Updated CLI

Author: Balearica

cli/cli.js

@@ -52,14 +52,14 @@ export const evalInternalCLI = async (files, options) => {
 
 /**
  *
- * @param {string} pdfFile - Path to PDF file.
+ * @param {string} inputFile - Path to PDF file.
  * @param {?string} [outputDir='.'] - Output directory.
  * @param {Object} [options]
  * @param {"pdf" | "hocr" | "docx" | "xlsx" | "txt" | "text" | "html"} [options.format]
  * @param {boolean} [options.reflow]
  */
-export const extractCLI = async (pdfFile, outputDir, options) => {
-  await extract(pdfFile, outputDir, options);
+export const extractCLI = async (inputFile, outputDir, options) => {
+  await extract(inputFile, outputDir, options);
   process.exitCode = 0;
 };
 

cli/extract.js

@@ -2,27 +2,34 @@ import fs from 'node:fs';
 import path from 'node:path';
 import scribe from '../scribe.js';
 
+// TODO: Consider whether this should exist and whether it should be combined into a larger CLI utility.
+// This was originally created to provide a simple interface to extract existing text from a PDF file,
+// however it now does other things, and this should likely be part of a larger `convert` utility.
+
 /**
  *
- * @param {string} pdfFile - Path to PDF file.
+ * @param {string} inputFile - Path to input file.
  * @param {?string} [output='.'] - Output file or directory.
  * @param {Object} [options]
  * @param {Parameters<typeof scribe.download>[0]} [options.format]
  * @param {boolean} [options.reflow]
  */
-export const extract = async (pdfFile, output, options) => {
+export const extract = async (inputFile, output, options) => {
   const format = options?.format || 'txt';
 
   output = output || '.';
   const outputDir = path.dirname(output);
-  const outputFile = outputDir === output ? `${path.basename(pdfFile).replace(/\.\w{1,6}$/i, `.${format}`)}` : path.basename(output);
+  const outputFile = outputDir === output ? `${path.basename(inputFile).replace(/\.\w{1,6}$/i, `.${format}`)}` : path.basename(output);
   const outputPath = `${outputDir}/${outputFile}`;
 
   scribe.opt.reflow = true;
   scribe.opt.extractText = true;
+  scribe.opt.displayMode = 'ebook';
 
-  await scribe.init();
-  await scribe.importFiles([pdfFile]);
+  // TODO: Fonts do not need to be loaded for .txt output, but are needed for .pdf output.
+  // so a more robust implementation would consider the arguments and only load fonts if necessary.
+  await scribe.init({ font: true });
+  await scribe.importFiles([inputFile]);
 
   if (outputDir) fs.mkdirSync(outputDir, { recursive: true });
 

cli/scribe.js

@@ -34,7 +34,7 @@ program
 
 program
   .command('extract')
-  .argument('<pdf_file>', 'Input PDF file.')
+  .argument('<input_file>', 'Input PDF file.')
   .argument('[output]', 'Output directory or file to save results.', '.')
   .addOption(new Option('-f, --format <ext>', 'Output format.').choices(['pdf', 'hocr', 'docx', 'xlsx', 'txt', 'text', 'html']).default('txt'))
   .option('-r, --reflow', 'Reflow text by combining lines into paragraphs.')