@opendataloader/pdf 1.0.3 → 1.0.4

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 CHANGED
@@ -40,13 +40,24 @@ AI-safety is enabled by default and automatically filters likely prompt-injectio
40
40
 
41
41
  ## 🚀 Upcoming Features
42
42
 
43
- - 🖨️ **OCR for scanned PDFs** — Extract data from image-only pages
44
- - 🧠 **Table AI option** — Higher accuracy for tables with borderless or merged cells
45
- - ⚡ **Performance Benchmarks** — Transparent evaluations with open datasets and metrics, reported regularly
46
- - 🛡️ **AI Red Teaming** — Transparent adversarial benchmarks with datasets and metrics, reported regularly
43
+ **Scheduled for November**
44
+ - 📄 **Tagged PDF** — Develop advanced data extraction technology based on Tagged PDF
45
+ - ⚡ **Performance Improvement** — Enhance the inference skill for greater accuracy and speed.
46
+ - 📊 **Benchmarks & Datasets** — Publish transparent evaluations using open datasets and standardized metrics.
47
+ - 🎯 **Metrics** — Publish the calculation methods to transparently share benchmark results.
47
48
 
48
49
  <br/>
49
50
 
51
+ **Scheduled for December**
52
+ - 🖨️ **OCR for scanned PDFs** — Extract data from image-only pages.
53
+ - 🧠 **Table AI option** — Higher accuracy for tables with borderless or merged cells.
54
+ <br/>
55
+
56
+ **Scheduled for 2026**
57
+ - 🛡️ **AI Red Teaming** — Transparent adversarial benchmarks with datasets and metrics, then reported regularly.
58
+
59
+ <br/>
60
+
50
61
  ## Prerequisites
51
62
 
52
63
  - Java 11 or higher must be installed and available in your system's PATH.
@@ -96,6 +107,7 @@ The main function to process PDFs.
96
107
  | `keep_line_breaks` | `bool` | No | `False` | If `True`, keeps line breaks in the output. |
97
108
  | `html_in_markdown` | `bool` | No | `False` | If `True`, uses HTML in the Markdown output. |
98
109
  | `add_image_to_markdown` | `bool` | No | `False` | If `True`, adds images to the Markdown output. |
110
+ | `no_json` | `bool` | No | `False` | If `True`, disables the JSON output. |
99
111
  | `debug` | `bool` | No | `False` | If `True`, prints CLI messages to the console during execution. |
100
112
 
101
113
  <br/>
@@ -167,6 +179,7 @@ The main function to process PDFs.
167
179
  | `keepLineBreaks` | `boolean` | `false` | If `true`, keeps line breaks in the output. |
168
180
  | `htmlInMarkdown` | `boolean` | `false` | If `true`, uses HTML in the Markdown output. |
169
181
  | `addImageToMarkdown` | `boolean` | `false` | If `true`, adds images to the Markdown output. |
182
+ | `noJson` | `boolean` | `false` | If `true`, disables the JSON output. |
170
183
  | `debug` | `boolean` | `false` | If `true`, prints CLI messages to the console during execution. |
171
184
 
172
185
  <br/>
@@ -314,6 +327,7 @@ Options:
314
327
  --markdown-with-images Sets the data extraction output format to Markdown with extracting images from the PDF and includes them as links
315
328
  --markdown Sets the data extraction output format to Markdown
316
329
  --html Sets the data extraction output format to HTML
330
+ --no-json Disables the JSON output format
317
331
  -p,--password <arg> Specifies the password for an encrypted PDF
318
332
  --pdf Generates a new PDF file where the extracted layout data is visualized as annotations
319
333
  --replace-invalid-chars <arg> Replaces invalid or unrecognized characters (e.g., �, \u0000) with the specified character
