & { isJavaExit?: boolean }).isJavaExit === true;\n if (!isJavaExit) {\n const message = err instanceof Error ? err.message : String(err);\n console.error(message);\n }\n return 1;\n }\n}\n\nmain().then((code) => {\n if (code !== 0) {\n process.exit(code);\n }\n});\n","import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { StringDecoder } from 'string_decoder';\nimport { fileURLToPath } from 'url';\n\n// Re-export types and utilities from auto-generated file\nexport type { ConvertOptions } from './convert-options.generated.js';\nexport { buildArgs } from './convert-options.generated.js';\nimport type { ConvertOptions } from './convert-options.generated.js';\nimport { buildArgs } from './convert-options.generated.js';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\ninterface JarExecutionOptions {\n /**\n * When true, forwards Java's stdout and stderr chunks to the parent\n * process in real time as well as accumulating them. Used by the bundled\n * CLI so long-running conversions show progress as it happens.\n */\n streamOutput?: boolean;\n}\n\nfunction executeJar(args: string[], executionOptions: JarExecutionOptions = {}): Promise<string> {\n const { streamOutput = false } = executionOptions;\n\n return new Promise((resolve, reject) => {\n const jarPath = path.join(__dirname, '..', 'lib', JAR_NAME);\n\n if (!fs.existsSync(jarPath)) {\n return reject(\n new Error(`JAR file not found at ${jarPath}. Please run the build script first.`),\n );\n }\n\n const command = 'java';\n // Force headless AWT so macOS doesn't surface a Dock icon (and steal focus)\n // every time the JVM touches ImageIO/PDFBox rendering. Safe on all OSes —\n // the CLI never opens a UI window, only manipulates BufferedImages.\n const commandArgs = [\n '-Djava.awt.headless=true',\n '-Dapple.awt.UIElement=true',\n '-jar',\n jarPath,\n ...args,\n ];\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n // StringDecoder buffers incomplete multi-byte UTF-8 sequences across\n // chunk boundaries — Buffer.toString() alone would emit replacement\n // characters when, e.g., a 3-byte Korean codepoint splits across two\n // 'data' events. One decoder per stream so they don't share state.\n const stdoutDecoder = new StringDecoder('utf8');\n const stderrDecoder = new StringDecoder('utf8');\n\n javaProcess.stdout.on('data', (data: Buffer) => {\n const chunk = stdoutDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n // Stream-only: don't also accumulate, or a multi-hour conversion\n // would buffer its entire (potentially gigabyte) stdout in memory\n // for no consumer.\n process.stdout.write(chunk);\n } else {\n stdout += chunk;\n }\n });\n\n javaProcess.stderr.on('data', (data: Buffer) => {\n const chunk = stderrDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n process.stderr.write(chunk);\n }\n // stderr is always accumulated (progress logs are small and we need\n // them for error messages on non-zero exit).\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n // Flush any trailing bytes the decoder is still holding (always emit\n // them — if we drop them on error paths, error messages with non-ASCII\n // characters lose their tail).\n const stdoutTail = stdoutDecoder.end();\n const stderrTail = stderrDecoder.end();\n if (stdoutTail.length > 0) {\n if (streamOutput) {\n process.stdout.write(stdoutTail);\n } else {\n stdout += stdoutTail;\n }\n }\n if (stderrTail.length > 0) {\n if (streamOutput) process.stderr.write(stderrTail);\n stderr += stderrTail;\n }\n\n if (code === 0) {\n resolve(stdout);\n } else {\n const errorOutput = stderr || stdout;\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${errorOutput}`,\n );\n // Tag so the CLI can suppress re-printing this message — Java's\n // stderr was already streamed live to the parent in CLI mode, and\n // re-printing risks leaking anything sensitive Java logged\n // (e.g. a --password value echoed by an underlying library).\n (error as Error & { isJavaExit?: boolean }).isJavaExit = true;\n reject(error);\n }\n });\n\n javaProcess.on('error', (err: Error) => {\n if (err.message.includes('ENOENT')) {\n reject(\n new Error(\n \"'java' command not found. Please ensure Java is installed and in your system's PATH.\",\n ),\n );\n } else {\n reject(err);\n }\n });\n });\n}\n\nfunction buildJarArgs(\n inputPaths: string | string[],\n options: ConvertOptions,\n): string[] | Error {\n const inputList = Array.isArray(inputPaths) ? inputPaths : [inputPaths];\n if (inputList.length === 0) {\n return new Error('At least one input path must be provided.');\n }\n\n for (const input of inputList) {\n if (!fs.existsSync(input)) {\n return new Error(`Input file or folder not found: ${input}`);\n }\n }\n\n return [...inputList, ...buildArgs(options)];\n}\n\nexport function convert(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<string> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n return Promise.reject(argsOrError);\n }\n // Library API: never streams to the parent process. Returns the full stdout\n // string so callers can do `const out = await convert(...)` without surprise\n // side-effects on process.stdout / process.stderr.\n return executeJar(argsOrError, { streamOutput: false });\n}\n\n/**\n * Internal entry point used by the bundled CLI. Streams Java's stdout and\n * stderr to the parent process in real time (so long-running conversions like\n * hybrid mode show progress as it happens) and resolves without a stdout\n * payload — preventing the caller from re-printing what was already streamed.\n *\n * Not part of the public API: do not import this from application code. Use\n * {@link convert} instead.\n *\n * @internal\n */\nexport async function _runForCli(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<void> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n throw argsOrError;\n }\n await executeJar(argsOrError, { streamOutput: true });\n}\n\n/**\n * @deprecated Use `convert()` and `ConvertOptions` instead. This function will be removed in a future version.\n */\nexport interface RunOptions {\n outputFolder?: string;\n password?: string;\n replaceInvalidChars?: string;\n generateMarkdown?: boolean;\n generateHtml?: boolean;\n generateAnnotatedPdf?: boolean;\n keepLineBreaks?: boolean;\n contentSafetyOff?: string;\n htmlInMarkdown?: boolean;\n addImageToMarkdown?: boolean;\n noJson?: boolean;\n debug?: boolean;\n useStructTree?: boolean;\n}\n\n/**\n * @deprecated Use `convert()` instead. This function will be removed in a future version.\n */\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n console.warn(\n 'Warning: run() is deprecated and will be removed in a future version. Use convert() instead.',\n );\n\n // Build format array based on legacy boolean options\n const formats: string[] = [];\n if (!options.noJson) {\n formats.push('json');\n }\n if (options.generateMarkdown) {\n if (options.addImageToMarkdown) {\n formats.push('markdown-with-images');\n } else if (options.htmlInMarkdown) {\n formats.push('markdown-with-html');\n } else {\n formats.push('markdown');\n }\n }\n if (options.generateHtml) {\n formats.push('html');\n }\n if (options.generateAnnotatedPdf) {\n formats.push('pdf');\n }\n\n return convert(inputPath, {\n outputDir: options.outputFolder,\n password: options.password,\n replaceInvalidChars: options.replaceInvalidChars,\n keepLineBreaks: options.keepLineBreaks,\n contentSafetyOff: options.contentSafetyOff,\n useStructTree: options.useStructTree,\n format: formats.length > 0 ? formats : undefined,\n quiet: !options.debug,\n });\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\n/**\n * Options for the convert function.\n */\nexport interface ConvertOptions {\n /** Directory where output files are written. Default: input file directory */\n outputDir?: string;\n /** Password for encrypted PDF files */\n password?: string;\n /** Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output. */\n format?: string | string[];\n /** Suppress console logging output */\n quiet?: boolean;\n /** Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background */\n contentSafetyOff?: string | string[];\n /** Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders */\n sanitize?: boolean;\n /** Preserve original line breaks in extracted text */\n keepLineBreaks?: boolean;\n /** Replacement character for invalid/unrecognized characters. Default: space */\n replaceInvalidChars?: string;\n /** Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called */\n useStructTree?: boolean;\n /** Table detection method. Values: default (border-based), cluster (border + cluster). Default: default */\n tableMethod?: string;\n /** Reading order algorithm. Values: off, xycut. Default: xycut */\n readingOrder?: string;\n /** Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none */\n markdownPageSeparator?: string;\n /** Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown. */\n markdownWithHtml?: boolean;\n /** Separator between pages in text output. Use %page-number% for page numbers. Default: none */\n textPageSeparator?: string;\n /** Separator between pages in HTML output. Use %page-number% for page numbers. Default: none */\n htmlPageSeparator?: string;\n /** Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external */\n imageOutput?: string;\n /** Output format for extracted images. Values: png, jpeg. Default: png */\n imageFormat?: string;\n /** Directory for extracted images (applies only with --image-output external) */\n imageDir?: string;\n /** Pages to extract (e.g., \"1,3,5-7\"). Default: all pages */\n pages?: string;\n /** Include page headers and footers in output */\n includeHeaderFooter?: boolean;\n /** Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental) */\n detectStrikethrough?: boolean;\n /** Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence) */\n hybrid?: string;\n /** Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend) */\n hybridMode?: string;\n /** Hybrid backend server URL (overrides default) */\n hybridUrl?: string;\n /** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */\n hybridTimeout?: string;\n /** Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50 */\n hybridChunkSize?: string;\n /** Opt in to Java fallback on hybrid backend error (default: disabled) */\n hybridFallback?: boolean;\n /** Write output to stdout instead of file (single format only) */\n toStdout?: boolean;\n /** Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode */\n threads?: string;\n /** Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0. */\n imageResolution?: string;\n /** Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17 */\n spaceRatio?: string;\n}\n\n/**\n * Options as parsed from CLI (all values are strings from commander).\n */\nexport interface CliOptions {\n outputDir?: string;\n password?: string;\n format?: string;\n quiet?: boolean;\n contentSafetyOff?: string;\n sanitize?: boolean;\n keepLineBreaks?: boolean;\n replaceInvalidChars?: string;\n useStructTree?: boolean;\n tableMethod?: string;\n readingOrder?: string;\n markdownPageSeparator?: string;\n markdownWithHtml?: boolean;\n textPageSeparator?: string;\n htmlPageSeparator?: string;\n imageOutput?: string;\n imageFormat?: string;\n imageDir?: string;\n pages?: string;\n includeHeaderFooter?: boolean;\n detectStrikethrough?: boolean;\n hybrid?: string;\n hybridMode?: string;\n hybridUrl?: string;\n hybridTimeout?: string;\n hybridChunkSize?: string;\n hybridFallback?: boolean;\n toStdout?: boolean;\n threads?: string;\n imageResolution?: string;\n spaceRatio?: string;\n}\n\n/**\n * Convert CLI options to ConvertOptions.\n */\nexport function buildConvertOptions(cliOptions: CliOptions): ConvertOptions {\n const convertOptions: ConvertOptions = {};\n\n if (cliOptions.outputDir) {\n convertOptions.outputDir = cliOptions.outputDir;\n }\n if (cliOptions.password) {\n convertOptions.password = cliOptions.password;\n }\n if (cliOptions.format) {\n convertOptions.format = cliOptions.format;\n }\n if (cliOptions.quiet) {\n convertOptions.quiet = true;\n }\n if (cliOptions.contentSafetyOff) {\n convertOptions.contentSafetyOff = cliOptions.contentSafetyOff;\n }\n if (cliOptions.sanitize) {\n convertOptions.sanitize = true;\n }\n if (cliOptions.keepLineBreaks) {\n convertOptions.keepLineBreaks = true;\n }\n if (cliOptions.replaceInvalidChars) {\n convertOptions.replaceInvalidChars = cliOptions.replaceInvalidChars;\n }\n if (cliOptions.useStructTree) {\n convertOptions.useStructTree = true;\n }\n if (cliOptions.tableMethod) {\n convertOptions.tableMethod = cliOptions.tableMethod;\n }\n if (cliOptions.readingOrder) {\n convertOptions.readingOrder = cliOptions.readingOrder;\n }\n if (cliOptions.markdownPageSeparator) {\n convertOptions.markdownPageSeparator = cliOptions.markdownPageSeparator;\n }\n if (cliOptions.markdownWithHtml) {\n convertOptions.markdownWithHtml = true;\n }\n if (cliOptions.textPageSeparator) {\n convertOptions.textPageSeparator = cliOptions.textPageSeparator;\n }\n if (cliOptions.htmlPageSeparator) {\n convertOptions.htmlPageSeparator = cliOptions.htmlPageSeparator;\n }\n if (cliOptions.imageOutput) {\n convertOptions.imageOutput = cliOptions.imageOutput;\n }\n if (cliOptions.imageFormat) {\n convertOptions.imageFormat = cliOptions.imageFormat;\n }\n if (cliOptions.imageDir) {\n convertOptions.imageDir = cliOptions.imageDir;\n }\n if (cliOptions.pages) {\n convertOptions.pages = cliOptions.pages;\n }\n if (cliOptions.includeHeaderFooter) {\n convertOptions.includeHeaderFooter = true;\n }\n if (cliOptions.detectStrikethrough) {\n convertOptions.detectStrikethrough = true;\n }\n if (cliOptions.hybrid) {\n convertOptions.hybrid = cliOptions.hybrid;\n }\n if (cliOptions.hybridMode) {\n convertOptions.hybridMode = cliOptions.hybridMode;\n }\n if (cliOptions.hybridUrl) {\n convertOptions.hybridUrl = cliOptions.hybridUrl;\n }\n if (cliOptions.hybridTimeout) {\n convertOptions.hybridTimeout = cliOptions.hybridTimeout;\n }\n if (cliOptions.hybridChunkSize) {\n convertOptions.hybridChunkSize = cliOptions.hybridChunkSize;\n }\n if (cliOptions.hybridFallback) {\n convertOptions.hybridFallback = true;\n }\n if (cliOptions.toStdout) {\n convertOptions.toStdout = true;\n }\n if (cliOptions.threads) {\n convertOptions.threads = cliOptions.threads;\n }\n if (cliOptions.imageResolution) {\n convertOptions.imageResolution = cliOptions.imageResolution;\n }\n if (cliOptions.spaceRatio) {\n convertOptions.spaceRatio = cliOptions.spaceRatio;\n }\n\n return convertOptions;\n}\n\n/**\n * Build CLI arguments array from ConvertOptions.