officeparser 5.2.0 → 5.2.2

package/officeParser.js

@@ -8,9 +8,6 @@ const fileType      = require('file-type');
 const fs            = require('fs');
 const yauzl         = require('yauzl');
 
-/** Load pdfjs-dist once at module scope. This returns a Promise that resolves to the module. */
-const pdfjsPromise = import('pdfjs-dist/legacy/build/pdf.mjs');
-
 /** Header for error messages */
 const ERRORHEADER = "[OfficeParser]: ";
 /** Error messages */
@@ -238,12 +235,19 @@ function parseExcel(file, callback, config) {
                     && cNode.getElementsByTagName("v")[0].childNodes[0].nodeValue != ''
             }
 
-            /** Find text nodes with t tags in sharedStrings xml file. If the sharedStringsFile is not present, we return an empty array. */
-            const sharedStringsXmlTNodesList = xmlContentFilesObject.sharedStringsFile != undefined ? parseString(xmlContentFilesObject.sharedStringsFile).getElementsByTagName("t")
-                                                                                                    : [];
+            /** Find text nodes with t tags in sharedStrings.xml file. If the sharedStringsFile is not present, we return an empty array. */
+            const sharedStringsXmlSiNodesList = xmlContentFilesObject.sharedStringsFile != undefined
+                ? parseString(xmlContentFilesObject.sharedStringsFile).getElementsByTagName("si")
+                : [];
+
             /** Create shared string array. This will be used as a map to get strings from within sheet files. */
-            const sharedStrings = Array.from(sharedStringsXmlTNodesList)
-                                    .map(tNode => tNode.childNodes[0]?.nodeValue ?? '');
+            const sharedStrings = Array.from(sharedStringsXmlSiNodesList)
+                .map(siNode => {
+                    // Concatenate all <t> nodes within the <si> node
+                    return Array.from(siNode.getElementsByTagName("t"))
+                        .map(tNode => tNode.childNodes[0]?.nodeValue ?? '') // Extract text content from each <t> node
+                        .join(''); // Combine all <t> node text into a single string
+                });
 
             // Parse Sheet files
             xmlContentFilesObject.sheetFiles.forEach(sheetXmlContent => {
@@ -264,13 +268,14 @@ function parseExcel(file, callback, config) {
                                 /** Flag whether this node's value represents an index in the shared string array */
                                 const isIndexInSharedStrings = cNode.getAttribute("t") == "s";
                                 /** Find value nodes represented by v tags */
-                                const value = parseInt(cNode.getElementsByTagName("v")[0].childNodes[0].nodeValue, 10);
+                                const value = cNode.getElementsByTagName("v")[0].childNodes[0].nodeValue;
+                                const valueAsIndex = Number(value);
                                 // Validate text
-                                if (isIndexInSharedStrings && value >= sharedStrings.length)
+                                if (isIndexInSharedStrings && (valueAsIndex != parseInt(value, 10) || valueAsIndex >= sharedStrings.length))
                                     throw ERRORMSG.fileCorrupted(file);
 
                                 return isIndexInSharedStrings
-                                        ? sharedStrings[value]
+                                        ? sharedStrings[valueAsIndex]
                                         : value;
                             }
                             // Should not reach here. If we do, it means we are not filtering out items that we are not ready to process.
@@ -375,13 +380,19 @@ function parseOpenOffice(file, callback, config) {
             function traversal(node, xmlTextArray, isFirstRecursion) {
                 if (!node.childNodes || node.childNodes.length == 0) {
                     if (node.parentNode.tagName.indexOf('text') == 0 && node.nodeValue) {
+                        // If the corresponding value is of type float, we take the value from office:value attribute.
+                        // However, it is not on the parentNode but rather grandparentNode.
+                        const value = node.parentNode.parentNode?.getAttribute('office:value-type') == 'float'
+                                        ? Number(node.parentNode.parentNode.getAttribute('office:value'))
+                                        : node.nodeValue;
+
                         if (isNotesNode(node.parentNode) && (config.putNotesAtLast || config.ignoreNotes)) {
-                            notesText.push(node.nodeValue);
+                            notesText.push(value);
                             if (allowedTextTags.includes(node.parentNode.tagName) && !isFirstRecursion)
                                 notesText.push(config.newlineDelimiter ?? "\n");
                         }
                         else {
-                            xmlTextArray.push(node.nodeValue);
+                            xmlTextArray.push(value);
                             if (allowedTextTags.includes(node.parentNode.tagName) && !isFirstRecursion)
                                 xmlTextArray.push(config.newlineDelimiter ?? "\n");
                         }
@@ -450,7 +461,8 @@ function parseOpenOffice(file, callback, config) {
  */
 async function parsePdf(file, callback, config) {
     // Wait for pdfjs module to be loaded once
-    const pdfjs = await pdfjsPromise;
+    // Lazy import pdfjs to avoid Node startup issues for environments that don't use PDF parsing
+    const pdfjs = await import('pdfjs-dist/legacy/build/pdf.mjs');
 
     // Get the pdfjs document for the filepath or Uint8Array buffers.
     // pdfjs does not accept Buffers directly, so we convert them to Uint8Array.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "officeparser",
-  "version": "5.2.0",
+  "version": "5.2.2",
   "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": [

package/typings/officeParser.d.ts

@@ -18,16 +18,16 @@ export type OfficeParserConfig = {
     putNotesAtLast?: boolean;
 };
 /** Main async function with callback to execute parseOffice for supported files
- * @param {string | Buffer}    file        File path or file buffers
- * @param {function}           callback    Callback function that returns value or error
- * @param {OfficeParserConfig} [config={}] [OPTIONAL]: Config Object for officeParser
+ * @param {string | Buffer | ArrayBuffer} srcFile      File path or file buffers or Javascript ArrayBuffer
+ * @param {function}                      callback     Callback function that returns value or error
+ * @param {OfficeParserConfig}            [config={}]  [OPTIONAL]: Config Object for officeParser
  * @returns {void}
  */
-export function parseOffice(file: string | Buffer, callback: Function, config?: OfficeParserConfig): void;
+export function parseOffice(srcFile: string | Buffer | ArrayBuffer, callback: Function, config?: OfficeParserConfig): void;
 /** Main async function that can be used with await to execute parseOffice. Or it can be used with promises.
- * @param {string | Buffer}    file        File path or file buffers
- * @param {OfficeParserConfig} [config={}] [OPTIONAL]: Config Object for officeParser
+ * @param {string | Buffer | ArrayBuffer} srcFile     File path or file buffers or Javascript ArrayBuffer
+ * @param {OfficeParserConfig}            [config={}] [OPTIONAL]: Config Object for officeParser
  * @returns {Promise<string>}
  */
-export function parseOfficeAsync(file: string | Buffer, config?: OfficeParserConfig): Promise<string>;
+export function parseOfficeAsync(srcFile: string | Buffer | ArrayBuffer, config?: OfficeParserConfig): Promise<string>;
 //# sourceMappingURL=officeParser.d.ts.map