package/dist/index.cjs CHANGED
@@ -85,6 +85,9 @@ function run(inputPath, options = {}) {
85
85
  if (options.addImageToMarkdown) {
86
86
  args.push("--markdown-with-images");
87
87
  }
88
+ if (options.noJson) {
89
+ args.push("--no-json");
90
+ }
88
91
  args.push(inputPath);
89
92
  const jarPath = path.join(__dirname, "..", "lib", JAR_NAME);
90
93
  if (!fs.existsSync(jarPath)) {
@@ -1 +1 @@
1
- {"version":3,"sources":["../src/index.ts"],"sourcesContent":["import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { fileURLToPath } from 'url';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\nfunction getRedactedCommandString(command: string, commandArgs: string[]): string {\n const commandArgsForLogging = [...commandArgs];\n const passwordIndex = commandArgsForLogging.indexOf('--password');\n if (passwordIndex > -1 && passwordIndex + 1 < commandArgsForLogging.length) {\n commandArgsForLogging[passwordIndex + 1] = '[REDACTED]';\n }\n return `${command} ${commandArgsForLogging.join(' ')}`;\n}\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 debug?: boolean;\n}\n\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n return new Promise((resolve, reject) => {\n if (!fs.existsSync(inputPath)) {\n return reject(new Error(`Input file or folder not found: ${inputPath}`));\n }\n\n const args: string[] = [];\n if (options.outputFolder) {\n args.push('--output-dir', options.outputFolder);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.generateMarkdown) {\n args.push('--markdown');\n }\n if (options.generateHtml) {\n args.push('--html');\n }\n if (options.generateAnnotatedPdf) {\n args.push('--pdf');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.contentSafetyOff) {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n if (options.htmlInMarkdown) {\n args.push('--markdown-with-html');\n }\n if (options.addImageToMarkdown) {\n args.push('--markdown-with-images');\n }\n\n args.push(inputPath);\n\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 const commandArgs = ['-jar', jarPath, ...args];\n\n if (options.debug) {\n console.error(`Running command: ${getRedactedCommandString(command, commandArgs)}`);\n }\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n\n javaProcess.stdout.on('data', (data) => {\n const chunk = data.toString();\n if (options.debug) {\n process.stdout.write(chunk);\n }\n stdout += chunk;\n });\n\n javaProcess.stderr.on('data', (data) => {\n const chunk = data.toString();\n if (options.debug) {\n process.stderr.write(chunk);\n }\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n if (code === 0) {\n resolve(stdout);\n } else {\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${stderr}`,\n );\n reject(error);\n }\n });\n\n javaProcess.on('error', (err) => {\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}"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,2BAAsB;AACtB,WAAsB;AACtB,SAAoB;AACpB,iBAA8B;AAH9B;AAKA,IAAM,iBAAa,0BAAc,YAAY,GAAG;AAChD,IAAM,YAAiB,aAAQ,UAAU;AAEzC,IAAM,WAAW;AAEjB,SAAS,yBAAyB,SAAiB,aAA+B;AAChF,QAAM,wBAAwB,CAAC,GAAG,WAAW;AAC7C,QAAM,gBAAgB,sBAAsB,QAAQ,YAAY;AAChE,MAAI,gBAAgB,MAAM,gBAAgB,IAAI,sBAAsB,QAAQ;AAC1E,0BAAsB,gBAAgB,CAAC,IAAI;AAAA,EAC7C;AACA,SAAO,GAAG,OAAO,IAAI,sBAAsB,KAAK,GAAG,CAAC;AACtD;AAgBO,SAAS,IAAI,WAAmB,UAAsB,CAAC,GAAoB;AAChF,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,QAAI,CAAI,cAAW,SAAS,GAAG;AAC7B,aAAO,OAAO,IAAI,MAAM,mCAAmC,SAAS,EAAE,CAAC;AAAA,IACzE;AAEA,UAAM,OAAiB,