\n */\nexport function buildArgs(options: ConvertOptions): string[] {\n const args: string[] = [];\n\n if (options.outputDir) {\n args.push('--output-dir', options.outputDir);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.format) {\n if (Array.isArray(options.format)) {\n if (options.format.length > 0) {\n args.push('--format', options.format.join(','));\n }\n } else {\n args.push('--format', options.format);\n }\n }\n if (options.quiet) {\n args.push('--quiet');\n }\n if (options.contentSafetyOff) {\n if (Array.isArray(options.contentSafetyOff)) {\n if (options.contentSafetyOff.length > 0) {\n args.push('--content-safety-off', options.contentSafetyOff.join(','));\n }\n } else {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n }\n if (options.sanitize) {\n args.push('--sanitize');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.useStructTree) {\n args.push('--use-struct-tree');\n }\n if (options.tableMethod) {\n args.push('--table-method', options.tableMethod);\n }\n if (options.readingOrder) {\n args.push('--reading-order', options.readingOrder);\n }\n if (options.markdownPageSeparator) {\n args.push('--markdown-page-separator', options.markdownPageSeparator);\n }\n if (options.markdownWithHtml) {\n args.push('--markdown-with-html');\n }\n if (options.textPageSeparator) {\n args.push('--text-page-separator', options.textPageSeparator);\n }\n if (options.htmlPageSeparator) {\n args.push('--html-page-separator', options.htmlPageSeparator);\n }\n if (options.imageOutput) {\n args.push('--image-output', options.imageOutput);\n }\n if (options.imageFormat) {\n args.push('--image-format', options.imageFormat);\n }\n if (options.imageDir) {\n args.push('--image-dir', options.imageDir);\n }\n if (options.pages) {\n args.push('--pages', options.pages);\n }\n if (options.includeHeaderFooter) {\n args.push('--include-header-footer');\n }\n if (options.detectStrikethrough) {\n args.push('--detect-strikethrough');\n }\n if (options.hybrid) {\n args.push('--hybrid', options.hybrid);\n }\n if (options.hybridMode) {\n args.push('--hybrid-mode', options.hybridMode);\n }\n if (options.hybridUrl) {\n args.push('--hybrid-url', options.hybridUrl);\n }\n if (options.hybridTimeout) {\n args.push('--hybrid-timeout', options.hybridTimeout);\n }\n if (options.hybridChunkSize) {\n args.push('--hybrid-chunk-size', options.hybridChunkSize);\n }\n if (options.hybridFallback) {\n args.push('--hybrid-fallback');\n }\n if (options.toStdout) {\n args.push('--to-stdout');\n }\n if (options.threads) {\n args.push('--threads', options.threads);\n }\n if (options.imageResolution) {\n args.push('--image-resolution', options.imageResolution);\n }\n if (options.spaceRatio) {\n args.push('--space-ratio', options.spaceRatio);\n }\n\n return args;\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\nimport { Command } from 'commander';\n\n/**\n * Register all CLI options on the given Commander program.\n */\nexport function registerCliOptions(program: Command): void {\n program.option('-o, --output-dir <value>', 'Directory where output files are written. Default: input file directory');\n program.option('-p, --password <value>', 'Password for encrypted PDF files');\n program.option('-f, --format <value>', 'Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output.');\n program.option('-q, --quiet', 'Suppress console logging output');\n program.option('--content-safety-off <value>', 'Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background');\n program.option('--sanitize', 'Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders');\n program.option('--keep-line-breaks', 'Preserve original line breaks in extracted text');\n program.option('--replace-invalid-chars <value>', 'Replacement character for invalid/unrecognized characters. Default: space');\n program.option('--use-struct-tree', 'Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called');\n program.option('--table-method <value>', 'Table detection method. Values: default (border-based), cluster (border + cluster). Default: default');\n program.option('--reading-order <value>', 'Reading order algorithm. Values: off, xycut. Default: xycut');\n program.option('--markdown-page-separator <value>', 'Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none');\n program.option('--markdown-with-html', 'Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown.');\n program.option('--text-page-separator <value>', 'Separator between pages in text output. Use %page-number% for page numbers. Default: none');\n program.option('--html-page-separator <value>', 'Separator between pages in HTML output. Use %page-number% for page numbers. Default: none');\n program.option('--image-output <value>', 'Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external');\n program.option('--image-format <value>', 'Output format for extracted images. Values: png, jpeg. Default: png');\n program.option('--image-dir <value>', 'Directory for extracted images (applies only with --image-output external)');\n program.option('--pages <value>', 'Pages to extract (e.g., \"1,3,5-7\"). Default: all pages');\n program.option('--include-header-footer', 'Include page headers and footers in output');\n program.option('--detect-strikethrough', 'Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental)');\n program.option('--hybrid <value>', 'Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence)');\n program.option('--hybrid-mode <value>', 'Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend)');\n program.option('--hybrid-url <value>', 'Hybrid backend server URL (overrides default)');\n program.option('--hybrid-timeout <value>', 'Hybrid backend request timeout in milliseconds (0 = use the backend\\'s own default). Default: 0');\n program.option('--hybrid-chunk-size <value>', 'Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50');\n program.option('--hybrid-fallback', 'Opt in to Java fallback on hybrid backend error (default: disabled)');\n program.option('--to-stdout', 'Write output to stdout instead of file (single format only)');\n program.option('--threads <value>', 'Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode');\n program.option('--image-resolution <value>', 'Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0.');\n program.option('--space-ratio <value>', 'Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17');\n}\n"],"mappings":";;;AACA,SAAS,SAAS,sBAAsB;;;ACDxC,SAAS,aAAa;AACtB,YAAY,UAAU;AACtB,YAAY,QAAQ;AACpB,SAAS,qBAAqB;AAC9B,SAAS,qBAAqB;;;AC2GvB,SAAS,oBAAoB,YAAwC;AAC1E,QAAM,iBAAiC,CAAC;AAExC,MAAI,WAAW,WAAW;AACxB,mBAAe,YAAY,WAAW;AAAA,EACxC;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW,WAAW;AAAA,EACvC;AACA,MAAI,WAAW,QAAQ;AACrB,mBAAe,SAAS,WAAW;AAAA,EACrC;AACA,MAAI,WAAW,OAAO;AACpB,mBAAe,QAAQ;AAAA,EACzB;AACA,MAAI,WAAW,kBAAkB;AAC/B,mBAAe,mBAAmB,WAAW;AAAA,EAC/C;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW;AAAA,EAC5B;AACA,MAAI,WAAW,gBAAgB;AAC7B,mBAAe,iBAAiB;AAAA,EAClC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB,WAAW;AAAA,EAClD;AACA,MAAI,WAAW,eAAe;AAC5B,mBAAe,gBAAgB;AAAA,EACjC;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,cAAc;AAC3B,mBAAe,eAAe,WAAW;AAAA,EAC3C;AACA,MAAI,WAAW,uBAAuB;AACpC,mBAAe,wBAAwB,WAAW;AAAA,EACpD;AACA,MAAI,WAAW,kBAAkB;AAC/B,mBAAe,mBAAmB;AAAA,EACpC;AACA,MAAI,WAAW,mBAAmB;AAChC,mBAAe,oBAAoB,WAAW;AAAA,EAChD;AACA,MAAI,WAAW,mBAAmB;AAChC,mBAAe,oBAAoB,WAAW;AAAA,EAChD;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,aAAa;AAC1B,mBAAe,cAAc,WAAW;AAAA,EAC1C;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW,WAAW;AAAA,EACvC;AACA,MAAI,WAAW,OAAO;AACpB,mBAAe,QAAQ,WAAW;AAAA,EACpC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB;AAAA,EACvC;AACA,MAAI,WAAW,qBAAqB;AAClC,mBAAe,sBAAsB;AAAA,EACvC;AACA,MAAI,WAAW,QAAQ;AACrB,mBAAe,SAAS,WAAW;AAAA,EACrC;AACA,MAAI,WAAW,YAAY;AACzB,mBAAe,aAAa,WAAW;AAAA,EACzC;AACA,MAAI,WAAW,WAAW;AACxB,mBAAe,YAAY,WAAW;AAAA,EACxC;AACA,MAAI,WAAW,eAAe;AAC5B,mBAAe,gBAAgB,WAAW;AAAA,EAC5C;AACA,MAAI,WAAW,iBAAiB;AAC9B,mBAAe,kBAAkB,WAAW;AAAA,EAC9C;AACA,MAAI,WAAW,gBAAgB;AAC7B,mBAAe,iBAAiB;AAAA,EAClC;AACA,MAAI,WAAW,UAAU;AACvB,mBAAe,WAAW;AAAA,EAC5B;AACA,MAAI,WAAW,SAAS;AACtB,mBAAe,UAAU,WAAW;AAAA,EACtC;AACA,MAAI,WAAW,iBAAiB;AAC9B,mBAAe,kBAAkB,WAAW;AAAA,EAC9C;AACA,MAAI,WAAW,YAAY;AACzB,mBAAe,aAAa,WAAW;AAAA,EACzC;AAEA,SAAO;AACT;AAKO,SAAS,UAAU,SAAmC;AAC3D,QAAM,OAAiB,CAAC;AAExB,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,EAC1C;AACA,MAAI,QAAQ,QAAQ;AAClB,QAAI,MAAM,QAAQ,QAAQ,MAAM,GAAG;AACjC,UAAI,QAAQ,OAAO,SAAS,GAAG;AAC7B,aAAK,KAAK,YAAY,QAAQ,OAAO,KAAK,GAAG,CAAC;AAAA,MAChD;AAAA,IACF,OAAO;AACL,WAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,IACtC;AAAA,EACF;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,SAAS;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,MAAM,QAAQ,QAAQ,gBAAgB,GAAG;AAC3C,UAAI,QAAQ,iBAAiB,SAAS,GAAG;AACvC,aAAK,KAAK,wBAAwB,QAAQ,iBAAiB,KAAK,GAAG,CAAC;AAAA,MACtE;AAAA,IACF,OAAO;AACL,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AAAA,EACF;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,YAAY;AAAA,EACxB;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,oBAAoB;AAAA,EAChC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,EAClE;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,cAAc;AACxB,SAAK,KAAK,mBAAmB,QAAQ,YAAY;AAAA,EACnD;AACA,MAAI,QAAQ,uBAAuB;AACjC,SAAK,KAAK,6BAA6B,QAAQ,qBAAqB;AAAA,EACtE;AACA,MAAI,QAAQ,kBAAkB;AAC5B,SAAK,KAAK,sBAAsB;AAAA,EAClC;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,eAAe,QAAQ,QAAQ;AAAA,EAC3C;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,WAAW,QAAQ,KAAK;AAAA,EACpC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,yBAAyB;AAAA,EACrC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,wBAAwB;AAAA,EACpC;AACA,MAAI,QAAQ,QAAQ;AAClB,SAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,EACtC;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AACA,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,oBAAoB,QAAQ,aAAa;AAAA,EACrD;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,uBAAuB,QAAQ,eAAe;AAAA,EAC1D;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,aAAa;AAAA,EACzB;AACA,MAAI,QAAQ,SAAS;AACnB,SAAK,KAAK,aAAa,QAAQ,OAAO;AAAA,EACxC;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,sBAAsB,QAAQ,eAAe;AAAA,EACzD;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AAEA,SAAO;AACT;;;ADxTA,IAAMA,cAAa,cAAc,YAAY,GAAG;AAChD,IAAMC,aAAiB,aAAQD,WAAU;AAEzC,IAAM,WAAW;AAWjB,SAAS,WAAW,MAAgB,mBAAwC,CAAC,GAAoB;AAC/F,QAAM,EAAE,eAAe,MAAM,IAAI;AAEjC,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,UAAe,UAAKC,YAAW,MAAM,OAAO,QAAQ;AAE1D,QAAI,CAAI,cAAW,OAAO,GAAG;AAC3B,aAAO;AAAA,QACL,IAAI,MAAM,yBAAyB,OAAO,sCAAsC;AAAA,MAClF;AAAA,IACF;AAEA,UAAM,UAAU;AAIhB,UAAM,cAAc;AAAA,MAClB;AAAA,MACA;AAAA,MACA;AAAA,MACA;AAAA,MACA,GAAG;AAAA,IACL;AAEA,UAAM,cAAc,MAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAKb,UAAM,gBAAgB,IAAI,cAAc,MAAM;AAC9C,UAAM,gBAAgB,IAAI,cAAc,MAAM;AAE9C,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAIhB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B,OAAO;AACL,kBAAU;AAAA,MACZ;AAAA,IACF,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAChB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AAGA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAIhC,YAAM,aAAa,cAAc,IAAI;AACrC,YAAM,aAAa,cAAc,IAAI;AACrC,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,cAAc;AAChB,kBAAQ,OAAO,MAAM,UAAU;AAAA,QACjC,OAAO;AACL,oBAAU;AAAA,QACZ;AAAA,MACF;AACA,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,aAAc,SAAQ,OAAO,MAAM,UAAU;AACjD,kBAAU;AAAA,MACZ;AAEA,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,cAAc,UAAU;AAC9B,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,WAAW;AAAA,QACxE;AAKA,QAAC,MAA2C,aAAa;AACzD,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAe;AACtC,UAAI,IAAI,QAAQ,SAAS,QAAQ,GAAG;AAClC;AAAA,UACE,IAAI;AAAA,YACF;AAAA,UACF;AAAA,QACF;AAAA,MACF,OAAO;AACL,eAAO,GAAG;AAAA,MACZ;AAAA,IACF,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,aACP,YACA,SACkB;AAClB,QAAM,YAAY,MAAM,QAAQ,UAAU,IAAI,aAAa,CAAC,UAAU;AACtE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO,IAAI,MAAM,2CAA2C;AAAA,EAC9D;AAEA,aAAW,SAAS,WAAW;AAC7B,QAAI,CAAI,cAAW,KAAK,GAAG;AACzB,aAAO,IAAI,MAAM,mCAAmC,KAAK,EAAE;AAAA,IAC7D;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,WAAW,GAAG,UAAU,OAAO,CAAC;AAC7C;AA2BA,eAAsB,WACpB,YACA,UAA0B,CAAC,GACZ;AACf,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,UAAM;AAAA,EACR;AACA,QAAM,WAAW,aAAa,EAAE,cAAc,KAAK,CAAC;AACtD;;;AEjLO,SAAS,mBAAmB,SAAwB;AACzD,UAAQ,OAAO,4BAA4B,yEAAyE;AACpH,UAAQ,OAAO,0BAA0B,kCAAkC;AAC3E,UAAQ,OAAO,wBAAwB,2MAA2M;AAClP,UAAQ,OAAO,eAAe,iCAAiC;AAC/D,UAAQ,OAAO,gCAAgC,kGAAkG;AACjJ,UAAQ,OAAO,cAAc,mHAAmH;AAChJ,UAAQ,OAAO,sBAAsB,iDAAiD;AACtF,UAAQ,OAAO,mCAAmC,2EAA2E;AAC7H,UAAQ,OAAO,qBAAqB,yPAAyP;AAC7R,UAAQ,OAAO,0BAA0B,sGAAsG;AAC/I,UAAQ,OAAO,2BAA2B,6DAA6D;AACvG,UAAQ,OAAO,qCAAqC,+FAA+F;AACnJ,UAAQ,OAAO,wBAAwB,yHAAyH;AAChK,UAAQ,OAAO,iCAAiC,2FAA2F;AAC3I,UAAQ,OAAO,iCAAiC,2FAA2F;AAC3I,UAAQ,OAAO,0BAA0B,wHAAwH;AACjK,UAAQ,OAAO,0BAA0B,qEAAqE;AAC9G,UAAQ,OAAO,uBAAuB,4EAA4E;AAClH,UAAQ,OAAO,mBAAmB,wDAAwD;AAC1F,UAAQ,OAAO,2BAA2B,4CAA4C;AACtF,UAAQ,OAAO,0BAA0B,gHAAgH;AACzJ,UAAQ,OAAO,oBAAoB,+SAA+S;AAClV,UAAQ,OAAO,yBAAyB,sGAAsG;AAC9I,UAAQ,OAAO,wBAAwB,+CAA+C;AACtF,UAAQ,OAAO,4BAA4B,gGAAiG;AAC5I,UAAQ,OAAO,+BAA+B,yOAAyO;AACvR,UAAQ,OAAO,qBAAqB,qEAAqE;AACzG,UAAQ,OAAO,eAAe,6DAA6D;AAC3F,UAAQ,OAAO,qBAAqB,iTAAiT;AACrV,UAAQ,OAAO,8BAA8B,iPAAiP;AAC9R,UAAQ,OAAO,yBAAyB,kRAAkR;AAC5T;;;AHlCA,SAAS,gBAAyB;AAChC,QAAM,UAAU,IAAI,QAAQ;AAE5B,UACG,KAAK,oBAAoB,EACzB,MAAM,sBAAsB,EAC5B,YAAY,4CAA4C,EACxD,mBAAmB,wCAAwC,EAC3D,yBAAyB,KAAK,EAC9B,SAAS,cAAc,uCAAuC;AAGjE,qBAAmB,OAAO;AAE1B,UAAQ,gBAAgB;AAAA,IACtB,UAAU,CAAC,QAAQ;AACjB,cAAQ,MAAM,IAAI,QAAQ,CAAC;AAAA,IAC7B;AAAA,IACA,aAAa,CAAC,KAAK,UAAU;AAC3B,YAAM,GAAG;AAAA,IACX;AAAA,EACF,CAAC;AAED,SAAO;AACT;AAEA,eAAe,OAAwB;AACrC,QAAM,UAAU,cAAc;AAE9B,UAAQ,aAAa;AAErB,MAAI;AACF,YAAQ,MAAM,QAAQ,IAAI;AAAA,EAC5B,SAAS,KAAK;AACZ,QAAI,eAAe,gBAAgB;AACjC,UAAI,IAAI,SAAS,2BAA2B;AAC1C,eAAO;AAAA,MACT;AACA,aAAO,IAAI,YAAY;AAAA,IACzB;AAEA,UAAM,UAAU,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC/D,YAAQ,MAAM,OAAO;AACrB,YAAQ,MAAM,wCAAwC;AACtD,WAAO;AAAA,EACT;AAEA,QAAM,aAAa,QAAQ,KAAiB;AAC5C,QAAM,aAAa,QAAQ;AAC3B,QAAM,iBAAiB,oBAAoB,UAAU;AAErD,MAAI;AAGF,UAAM,WAAW,YAAY,cAAc;AAC3C,WAAO;AAAA,EACT,SAAS,KAAK;AAMZ,UAAM,aACJ,eAAe,SAAU,IAAyC,eAAe;AACnF,QAAI,CAAC,YAAY;AACf,YAAM,UAAU,eAAe,QAAQ,IAAI,UAAU,OAAO,GAAG;AAC/D,cAAQ,MAAM,OAAO;AAAA,IACvB;AACA,WAAO;AAAA,EACT;AACF;AAEA,KAAK,EAAE,KAAK,CAAC,SAAS;AACpB,MAAI,SAAS,GAAG;AACd,YAAQ,KAAK,IAAI;AAAA,EACnB;AACF,CAAC;","names":["__filename","__dirname"]}
|
package/dist/index.cjs
CHANGED
|
@@ -138,6 +138,9 @@ function buildArgs(options) {
|
|
|
138
138
|
if (options.hybridTimeout) {
|
|
139
139
|
args.push("--hybrid-timeout", options.hybridTimeout);
|
|
140
140
|
}
|
|
141
|
+
if (options.hybridChunkSize) {
|
|
142
|
+
args.push("--hybrid-chunk-size", options.hybridChunkSize);
|
|
143
|
+
}
|
|
141
144
|
if (options.hybridFallback) {
|
|
142
145
|
args.push("--hybrid-fallback");
|
|
143
146
|
}
|
