officeparser 4.0.7 → 4.1.0

package/officeParser.js

@@ -4,7 +4,7 @@ const decompress    = require('decompress');
 const fs            = require('fs');
 const rimraf        = require('rimraf');
 const fileType      = require('file-type');
-const pdfParse      = require('pdf-parse');
+const pdfjs         = require('./pdfjs-dist-build/pdf.js');
 const { DOMParser } = require('@xmldom/xmldom');
 
 /** Header for error messages */
@@ -59,25 +59,23 @@ const parseString = (xml) => {
  */
 function parseWord(filepath, callback, config) {
     /** The target content xml file for the docx file. */
-    const mainContentFile = 'word/document.xml';
-    const footnotesFile   = 'word/footnotes.xml';
-    const endnotesFile    = 'word/endnotes.xml';
+    const mainContentFileRegex = /word\/document[\d+]?.xml/g;
+    const footnotesFileRegex   = /word\/footnotes[\d+]?.xml/g;
+    const endnotesFileRegex    = /word\/endnotes[\d+]?.xml/g;
     /** The decompress location which contains the filename in it */
     const decompressLocation = `${config.tempFilesLocation}/${filepath.split("/").pop()}`;
     decompress(filepath,
         decompressLocation,
-        { filter: x => [mainContentFile, footnotesFile, endnotesFile].includes(x.path) }
+        { filter: x => [mainContentFileRegex, footnotesFileRegex, endnotesFileRegex].some(fileRegex => x.path.match(fileRegex)) }
     )
     .then(files => {
         // Verify if atleast the document xml file exists in the extracted files list.
-        if (!files.map(file => file.path).includes(mainContentFile))
+        if (!files.some(file => file.path.match(mainContentFileRegex)))
             throw ERRORMSG.fileCorrupted(filepath);
 
-        return [...files.filter(file => file.path == mainContentFile),
-                    ...files.filter(file => file.path == footnotesFile),
-                    ...files.filter(file => file.path == endnotesFile)
-                    ]
-                    .map(file => fs.readFileSync(`${decompressLocation}/${file.path}`, 'utf8'));
+            return files
+                .filter(file => file.path.match(mainContentFileRegex) || file.path.match(footnotesFileRegex) || file.path.match(endnotesFileRegex))
+                .map(file => fs.readFileSync(`${decompressLocation}/${file.path}`, 'utf8'));
     })
     // ************************************* word xml files explanation *************************************
     // Structure of xmlContent of a word file is simple.
@@ -204,7 +202,7 @@ function parseExcel(filepath, callback, config) {
     const decompressLocation = `${config.tempFilesLocation}/${filepath.split("/").pop()}`;
     decompress(filepath,
         decompressLocation,
-        { filter: x => ([sheetsRegex, drawingsRegex, chartsRegex].findIndex(fileRegex => x.path.match(fileRegex)) > -1) || (x.path == stringsFilePath )}
+        { filter: x => [sheetsRegex, drawingsRegex, chartsRegex].some(fileRegex => x.path.match(fileRegex)) || x.path == stringsFilePath }
     )
     .then(files => {
         // Verify if atleast the slides xml files exist in the extracted files list.
@@ -433,9 +431,6 @@ function parseOpenOffice(filepath, callback, config) {
     .catch(e => callback(undefined, e));
 }
 
-/** Header for error messages */
-const PDFPARSEERRORHEADER = "[pdf-parse]: ";
-
 /** Main function for parsing text from pdf files
  * @param {string}             filepath File path
  * @param {function}           callback Callback function that returns value or error
@@ -443,17 +438,35 @@ const PDFPARSEERRORHEADER = "[pdf-parse]: ";
  * @returns {void}
  */
 function parsePdf(filepath, callback, config) {
-    // Get the data buffer for the given file path.
-    const dataBuffer = fs.readFileSync(filepath);
-
-    pdfParse(dataBuffer)
-        .then(data => {
-            let text = data.text;
-            if (!!config.newlineDelimiter && config.newlineDelimiter != "\n")
-                text = text.replaceAll("\n", config.newlineDelimiter)
-            callback(text, undefined);
-        })
-        .catch(e => callback(undefined, PDFPARSEERRORHEADER + e));
+    // Get the pdfjs document for the filepath.
+    pdfjs.getDocument(filepath).promise
+    // We go through each page and build our text content promise array.
+    .then(document => Promise.all(Array.from({ length: document.numPages }, (_, index) => index + 1).map(pageNr => document.getPage(pageNr).then(page => page.getTextContent()))))
+    // Each textContent item has property 'items' which is an array of objects.
+    // Each object element in the array has text stored in their 'str' key.
+    // The concatenation of str is what makes our pdf content.
+    // str already contains any space that was in the text.
+    // So, we only care about when to add the new line.
+    // That we determine using transform[5] value which is the y-coordinate of the item object.
+    // So, if there is a mismatch in the transform[5] value between the current item and the previous item, we put a line break.
+    .then(textContentArray => {
+        /** Store all the text content to respond */
+        const responseText = textContentArray
+                                .map(textContent => textContent.items)      // Get all the items
+                                .flat()                                     // Flatten all the items object
+                                .reduce((a, v) => (
+                                    {
+                                        text: a.text + (v.transform[5] != a.transform5 ? (config.newlineDelimiter ?? "\n") : '') + v.str,
+                                        transform5: v.transform[5]
+                                    }),
+                                    {
+                                        text: '',
+                                        transform5: undefined
+                                    }).text;
+        
+        callback(responseText, undefined);
+    })
+    .catch(e => callback(undefined, e));
 }
 
 /** Main async function with callback to execute parseOffice for supported files
@@ -586,7 +599,7 @@ let globalFileNameIterator = 0;
  * to allow the files to be sorted in chronological order
  * @param {string} tempFilesLocation Directory whether this new file needs to be stored
  * @param {string} ext               File extension for this new generated file name
- * @returns {string} 
+ * @returns {string}
  */
 function getNewFileName(tempFilesLocation, ext) {
     // Get the iterator part of the file name

package/package.json

@@ -1,11 +1,12 @@
 {
   "name": "officeparser",
-  "version": "4.0.7",
+  "version": "4.1.0",
   "description": "A Node.js library to parse text out of any office file. Currently supports docx, pptx, xlsx, odt, odp, ods, pdf files.",
   "main": "officeParser.js",
   "files": [
     "officeParser.js",
-    "typings/officeParser.d.ts"
+    "typings/officeParser.d.ts",
+    "pdfjs-dist-build/*"
   ],
   "types": "typings/officeParser.d.ts",
   "scripts": {
@@ -45,7 +46,7 @@
     "@xmldom/xmldom": "^0.8.10",
     "decompress": "^4.2.0",
     "file-type": "^16.5.4",
-    "pdf-parse": "^1.1.1",
+    "node-ensure": "^0.0.0",
     "rimraf": "^2.6.3"
   },
   "devDependencies": {