CAAC;AACxB,QAAI,QAAQ,cAAc;AACxB,WAAK,KAAK,gBAAgB,QAAQ,YAAY;AAAA,IAChD;AACA,QAAI,QAAQ,UAAU;AACpB,WAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,IAC1C;AACA,QAAI,QAAQ,qBAAqB;AAC/B,WAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,IAClE;AACA,QAAI,QAAQ,kBAAkB;AAC5B,WAAK,KAAK,YAAY;AAAA,IACxB;AACA,QAAI,QAAQ,cAAc;AACxB,WAAK,KAAK,QAAQ;AAAA,IACpB;AACA,QAAI,QAAQ,sBAAsB;AAChC,WAAK,KAAK,OAAO;AAAA,IACnB;AACA,QAAI,QAAQ,gBAAgB;AAC1B,WAAK,KAAK,oBAAoB;AAAA,IAChC;AACA,QAAI,QAAQ,kBAAkB;AAC5B,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AACA,QAAI,QAAQ,gBAAgB;AAC1B,WAAK,KAAK,sBAAsB;AAAA,IAClC;AACA,QAAI,QAAQ,oBAAoB;AAC9B,WAAK,KAAK,wBAAwB;AAAA,IACpC;AAEA,SAAK,KAAK,SAAS;AAEnB,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;AAChB,UAAM,cAAc,CAAC,QAAQ,SAAS,GAAG,IAAI;AAE7C,QAAI,QAAQ,OAAO;AACjB,cAAQ,MAAM,oBAAoB,yBAAyB,SAAS,WAAW,CAAC,EAAE;AAAA,IACpF;AAEA,UAAM,kBAAc,4BAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAEb,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAS;AACtC,YAAM,QAAQ,KAAK,SAAS;AAC5B,UAAI,QAAQ,OAAO;AACjB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AACA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAS;AACtC,YAAM,QAAQ,KAAK,SAAS;AAC5B,UAAI,QAAQ,OAAO;AACjB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AACA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAChC,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,MAAM;AAAA,QACnE;AACA,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAQ;AAC/B,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;","names":[]}
1
+ {"version":3,"sources":["../src/index.ts"],"sourcesContent":["import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { fileURLToPath } from 'url';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\nfunction getRedactedCommandString(command: string, commandArgs: string[]): string {\n const commandArgsForLogging = [...commandArgs];\n const passwordIndex = commandArgsForLogging.indexOf('--password');\n if (passwordIndex > -1 && passwordIndex + 1 < commandArgsForLogging.length) {\n commandArgsForLogging[passwordIndex + 1] = '[REDACTED]';\n }\n return `${command} ${commandArgsForLogging.join(' ')}`;\n}\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}\n\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n return new Promise((resolve, reject) => {\n if (!fs.existsSync(inputPath)) {\n return reject(new Error(`Input file or folder not found: ${inputPath}`));\n }\n\n const args: string[] = [];\n if (options.outputFolder) {\n args.push('--output-dir', options.outputFolder);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.generateMarkdown) {\n args.push('--markdown');\n }\n if (options.generateHtml) {\n args.push('--html');\n }\n if (options.generateAnnotatedPdf) {\n args.push('--pdf');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.contentSafetyOff) {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n if (options.htmlInMarkdown) {\n args.push('--markdown-with-html');\n }\n if (options.addImageToMarkdown) {\n args.push('--markdown-with-images');\n }\n if (options.noJson) {\n args.push('--no-json');\n }\n\n args.push(inputPath);\n\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 const commandArgs = ['-jar', jarPath, ...args];\n\n if (options.debug) {\n console.error(`Running command: ${getRedactedCommandString(command, commandArgs)}`);\n }\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n\n javaProcess.stdout.on('data', (data) => {\n