package/dist/index.cjs.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"sources":["../src/index.ts","../node_modules/.pnpm/tsup@8.5.1_postcss@8.5.28_typescript@6.0.3/node_modules/tsup/assets/cjs_shims.js","../src/convert-options.generated.ts"],"sourcesContent":["import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { StringDecoder } from 'string_decoder';\nimport { fileURLToPath } from 'url';\n\n// Re-export types and utilities from auto-generated file\nexport type { ConvertOptions } from './convert-options.generated.js';\nexport { buildArgs } from './convert-options.generated.js';\nimport type { ConvertOptions } from './convert-options.generated.js';\nimport { buildArgs } from './convert-options.generated.js';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\ninterface JarExecutionOptions {\n /**\n * When true, forwards Java's stdout and stderr chunks to the parent\n * process in real time as well as accumulating them. Used by the bundled\n * CLI so long-running conversions show progress as it happens.\n */\n streamOutput?: boolean;\n}\n\nfunction executeJar(args: string[], executionOptions: JarExecutionOptions = {}): Promise<string> {\n const { streamOutput = false } = executionOptions;\n\n return new Promise((resolve, reject) => {\n const jarPath = path.join(__dirname, '..', 'lib', JAR_NAME);\n\n if (!fs.existsSync(jarPath)) {\n return reject(\n new Error(`JAR file not found at ${jarPath}. Please run the build script first.`),\n );\n }\n\n const command = 'java';\n // Force headless AWT so macOS doesn't surface a Dock icon (and steal focus)\n // every time the JVM touches ImageIO/PDFBox rendering. Safe on all OSes —\n // the CLI never opens a UI window, only manipulates BufferedImages.\n const commandArgs = [\n '-Djava.awt.headless=true',\n '-Dapple.awt.UIElement=true',\n '-jar',\n jarPath,\n ...args,\n ];\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n // StringDecoder buffers incomplete multi-byte UTF-8 sequences across\n // chunk boundaries — Buffer.toString() alone would emit replacement\n // characters when, e.g., a 3-byte Korean codepoint splits across two\n // 'data' events. One decoder per stream so they don't share state.\n const stdoutDecoder = new StringDecoder('utf8');\n const stderrDecoder = new StringDecoder('utf8');\n\n javaProcess.stdout.on('data', (data: Buffer) => {\n const chunk = stdoutDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n // Stream-only: don't also accumulate, or a multi-hour conversion\n // would buffer its entire (potentially gigabyte) stdout in memory\n // for no consumer.\n process.stdout.write(chunk);\n } else {\n stdout += chunk;\n }\n });\n\n javaProcess.stderr.on('data', (data: Buffer) => {\n const chunk = stderrDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n process.stderr.write(chunk);\n }\n // stderr is always accumulated (progress logs are small and we need\n // them for error messages on non-zero exit).\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n // Flush any trailing bytes the decoder is still holding (always emit\n // them — if we drop them on error paths, error messages with non-ASCII\n // characters lose their tail).\n const stdoutTail = stdoutDecoder.end();\n const stderrTail = stderrDecoder.end();\n if (stdoutTail.length > 0) {\n if (streamOutput) {\n process.stdout.write(stdoutTail);\n } else {\n stdout += stdoutTail;\n }\n }\n if (stderrTail.length > 0) {\n if (streamOutput) process.stderr.write(stderrTail);\n stderr += stderrTail;\n }\n\n if (code === 0) {\n resolve(stdout);\n } else {\n const errorOutput = stderr || stdout;\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${errorOutput}`,\n );\n // Tag so the CLI can suppress re-printing this message — Java's\n // stderr was already streamed live to the parent in CLI mode, and\n // re-printing risks leaking anything sensitive Java logged\n // (e.g. a --password value echoed by an underlying library).\n (error as Error & { isJavaExit?: boolean }).isJavaExit = true;\n reject(error);\n }\n });\n\n javaProcess.on('error', (err: Error) => {\n if (err.message.includes('ENOENT')) {\n reject(\n new Error(\n \"'java' command not found. Please ensure Java is installed and in your system's PATH.\",\n ),\n );\n } else {\n reject(err);\n }\n });\n });\n}\n\nfunction buildJarArgs(\n inputPaths: string | string[],\n options: ConvertOptions,\n): string[] | Error {\n const inputList = Array.isArray(inputPaths) ? inputPaths : [inputPaths];\n if (inputList.length === 0) {\n return new Error('At least one input path must be provided.');\n }\n\n for (const input of inputList) {\n if (!fs.existsSync(input)) {\n return new Error(`Input file or folder not found: ${input}`);\n }\n }\n\n return [...inputList, ...buildArgs(options)];\n}\n\nexport function convert(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<string> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n return Promise.reject(argsOrError);\n }\n // Library API: never streams to the parent process. Returns the full stdout\n // string so callers can do `const out = await convert(...)` without surprise\n // side-effects on process.stdout / process.stderr.\n return executeJar(argsOrError, { streamOutput: false });\n}\n\n/**\n * Internal entry point used by the bundled CLI. Streams Java's stdout and\n * stderr to the parent process in real time (so long-running conversions like\n * hybrid mode show progress as it happens) and resolves without a stdout\n * payload — preventing the caller from re-printing what was already streamed.\n *\n * Not part of the public API: do not import this from application code. Use\n * {@link convert} instead.\n *\n * @internal\n */\nexport async function _runForCli(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<void> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n throw argsOrError;\n }\n await executeJar(argsOrError, { streamOutput: true });\n}\n\n/**\n * @deprecated Use `convert()` and `ConvertOptions` instead. This function will be removed in a future version.\n */\nexport interface RunOptions {\n outputFolder?: string;\n password?: string;\n replaceInvalidChars?: string;\n generateMarkdown?: boolean;\n generateHtml?: boolean;\n generateAnnotatedPdf?: boolean;\n keepLineBreaks?: boolean;\n contentSafetyOff?: string;\n htmlInMarkdown?: boolean;\n addImageToMarkdown?: boolean;\n noJson?: boolean;\n debug?: boolean;\n useStructTree?: boolean;\n}\n\n/**\n * @deprecated Use `convert()` instead. This function will be removed in a future version.\n */\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n console.warn(\n 'Warning: run() is deprecated and will be removed in a future version. Use convert() instead.',\n );\n\n // Build format array based on legacy boolean options\n const formats: string[] = [];\n if (!options.noJson) {\n formats.push('json');\n }\n if (options.generateMarkdown) {\n if (options.addImageToMarkdown) {\n formats.push('markdown-with-images');\n } else if (options.htmlInMarkdown) {\n formats.push('markdown-with-html');\n } else {\n formats.push('markdown');\n }\n }\n if (options.generateHtml) {\n formats.push('html');\n }\n if (options.generateAnnotatedPdf) {\n formats.push('pdf');\n }\n\n return convert(inputPath, {\n outputDir: options.outputFolder,\n password: options.password,\n replaceInvalidChars: options.replaceInvalidChars,\n keepLineBreaks: options.keepLineBreaks,\n contentSafetyOff: options.contentSafetyOff,\n useStructTree: options.useStructTree,\n format: formats.length > 0 ? formats : undefined,\n quiet: !options.debug,\n });\n}\n","// Shim globals in cjs bundle\n// There's a weird bug that esbuild will always inject importMetaUrl\n// if we export it as `const importMetaUrl = ... __filename ...`\n// But using a function will not cause this issue\n\nconst getImportMetaUrl = () => \n typeof document === \"undefined\" \n ? new URL(`file:${__filename}`).href \n : (document.currentScript && document.currentScript.tagName.toUpperCase() === 'SCRIPT') \n ? document.currentScript.src \n : new URL(\"main.js\", document.baseURI).href;\n\nexport const importMetaUrl = /* @__PURE__ */ getImportMetaUrl()\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\n/**\n * Options for the convert function.\n */\nexport interface ConvertOptions {\n /** Directory where output files are written. Default: input file directory */\n outputDir?: string;\n /** Password for encrypted PDF files */\n password?: string;\n /** Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output. */\n format?: string | string[];\n /** Suppress console logging output */\n quiet?: boolean;\n /** Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background */\n contentSafetyOff?: string | string[];\n /** Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders */\n sanitize?: boolean;\n /** Preserve original line breaks in extracted text */\n keepLineBreaks?: boolean;\n /** Replacement character for invalid/unrecognized characters. Default: space */\n replaceInvalidChars?: string;\n /** Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called */\n useStructTree?: boolean;\n /** Table detection method. Values: default (border-based), cluster (border + cluster). Default: default */\n tableMethod?: string;\n /** Reading order algorithm. Values: off, xycut. Default: xycut */\n readingOrder?: string;\n /** Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none */\n markdownPageSeparator?: string;\n /** Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown. */\n markdownWithHtml?: boolean;\n /** Separator between pages in text output. Use %page-number% for page numbers. Default: none */\n textPageSeparator?: string;\n /** Separator between pages in HTML output. Use %page-number% for page numbers. Default: none */\n htmlPageSeparator?: string;\n /** Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external */\n imageOutput?: string;\n /** Output format for extracted images. Values: png, jpeg. Default: png */\n imageFormat?: string;\n /** Directory for extracted images (applies only with --image-output external) */\n imageDir?: string;\n /** Pages to extract (e.g., \"1,3,5-7\"). Default: all pages */\n pages?: string;\n /** Include page headers and footers in output */\n includeHeaderFooter?: boolean;\n /** Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental) */\n detectStrikethrough?: boolean;\n /** Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence) */\n hybrid?: string;\n /** Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend) */\n hybridMode?: string;\n /** Hybrid backend server URL (overrides default) */\n hybridUrl?: string;\n /** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */\n hybridTimeout?: string;\n /** Opt in to Java fallback on hybrid backend error (default: disabled) */\n hybridFallback?: boolean;\n /** Write output to stdout instead of file (single format only) */\n toStdout?: boolean;\n /** Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode */\n threads?: string;\n /** Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0. */\n imageResolution?: string;\n /** Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17 */\n spaceRatio?: string;\n}\n\n/**\n * Options as parsed from CLI (all values are strings from commander).\n */\nexport interface CliOptions {\n outputDir?: string;\n password?: string;\n format?: string;\n quiet?: boolean;\n contentSafetyOff?: string;\n sanitize?: boolean;\n keepLineBreaks?: boolean;\n replaceInvalidChars?: string;\n useStructTree?: boolean;\n tableMethod?: string;\n readingOrder?: string;\n markdownPageSeparator?: string;\n markdownWithHtml?: boolean;\n textPageSeparator?: string;\n htmlPageSeparator?: string;\n imageOutput?: string;\n imageFormat?: string;\n imageDir?: string;\n pages?: string;\n includeHeaderFooter?: boolean;\n detectStrikethrough?: boolean;\n hybrid?: string;\n hybridMode?: string;\n hybridUrl?: string;\n hybridTimeout?: string;\n hybridFallback?: boolean;\n toStdout?: boolean;\n threads?: string;\n imageResolution?: string;\n spaceRatio?: string;\n}\n\n/**\n * Convert CLI options to ConvertOptions.\n */\nexport function buildConvertOptions(cliOptions: CliOptions): ConvertOptions {\n const convertOptions: ConvertOptions = {};\n\n if (cliOptions.outputDir) {\n convertOptions.outputDir = cliOptions.outputDir;\n }\n if (cliOptions.password) {\n convertOptions.password = cliOptions.password;\n }\n if (cliOptions.format) {\n convertOptions.format = cliOptions.format;\n }\n if (cliOptions.quiet) {\n convertOptions.quiet = true;\n }\n if (cliOptions.contentSafetyOff) {\n convertOptions.contentSafetyOff = cliOptions.contentSafetyOff;\n }\n if (cliOptions.sanitize) {\n convertOptions.sanitize = true;\n }\n if (cliOptions.keepLineBreaks) {\n convertOptions.keepLineBreaks = true;\n }\n if (cliOptions.replaceInvalidChars) {\n convertOptions.replaceInvalidChars = cliOptions.replaceInvalidChars;\n }\n if (cliOptions.useStructTree) {\n convertOptions.useStructTree = true;\n }\n if (cliOptions.tableMethod) {\n convertOptions.tableMethod = cliOptions.tableMethod;\n }\n if (cliOptions.readingOrder) {\n convertOptions.readingOrder = cliOptions.readingOrder;\n }\n if (cliOptions.markdownPageSeparator) {\n convertOptions.markdownPageSeparator = cliOptions.markdownPageSeparator;\n }\n if (cliOptions.markdownWithHtml) {\n convertOptions.markdownWithHtml = true;\n }\n if (cliOptions.textPageSeparator) {\n convertOptions.textPageSeparator = cliOptions.textPageSeparator;\n }\n if (cliOptions.htmlPageSeparator) {\n convertOptions.htmlPageSeparator = cliOptions.htmlPageSeparator;\n }\n if (cliOptions.imageOutput) {\n convertOptions.imageOutput = cliOptions.imageOutput;\n }\n if (cliOptions.imageFormat) {\n convertOptions.imageFormat = cliOptions.imageFormat;\n }\n if (cliOptions.imageDir) {\n convertOptions.imageDir = cliOptions.imageDir;\n }\n if (cliOptions.pages) {\n convertOptions.pages = cliOptions.pages;\n }\n if (cliOptions.includeHeaderFooter) {\n convertOptions.includeHeaderFooter = true;\n }\n if (cliOptions.detectStrikethrough) {\n convertOptions.detectStrikethrough = true;\n }\n if (cliOptions.hybrid) {\n convertOptions.hybrid = cliOptions.hybrid;\n }\n if (cliOptions.hybridMode) {\n convertOptions.hybridMode = cliOptions.hybridMode;\n }\n if (cliOptions.hybridUrl) {\n convertOptions.hybridUrl = cliOptions.hybridUrl;\n }\n if (cliOptions.hybridTimeout) {\n convertOptions.hybridTimeout = cliOptions.hybridTimeout;\n }\n if (cliOptions.hybridFallback) {\n convertOptions.hybridFallback = true;\n }\n if (cliOptions.toStdout) {\n convertOptions.toStdout = true;\n }\n if (cliOptions.threads) {\n convertOptions.threads = cliOptions.threads;\n }\n if (cliOptions.imageResolution) {\n convertOptions.imageResolution = cliOptions.imageResolution;\n }\n if (cliOptions.spaceRatio) {\n convertOptions.spaceRatio = cliOptions.spaceRatio;\n }\n\n return convertOptions;\n}\n\n/**\n * Build CLI arguments array from ConvertOptions.