Desktop: Rich Text: Do not converts to markdown links URLs that would be linkified

.eslintignore

@@ -70,6 +70,9 @@ packages/app-cli/app/services/plugins/PluginRunner.js.map
 packages/app-cli/tests/EnexToMd.d.ts
 packages/app-cli/tests/EnexToMd.js
 packages/app-cli/tests/EnexToMd.js.map
+packages/app-cli/tests/HtmlToMd.d.ts
+packages/app-cli/tests/HtmlToMd.js
+packages/app-cli/tests/HtmlToMd.js.map
 packages/app-cli/tests/InMemoryCache.d.ts
 packages/app-cli/tests/InMemoryCache.js
 packages/app-cli/tests/InMemoryCache.js.map

.gitignore

@@ -59,6 +59,9 @@ packages/app-cli/app/services/plugins/PluginRunner.js.map
 packages/app-cli/tests/EnexToMd.d.ts
 packages/app-cli/tests/EnexToMd.js
 packages/app-cli/tests/EnexToMd.js.map
+packages/app-cli/tests/HtmlToMd.d.ts
+packages/app-cli/tests/HtmlToMd.js
+packages/app-cli/tests/HtmlToMd.js.map
 packages/app-cli/tests/InMemoryCache.d.ts
 packages/app-cli/tests/InMemoryCache.js
 packages/app-cli/tests/InMemoryCache.js.map

packages/app-cli/tests/HtmlToMd.ts

@@ -1,24 +1,15 @@
-/* eslint-disable no-unused-vars */
+import shim from '@joplin/lib/shim';
-
-
 const os = require('os');
-const time = require('@joplin/lib/time').default;
 const { filename } = require('@joplin/lib/path-utils');
-const { fileContentEqual, setupDatabase, setupDatabaseAndSynchronizer, db, synchronizer, fileApi, sleep, clearDatabase, switchClient, syncTargetId, objectsEqual, checkThrowAsync } = require('./test-utils.js');
-const Folder = require('@joplin/lib/models/Folder.js');
-const Note = require('@joplin/lib/models/Note.js');
-const BaseModel = require('@joplin/lib/BaseModel').default;
-const shim = require('@joplin/lib/shim').default;
 const HtmlToMd = require('@joplin/lib/HtmlToMd');
-const { enexXmlToMd } = require('@joplin/lib/import-enex-md-gen.js');
 
 describe('HtmlToMd', function() {
 
-	beforeEach(async (done) => {
+	// beforeEach(async (done) => {
-		await setupDatabaseAndSynchronizer(1);
+	// 	await setupDatabaseAndSynchronizer(1);
-		await switchClient(1);
+	// 	await switchClient(1);
-		done();
+	// 	done();
-	});
+	// });
 
 	it('should convert from Html to Markdown', (async () => {
 		const basePath = `${__dirname}/html_to_md`;
@@ -32,11 +23,11 @@ describe('HtmlToMd', function() {
 			const htmlPath = `${basePath}/${htmlFilename}`;
 			const mdPath = `${basePath}/${filename(htmlFilename)}.md`;
 
-			// if (htmlFilename !== 'code_3.html') continue;
+			// if (htmlFilename !== 'anchor_same_title_and_url.html') continue;
 
 			// if (htmlFilename.indexOf('image_preserve_size') !== 0) continue;
 
-			const htmlToMdOptions = {};
+			const htmlToMdOptions: any = {};
 
 			if (htmlFilename === 'anchor_local.html') {
 				// Normally the list of anchor names in the document are retrieved from the HTML code
@@ -66,15 +57,14 @@ describe('HtmlToMd', function() {
 				result.push('');
 				result.push(`Error converting file: ${htmlFilename}`);
 				result.push('--------------------------------- Got:');
-				result.push(actualMd.split('\n').map(l => `"${l}"`).join('\n'));
+				result.push(actualMd.split('\n').map((l: string) => `"${l}"`).join('\n'));
 				result.push('--------------------------------- Expected:');
-				result.push(expectedMd.split('\n').map(l => `"${l}"`).join('\n'));
+				result.push(expectedMd.split('\n').map((l: string) => `"${l}"`).join('\n'));
 				result.push('--------------------------------------------');
 				result.push('');
 
 				console.info(result.join('\n'));
 
-
 				// console.info('');
 				// console.info(`Error converting file: ${htmlFilename}`);
 				// console.info('--------------------------------- Got:');

packages/app-cli/tests/html_to_md/anchor_same_title_and_url.html

@@ -0,0 +1,8 @@
+<ul>
+	<li><a href="https://example.com"/>https://example.com</a></li>
+	<li><a href="http://example.com"/>http://example.com</a></li>
+	<li><a href="file:///mnt/c/test.txt"/>file:///mnt/c/test.txt</a></li>
+	<li><a href="https://example.com" title="with title"/>https://example.com</a></li>
+	<li><a href="example.com"/>example.com</a></li>
+	<li><a href="test@example.com"/>test@example.com</a></li>
+</ul>

packages/app-cli/tests/html_to_md/anchor_same_title_and_url.md

@@ -0,0 +1,6 @@
+- https://example.com
+- http://example.com
+- file:///mnt/c/test.txt
+- [https://example.com](https://example.com "with title")
+- [example.com](example.com)
+- [test@example.com](test@example.com)

packages/turndown/src/commonmark-rules.js

@@ -216,6 +216,10 @@ function getNamedAnchorFromLink(node, options) {
   }
 }
 
+function isLinkifiedUrl(url) {
+  return url.indexOf('http://') === 0 || url.indexOf('https://') === 0 || url.indexOf('file://') === 0;
+}
+
 rules.inlineLink = {
   filter: function (node, options) {
     return (
@@ -227,12 +231,32 @@ rules.inlineLink = {
 
   replacement: function (content, node, options) {
     var href = filterLinkHref(node.getAttribute('href'))
+
     if (!href) {
       return getNamedAnchorFromLink(node, options) + filterLinkContent(content)
     } else {
       var title = node.title && node.title !== href ? ' "' + node.title + '"' : ''
       if (!href) title = ''
-      return getNamedAnchorFromLink(node, options) + '[' + filterLinkContent(content) + '](' + href + title + ')'
+      let output = getNamedAnchorFromLink(node, options) + '[' + filterLinkContent(content) + '](' + href + title + ')'
+
+      // If the URL is automatically linkified by Joplin, and the title is
+      // the same as the URL, there is no need to make it a link here. That
+      // will prevent URsL from the rich text editor to be needlessly
+      // converted from this:
+      //
+      // <a href="https://example.com">https://example.com</a>
+      //
+      // to this:
+      //
+      // [https://example.com](https://example.com)
+      //
+      // It means cleaner Markdown will also be generated by the web
+      // clipper.
+      if (isLinkifiedUrl(href)) {
+        if (output === '[' + href + '](' + href + ')') return href;
+      }
+
+      return output;
     }
   }
 }