const chunk = data.toString();\n if (options.debug) {\n process.stdout.write(chunk);\n }\n stdout += chunk;\n });\n\n javaProcess.stderr.on('data', (data) => {\n const chunk = data.toString();\n if (options.debug) {\n process.stderr.write(chunk);\n }\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n if (code === 0) {\n resolve(stdout);\n } else {\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${stderr}`,\n );\n reject(error);\n }\n });\n\n javaProcess.on('error', (err) => {\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"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,2BAAsB;AACtB,WAAsB;AACtB,SAAoB;AACpB,iBAA8B;AAH9B;AAKA,IAAM,iBAAa,0BAAc,YAAY,GAAG;AAChD,IAAM,YAAiB,aAAQ,UAAU;AAEzC,IAAM,WAAW;AAEjB,SAAS,yBAAyB,SAAiB,aAA+B;AAChF,QAAM,wBAAwB,CAAC,GAAG,WAAW;AAC7C,QAAM,gBAAgB,sBAAsB,QAAQ,YAAY;AAChE,MAAI,gBAAgB,MAAM,gBAAgB,IAAI,sBAAsB,QAAQ;AAC1E,0BAAsB,gBAAgB,CAAC,IAAI;AAAA,EAC7C;AACA,SAAO,GAAG,OAAO,IAAI,sBAAsB,KAAK,GAAG,CAAC;AACtD;AAiBO,SAAS,IAAI,WAAmB,UAAsB,CAAC,GAAoB;AAChF,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,QAAI,CAAI,cAAW,SAAS,GAAG;AAC7B,aAAO,OAAO,IAAI,MAAM,mCAAmC,SAAS,EAAE,CAAC;AAAA,IACzE;AAEA,UAAM,OAAiB,CAAC;AACxB,QAAI,QAAQ,cAAc;AACxB,WAAK,KAAK,gBAAgB,QAAQ,YAAY;AAAA,IAChD;AACA,QAAI,QAAQ,UAAU;AACpB,WAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,IAC1C;AACA,QAAI,QAAQ,qBAAqB;AAC/B,WAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,IAClE;AACA,QAAI,QAAQ,kBAAkB;AAC5B,WAAK,KAAK,YAAY;AAAA,IACxB;AACA,QAAI,QAAQ,cAAc;AACxB,WAAK,KAAK,QAAQ;AAAA,IACpB;AACA,QAAI,QAAQ,sBAAsB;AAChC,WAAK,KAAK,OAAO;AAAA,IACnB;AACA,QAAI,QAAQ,gBAAgB;AAC1B,WAAK,KAAK,oBAAoB;AAAA,IAChC;AACA,QAAI,QAAQ,kBAAkB;AAC5B,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AACA,QAAI,QAAQ,gBAAgB;AAC1B,WAAK,KAAK,sBAAsB;AAAA,IAClC;AACA,QAAI,QAAQ,oBAAoB;AAC9B,WAAK,KAAK,wBAAwB;AAAA,IACpC;AACA,QAAI,QAAQ,QAAQ;AAClB,WAAK,KAAK,WAAW;AAAA,IACvB;AAEA,SAAK,KAAK,SAAS;AAEnB,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;AAChB,UAAM,cAAc,CAAC,QAAQ,SAAS,GAAG,IAAI;AAE7C,QAAI,QAAQ,OAAO;AACjB,cAAQ,MAAM,oBAAoB,yBAAyB,SAAS,WAAW,CAAC,EAAE;AAAA,IACpF;AAEA,UAAM,kBAAc,4BAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAEb,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAS;AACtC,YAAM,QAAQ,KAAK,SAAS;AAC5B,UAAI,QAAQ,OAAO;AACjB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AACA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAS;AACtC,YAAM,QAAQ,KAAK,SAAS;AAC5B,UAAI,QAAQ,OAAO;AACjB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AACA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAChC,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,MAAM;AAAA,QACnE;AACA,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAQ;AAC/B,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;","names":[]}
package/dist/index.d.cts CHANGED
@@ -9,6 +9,7 @@ interface RunOptions {
9
9
  contentSafetyOff?: string;
10
10
  htmlInMarkdown?: boolean;
11
11
  addImageToMarkdown?: boolean;
12
+ noJson?: boolean;
12
13
  debug?: boolean;
13
14
  }
14
15
  declare function run(inputPath: string, options?: RunOptions): Promise<string>;
package/dist/index.d.ts CHANGED
@@ -9,6 +9,7 @@ interface RunOptions {
9
9
  contentSafetyOff?: string;
10
10
  htmlInMarkdown?: boolean;
11
11
  addImageToMarkdown?: boolean;
12
+ noJson?: boolean;
12
13
  debug?: boolean;
13
14
  }
14
15
  declare function run(inputPath: string, options?: RunOptions): Promise<string>;
package/dist/index.js CHANGED