\n */\nexport function buildArgs(options: ConvertOptions): string[] {\n const args: string[] = [];\n\n if (options.outputDir) {\n args.push('--output-dir', options.outputDir);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.format) {\n if (Array.isArray(options.format)) {\n if (options.format.length > 0) {\n args.push('--format', options.format.join(','));\n }\n } else {\n args.push('--format', options.format);\n }\n }\n if (options.quiet) {\n args.push('--quiet');\n }\n if (options.contentSafetyOff) {\n if (Array.isArray(options.contentSafetyOff)) {\n if (options.contentSafetyOff.length > 0) {\n args.push('--content-safety-off', options.contentSafetyOff.join(','));\n }\n } else {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n }\n if (options.sanitize) {\n args.push('--sanitize');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.useStructTree) {\n args.push('--use-struct-tree');\n }\n if (options.tableMethod) {\n args.push('--table-method', options.tableMethod);\n }\n if (options.readingOrder) {\n args.push('--reading-order', options.readingOrder);\n }\n if (options.markdownPageSeparator) {\n args.push('--markdown-page-separator', options.markdownPageSeparator);\n }\n if (options.markdownWithHtml) {\n args.push('--markdown-with-html');\n }\n if (options.textPageSeparator) {\n args.push('--text-page-separator', options.textPageSeparator);\n }\n if (options.htmlPageSeparator) {\n args.push('--html-page-separator', options.htmlPageSeparator);\n }\n if (options.imageOutput) {\n args.push('--image-output', options.imageOutput);\n }\n if (options.imageFormat) {\n args.push('--image-format', options.imageFormat);\n }\n if (options.imageDir) {\n args.push('--image-dir', options.imageDir);\n }\n if (options.pages) {\n args.push('--pages', options.pages);\n }\n if (options.includeHeaderFooter) {\n args.push('--include-header-footer');\n }\n if (options.detectStrikethrough) {\n args.push('--detect-strikethrough');\n }\n if (options.hybrid) {\n args.push('--hybrid', options.hybrid);\n }\n if (options.hybridMode) {\n args.push('--hybrid-mode', options.hybridMode);\n }\n if (options.hybridUrl) {\n args.push('--hybrid-url', options.hybridUrl);\n }\n if (options.hybridTimeout) {\n args.push('--hybrid-timeout', options.hybridTimeout);\n }\n if (options.hybridFallback) {\n args.push('--hybrid-fallback');\n }\n if (options.toStdout) {\n args.push('--to-stdout');\n }\n if (options.threads) {\n args.push('--threads', options.threads);\n }\n if (options.imageResolution) {\n args.push('--image-resolution', options.imageResolution);\n }\n if (options.spaceRatio) {\n args.push('--space-ratio', options.spaceRatio);\n }\n\n return args;\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;;;ACKA,IAAM,mBAAmB,MACvB,OAAO,aAAa,cAChB,IAAI,IAAI,QAAQ,UAAU,EAAE,EAAE,OAC7B,SAAS,iBAAiB,SAAS,cAAc,QAAQ,YAAY,MAAM,WAC1E,SAAS,cAAc,MACvB,IAAI,IAAI,WAAW,SAAS,OAAO,EAAE;AAEtC,IAAM,gBAAgC,iCAAiB;;;ADZ9D,2BAAsB;AACtB,WAAsB;AACtB,SAAoB;AACpB,4BAA8B;AAC9B,iBAA8B;;;AE4MvB,SAAS,UAAU,SAAmC;AAC3D,QAAM,OAAiB,CAAC;AAExB,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,EAC1C;AACA,MAAI,QAAQ,QAAQ;AAClB,QAAI,MAAM,QAAQ,QAAQ,MAAM,GAAG;AACjC,UAAI,QAAQ,OAAO,SAAS,GAAG;AAC7B,aAAK,KAAK,YAAY,QAAQ,OAAO,KAAK,GAAG,CAAC;AAAA,MAChD;AAAA,IACF,OAAO;AACL,WAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,IACtC;AAAA,EACF;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,SAAS;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,MAAM,QAAQ,QAAQ,gBAAgB,GAAG;AAC3C,UAAI,QAAQ,iBAAiB,SAAS,GAAG;AACvC,aAAK,KAAK,wBAAwB,QAAQ,iBAAiB,KAAK,GAAG,CAAC;AAAA,MACtE;AAAA,IACF,OAAO;AACL,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AAAA,EACF;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,YAAY;AAAA,EACxB;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,oBAAoB;AAAA,EAChC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,EAClE;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,cAAc;AACxB,SAAK,KAAK,mBAAmB,QAAQ,YAAY;AAAA,EACnD;AACA,MAAI,QAAQ,uBAAuB;AACjC,SAAK,KAAK,6BAA6B,QAAQ,qBAAqB;AAAA,EACtE;AACA,MAAI,QAAQ,kBAAkB;AAC5B,SAAK,KAAK,sBAAsB;AAAA,EAClC;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,eAAe,QAAQ,QAAQ;AAAA,EAC3C;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,WAAW,QAAQ,KAAK;AAAA,EACpC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,yBAAyB;AAAA,EACrC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,wBAAwB;AAAA,EACpC;AACA,MAAI,QAAQ,QAAQ;AAClB,SAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,EACtC;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AACA,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,oBAAoB,QAAQ,aAAa;AAAA,EACrD;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,aAAa;AAAA,EACzB;AACA,MAAI,QAAQ,SAAS;AACnB,SAAK,KAAK,aAAa,QAAQ,OAAO;AAAA,EACxC;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,sBAAsB,QAAQ,eAAe;AAAA,EACzD;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AAEA,SAAO;AACT;;;AF/SA,IAAMA,kBAAa,0BAAc,aAAe;AAChD,IAAM,YAAiB,aAAQA,WAAU;AAEzC,IAAM,WAAW;AAWjB,SAAS,WAAW,MAAgB,mBAAwC,CAAC,GAAoB;AAC/F,QAAM,EAAE,eAAe,MAAM,IAAI;AAEjC,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,UAAe,UAAK,WAAW,MAAM,OAAO,QAAQ;AAE1D,QAAI,CAAI,cAAW,OAAO,GAAG;AAC3B,aAAO;AAAA,QACL,IAAI,MAAM,yBAAyB,OAAO,sCAAsC;AAAA,MAClF;AAAA,IACF;AAEA,UAAM,UAAU;AAIhB,UAAM,cAAc;AAAA,MAClB;AAAA,MACA;AAAA,MACA;AAAA,MACA;AAAA,MACA,GAAG;AAAA,IACL;AAEA,UAAM,kBAAc,4BAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAKb,UAAM,gBAAgB,IAAI,oCAAc,MAAM;AAC9C,UAAM,gBAAgB,IAAI,oCAAc,MAAM;AAE9C,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAIhB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B,OAAO;AACL,kBAAU;AAAA,MACZ;AAAA,IACF,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAChB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AAGA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAIhC,YAAM,aAAa,cAAc,IAAI;AACrC,YAAM,aAAa,cAAc,IAAI;AACrC,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,cAAc;AAChB,kBAAQ,OAAO,MAAM,UAAU;AAAA,QACjC,OAAO;AACL,oBAAU;AAAA,QACZ;AAAA,MACF;AACA,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,aAAc,SAAQ,OAAO,MAAM,UAAU;AACjD,kBAAU;AAAA,MACZ;AAEA,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,cAAc,UAAU;AAC9B,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,WAAW;AAAA,QACxE;AAKA,QAAC,MAA2C,aAAa;AACzD,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAe;AACtC,UAAI,IAAI,QAAQ,SAAS,QAAQ,GAAG;AAClC;AAAA,UACE,IAAI;AAAA,YACF;AAAA,UACF;AAAA,QACF;AAAA,MACF,OAAO;AACL,eAAO,GAAG;AAAA,MACZ;AAAA,IACF,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,aACP,YACA,SACkB;AAClB,QAAM,YAAY,MAAM,QAAQ,UAAU,IAAI,aAAa,CAAC,UAAU;AACtE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO,IAAI,MAAM,2CAA2C;AAAA,EAC9D;AAEA,aAAW,SAAS,WAAW;AAC7B,QAAI,CAAI,cAAW,KAAK,GAAG;AACzB,aAAO,IAAI,MAAM,mCAAmC,KAAK,EAAE;AAAA,IAC7D;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,WAAW,GAAG,UAAU,OAAO,CAAC;AAC7C;AAEO,SAAS,QACd,YACA,UAA0B,CAAC,GACV;AACjB,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,WAAO,QAAQ,OAAO,WAAW;AAAA,EACnC;AAIA,SAAO,WAAW,aAAa,EAAE,cAAc,MAAM,CAAC;AACxD;AAaA,eAAsB,WACpB,YACA,UAA0B,CAAC,GACZ;AACf,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,UAAM;AAAA,EACR;AACA,QAAM,WAAW,aAAa,EAAE,cAAc,KAAK,CAAC;AACtD;AAwBO,SAAS,IAAI,WAAmB,UAAsB,CAAC,GAAoB;AAChF,UAAQ;AAAA,IACN;AAAA,EACF;AAGA,QAAM,UAAoB,CAAC;AAC3B,MAAI,CAAC,QAAQ,QAAQ;AACnB,YAAQ,KAAK,MAAM;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,QAAQ,oBAAoB;AAC9B,cAAQ,KAAK,sBAAsB;AAAA,IACrC,WAAW,QAAQ,gBAAgB;AACjC,cAAQ,KAAK,oBAAoB;AAAA,IACnC,OAAO;AACL,cAAQ,KAAK,UAAU;AAAA,IACzB;AAAA,EACF;AACA,MAAI,QAAQ,cAAc;AACxB,YAAQ,KAAK,MAAM;AAAA,EACrB;AACA,MAAI,QAAQ,sBAAsB;AAChC,YAAQ,KAAK,KAAK;AAAA,EACpB;AAEA,SAAO,QAAQ,WAAW;AAAA,IACxB,WAAW,QAAQ;AAAA,IACnB,UAAU,QAAQ;AAAA,IAClB,qBAAqB,QAAQ;AAAA,IAC7B,gBAAgB,QAAQ;AAAA,IACxB,kBAAkB,QAAQ;AAAA,IAC1B,eAAe,QAAQ;AAAA,IACvB,QAAQ,QAAQ,SAAS,IAAI,UAAU;AAAA,IACvC,OAAO,CAAC,QAAQ;AAAA,EAClB,CAAC;AACH;","names":["__filename"]}
|
|
1
|
+
{"version":3,"sources":["../src/index.ts","../node_modules/.pnpm/tsup@8.5.1_postcss@8.5.28_typescript@6.0.3/node_modules/tsup/assets/cjs_shims.js","../src/convert-options.generated.ts"],"sourcesContent":["import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { StringDecoder } from 'string_decoder';\nimport { fileURLToPath } from 'url';\n\n// Re-export types and utilities from auto-generated file\nexport type { ConvertOptions } from './convert-options.generated.js';\nexport { buildArgs } from './convert-options.generated.js';\nimport type { ConvertOptions } from './convert-options.generated.js';\nimport { buildArgs } from './convert-options.generated.js';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\ninterface JarExecutionOptions {\n /**\n * When true, forwards Java's stdout and stderr chunks to the parent\n * process in real time as well as accumulating them. Used by the bundled\n * CLI so long-running conversions show progress as it happens.\n */\n streamOutput?: boolean;\n}\n\nfunction executeJar(args: string[], executionOptions: JarExecutionOptions = {}): Promise<string> {\n const { streamOutput = false } = executionOptions;\n\n return new Promise((resolve, reject) => {\n const jarPath = path.join(__dirname, '..', 'lib', JAR_NAME);\n\n if (!fs.existsSync(jarPath)) {\n return reject(\n new Error(`JAR file not found at ${jarPath}. Please run the build script first.`),\n );\n }\n\n const command = 'java';\n // Force headless AWT so macOS doesn't surface a Dock icon (and steal focus)\n // every time the JVM touches ImageIO/PDFBox rendering. Safe on all OSes —\n // the CLI never opens a UI window, only manipulates BufferedImages.\n const commandArgs = [\n '-Djava.awt.headless=true',\n '-Dapple.awt.UIElement=true',\n '-jar',\n jarPath,\n ...args,\n ];\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n // StringDecoder buffers incomplete multi-byte UTF-8 sequences across\n // chunk boundaries — Buffer.toString() alone would emit replacement\n // characters when, e.g., a 3-byte Korean codepoint splits across two\n // 'data' events. One decoder per stream so they don't share state.\n const stdoutDecoder = new StringDecoder('utf8');\n const stderrDecoder = new StringDecoder('utf8');\n\n javaProcess.stdout.on('data', (data: Buffer) => {\n const chunk = stdoutDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n // Stream-only: don't also accumulate, or a multi-hour conversion\n // would buffer its entire (potentially gigabyte) stdout in memory\n // for no consumer.\n process.stdout.write(chunk);\n } else {\n stdout += chunk;\n }\n });\n\n javaProcess.stderr.on('data', (data: Buffer) => {\n const chunk = stderrDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n process.stderr.write(chunk);\n }\n // stderr is always accumulated (progress logs are small and we need\n // them for error messages on non-zero exit).\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n // Flush any trailing bytes the decoder is still holding (always emit\n // them — if we drop them on error paths, error messages with non-ASCII\n // characters lose their tail).\n const stdoutTail = stdoutDecoder.end();\n const stderrTail = stderrDecoder.end();\n if (stdoutTail.length > 0) {\n if (streamOutput) {\n process.stdout.write(stdoutTail);\n } else {\n stdout += stdoutTail;\n }\n }\n if (stderrTail.length > 0) {\n if (streamOutput) process.stderr.write(stderrTail);\n stderr += stderrTail;\n }\n\n if (code === 0) {\n resolve(stdout);\n } else {\n const errorOutput = stderr || stdout;\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${errorOutput}`,\n );\n // Tag so the CLI can suppress re-printing this message — Java's\n // stderr was already streamed live to the parent in CLI mode, and\n // re-printing risks leaking anything sensitive Java logged\n // (e.g. a --password value echoed by an underlying library).\n (error as Error & { isJavaExit?: boolean }).isJavaExit = true;\n reject(error);\n }\n });\n\n javaProcess.on('error', (err: Error) => {\n if (err.message.includes('ENOENT')) {\n reject(\n new Error(\n \"'java' command not found. Please ensure Java is installed and in your system's PATH.\",\n ),\n );\n } else {\n reject(err);\n }\n });\n });\n}\n\nfunction buildJarArgs(\n inputPaths: string | string[],\n options: ConvertOptions,\n): string[] | Error {\n const inputList = Array.isArray(inputPaths) ? inputPaths : [inputPaths];\n if (inputList.length === 0) {\n return new Error('At least one input path must be provided.');\n }\n\n for (const input of inputList) {\n if (!fs.existsSync(input)) {\n return new Error(`Input file or folder not found: ${input}`);\n }\n }\n\n return [...inputList, ...buildArgs(options)];\n}\n\nexport function convert(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<string> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n return Promise.reject(argsOrError);\n }\n // Library API: never streams to the parent process. Returns the full stdout\n // string so callers can do `const out = await convert(...)` without surprise\n // side-effects on process.stdout / process.stderr.\n return executeJar(argsOrError, { streamOutput: false });\n}\n\n/**\n * Internal entry point used by the bundled CLI. Streams Java's stdout and\n * stderr to the parent process in real time (so long-running conversions like\n * hybrid mode show progress as it happens) and resolves without a stdout\n * payload — preventing the caller from re-printing what was already streamed.\n *\n * Not part of the public API: do not import this from application code. Use\n * {@link convert} instead.\n *\n * @internal\n */\nexport async function _runForCli(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<void> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n throw argsOrError;\n }\n await executeJar(argsOrError, { streamOutput: true });\n}\n\n/**\n * @deprecated Use `convert()` and `ConvertOptions` instead. This function will be removed in a future version.\n */\nexport interface RunOptions {\n outputFolder?: string;\n password?: string;\n replaceInvalidChars?: string;\n generateMarkdown?: boolean;\n generateHtml?: boolean;\n generateAnnotatedPdf?: boolean;\n keepLineBreaks?: boolean;\n contentSafetyOff?: string;\n htmlInMarkdown?: boolean;\n addImageToMarkdown?: boolean;\n noJson?: boolean;\n debug?: boolean;\n useStructTree?: boolean;\n}\n\n/**\n * @deprecated Use `convert()` instead. This function will be removed in a future version.\n */\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n console.warn(\n 'Warning: run() is deprecated and will be removed in a future version. Use convert() instead.',\n );\n\n // Build format array based on legacy boolean options\n const formats: string[] = [];\n if (!options.noJson) {\n formats.push('json');\n }\n if (options.generateMarkdown) {\n if (options.addImageToMarkdown) {\n formats.push('markdown-with-images');\n } else if (options.htmlInMarkdown) {\n formats.push('markdown-with-html');\n } else {\n formats.push('markdown');\n }\n }\n if (options.generateHtml) {\n formats.push('html');\n }\n if (options.generateAnnotatedPdf) {\n formats.push('pdf');\n }\n\n return convert(inputPath, {\n outputDir: options.outputFolder,\n password: options.password,\n replaceInvalidChars: options.replaceInvalidChars,\n keepLineBreaks: options.keepLineBreaks,\n contentSafetyOff: options.contentSafetyOff,\n useStructTree: options.useStructTree,\n format: formats.length > 0 ? formats : undefined,\n quiet: !options.debug,\n });\n}\n","// Shim globals in cjs bundle\n// There's a weird bug that esbuild will always inject importMetaUrl\n// if we export it as `const importMetaUrl = ... __filename ...`\n// But using a function will not cause this issue\n\nconst getImportMetaUrl = () => \n typeof document === \"undefined\" \n ? new URL(`file:${__filename}`).href \n : (document.currentScript && document.currentScript.tagName.toUpperCase() === 'SCRIPT') \n ? document.currentScript.src \n : new URL(\"main.js\", document.baseURI).href;\n\nexport const importMetaUrl = /* @__PURE__ */ getImportMetaUrl()\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\n/**\n * Options for the convert function.