fix: use html-minifier-terser instead of regex. feat: add option to s…

…kip pre-processing (#51) * feat: add option to skip pre-processing * fix: remove readme issue * feat: initial implementation of html-minifier-terser * chore: terser improvements
TurboDocx · Oct 4, 2024 · ce2e9d8 · ce2e9d8
1 parent 52dd6bd
commit ce2e9d8
Show file tree

Hide file tree

Showing 10 changed files with 343 additions and 143 deletions.
diff --git a/README.md b/README.md
@@ -81,6 +81,8 @@ full fledged examples can be found under `example/`
     - `defaultOrderedListStyleType` <?[String]> default ordered list style type. Defaults to `decimal`.
   - `decodeUnicode` <?[Boolean]> flag to enable unicode decoding of header, body and footer. Defaults to `false`.
   - `lang` <?[String]> language localization code for spell checker to work properly. Defaults to `en-US`.
+  - `preProcessing` <?[Object]>
+    - `skipHTMLMinify` <?[Boolean]> flag to skip minification of HTML. Defaults to `false`.
 - `footerHTMLString` <[String]> clean html string equivalent of footer. Defaults to `<p></p>` if footer flag is `true`.
 
 ### Returns

diff --git a/example/example-node.js b/example/example-node.js
@@ -1654,6 +1654,7 @@ const htmlString = `<!DOCTYPE html>
         table: { row: { cantSplit: true } },
         footer: true,
         pageNumber: true,
+        preprocessing: { skipHTMLMinify: false }
     });
 
     fs.writeFile(filePath, fileBuffer, (error) => {

diff --git a/example/example.js b/example/example.js
@@ -1627,6 +1627,7 @@ const htmlString = `<!DOCTYPE html>
         table: { row: { cantSplit: true } },
         footer: true,
         pageNumber: true,
+        preprocessing: { skipHTMLMinify: false }
     });
 
     fs.writeFile(filePath, fileBuffer, (error) => {

diff --git a/example/react-example/src/App.js b/example/react-example/src/App.js
@@ -1625,6 +1625,7 @@ function App() {
       table: { row: { cantSplit: true } },
       footer: true,
       pageNumber: true,
+      preprocessing: { skipHTMLMinify: false }
     });
 
     saveAs(fileBuffer, 'html-to-docx.docx');

diff --git a/index.js b/index.js
@@ -1,18 +1,17 @@
 /* eslint-disable no-useless-escape */
 import JSZip from 'jszip';
+import { minify } from 'html-minifier-terser';
+
+import createDocumentOptionsAndMergeWithDefaults from './src/utils/options-utils';
 import addFilesToContainer from './src/html-to-docx';
 
-const minifyHTMLString = (htmlString) => {
+const minifyHTMLString = async (htmlString) => {
   try {
     if (typeof htmlString === 'string' || htmlString instanceof String) {
-      const minifiedHTMLString = htmlString
-        .replace(/\n/g, ' ')
-        .replace(/\r/g, ' ')
-        .replace(/\r\n/g, ' ')
-        .replace(/[\t]+\</g, '<')
-        .replace(/\>[\t ]+\</g, '><')
-        .replace(/\>[\t ]+$/g, '>');
-
+      const minifiedHTMLString = await minify(htmlString, {
+        collapseWhitespace: true,
+        removeComments: true,
+      });
       return minifiedHTMLString;
     }
 
@@ -30,20 +29,22 @@ async function generateContainer(
 ) {
   const zip = new JSZip();
 
+  const normalizedDocumentOptions = createDocumentOptionsAndMergeWithDefaults(documentOptions);
+
   let contentHTML = htmlString;
   let headerHTML = headerHTMLString;
   let footerHTML = footerHTMLString;
-  if (htmlString) {
-    contentHTML = minifyHTMLString(contentHTML);
+  if (htmlString && !normalizedDocumentOptions['preprocessing']['skipHTMLMinify']) {
+    contentHTML = await minifyHTMLString(contentHTML);
   }
-  if (headerHTMLString) {
-    headerHTML = minifyHTMLString(headerHTML);
+  if (headerHTMLString && !normalizedDocumentOptions['preprocessing']['skipHTMLMinify']) {
+    headerHTML = await minifyHTMLString(headerHTML);
   }
-  if (footerHTMLString) {
-    footerHTML = minifyHTMLString(footerHTML);
+  if (footerHTMLString && !normalizedDocumentOptions['preprocessing']['skipHTMLMinify']) {
+    footerHTML = await minifyHTMLString(footerHTML);
   }
 
-  await addFilesToContainer(zip, contentHTML, documentOptions, headerHTML, footerHTML);
+  await addFilesToContainer(zip, contentHTML, normalizedDocumentOptions, headerHTML, footerHTML);
 
   const buffer = await zip.generateAsync({ type: 'arraybuffer' });
   if (Object.prototype.hasOwnProperty.call(global, 'Buffer')) {