@@ -50,6 +50,9 @@ function run(inputPath, options = {}) {
50
50
  if (options.addImageToMarkdown) {
51
51
  args.push("--markdown-with-images");
52
52
  }
53
+ if (options.noJson) {
54
+ args.push("--no-json");
55
+ }
53
56
  args.push(inputPath);
54
57
  const jarPath = path.join(__dirname, "..", "lib", JAR_NAME);
55
58
  if (!fs.existsSync(jarPath)) {
package/dist/index.js.map CHANGED
@@ -1 +1 @@
1
- {"version":3,"sources":["../src/index.ts"],"sourcesContent":["import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { fileURLToPath } from 'url';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\nfunction getRedactedCommandString(command: string, commandArgs: string[]): string {\n const commandArgsForLogging = [...commandArgs];\n const passwordIndex = commandArgsForLogging.indexOf('--password');\n if (passwordIndex > -1 && passwordIndex + 1 < commandArgsForLogging.length) {\n commandArgsForLogging[passwordIndex + 1] = '[REDACTED]';\n }\n return `${command} ${commandArgsForLogging.join(' ')}`;\n}\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 debug?: boolean;\n}\n\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n return new Promise((resolve, reject) => {\n if (!fs.existsSync(inputPath)) {\n return reject(new Error(`Input file or folder not found: ${inputPath}`));\n }\n\n const args: string[] = [];\n if (options.outputFolder) {\n args.push('--output-dir', options.outputFolder);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.generateMarkdown) {\n args.push('--markdown');\n }\n if (options.generateHtml) {\n args.push('--html');\n }\n if (options.generateAnnotatedPdf) {\n args.push('--pdf');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.contentSafetyOff) {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n if (options.htmlInMarkdown) {\n args.push('--markdown-with-html');\n }\n if (options.addImageToMarkdown) {\n args.push('--markdown-with-images');\n }\n\n args.push(inputPath);\n\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 const commandArgs = ['-jar', jarPath, ...args];\n\n if (options.debug) {\n console.error(`Running command: ${getRedactedCommandString(command, commandArgs)}`);\n }\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n\n javaProcess.stdout.on('data', (data) => {\n const chunk = data.toString();\n if (options.debug) {\n process.stdout.write(chunk);\n }\n stdout += chunk;\n });\n\n javaProcess.stderr.on('data', (data) => {\n const chunk = data.toString();\n if (options.debug) {\n process.stderr.write(chunk);\n }\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n if (code === 0) {\n resolve(stdout);\n } else {\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${stderr}`,\n );\n reject(error);\n }\n });\n\n javaProcess.on('error', (err) => {\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}"],"mappings":";AAAA,SAAS,aAAa;AACtB,YAAY,UAAU;AACtB,YAAY,QAAQ;AACpB,SAAS,qBAAqB;AAE9B,IAAM,aAAa,cAAc,YAAY,GAAG;AAChD,IAAM,YAAiB,aAAQ,UAAU;AAEzC,IAAM,WAAW;AAEjB,SAAS,yBAAyB,SAAiB,aAA+B;AAChF,QAAM,wBAAwB,CAAC,GAAG,WAAW;AAC7C,QAAM,gBAAgB,sBAAsB,QAAQ,YAAY;AAChE,MAAI,gBAAgB,MAAM,gBAAgB,IAAI,sBAAsB,QAAQ;AAC1E,0BAAsB,gBAAgB,CAAC,IAAI;AAAA,EAC7C;AACA,SAAO,GAAG,OAAO,IAAI,sBAAsB,KAAK,GAAG,CAAC;AACtD;AAgBO,SAAS,IAAI,WAAmB,UAAsB,CAAC,GAAoB;AAChF,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,QAAI,CAAI,cAAW,SAAS,GAAG;AAC7B,aAAO,OAAO,IAAI,MAAM,mCAAmC,SAAS,EAAE,CAAC;AAAA,IACzE;AAEA,UAAM,OAAiB,CAAC;AACxB,QAAI,QAAQ,cAAc;AACxB,WAAK,KAAK,gBAAgB,QAAQ,YAAY;AAAA,IAChD;AACA,QAAI,QAAQ,UAAU;AACpB,WAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,IAC1C;AACA,QAAI,QAAQ,qBAAqB;AAC/B,WAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,IAClE;AACA,QAAI,QAAQ,kBAAkB;AAC5B,WAAK,KAAK,YAAY;AAAA,IACxB;AACA,QAAI,QAAQ,cAAc;AACxB,WAAK,KAAK,QAAQ;AAAA,IACpB;AACA,QAAI,QAAQ,sBAAsB;AAChC,WAAK,KAAK,OAAO;AAAA,IACnB;AACA,QAAI,QAAQ,gBAAgB;AAC1B,WAAK,KAAK,oBAAoB;AAAA,IAChC;AACA,QAAI,QAAQ,kBAAkB;AAC5B,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AACA,QAAI,QAAQ,gBAAgB;AAC1B,WAAK,KAAK,sBAAsB;AAAA,IAClC;AACA,QAAI,QAAQ,oBAAoB;AAC9B,WAAK,KAAK,wBAAwB;AAAA,IACpC;AAEA,SAAK,KAAK,SAAS;AAEnB,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;AAChB,UAAM,cAAc,CAAC,QAAQ,SAAS,GAAG,IAAI;AAE7C,QAAI,QAAQ,OAAO;AACjB,cAAQ,MAAM,oBAAoB,yBAAyB,SAAS,WAAW,CAAC,EAAE;AAAA,IACpF;AAEA,UAAM,cAAc,MAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAEb,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAS;AACtC,YAAM,QAAQ,KAAK,SAAS;AAC5B,UAAI,QAAQ,OAAO;AACjB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AACA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAS;AACtC,YAAM,QAAQ,KAAK,SAAS;AAC5B,UAAI,QAAQ,OAAO;AACjB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AACA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAChC,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,MAAM;AAAA,QACnE;AACA,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAQ;AAC/B,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;","names":[]}
1
+ {"version":3,"sources":["../src/index.ts"],"sourcesContent":["import { spawn } from 'child_process';\nimport * as path from 'path';\nimport * as fs from 'fs';\nimport { fileURLToPath } from 'url';\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = path.dirname(__filename);\n\nconst JAR_NAME = 'opendataloader-pdf-cli.jar';\n\nfunction getRedactedCommandString(command: string, commandArgs: string[]): string {\n const commandArgsForLogging = [...commandArgs];\n const passwordIndex = commandArgsForLogging.indexOf('--password');\n if (passwordIndex > -1 && passwordIndex + 1 < commandArgsForLogging.length) {\n commandArgsForLogging[passwordIndex + 1] = '[REDACTED]';\n }\n return `${command} ${commandArgsForLogging.join(' ')}`;\n}\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}\n\nexport function run(inputPath: string, options: RunOptions = {}): Promise<string> {\n return new Promise((resolve, reject) => {\n if (!fs.existsSync(inputPath)) {\n return reject(new Error(`Input file or folder not found: ${inputPath}`));\n }\n\n const args: string[] = [];\n if (options.outputFolder) {\n args.push('--output-dir', options.outputFolder);\n }\n if (options.password) {\n args.push('--password', options.password);\n }\n if (options.replaceInvalidChars) {\n args.push('--replace-invalid-chars', options.replaceInvalidChars);\n }\n if (options.generateMarkdown) {\n args.push('--markdown');\n }\n if (options.generateHtml) {\n args.push('--html');\n }\n if (options.generateAnnotatedPdf) {\n args.push('--pdf');\n }\n if (options.keepLineBreaks) {\n args.push('--keep-line-breaks');\n }\n if (options.contentSafetyOff) {\n args.push('--content-safety-off', options.contentSafetyOff);\n }\n if (options.htmlInMarkdown) {\n args.push('--markdown-with-html');\n }\n if (options.addImageToMarkdown) {\n args.push('--markdown-with-images');\n }\n if (options.noJson) {\n args.push('--no-json');\n }\n\n