\n */\nexport interface ConvertOptions {\n /** Directory where output files are written. Default: input file directory */\n outputDir?: string;\n /** Password for encrypted PDF files */\n password?: string;\n /** Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output. */\n format?: string | string[];\n /** Suppress console logging output */\n quiet?: boolean;\n /** Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background */\n contentSafetyOff?: string | string[];\n /** Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders */\n sanitize?: boolean;\n /** Preserve original line breaks in extracted text */\n keepLineBreaks?: boolean;\n /** Replacement character for invalid/unrecognized characters. Default: space */\n replaceInvalidChars?: string;\n /** Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called */\n useStructTree?: boolean;\n /** Table detection method. Values: default (border-based), cluster (border + cluster). Default: default */\n tableMethod?: string;\n /** Reading order algorithm. Values: off, xycut. Default: xycut */\n readingOrder?: string;\n /** Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none */\n markdownPageSeparator?: string;\n /** Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown. */\n markdownWithHtml?: boolean;\n /** Separator between pages in text output. Use %page-number% for page numbers. Default: none */\n textPageSeparator?: string;\n /** Separator between pages in HTML output. Use %page-number% for page numbers. Default: none */\n htmlPageSeparator?: string;\n /** Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external */\n imageOutput?: string;\n /** Output format for extracted images. Values: png, jpeg. Default: png */\n imageFormat?: string;\n /** Directory for extracted images (applies only with --image-output external) */\n imageDir?: string;\n /** Pages to extract (e.g., \"1,3,5-7\"). Default: all pages */\n pages?: string;\n /** Include page headers and footers in output */\n includeHeaderFooter?: boolean;\n /** Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental) */\n detectStrikethrough?: boolean;\n /** Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence) */\n hybrid?: string;\n /** Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend) */\n hybridMode?: string;\n /** Hybrid backend server URL (overrides default) */\n hybridUrl?: string;\n /** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */\n hybridTimeout?: string;\n /** Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50 */\n hybridChunkSize?: string;\n /** Opt in to Java fallback on hybrid backend error (default: disabled) */\n hybridFallback?: boolean;\n /** Write output to stdout instead of file (single format only) */\n toStdout?: boolean;\n /** Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode */\n threads?: string;\n /** Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0. */\n imageResolution?: string;\n /** Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17 */\n spaceRatio?: string;\n}\n\n/**\n * Options as parsed from CLI (all values are strings from commander).\n */\nexport interface CliOptions {\n outputDir?: string;\n password?: string;\n format?: string;\n quiet?: boolean;\n contentSafetyOff?: string;\n sanitize?: boolean;\n keepLineBreaks?: boolean;\n replaceInvalidChars?: string;\n useStructTree?: boolean;\n tableMethod?: string;\n readingOrder?: string;\n markdownPageSeparator?: string;\n markdownWithHtml?: boolean;\n textPageSeparator?: string;\n htmlPageSeparator?: string;\n imageOutput?: string;\n imageFormat?: string;\n imageDir?: string;\n pages?: string;\n includeHeaderFooter?: boolean;\n detectStrikethrough?: boolean;\n hybrid?: string;\n hybridMode?: string;\n hybridUrl?: string;\n hybridTimeout?: string;\n hybridChunkSize?: string;\n hybridFallback?: boolean;\n toStdout?: boolean;\n threads?: string;\n imageResolution?: string;\n spaceRatio?: string;\n}\n\n/**\n * Convert CLI options to ConvertOptions.\n */\nexport function buildConvertOptions(cliOptions: CliOptions): ConvertOptions {\n const convertOptions: ConvertOptions = {};\n\n if (cliOptions.outputDir) {\n convertOptions.outputDir = cliOptions.outputDir;\n }\n if (cliOptions.password) {\n convertOptions.password = cliOptions.password;\n }\n if (cliOptions.format) {\n convertOptions.format = cliOptions.format;\n }\n if (cliOptions.quiet) {\n convertOptions.quiet = true;\n }\n if (cliOptions.contentSafetyOff) {\n convertOptions.contentSafetyOff = cliOptions.contentSafetyOff;\n }\n if (cliOptions.sanitize) {\n convertOptions.sanitize = true;\n }\n if (cliOptions.keepLineBreaks) {\n convertOptions.keepLineBreaks = true;\n }\n if (cliOptions.replaceInvalidChars) {\n convertOptions.replaceInvalidChars = cliOptions.replaceInvalidChars;\n }\n if (cliOptions.useStructTree) {\n convertOptions.useStructTree = true;\n }\n if (cliOptions.tableMethod) {\n convertOptions.tableMethod = cliOptions.tableMethod;\n }\n if (cliOptions.readingOrder) {\n convertOptions.readingOrder = cliOptions.readingOrder;\n }\n if (cliOptions.markdownPageSeparator) {\n convertOptions.markdownPageSeparator = cliOptions.markdownPageSeparator;\n }\n if (cliOptions.markdownWithHtml) {\n convertOptions.markdownWithHtml = true;\n }\n if (cliOptions.textPageSeparator) {\n convertOptions.textPageSeparator = cliOptions.textPageSeparator;\n }\n if (cliOptions.htmlPageSeparator) {\n convertOptions.htmlPageSeparator = cliOptions.htmlPageSeparator;\n }\n if (cliOptions.imageOutput) {\n convertOptions.imageOutput = cliOptions.imageOutput;\n }\n if (cliOptions.imageFormat) {\n convertOptions.imageFormat = cliOptions.imageFormat;\n }\n if (cliOptions.imageDir) {\n convertOptions.imageDir = cliOptions.imageDir;\n }\n if (cliOptions.pages) {\n convertOptions.pages = cliOptions.pages;\n }\n if (cliOptions.includeHeaderFooter) {\n convertOptions.includeHeaderFooter = true;\n }\n if (cliOptions.detectStrikethrough) {\n convertOptions.detectStrikethrough = true;\n }\n if (cliOptions.hybrid) {\n convertOptions.hybrid = cliOptions.hybrid;\n }\n if (cliOptions.hybridMode) {\n convertOptions.hybridMode = cliOptions.hybridMode;\n }\n if (cliOptions.hybridUrl) {\n convertOptions.hybridUrl = cliOptions.hybridUrl;\n }\n if (cliOptions.hybridTimeout) {\n convertOptions.hybridTimeout = cliOptions.hybridTimeout;\n }\n if (cliOptions.hybridChunkSize) {\n convertOptions.hybridChunkSize = cliOptions.hybridChunkSize;\n }\n if (cliOptions.hybridFallback) {\n convertOptions.hybridFallback = true;\n }\n if (cliOptions.toStdout) {\n convertOptions.toStdout = true;\n }\n if (cliOptions.threads) {\n convertOptions.threads = cliOptions.threads;\n }\n if (cliOptions.imageResolution) {\n convertOptions.imageResolution = cliOptions.imageResolution;\n }\n if (cliOptions.spaceRatio) {\n convertOptions.spaceRatio = cliOptions.spaceRatio;\n }\n\n return convertOptions;\n}\n\n/**\n * Build CLI arguments array from ConvertOptions.\n */\nexport function buildArgs(options: ConvertOptions): string[] {\n const args: string[] = [];\n\n if (options.outputDir) {\n args.push('--output-dir', options.outputDir);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.format) {\n if (Array.isArray(options.format)) {\n if (options.format.length > 0) {\n args.push('--format', options.format.join(','));\n }\n } else {\n args.push('--format', options.format);\n }\n }\n if (options.quiet) {\n args.push('--quiet');\n }\n if (options.contentSafetyOff) {\n if (Array.isArray(options.contentSafetyOff)) {\n if (options.contentSafetyOff.length > 0) {\n args.push('--content-safety-off', options.contentSafetyOff.join(','));\n }\n } else {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n }\n if (options.sanitize) {\n args.push('--sanitize');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.useStructTree) {\n args.push('--use-struct-tree');\n }\n if (options.tableMethod) {\n args.push('--table-method', options.tableMethod);\n }\n if (options.readingOrder) {\n args.push('--reading-order', options.readingOrder);\n }\n if (options.markdownPageSeparator) {\n args.push('--markdown-page-separator', options.markdownPageSeparator);\n }\n if (options.markdownWithHtml) {\n args.push('--markdown-with-html');\n }\n if (options.textPageSeparator) {\n args.push('--text-page-separator', options.textPageSeparator);\n }\n if (options.htmlPageSeparator) {\n args.push('--html-page-separator', options.htmlPageSeparator);\n }\n if (options.imageOutput) {\n args.push('--image-output', options.imageOutput);\n }\n if (options.imageFormat) {\n args.push('--image-format', options.imageFormat);\n }\n if (options.imageDir) {\n args.push('--image-dir', options.imageDir);\n }\n if (options.pages) {\n args.push('--pages', options.pages);\n }\n if (options.includeHeaderFooter) {\n args.push('--include-header-footer');\n }\n if (options.detectStrikethrough) {\n args.push('--detect-strikethrough');\n }\n if (options.hybrid) {\n args.push('--hybrid', options.hybrid);\n }\n if (options.hybridMode) {\n args.push('--hybrid-mode', options.hybridMode);\n }\n if (options.hybridUrl) {\n args.push('--hybrid-url', options.hybridUrl);\n }\n if (options.hybridTimeout) {\n args.push('--hybrid-timeout', options.hybridTimeout);\n }\n if (options.hybridChunkSize) {\n args.push('--hybrid-chunk-size', options.hybridChunkSize);\n }\n if (options.hybridFallback) {\n args.push('--hybrid-fallback');\n }\n if (options.toStdout) {\n args.push('--to-stdout');\n }\n if (options.threads) {\n args.push('--threads', options.threads);\n }\n if (options.imageResolution) {\n args.push('--image-resolution', options.imageResolution);\n }\n if (options.spaceRatio) {\n args.push('--space-ratio', options.spaceRatio);\n }\n\n return args;\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;;;ACKA,IAAM,mBAAmB,MACvB,OAAO,aAAa,cAChB,IAAI,IAAI,QAAQ,UAAU,EAAE,EAAE,OAC7B,SAAS,iBAAiB,SAAS,cAAc,QAAQ,YAAY,MAAM,WAC1E,SAAS,cAAc,MACvB,IAAI,IAAI,WAAW,SAAS,OAAO,EAAE;AAEtC,IAAM,gBAAgC,iCAAiB;;;ADZ9D,2BAAsB;AACtB,WAAsB;AACtB,SAAoB;AACpB,4BAA8B;AAC9B,iBAA8B;;;AEkNvB,SAAS,UAAU,SAAmC;AAC3D,QAAM,OAAiB,CAAC;AAExB,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,EAC1C;AACA,MAAI,QAAQ,QAAQ;AAClB,QAAI,MAAM,QAAQ,QAAQ,MAAM,GAAG;AACjC,UAAI,QAAQ,OAAO,SAAS,GAAG;AAC7B,aAAK,KAAK,YAAY,QAAQ,OAAO,KAAK,GAAG,CAAC;AAAA,MAChD;AAAA,IACF,OAAO;AACL,WAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,IACtC;AAAA,EACF;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,SAAS;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,MAAM,QAAQ,QAAQ,gBAAgB,GAAG;AAC3C,UAAI,QAAQ,iBAAiB,SAAS,GAAG;AACvC,aAAK,KAAK,wBAAwB,QAAQ,iBAAiB,KAAK,GAAG,CAAC;AAAA,MACtE;AAAA,IACF,OAAO;AACL,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AAAA,EACF;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,YAAY;AAAA,EACxB;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,oBAAoB;AAAA,EAChC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,EAClE;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,cAAc;AACxB,SAAK,KAAK,mBAAmB,QAAQ,YAAY;AAAA,EACnD;AACA,MAAI,QAAQ,uBAAuB;AACjC,SAAK,KAAK,6BAA6B,QAAQ,qBAAqB;AAAA,EACtE;AACA,MAAI,QAAQ,kBAAkB;AAC5B,SAAK,KAAK,sBAAsB;AAAA,EAClC;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,eAAe,QAAQ,QAAQ;AAAA,EAC3C;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,WAAW,QAAQ,KAAK;AAAA,EACpC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,yBAAyB;AAAA,EACrC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,wBAAwB;AAAA,EACpC;AACA,MAAI,QAAQ,QAAQ;AAClB,SAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,EACtC;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AACA,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,oBAAoB,QAAQ,aAAa;AAAA,EACrD;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,uBAAuB,QAAQ,eAAe;AAAA,EAC1D;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,aAAa;AAAA,EACzB;AACA,MAAI,QAAQ,SAAS;AACnB,SAAK,KAAK,aAAa,QAAQ,OAAO;AAAA,EACxC;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,sBAAsB,QAAQ,eAAe;AAAA,EACzD;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AAEA,SAAO;AACT;;;AFxTA,IAAMA,kBAAa,0BAAc,aAAe;AAChD,IAAM,YAAiB,aAAQA,WAAU;AAEzC,IAAM,WAAW;AAWjB,SAAS,WAAW,MAAgB,mBAAwC,CAAC,GAAoB;AAC/F,QAAM,EAAE,eAAe,MAAM,IAAI;AAEjC,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,UAAe,UAAK,WAAW,MAAM,OAAO,QAAQ;AAE1D,QAAI,CAAI,cAAW,OAAO,GAAG;AAC3B,aAAO;AAAA,QACL,IAAI,MAAM,yBAAyB,OAAO,sCAAsC;AAAA,MAClF;AAAA,IACF;AAEA,UAAM,UAAU;AAIhB,UAAM,cAAc;AAAA,MAClB;AAAA,MACA;AAAA,MACA;AAAA,MACA;AAAA,MACA,GAAG;AAAA,IACL;AAEA,UAAM,kBAAc,4BAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAKb,UAAM,gBAAgB,IAAI,oCAAc,MAAM;AAC9C,UAAM,gBAAgB,IAAI,oCAAc,MAAM;AAE9C,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAIhB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B,OAAO;AACL,kBAAU;AAAA,MACZ;AAAA,IACF,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAChB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AAGA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAIhC,YAAM,aAAa,cAAc,IAAI;AACrC,YAAM,aAAa,cAAc,IAAI;AACrC,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,cAAc;AAChB,kBAAQ,OAAO,MAAM,UAAU;AAAA,QACjC,OAAO;AACL,oBAAU;AAAA,QACZ;AAAA,MACF;AACA,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,aAAc,SAAQ,OAAO,MAAM,UAAU;AACjD,kBAAU;AAAA,MACZ;AAEA,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,cAAc,UAAU;AAC9B,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,WAAW;AAAA,QACxE;AAKA,QAAC,MAA2C,aAAa;AACzD,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAe;AACtC,UAAI,IAAI,QAAQ,SAAS,QAAQ,GAAG;AAClC;AAAA,UACE,IAAI;AAAA,YACF;AAAA,UACF;AAAA,QACF;AAAA,MACF,OAAO;AACL,eAAO,GAAG;AAAA,MACZ;AAAA,IACF,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,aACP,YACA,SACkB;AAClB,QAAM,YAAY,MAAM,QAAQ,UAAU,IAAI,aAAa,CAAC,UAAU;AACtE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO,IAAI,MAAM,2CAA2C;AAAA,EAC9D;AAEA,aAAW,SAAS,WAAW;AAC7B,QAAI,CAAI,cAAW,KAAK,GAAG;AACzB,aAAO,IAAI,MAAM,mCAAmC,KAAK,EAAE;AAAA,IAC7D;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,WAAW,GAAG,UAAU,OAAO,CAAC;AAC7C;AAEO,SAAS,QACd,YACA,UAA0B,CAAC,GACV;AACjB,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,WAAO,QAAQ,OAAO,WAAW;AAAA,EACnC;AAIA,SAAO,WAAW,aAAa,EAAE,cAAc,MAAM,CAAC;AACxD;AAaA,eAAsB,WACpB,YACA,UAA0B,CAAC,GACZ;AACf,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,UAAM;AAAA,EACR;AACA,QAAM,WAAW,aAAa,EAAE,cAAc,KAAK,CAAC;AACtD;AAwBO,SAAS,IAAI,WAAmB,UAAsB,CAAC,GAAoB;AAChF,UAAQ;AAAA,IACN;AAAA,EACF;AAGA,QAAM,UAAoB,CAAC;AAC3B,MAAI,CAAC,QAAQ,QAAQ;AACnB,YAAQ,KAAK,MAAM;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,QAAQ,oBAAoB;AAC9B,cAAQ,KAAK,sBAAsB;AAAA,IACrC,WAAW,QAAQ,gBAAgB;AACjC,cAAQ,KAAK,oBAAoB;AAAA,IACnC,OAAO;AACL,cAAQ,KAAK,UAAU;AAAA,IACzB;AAAA,EACF;AACA,MAAI,QAAQ,cAAc;AACxB,YAAQ,KAAK,MAAM;AAAA,EACrB;AACA,MAAI,QAAQ,sBAAsB;AAChC,YAAQ,KAAK,KAAK;AAAA,EACpB;AAEA,SAAO,QAAQ,WAAW;AAAA,IACxB,WAAW,QAAQ;AAAA,IACnB,UAAU,QAAQ;AAAA,IAClB,qBAAqB,QAAQ;AAAA,IAC7B,gBAAgB,QAAQ;AAAA,IACxB,kBAAkB,QAAQ;AAAA,IAC1B,eAAe,QAAQ;AAAA,IACvB,QAAQ,QAAQ,SAAS,IAAI,UAAU;AAAA,IACvC,OAAO,CAAC,QAAQ;AAAA,EAClB,CAAC;AACH;","names":["__filename"]}
|
package/dist/index.d.cts
CHANGED
|
@@ -52,6 +52,8 @@ interface ConvertOptions {
|
|
|
52
52
|
hybridUrl?: string;
|
|
53
53
|
/** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */
|
|
54
54
|
hybridTimeout?: string;
|
|
55
|
+
/** Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50 */
|
|
56
|
+
hybridChunkSize?: string;
|
|
55
57
|
/** Opt in to Java fallback on hybrid backend error (default: disabled) */
|
|
56
58
|
hybridFallback?: boolean;
|
|
57
59
|
/** Write output to stdout instead of file (single format only) */
|
package/dist/index.d.ts
CHANGED
|
@@ -52,6 +52,8 @@ interface ConvertOptions {
|
|
|
52
52
|
hybridUrl?: string;
|
|
53
53
|
/** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */
|
|
54
54
|
hybridTimeout?: string;
|
|
55
|
+
/** Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50 */
|
|
56
|
+
hybridChunkSize?: string;
|
|
55
57
|
/** Opt in to Java fallback on hybrid backend error (default: disabled) */
|
|
56
58
|
hybridFallback?: boolean;
|
|
57
59
|
/** Write output to stdout instead of file (single format only) */
|
package/dist/index.js
CHANGED
|
@@ -95,6 +95,9 @@ function buildArgs(options) {
|
|
|
95
95
|
if (options.hybridTimeout) {
|
|
96
96
|
args.push("--hybrid-timeout", options.hybridTimeout);
|
|
97
97
|
}
|
|
98
|
+
if (options.hybridChunkSize) {
|
|
99
|
+
args.push("--hybrid-chunk-size", options.hybridChunkSize);
|
|
100
|
+
}
|
|
98
101
|
if (options.hybridFallback) {
|
|
99
102
|
args.push("--hybrid-fallback");
|
|
100
103
|
}
|