args.push(inputPath);\n\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 const commandArgs = ['-jar', jarPath, ...args];\n\n if (options.debug) {\n console.error(`Running command: ${getRedactedCommandString(command, commandArgs)}`);\n }\n\n const javaProcess = spawn(command, commandArgs);\n\n let stdout = '';\n let stderr = '';\n\n javaProcess.stdout.on('data', (data) => {\n const chunk = data.toString();\n if (options.debug) {\n process.stdout.write(chunk);\n }\n stdout += chunk;\n });\n\n javaProcess.stderr.on('data', (data) => {\n const chunk = data.toString();\n if (options.debug) {\n process.stderr.write(chunk);\n }\n stderr += chunk;\n });\n\n javaProcess.on('close', (code) => {\n if (code === 0) {\n resolve(stdout);\n } else {\n const error = new Error(\n `The opendataloader-pdf CLI exited with code ${code}.\\n\\n${stderr}`,\n );\n reject(error);\n }\n });\n\n javaProcess.on('error', (err) => {\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"],"mappings":";AAAA,SAAS,aAAa;AACtB,YAAY,UAAU;AACtB,YAAY,QAAQ;AACpB,SAAS,qBAAqB;AAE9B,IAAM,aAAa,cAAc,YAAY,GAAG;AAChD,IAAM,YAAiB,aAAQ,UAAU;AAEzC,IAAM,WAAW;AAEjB,SAAS,yBAAyB,SAAiB,aAA+B;AAChF,QAAM,wBAAwB,CAAC,GAAG,WAAW;AAC7C,QAAM,gBAAgB,sBAAsB,QAAQ,YAAY;AAChE,MAAI,gBAAgB,MAAM,gBAAgB,IAAI,sBAAsB,QAAQ;AAC1E,0BAAsB,gBAAgB,CAAC,IAAI;AAAA,EAC7C;AACA,SAAO,GAAG,OAAO,IAAI,sBAAsB,KAAK,GAAG,CAAC;AACtD;AAiBO,SAAS,IAAI,WAAmB,UAAsB,CAAC,GAAoB;AAChF,SAAO,IAAI,QAAQ,CAAC,SAAS,WAAW;AACtC,QAAI,CAAI,cAAW,SAAS,GAAG;AAC7B,aAAO,OAAO,IAAI,MAAM,mCAAmC,SAAS,EAAE,CAAC;AAAA,IACzE;AAEA,UAAM,OAAiB,CAAC;AACxB,QAAI,QAAQ,cAAc;AACxB,WAAK,KAAK,gBAAgB,QAAQ,YAAY;AAAA,IAChD;AACA,QAAI,QAAQ,UAAU;AACpB,WAAK,KAAK,cAAc,QAAQ,QAAQ;AAAA,IAC1C;AACA,QAAI,QAAQ,qBAAqB;AAC/B,WAAK,KAAK,2BAA2B,QAAQ,mBAAmB;AAAA,IAClE;AACA,QAAI,QAAQ,kBAAkB;AAC5B,WAAK,KAAK,YAAY;AAAA,IACxB;AACA,QAAI,QAAQ,cAAc;AACxB,WAAK,KAAK,QAAQ;AAAA,IACpB;AACA,QAAI,QAAQ,sBAAsB;AAChC,WAAK,KAAK,OAAO;AAAA,IACnB;AACA,QAAI,QAAQ,gBAAgB;AAC1B,WAAK,KAAK,oBAAoB;AAAA,IAChC;AACA,QAAI,QAAQ,kBAAkB;AAC5B,WAAK,KAAK,wBAAwB,QAAQ,gBAAgB;AAAA,IAC5D;AACA,QAAI,QAAQ,gBAAgB;AAC1B,WAAK,KAAK,sBAAsB;AAAA,IAClC;AACA,QAAI,QAAQ,oBAAoB;AAC9B,WAAK,KAAK,wBAAwB;AAAA,IACpC;AACA,QAAI,QAAQ,QAAQ;AAClB,WAAK,KAAK,WAAW;AAAA,IACvB;AAEA,SAAK,KAAK,SAAS;AAEnB,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;AAChB,UAAM,cAAc,CAAC,QAAQ,SAAS,GAAG,IAAI;AAE7C,QAAI,QAAQ,OAAO;AACjB,cAAQ,MAAM,oBAAoB,yBAAyB,SAAS,WAAW,CAAC,EAAE;AAAA,IACpF;AAEA,UAAM,cAAc,MAAM,SAAS,WAAW;AAE9C,QAAI,SAAS;AACb,QAAI,SAAS;AAEb,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAS;AACtC,YAAM,QAAQ,KAAK,SAAS;AAC5B,UAAI,QAAQ,OAAO;AACjB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AACA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,OAAO,GAAG,QAAQ,CAAC,SAAS;AACtC,YAAM,QAAQ,KAAK,SAAS;AAC5B,UAAI,QAAQ,OAAO;AACjB,gBAAQ,OAAO,MAAM,KAAK;AAAA,MAC5B;AACA,gBAAU;AAAA,IACZ,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,SAAS;AAChC,UAAI,SAAS,GAAG;AACd,gBAAQ,MAAM;AAAA,MAChB,OAAO;AACL,cAAM,QAAQ,IAAI;AAAA,UAChB,+CAA+C,IAAI;AAAA;AAAA,EAAQ,MAAM;AAAA,QACnE;AACA,eAAO,KAAK;AAAA,MACd;AAAA,IACF,CAAC;AAED,gBAAY,GAAG,SAAS,CAAC,QAAQ;AAC/B,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;","names":[]}
Binary file
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@opendataloader/pdf",
3
- "version": "1.0.3",
3
+ "version": "1.0.4",
4
4
  "description": "A Node.js wrapper for the opendataloader-pdf Java CLI.",
5
5
  "main": "./dist/index.cjs",
6
6
  "module": "./dist/index.js",