package/dist/index.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"sources":["../src/index.ts","../src/convert-options.generated.ts"],"sourcesContent":["import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { StringDecoder } from 'string_decoder';\nimport { fileURLToPath } from 'url';\n\n// Re-export types and utilities from auto-generated file\nexport type { ConvertOptions } from './convert-options.generated.js';\nexport { buildArgs } from './convert-options.generated.js';\nimport type { ConvertOptions } from './convert-options.generated.js';\nimport { buildArgs } from './convert-options.generated.js';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\ninterface JarExecutionOptions {\n /**\n * When true, forwards Java's stdout and stderr chunks to the parent\n * process in real time as well as accumulating them. Used by the bundled\n * CLI so long-running conversions show progress as it happens.\n */\n streamOutput?: boolean;\n}\n\nfunction executeJar(args: string[], executionOptions: JarExecutionOptions = {}): Promise<string> {\n const { streamOutput = false } = executionOptions;\n\n return new Promise((resolve, reject) => {\n const jarPath = path.join(__dirname, '..', 'lib', JAR_NAME);\n\n if (!fs.existsSync(jarPath)) {\n return reject(\n new Error(`JAR file not found at ${jarPath}. Please run the build script first.`),\n );\n }\n\n const command = 'java';\n // Force headless AWT so macOS doesn't surface a Dock icon (and steal focus)\n // every time the JVM touches ImageIO/PDFBox rendering. Safe on all OSes —\n // the CLI never opens a UI window, only manipulates BufferedImages.\n const commandArgs = [\n '-Djava.awt.headless=true',\n '-Dapple.awt.UIElement=true',\n '-jar',\n jarPath,\n ...args,\n ];\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n // StringDecoder buffers incomplete multi-byte UTF-8 sequences across\n // chunk boundaries — Buffer.toString() alone would emit replacement\n // characters when, e.g., a 3-byte Korean codepoint splits across two\n // 'data' events. One decoder per stream so they don't share state.\n const stdoutDecoder = new StringDecoder('utf8');\n const stderrDecoder = new StringDecoder('utf8');\n\n javaProcess.stdout.on('data', (data: Buffer) => {\n const chunk = stdoutDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n // Stream-only: don't also accumulate, or a multi-hour conversion\n // would buffer its entire (potentially gigabyte) stdout in memory\n // for no consumer.\n process.stdout.write(chunk);\n } else {\n stdout += chunk;\n }\n });\n\n javaProcess.stderr.on('data', (data: Buffer) => {\n const chunk = stderrDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n process.stderr.write(chunk);\n }\n // stderr is always accumulated (progress logs are small and we need\n // them for error messages on non-zero exit).\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n // Flush any trailing bytes the decoder is still holding (always emit\n // them — if we drop them on error paths, error messages with non-ASCII\n // characters lose their tail).\n const stdoutTail = stdoutDecoder.end();\n const stderrTail = stderrDecoder.end();\n if (stdoutTail.length > 0) {\n if (streamOutput) {\n process.stdout.write(stdoutTail);\n } else {\n stdout += stdoutTail;\n }\n }\n if (stderrTail.length > 0) {\n if (streamOutput) process.stderr.write(stderrTail);\n stderr += stderrTail;\n }\n\n if (code === 0) {\n resolve(stdout);\n } else {\n const errorOutput = stderr || stdout;\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${errorOutput}`,\n );\n // Tag so the CLI can suppress re-printing this message — Java's\n // stderr was already streamed live to the parent in CLI mode, and\n // re-printing risks leaking anything sensitive Java logged\n // (e.g. a --password value echoed by an underlying library).\n (error as Error & { isJavaExit?: boolean }).isJavaExit = true;\n reject(error);\n }\n });\n\n javaProcess.on('error', (err: Error) => {\n if (err.message.includes('ENOENT')) {\n reject(\n new Error(\n \"'java' command not found. Please ensure Java is installed and in your system's PATH.\",\n ),\n );\n } else {\n reject(err);\n }\n });\n });\n}\n\nfunction buildJarArgs(\n inputPaths: string | string[],\n options: ConvertOptions,\n): string[] | Error {\n const inputList = Array.isArray(inputPaths) ? inputPaths : [inputPaths];\n if (inputList.length === 0) {\n return new Error('At least one input path must be provided.');\n }\n\n for (const input of inputList) {\n if (!fs.existsSync(input)) {\n return new Error(`Input file or folder not found: ${input}`);\n }\n }\n\n return [...inputList, ...buildArgs(options)];\n}\n\nexport function convert(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<string> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n return Promise.reject(argsOrError);\n }\n // Library API: never streams to the parent process. Returns the full stdout\n // string so callers can do `const out = await convert(...)` without surprise\n // side-effects on process.stdout / process.stderr.\n return executeJar(argsOrError, { streamOutput: false });\n}\n\n/**\n * Internal entry point used by the bundled CLI. Streams Java's stdout and\n * stderr to the parent process in real time (so long-running conversions like\n * hybrid mode show progress as it happens) and resolves without a stdout\n * payload — preventing the caller from re-printing what was already streamed.\n *\n * Not part of the public API: do not import this from application code. Use\n * {@link convert} instead.\n *\n * @internal\n */\nexport async function _runForCli(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<void> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n throw argsOrError;\n }\n await executeJar(argsOrError, { streamOutput: true });\n}\n\n/**\n * @deprecated Use `convert()` and `ConvertOptions` instead. This function will be removed in a future version.\n */\nexport interface RunOptions {\n outputFolder?: string;\n password?: string;\n replaceInvalidChars?: string;\n generateMarkdown?: boolean;\n generateHtml?: boolean;\n generateAnnotatedPdf?: boolean;\n keepLineBreaks?: boolean;\n contentSafetyOff?: string;\n htmlInMarkdown?: boolean;\n addImageToMarkdown?: boolean;\n noJson?: boolean;\n debug?: boolean;\n useStructTree?: boolean;\n}\n\n/**\n * @deprecated Use `convert()` instead. This function will be removed in a future version.\n */\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n console.warn(\n 'Warning: run() is deprecated and will be removed in a future version. Use convert() instead.',\n );\n\n // Build format array based on legacy boolean options\n const formats: string[] = [];\n if (!options.noJson) {\n formats.push('json');\n }\n if (options.generateMarkdown) {\n if (options.addImageToMarkdown) {\n formats.push('markdown-with-images');\n } else if (options.htmlInMarkdown) {\n formats.push('markdown-with-html');\n } else {\n formats.push('markdown');\n }\n }\n if (options.generateHtml) {\n formats.push('html');\n }\n if (options.generateAnnotatedPdf) {\n formats.push('pdf');\n }\n\n return convert(inputPath, {\n outputDir: options.outputFolder,\n password: options.password,\n replaceInvalidChars: options.replaceInvalidChars,\n keepLineBreaks: options.keepLineBreaks,\n contentSafetyOff: options.contentSafetyOff,\n useStructTree: options.useStructTree,\n format: formats.length > 0 ? formats : undefined,\n quiet: !options.debug,\n });\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\n/**\n * Options for the convert function.\n */\nexport interface ConvertOptions {\n /** Directory where output files are written. Default: input file directory */\n outputDir?: string;\n /** Password for encrypted PDF files */\n password?: string;\n /** Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output. */\n format?: string | string[];\n /** Suppress console logging output */\n quiet?: boolean;\n /** Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background */\n contentSafetyOff?: string | string[];\n /** Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders */\n sanitize?: boolean;\n /** Preserve original line breaks in extracted text */\n keepLineBreaks?: boolean;\n /** Replacement character for invalid/unrecognized characters. Default: space */\n replaceInvalidChars?: string;\n /** Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called */\n useStructTree?: boolean;\n /** Table detection method. Values: default (border-based), cluster (border + cluster). Default: default */\n tableMethod?: string;\n /** Reading order algorithm. Values: off, xycut. Default: xycut */\n readingOrder?: string;\n /** Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none */\n markdownPageSeparator?: string;\n /** Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown. */\n markdownWithHtml?: boolean;\n /** Separator between pages in text output. Use %page-number% for page numbers. Default: none */\n textPageSeparator?: string;\n /** Separator between pages in HTML output. Use %page-number% for page numbers. Default: none */\n htmlPageSeparator?: string;\n /** Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external */\n imageOutput?: string;\n /** Output format for extracted images. Values: png, jpeg. Default: png */\n imageFormat?: string;\n /** Directory for extracted images (applies only with --image-output external) */\n imageDir?: string;\n /** Pages to extract (e.g., \"1,3,5-7\"). Default: all pages */\n pages?: string;\n /** Include page headers and footers in output */\n includeHeaderFooter?: boolean;\n /** Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental) */\n detectStrikethrough?: boolean;\n /** Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence) */\n hybrid?: string;\n /** Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend) */\n hybridMode?: string;\n /** Hybrid backend server URL (overrides default) */\n hybridUrl?: string;\n /** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */\n hybridTimeout?: string;\n /** Opt in to Java fallback on hybrid backend error (default: disabled) */\n hybridFallback?: boolean;\n /** Write output to stdout instead of file (single format only) */\n toStdout?: boolean;\n /** Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode */\n threads?: string;\n /** Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0. */\n imageResolution?: string;\n /** Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17 */\n spaceRatio?: string;\n}\n\n/**\n * Options as parsed from CLI (all values are strings from commander).\n */\nexport interface CliOptions {\n outputDir?: string;\n password?: string;\n format?: string;\n quiet?: boolean;\n contentSafetyOff?: string;\n sanitize?: boolean;\n keepLineBreaks?: boolean;\n replaceInvalidChars?: string;\n useStructTree?: boolean;\n tableMethod?: string;\n readingOrder?: string;\n markdownPageSeparator?: string;\n markdownWithHtml?: boolean;\n textPageSeparator?: string;\n htmlPageSeparator?: string;\n imageOutput?: string;\n imageFormat?: string;\n imageDir?: string;\n pages?: string;\n includeHeaderFooter?: boolean;\n detectStrikethrough?: boolean;\n hybrid?: string;\n hybridMode?: string;\n hybridUrl?: string;\n hybridTimeout?: string;\n hybridFallback?: boolean;\n toStdout?: boolean;\n threads?: string;\n imageResolution?: string;\n spaceRatio?: string;\n}\n\n/**\n * Convert CLI options to ConvertOptions.\n */\nexport function buildConvertOptions(cliOptions: CliOptions): ConvertOptions {\n const convertOptions: ConvertOptions = {};\n\n if (cliOptions.outputDir) {\n convertOptions.outputDir = cliOptions.outputDir;\n }\n if (cliOptions.password) {\n convertOptions.password = cliOptions.password;\n }\n if (cliOptions.format) {\n convertOptions.format = cliOptions.format;\n }\n if (cliOptions.quiet) {\n convertOptions.quiet = true;\n }\n if (cliOptions.contentSafetyOff) {\n convertOptions.contentSafetyOff = cliOptions.contentSafetyOff;\n }\n if (cliOptions.sanitize) {\n convertOptions.sanitize = true;\n }\n if (cliOptions.keepLineBreaks) {\n convertOptions.keepLineBreaks = true;\n }\n if (cliOptions.replaceInvalidChars) {\n convertOptions.replaceInvalidChars = cliOptions.replaceInvalidChars;\n }\n if (cliOptions.useStructTree) {\n convertOptions.useStructTree = true;\n }\n if (cliOptions.tableMethod) {\n convertOptions.tableMethod = cliOptions.tableMethod;\n }\n if (cliOptions.readingOrder) {\n convertOptions.readingOrder = cliOptions.readingOrder;\n }\n if (cliOptions.markdownPageSeparator) {\n convertOptions.markdownPageSeparator = cliOptions.markdownPageSeparator;\n }\n if (cliOptions.markdownWithHtml) {\n convertOptions.markdownWithHtml = true;\n }\n if (cliOptions.textPageSeparator) {\n convertOptions.textPageSeparator = cliOptions.textPageSeparator;\n }\n if (cliOptions.htmlPageSeparator) {\n convertOptions.htmlPageSeparator = cliOptions.htmlPageSeparator;\n }\n if (cliOptions.imageOutput) {\n convertOptions.imageOutput = cliOptions.imageOutput;\n }\n if (cliOptions.imageFormat) {\n convertOptions.imageFormat = cliOptions.imageFormat;\n }\n if (cliOptions.imageDir) {\n convertOptions.imageDir = cliOptions.imageDir;\n }\n if (cliOptions.pages) {\n convertOptions.pages = cliOptions.pages;\n }\n if (cliOptions.includeHeaderFooter) {\n convertOptions.includeHeaderFooter = true;\n }\n if (cliOptions.detectStrikethrough) {\n convertOptions.detectStrikethrough = true;\n }\n if (cliOptions.hybrid) {\n convertOptions.hybrid = cliOptions.hybrid;\n }\n if (cliOptions.hybridMode) {\n convertOptions.hybridMode = cliOptions.hybridMode;\n }\n if (cliOptions.hybridUrl) {\n convertOptions.hybridUrl = cliOptions.hybridUrl;\n }\n if (cliOptions.hybridTimeout) {\n convertOptions.hybridTimeout = cliOptions.hybridTimeout;\n }\n if (cliOptions.hybridFallback) {\n convertOptions.hybridFallback = true;\n }\n if (cliOptions.toStdout) {\n convertOptions.toStdout = true;\n }\n if (cliOptions.threads) {\n convertOptions.threads = cliOptions.threads;\n }\n if (cliOptions.imageResolution) {\n convertOptions.imageResolution = cliOptions.imageResolution;\n }\n if (cliOptions.spaceRatio) {\n convertOptions.spaceRatio = cliOptions.spaceRatio;\n }\n\n return convertOptions;\n}\n\n/**\n * Build CLI arguments array from ConvertOptions.\n */\nexport function buildArgs(options: ConvertOptions): string[] {\n const args: string[] = [];\n\n if (options.outputDir) {\n args.push('--output-dir', options.outputDir);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.format) {\n if (Array.isArray(options.format)) {\n if (options.format.length > 0) {\n args.push('--format', options.format.join(','));\n }\n } else {\n args.push('--format', options.format);\n }\n }\n if (options.quiet) {\n args.push('--quiet');\n }\n if (options.contentSafetyOff) {\n if (Array.isArray(options.contentSafetyOff)) {\n if (options.contentSafetyOff.length > 0) {\n args.push('--content-safety-off', options.contentSafetyOff.join(','));\n }\n } else {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n }\n if (options.sanitize) {\n args.push('--sanitize');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.useStructTree) {\n args.push('--use-struct-tree');\n }\n if (options.tableMethod) {\n args.push('--table-method', options.tableMethod);\n }\n if (options.readingOrder) {\n args.push('--reading-order', options.readingOrder);\n }\n if (options.markdownPageSeparator) {\n args.push('--markdown-page-separator', options.markdownPageSeparator);\n }\n if (options.markdownWithHtml) {\n args.push('--markdown-with-html');\n }\n if (options.textPageSeparator) {\n args.push('--text-page-separator', options.textPageSeparator);\n }\n if (options.htmlPageSeparator) {\n args.push('--html-page-separator', options.htmlPageSeparator);\n }\n if (options.imageOutput) {\n args.push('--image-output', options.imageOutput);\n }\n if (options.imageFormat) {\n args.push('--image-format', options.imageFormat);\n }\n if (options.imageDir) {\n args.push('--image-dir', options.imageDir);\n }\n if (options.pages) {\n args.push('--pages', options.pages);\n }\n if (options.includeHeaderFooter) {\n args.push('--include-header-footer');\n }\n if (options.detectStrikethrough) {\n args.push('--detect-strikethrough');\n }\n if (options.hybrid) {\n args.push('--hybrid', options.hybrid);\n }\n if (options.hybridMode) {\n args.push('--hybrid-mode', options.hybridMode);\n }\n if (options.hybridUrl) {\n args.push('--hybrid-url', options.hybridUrl);\n }\n if (options.hybridTimeout) {\n args.push('--hybrid-timeout', options.hybridTimeout);\n }\n if (options.hybridFallback) {\n args.push('--hybrid-fallback');\n }\n if (options.toStdout) {\n args.push('--to-stdout');\n }\n if (options.threads) {\n args.push('--threads', options.threads);\n }\n if (options.imageResolution) {\n args.push('--image-resolution', options.imageResolution);\n }\n if (options.spaceRatio) {\n args.push('--space-ratio', options.spaceRatio);\n }\n\n return args;\n}\n"],"mappings":";AAAA,SAAS,aAAa;AACtB,YAAY,UAAU;AACtB,YAAY,QAAQ;AACpB,SAAS,qBAAqB;AAC9B,SAAS,qBAAqB;;;AC4MvB,SAAS,UAAU,SAAmC;AAC3D,QAAM,OAAiB,CAAC;AAExB,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,EAC1C;AACA,MAAI,QAAQ,QAAQ;AAClB,QAAI,MAAM,QAAQ,QAAQ,MAAM,GAAG;AACjC,UAAI,QAAQ,OAAO,SAAS,GAAG;AAC7B,aAAK,KAAK,YAAY,QAAQ,OAAO,KAAK,GAAG,CAAC;AAAA,MAChD;AAAA,IACF,OAAO;AACL,WAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,IACtC;AAAA,EACF;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,SAAS;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,MAAM,QAAQ,QAAQ,gBAAgB,GAAG;AAC3C,UAAI,QAAQ,iBAAiB,SAAS,GAAG;AACvC,aAAK,KAAK,wBAAwB,QAAQ,iBAAiB,KAAK,GAAG,CAAC;AAAA,MACtE;AAAA,IACF,OAAO;AACL,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AAAA,EACF;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,YAAY;AAAA,EACxB;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,oBAAoB;AAAA,EAChC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,EAClE;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,cAAc;AACxB,SAAK,KAAK,mBAAmB,QAAQ,YAAY;AAAA,EACnD;AACA,MAAI,QAAQ,uBAAuB;AACjC,SAAK,KAAK,6BAA6B,QAAQ,qBAAqB;AAAA,EACtE;AACA,MAAI,QAAQ,kBAAkB;AAC5B,SAAK,KAAK,sBAAsB;AAAA,EAClC;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,eAAe,QAAQ,QAAQ;AAAA,EAC3C;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,WAAW,QAAQ,KAAK;AAAA,EACpC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,yBAAyB;AAAA,EACrC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,wBAAwB;AAAA,EACpC;AACA,MAAI,QAAQ,QAAQ;AAClB,SAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,EACtC;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AACA,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,oBAAoB,QAAQ,aAAa;AAAA,EACrD;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,aAAa;AAAA,EACzB;AACA,MAAI,QAAQ,SAAS;AACnB,SAAK,KAAK,aAAa,QAAQ,OAAO;AAAA,EACxC;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,sBAAsB,QAAQ,eAAe;AAAA,EACzD;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AAEA,SAAO;AACT;;;AD/SA,IAAMA,cAAa,cAAc,YAAY,GAAG;AAChD,IAAMC,aAAiB,aAAQD,WAAU;AAEzC,IAAM,WAAW;AAWjB,SAAS,WAAW,MAAgB,mBAAwC,CAAC,GAAoB;AAC/F,QAAM,EAAE,eAAe,MAAM,IAAI;AAEjC,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,UAAe,UAAKC,YAAW,MAAM,OAAO,QAAQ;AAE1D,QAAI,CAAI,cAAW,OAAO,GAAG;AAC3B,aAAO;AAAA,QACL,IAAI,MAAM,yBAAyB,OAAO,sCAAsC;AAAA,MAClF;AAAA,IACF;AAEA,UAAM,UAAU;AAIhB,UAAM,cAAc;AAAA,MAClB;AAAA,MACA;AAAA,MACA;AAAA,MACA;AAAA,MACA,GAAG;AAAA,IACL;AAEA,UAAM,cAAc,MAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAKb,UAAM,gBAAgB,IAAI,cAAc,MAAM;AAC9C,UAAM,gBAAgB,IAAI,cAAc,MAAM;AAE9C,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAIhB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B,OAAO;AACL,kBAAU;AAAA,MACZ;AAAA,IACF,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAChB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AAGA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAIhC,YAAM,aAAa,cAAc,IAAI;AACrC,YAAM,aAAa,cAAc,IAAI;AACrC,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,cAAc;AAChB,kBAAQ,OAAO,MAAM,UAAU;AAAA,QACjC,OAAO;AACL,oBAAU;AAAA,QACZ;AAAA,MACF;AACA,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,aAAc,SAAQ,OAAO,MAAM,UAAU;AACjD,kBAAU;AAAA,MACZ;AAEA,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,cAAc,UAAU;AAC9B,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,WAAW;AAAA,QACxE;AAKA,QAAC,MAA2C,aAAa;AACzD,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAe;AACtC,UAAI,IAAI,QAAQ,SAAS,QAAQ,GAAG;AAClC;AAAA,UACE,IAAI;AAAA,YACF;AAAA,UACF;AAAA,QACF;AAAA,MACF,OAAO;AACL,eAAO,GAAG;AAAA,MACZ;AAAA,IACF,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,aACP,YACA,SACkB;AAClB,QAAM,YAAY,MAAM,QAAQ,UAAU,IAAI,aAAa,CAAC,UAAU;AACtE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO,IAAI,MAAM,2CAA2C;AAAA,EAC9D;AAEA,aAAW,SAAS,WAAW;AAC7B,QAAI,CAAI,cAAW,KAAK,GAAG;AACzB,aAAO,IAAI,MAAM,mCAAmC,KAAK,EAAE;AAAA,IAC7D;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,WAAW,GAAG,UAAU,OAAO,CAAC;AAC7C;AAEO,SAAS,QACd,YACA,UAA0B,CAAC,GACV;AACjB,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,WAAO,QAAQ,OAAO,WAAW;AAAA,EACnC;AAIA,SAAO,WAAW,aAAa,EAAE,cAAc,MAAM,CAAC;AACxD;AAaA,eAAsB,WACpB,YACA,UAA0B,CAAC,GACZ;AACf,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,UAAM;AAAA,EACR;AACA,QAAM,WAAW,aAAa,EAAE,cAAc,KAAK,CAAC;AACtD;AAwBO,SAAS,IAAI,WAAmB,UAAsB,CAAC,GAAoB;AAChF,UAAQ;AAAA,IACN;AAAA,EACF;AAGA,QAAM,UAAoB,CAAC;AAC3B,MAAI,CAAC,QAAQ,QAAQ;AACnB,YAAQ,KAAK,MAAM;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,QAAQ,oBAAoB;AAC9B,cAAQ,KAAK,sBAAsB;AAAA,IACrC,WAAW,QAAQ,gBAAgB;AACjC,cAAQ,KAAK,oBAAoB;AAAA,IACnC,OAAO;AACL,cAAQ,KAAK,UAAU;AAAA,IACzB;AAAA,EACF;AACA,MAAI,QAAQ,cAAc;AACxB,YAAQ,KAAK,MAAM;AAAA,EACrB;AACA,MAAI,QAAQ,sBAAsB;AAChC,YAAQ,KAAK,KAAK;AAAA,EACpB;AAEA,SAAO,QAAQ,WAAW;AAAA,IACxB,WAAW,QAAQ;AAAA,IACnB,UAAU,QAAQ;AAAA,IAClB,qBAAqB,QAAQ;AAAA,IAC7B,gBAAgB,QAAQ;AAAA,IACxB,kBAAkB,QAAQ;AAAA,IAC1B,eAAe,QAAQ;AAAA,IACvB,QAAQ,QAAQ,SAAS,IAAI,UAAU;AAAA,IACvC,OAAO,CAAC,QAAQ;AAAA,EAClB,CAAC;AACH;","names":["__filename","__dirname"]}
|
|
1
|
+
{"version":3,"sources":["../src/index.ts","../src/convert-options.generated.ts"],"sourcesContent":["import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { StringDecoder } from 'string_decoder';\nimport { fileURLToPath } from 'url';\n\n// Re-export types and utilities from auto-generated file\nexport type { ConvertOptions } from './convert-options.generated.js';\nexport { buildArgs } from './convert-options.generated.js';\nimport type { ConvertOptions } from './convert-options.generated.js';\nimport { buildArgs } from './convert-options.generated.js';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\ninterface JarExecutionOptions {\n /**\n * When true, forwards Java's stdout and stderr chunks to the parent\n * process in real time as well as accumulating them. Used by the bundled\n * CLI so long-running conversions show progress as it happens.\n */\n streamOutput?: boolean;\n}\n\nfunction executeJar(args: string[], executionOptions: JarExecutionOptions = {}): Promise<string> {\n const { streamOutput = false } = executionOptions;\n\n return new Promise((resolve, reject) => {\n const jarPath = path.join(__dirname, '..', 'lib', JAR_NAME);\n\n if (!fs.existsSync(jarPath)) {\n return reject(\n new Error(`JAR file not found at ${jarPath}. Please run the build script first.`),\n );\n }\n\n const command = 'java';\n // Force headless AWT so macOS doesn't surface a Dock icon (and steal focus)\n // every time the JVM touches ImageIO/PDFBox rendering. Safe on all OSes —\n // the CLI never opens a UI window, only manipulates BufferedImages.\n const commandArgs = [\n '-Djava.awt.headless=true',\n '-Dapple.awt.UIElement=true',\n '-jar',\n jarPath,\n ...args,\n ];\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n // StringDecoder buffers incomplete multi-byte UTF-8 sequences across\n // chunk boundaries — Buffer.toString() alone would emit replacement\n // characters when, e.g., a 3-byte Korean codepoint splits across two\n // 'data' events. One decoder per stream so they don't share state.\n const stdoutDecoder = new StringDecoder('utf8');\n const stderrDecoder = new StringDecoder('utf8');\n\n javaProcess.stdout.on('data', (data: Buffer) => {\n const chunk = stdoutDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n // Stream-only: don't also accumulate, or a multi-hour conversion\n // would buffer its entire (potentially gigabyte) stdout in memory\n // for no consumer.\n process.stdout.write(chunk);\n } else {\n stdout += chunk;\n }\n });\n\n javaProcess.stderr.on('data', (data: Buffer) => {\n const chunk = stderrDecoder.write(data);\n if (chunk.length === 0) return;\n if (streamOutput) {\n process.stderr.write(chunk);\n }\n // stderr is always accumulated (progress logs are small and we need\n // them for error messages on non-zero exit).\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n // Flush any trailing bytes the decoder is still holding (always emit\n // them — if we drop them on error paths, error messages with non-ASCII\n // characters lose their tail).\n const stdoutTail = stdoutDecoder.end();\n const stderrTail = stderrDecoder.end();\n if (stdoutTail.length > 0) {\n if (streamOutput) {\n process.stdout.write(stdoutTail);\n } else {\n stdout += stdoutTail;\n }\n }\n if (stderrTail.length > 0) {\n if (streamOutput) process.stderr.write(stderrTail);\n stderr += stderrTail;\n }\n\n if (code === 0) {\n resolve(stdout);\n } else {\n const errorOutput = stderr || stdout;\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${errorOutput}`,\n );\n // Tag so the CLI can suppress re-printing this message — Java's\n // stderr was already streamed live to the parent in CLI mode, and\n // re-printing risks leaking anything sensitive Java logged\n // (e.g. a --password value echoed by an underlying library).\n (error as Error & { isJavaExit?: boolean }).isJavaExit = true;\n reject(error);\n }\n });\n\n javaProcess.on('error', (err: Error) => {\n if (err.message.includes('ENOENT')) {\n reject(\n new Error(\n \"'java' command not found. Please ensure Java is installed and in your system's PATH.\",\n ),\n );\n } else {\n reject(err);\n }\n });\n });\n}\n\nfunction buildJarArgs(\n inputPaths: string | string[],\n options: ConvertOptions,\n): string[] | Error {\n const inputList = Array.isArray(inputPaths) ? inputPaths : [inputPaths];\n if (inputList.length === 0) {\n return new Error('At least one input path must be provided.');\n }\n\n for (const input of inputList) {\n if (!fs.existsSync(input)) {\n return new Error(`Input file or folder not found: ${input}`);\n }\n }\n\n return [...inputList, ...buildArgs(options)];\n}\n\nexport function convert(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<string> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n return Promise.reject(argsOrError);\n }\n // Library API: never streams to the parent process. Returns the full stdout\n // string so callers can do `const out = await convert(...)` without surprise\n // side-effects on process.stdout / process.stderr.\n return executeJar(argsOrError, { streamOutput: false });\n}\n\n/**\n * Internal entry point used by the bundled CLI. Streams Java's stdout and\n * stderr to the parent process in real time (so long-running conversions like\n * hybrid mode show progress as it happens) and resolves without a stdout\n * payload — preventing the caller from re-printing what was already streamed.\n *\n * Not part of the public API: do not import this from application code. Use\n * {@link convert} instead.\n *\n * @internal\n */\nexport async function _runForCli(\n inputPaths: string | string[],\n options: ConvertOptions = {},\n): Promise<void> {\n const argsOrError = buildJarArgs(inputPaths, options);\n if (argsOrError instanceof Error) {\n throw argsOrError;\n }\n await executeJar(argsOrError, { streamOutput: true });\n}\n\n/**\n * @deprecated Use `convert()` and `ConvertOptions` instead. This function will be removed in a future version.\n */\nexport interface RunOptions {\n outputFolder?: string;\n password?: string;\n replaceInvalidChars?: string;\n generateMarkdown?: boolean;\n generateHtml?: boolean;\n generateAnnotatedPdf?: boolean;\n keepLineBreaks?: boolean;\n contentSafetyOff?: string;\n htmlInMarkdown?: boolean;\n addImageToMarkdown?: boolean;\n noJson?: boolean;\n debug?: boolean;\n useStructTree?: boolean;\n}\n\n/**\n * @deprecated Use `convert()` instead. This function will be removed in a future version.\n */\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n console.warn(\n 'Warning: run() is deprecated and will be removed in a future version. Use convert() instead.',\n );\n\n // Build format array based on legacy boolean options\n const formats: string[] = [];\n if (!options.noJson) {\n formats.push('json');\n }\n if (options.generateMarkdown) {\n if (options.addImageToMarkdown) {\n formats.push('markdown-with-images');\n } else if (options.htmlInMarkdown) {\n formats.push('markdown-with-html');\n } else {\n formats.push('markdown');\n }\n }\n if (options.generateHtml) {\n formats.push('html');\n }\n if (options.generateAnnotatedPdf) {\n formats.push('pdf');\n }\n\n return convert(inputPath, {\n outputDir: options.outputFolder,\n password: options.password,\n replaceInvalidChars: options.replaceInvalidChars,\n keepLineBreaks: options.keepLineBreaks,\n contentSafetyOff: options.contentSafetyOff,\n useStructTree: options.useStructTree,\n format: formats.length > 0 ? formats : undefined,\n quiet: !options.debug,\n });\n}\n","// AUTO-GENERATED FROM options.json - DO NOT EDIT DIRECTLY\n// Run `npm run generate-options` to regenerate\n\n/**\n * Options for the convert function.\n */\nexport interface ConvertOptions {\n /** Directory where output files are written. Default: input file directory */\n outputDir?: string;\n /** Password for encrypted PDF files */\n password?: string;\n /** Output formats (comma-separated). Values: json, text, html, pdf, markdown, tagged-pdf. Default: json. For HTML inside Markdown use --markdown-with-html. For image extraction control use --image-output. */\n format?: string | string[];\n /** Suppress console logging output */\n quiet?: boolean;\n /** Disable content safety filters. Values: all, hidden-text, off-page, tiny, hidden-ocg, background */\n contentSafetyOff?: string | string[];\n /** Enable sensitive data sanitization. Replaces emails, phone numbers, IPs, credit cards, and URLs with placeholders */\n sanitize?: boolean;\n /** Preserve original line breaks in extracted text */\n keepLineBreaks?: boolean;\n /** Replacement character for invalid/unrecognized characters. Default: space */\n replaceInvalidChars?: string;\n /** Use PDF structure tree (tagged PDF) for reading order and semantic structure. Output quality depends on tag quality. Takes precedence over --hybrid: when both are set on a tagged PDF, the structure tree is used and the hybrid backend is not called */\n useStructTree?: boolean;\n /** Table detection method. Values: default (border-based), cluster (border + cluster). Default: default */\n tableMethod?: string;\n /** Reading order algorithm. Values: off, xycut. Default: xycut */\n readingOrder?: string;\n /** Separator between pages in Markdown output. Use %page-number% for page numbers. Default: none */\n markdownPageSeparator?: string;\n /** Allow HTML tags inside Markdown output for complex structures such as multi-row-span tables. Implies --format markdown. */\n markdownWithHtml?: boolean;\n /** Separator between pages in text output. Use %page-number% for page numbers. Default: none */\n textPageSeparator?: string;\n /** Separator between pages in HTML output. Use %page-number% for page numbers. Default: none */\n htmlPageSeparator?: string;\n /** Image output mode. Values: off (no images), embedded (Base64 data URIs), external (file references). Default: external */\n imageOutput?: string;\n /** Output format for extracted images. Values: png, jpeg. Default: png */\n imageFormat?: string;\n /** Directory for extracted images (applies only with --image-output external) */\n imageDir?: string;\n /** Pages to extract (e.g., \"1,3,5-7\"). Default: all pages */\n pages?: string;\n /** Include page headers and footers in output */\n includeHeaderFooter?: boolean;\n /** Detect strikethrough text and wrap with ~~ in Markdown output or <del></del> tag in HTML output (experimental) */\n detectStrikethrough?: boolean;\n /** Hybrid backend (requires a running server). Quick start: pip install \"opendataloader-pdf[hybrid]\" && opendataloader-pdf-hybrid --port 5002. For remote servers use --hybrid-url. Values: off (default), docling-fast. Ignored when --use-struct-tree is set on a tagged PDF (structure tree takes precedence) */\n hybrid?: string;\n /** Hybrid triage mode. Values: auto (default, dynamic triage), full (skip triage, all pages to backend) */\n hybridMode?: string;\n /** Hybrid backend server URL (overrides default) */\n hybridUrl?: string;\n /** Hybrid backend request timeout in milliseconds (0 = use the backend's own default). Default: 0 */\n hybridTimeout?: string;\n /** Maximum number of pages to send to the hybrid backend in a single request. Large documents are split into batches of this size; smaller values increase the number of backend requests (the full PDF is re-sent per batch). Default: 50 */\n hybridChunkSize?: string;\n /** Opt in to Java fallback on hybrid backend error (default: disabled) */\n hybridFallback?: boolean;\n /** Write output to stdout instead of file (single format only) */\n toStdout?: boolean;\n /** Number of worker threads for per-page processing. Default: 1 (sequential, stable). Values >1 (experimental) run pages in parallel for faster throughput; output may vary slightly on some PDFs. Capped at the number of available CPU cores. Applies to the native Java pipeline only; ignored in --hybrid mode */\n threads?: string;\n /** Set the rendering resolution for images in DPI. Higher values improve image quality but increase memory consumption; lower values reduce memory usage at the cost of detail. Accepts positive decimal DPI values (e.g., 144.0). Default: 144.0. */\n imageResolution?: string;\n /** Set the ratio used to calculate the automatic space-insertion threshold (threshold = space-ratio * font size). If the horizontal gap between two adjacent symbols exceeds this threshold, an extra space is inserted to text value. Accepts decimals (e.g., 0.17). Default: 0.17 */\n spaceRatio?: string;\n}\n\n/**\n * Options as parsed from CLI (all values are strings from commander).\n */\nexport interface CliOptions {\n outputDir?: string;\n password?: string;\n format?: string;\n quiet?: boolean;\n contentSafetyOff?: string;\n sanitize?: boolean;\n keepLineBreaks?: boolean;\n replaceInvalidChars?: string;\n useStructTree?: boolean;\n tableMethod?: string;\n readingOrder?: string;\n markdownPageSeparator?: string;\n markdownWithHtml?: boolean;\n textPageSeparator?: string;\n htmlPageSeparator?: string;\n imageOutput?: string;\n imageFormat?: string;\n imageDir?: string;\n pages?: string;\n includeHeaderFooter?: boolean;\n detectStrikethrough?: boolean;\n hybrid?: string;\n hybridMode?: string;\n hybridUrl?: string;\n hybridTimeout?: string;\n hybridChunkSize?: string;\n hybridFallback?: boolean;\n toStdout?: boolean;\n threads?: string;\n imageResolution?: string;\n spaceRatio?: string;\n}\n\n/**\n * Convert CLI options to ConvertOptions.\n */\nexport function buildConvertOptions(cliOptions: CliOptions): ConvertOptions {\n const convertOptions: ConvertOptions = {};\n\n if (cliOptions.outputDir) {\n convertOptions.outputDir = cliOptions.outputDir;\n }\n if (cliOptions.password) {\n convertOptions.password = cliOptions.password;\n }\n if (cliOptions.format) {\n convertOptions.format = cliOptions.format;\n }\n if (cliOptions.quiet) {\n convertOptions.quiet = true;\n }\n if (cliOptions.contentSafetyOff) {\n convertOptions.contentSafetyOff = cliOptions.contentSafetyOff;\n }\n if (cliOptions.sanitize) {\n convertOptions.sanitize = true;\n }\n if (cliOptions.keepLineBreaks) {\n convertOptions.keepLineBreaks = true;\n }\n if (cliOptions.replaceInvalidChars) {\n convertOptions.replaceInvalidChars = cliOptions.replaceInvalidChars;\n }\n if (cliOptions.useStructTree) {\n convertOptions.useStructTree = true;\n }\n if (cliOptions.tableMethod) {\n convertOptions.tableMethod = cliOptions.tableMethod;\n }\n if (cliOptions.readingOrder) {\n convertOptions.readingOrder = cliOptions.readingOrder;\n }\n if (cliOptions.markdownPageSeparator) {\n convertOptions.markdownPageSeparator = cliOptions.markdownPageSeparator;\n }\n if (cliOptions.markdownWithHtml) {\n convertOptions.markdownWithHtml = true;\n }\n if (cliOptions.textPageSeparator) {\n convertOptions.textPageSeparator = cliOptions.textPageSeparator;\n }\n if (cliOptions.htmlPageSeparator) {\n convertOptions.htmlPageSeparator = cliOptions.htmlPageSeparator;\n }\n if (cliOptions.imageOutput) {\n convertOptions.imageOutput = cliOptions.imageOutput;\n }\n if (cliOptions.imageFormat) {\n convertOptions.imageFormat = cliOptions.imageFormat;\n }\n if (cliOptions.imageDir) {\n convertOptions.imageDir = cliOptions.imageDir;\n }\n if (cliOptions.pages) {\n convertOptions.pages = cliOptions.pages;\n }\n if (cliOptions.includeHeaderFooter) {\n convertOptions.includeHeaderFooter = true;\n }\n if (cliOptions.detectStrikethrough) {\n convertOptions.detectStrikethrough = true;\n }\n if (cliOptions.hybrid) {\n convertOptions.hybrid = cliOptions.hybrid;\n }\n if (cliOptions.hybridMode) {\n convertOptions.hybridMode = cliOptions.hybridMode;\n }\n if (cliOptions.hybridUrl) {\n convertOptions.hybridUrl = cliOptions.hybridUrl;\n }\n if (cliOptions.hybridTimeout) {\n convertOptions.hybridTimeout = cliOptions.hybridTimeout;\n }\n if (cliOptions.hybridChunkSize) {\n convertOptions.hybridChunkSize = cliOptions.hybridChunkSize;\n }\n if (cliOptions.hybridFallback) {\n convertOptions.hybridFallback = true;\n }\n if (cliOptions.toStdout) {\n convertOptions.toStdout = true;\n }\n if (cliOptions.threads) {\n convertOptions.threads = cliOptions.threads;\n }\n if (cliOptions.imageResolution) {\n convertOptions.imageResolution = cliOptions.imageResolution;\n }\n if (cliOptions.spaceRatio) {\n convertOptions.spaceRatio = cliOptions.spaceRatio;\n }\n\n return convertOptions;\n}\n\n/**\n * Build CLI arguments array from ConvertOptions.\n */\nexport function buildArgs(options: ConvertOptions): string[] {\n const args: string[] = [];\n\n if (options.outputDir) {\n args.push('--output-dir', options.outputDir);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.format) {\n if (Array.isArray(options.format)) {\n if (options.format.length > 0) {\n args.push('--format', options.format.join(','));\n }\n } else {\n args.push('--format', options.format);\n }\n }\n if (options.quiet) {\n args.push('--quiet');\n }\n if (options.contentSafetyOff) {\n if (Array.isArray(options.contentSafetyOff)) {\n if (options.contentSafetyOff.length > 0) {\n args.push('--content-safety-off', options.contentSafetyOff.join(','));\n }\n } else {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n }\n if (options.sanitize) {\n args.push('--sanitize');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.useStructTree) {\n args.push('--use-struct-tree');\n }\n if (options.tableMethod) {\n args.push('--table-method', options.tableMethod);\n }\n if (options.readingOrder) {\n args.push('--reading-order', options.readingOrder);\n }\n if (options.markdownPageSeparator) {\n args.push('--markdown-page-separator', options.markdownPageSeparator);\n }\n if (options.markdownWithHtml) {\n args.push('--markdown-with-html');\n }\n if (options.textPageSeparator) {\n args.push('--text-page-separator', options.textPageSeparator);\n }\n if (options.htmlPageSeparator) {\n args.push('--html-page-separator', options.htmlPageSeparator);\n }\n if (options.imageOutput) {\n args.push('--image-output', options.imageOutput);\n }\n if (options.imageFormat) {\n args.push('--image-format', options.imageFormat);\n }\n if (options.imageDir) {\n args.push('--image-dir', options.imageDir);\n }\n if (options.pages) {\n args.push('--pages', options.pages);\n }\n if (options.includeHeaderFooter) {\n args.push('--include-header-footer');\n }\n if (options.detectStrikethrough) {\n args.push('--detect-strikethrough');\n }\n if (options.hybrid) {\n args.push('--hybrid', options.hybrid);\n }\n if (options.hybridMode) {\n args.push('--hybrid-mode', options.hybridMode);\n }\n if (options.hybridUrl) {\n args.push('--hybrid-url', options.hybridUrl);\n }\n if (options.hybridTimeout) {\n args.push('--hybrid-timeout', options.hybridTimeout);\n }\n if (options.hybridChunkSize) {\n args.push('--hybrid-chunk-size', options.hybridChunkSize);\n }\n if (options.hybridFallback) {\n args.push('--hybrid-fallback');\n }\n if (options.toStdout) {\n args.push('--to-stdout');\n }\n if (options.threads) {\n args.push('--threads', options.threads);\n }\n if (options.imageResolution) {\n args.push('--image-resolution', options.imageResolution);\n }\n if (options.spaceRatio) {\n args.push('--space-ratio', options.spaceRatio);\n }\n\n return args;\n}\n"],"mappings":";AAAA,SAAS,aAAa;AACtB,YAAY,UAAU;AACtB,YAAY,QAAQ;AACpB,SAAS,qBAAqB;AAC9B,SAAS,qBAAqB;;;ACkNvB,SAAS,UAAU,SAAmC;AAC3D,QAAM,OAAiB,CAAC;AAExB,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,EAC1C;AACA,MAAI,QAAQ,QAAQ;AAClB,QAAI,MAAM,QAAQ,QAAQ,MAAM,GAAG;AACjC,UAAI,QAAQ,OAAO,SAAS,GAAG;AAC7B,aAAK,KAAK,YAAY,QAAQ,OAAO,KAAK,GAAG,CAAC;AAAA,MAChD;AAAA,IACF,OAAO;AACL,WAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,IACtC;AAAA,EACF;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,SAAS;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,MAAM,QAAQ,QAAQ,gBAAgB,GAAG;AAC3C,UAAI,QAAQ,iBAAiB,SAAS,GAAG;AACvC,aAAK,KAAK,wBAAwB,QAAQ,iBAAiB,KAAK,GAAG,CAAC;AAAA,MACtE;AAAA,IACF,OAAO;AACL,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AAAA,EACF;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,YAAY;AAAA,EACxB;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,oBAAoB;AAAA,EAChC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,EAClE;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,cAAc;AACxB,SAAK,KAAK,mBAAmB,QAAQ,YAAY;AAAA,EACnD;AACA,MAAI,QAAQ,uBAAuB;AACjC,SAAK,KAAK,6BAA6B,QAAQ,qBAAqB;AAAA,EACtE;AACA,MAAI,QAAQ,kBAAkB;AAC5B,SAAK,KAAK,sBAAsB;AAAA,EAClC;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,mBAAmB;AAC7B,SAAK,KAAK,yBAAyB,QAAQ,iBAAiB;AAAA,EAC9D;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,aAAa;AACvB,SAAK,KAAK,kBAAkB,QAAQ,WAAW;AAAA,EACjD;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,eAAe,QAAQ,QAAQ;AAAA,EAC3C;AACA,MAAI,QAAQ,OAAO;AACjB,SAAK,KAAK,WAAW,QAAQ,KAAK;AAAA,EACpC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,yBAAyB;AAAA,EACrC;AACA,MAAI,QAAQ,qBAAqB;AAC/B,SAAK,KAAK,wBAAwB;AAAA,EACpC;AACA,MAAI,QAAQ,QAAQ;AAClB,SAAK,KAAK,YAAY,QAAQ,MAAM;AAAA,EACtC;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AACA,MAAI,QAAQ,WAAW;AACrB,SAAK,KAAK,gBAAgB,QAAQ,SAAS;AAAA,EAC7C;AACA,MAAI,QAAQ,eAAe;AACzB,SAAK,KAAK,oBAAoB,QAAQ,aAAa;AAAA,EACrD;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,uBAAuB,QAAQ,eAAe;AAAA,EAC1D;AACA,MAAI,QAAQ,gBAAgB;AAC1B,SAAK,KAAK,mBAAmB;AAAA,EAC/B;AACA,MAAI,QAAQ,UAAU;AACpB,SAAK,KAAK,aAAa;AAAA,EACzB;AACA,MAAI,QAAQ,SAAS;AACnB,SAAK,KAAK,aAAa,QAAQ,OAAO;AAAA,EACxC;AACA,MAAI,QAAQ,iBAAiB;AAC3B,SAAK,KAAK,sBAAsB,QAAQ,eAAe;AAAA,EACzD;AACA,MAAI,QAAQ,YAAY;AACtB,SAAK,KAAK,iBAAiB,QAAQ,UAAU;AAAA,EAC/C;AAEA,SAAO;AACT;;;ADxTA,IAAMA,cAAa,cAAc,YAAY,GAAG;AAChD,IAAMC,aAAiB,aAAQD,WAAU;AAEzC,IAAM,WAAW;AAWjB,SAAS,WAAW,MAAgB,mBAAwC,CAAC,GAAoB;AAC/F,QAAM,EAAE,eAAe,MAAM,IAAI;AAEjC,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,UAAM,UAAe,UAAKC,YAAW,MAAM,OAAO,QAAQ;AAE1D,QAAI,CAAI,cAAW,OAAO,GAAG;AAC3B,aAAO;AAAA,QACL,IAAI,MAAM,yBAAyB,OAAO,sCAAsC;AAAA,MAClF;AAAA,IACF;AAEA,UAAM,UAAU;AAIhB,UAAM,cAAc;AAAA,MAClB;AAAA,MACA;AAAA,MACA;AAAA,MACA;AAAA,MACA,GAAG;AAAA,IACL;AAEA,UAAM,cAAc,MAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAKb,UAAM,gBAAgB,IAAI,cAAc,MAAM;AAC9C,UAAM,gBAAgB,IAAI,cAAc,MAAM;AAE9C,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAIhB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B,OAAO;AACL,kBAAU;AAAA,MACZ;AAAA,IACF,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAiB;AAC9C,YAAM,QAAQ,cAAc,MAAM,IAAI;AACtC,UAAI,MAAM,WAAW,EAAG;AACxB,UAAI,cAAc;AAChB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AAGA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAIhC,YAAM,aAAa,cAAc,IAAI;AACrC,YAAM,aAAa,cAAc,IAAI;AACrC,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,cAAc;AAChB,kBAAQ,OAAO,MAAM,UAAU;AAAA,QACjC,OAAO;AACL,oBAAU;AAAA,QACZ;AAAA,MACF;AACA,UAAI,WAAW,SAAS,GAAG;AACzB,YAAI,aAAc,SAAQ,OAAO,MAAM,UAAU;AACjD,kBAAU;AAAA,MACZ;AAEA,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,cAAc,UAAU;AAC9B,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,WAAW;AAAA,QACxE;AAKA,QAAC,MAA2C,aAAa;AACzD,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAe;AACtC,UAAI,IAAI,QAAQ,SAAS,QAAQ,GAAG;AAClC;AAAA,UACE,IAAI;AAAA,YACF;AAAA,UACF;AAAA,QACF;AAAA,MACF,OAAO;AACL,eAAO,GAAG;AAAA,MACZ;AAAA,IACF,CAAC;AAAA,EACH,CAAC;AACH;AAEA,SAAS,aACP,YACA,SACkB;AAClB,QAAM,YAAY,MAAM,QAAQ,UAAU,IAAI,aAAa,CAAC,UAAU;AACtE,MAAI,UAAU,WAAW,GAAG;AAC1B,WAAO,IAAI,MAAM,2CAA2C;AAAA,EAC9D;AAEA,aAAW,SAAS,WAAW;AAC7B,QAAI,CAAI,cAAW,KAAK,GAAG;AACzB,aAAO,IAAI,MAAM,mCAAmC,KAAK,EAAE;AAAA,IAC7D;AAAA,EACF;AAEA,SAAO,CAAC,GAAG,WAAW,GAAG,UAAU,OAAO,CAAC;AAC7C;AAEO,SAAS,QACd,YACA,UAA0B,CAAC,GACV;AACjB,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,WAAO,QAAQ,OAAO,WAAW;AAAA,EACnC;AAIA,SAAO,WAAW,aAAa,EAAE,cAAc,MAAM,CAAC;AACxD;AAaA,eAAsB,WACpB,YACA,UAA0B,CAAC,GACZ;AACf,QAAM,cAAc,aAAa,YAAY,OAAO;AACpD,MAAI,uBAAuB,OAAO;AAChC,UAAM;AAAA,EACR;AACA,QAAM,WAAW,aAAa,EAAE,cAAc,KAAK,CAAC;AACtD;AAwBO,SAAS,IAAI,WAAmB,UAAsB,CAAC,GAAoB;AAChF,UAAQ;AAAA,IACN;AAAA,EACF;AAGA,QAAM,UAAoB,CAAC;AAC3B,MAAI,CAAC,QAAQ,QAAQ;AACnB,YAAQ,KAAK,MAAM;AAAA,EACrB;AACA,MAAI,QAAQ,kBAAkB;AAC5B,QAAI,QAAQ,oBAAoB;AAC9B,cAAQ,KAAK,sBAAsB;AAAA,IACrC,WAAW,QAAQ,gBAAgB;AACjC,cAAQ,KAAK,oBAAoB;AAAA,IACnC,OAAO;AACL,cAAQ,KAAK,UAAU;AAAA,IACzB;AAAA,EACF;AACA,MAAI,QAAQ,cAAc;AACxB,YAAQ,KAAK,MAAM;AAAA,EACrB;AACA,MAAI,QAAQ,sBAAsB;AAChC,YAAQ,KAAK,KAAK;AAAA,EACpB;AAEA,SAAO,QAAQ,WAAW;AAAA,IACxB,WAAW,QAAQ;AAAA,IACnB,UAAU,QAAQ;AAAA,IAClB,qBAAqB,QAAQ;AAAA,IAC7B,gBAAgB,QAAQ;AAAA,IACxB,kBAAkB,QAAQ;AAAA,IAC1B,eAAe,QAAQ;AAAA,IACvB,QAAQ,QAAQ,SAAS,IAAI,UAAU;AAAA,IACvC,OAAO,CAAC,QAAQ;AAAA,EAClB,CAAC;AACH;","names":["__filename","__dirname"]}
|
|
Binary file
|