From de34fed8346e5ede58018fda1df3c7970e10ec2d Mon Sep 17 00:00:00 2001
From: Deependra Solanki <deso@odoo.com>
Date: Wed, 20 Sep 2023 13:05:05 +0000
Subject: [PATCH] [FIX] web_editor: improve url regex

Before this commit:

There were certain issues with the URL regex, wherein it failed to validate
against domain list under certain circumstances involving URLs with two dots.
for eg. this.document.anything

After this commit:

The URL regex is modified to ensure that if a URL regex does not ends with
the domains in tldWhiteList, it must start with http://, https://, www. or
if it does not start with http://, https://, www., it must end with a
domain from the tldWhiteList.

task-3468646

Part-of: odoo/odoo#143250
---
 .../static/src/js/editor/odoo-editor/src/utils/utils.js       | 2 +-
 .../static/src/js/editor/odoo-editor/test/spec/link.test.js   | 2 +-
 .../src/js/editor/odoo-editor/test/spec/urlRegex.test.js      | 4 ++--
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/addons/web_editor/static/src/js/editor/odoo-editor/src/utils/utils.js b/addons/web_editor/static/src/js/editor/odoo-editor/src/utils/utils.js
index bbfe6d81b52..f4b9626cee5 100644
--- a/addons/web_editor/static/src/js/editor/odoo-editor/src/utils/utils.js
+++ b/addons/web_editor/static/src/js/editor/odoo-editor/src/utils/utils.js
@@ -50,7 +50,7 @@ const tldWhitelist = [
     'ug', 'uk', 'um', 'us', 'uy', 'uz', 'va', 'vc', 've', 'vg', 'vi', 'vn',
     'vu', 'wf', 'ws', 'ye', 'yt', 'yu', 'za', 'zm', 'zr', 'zw', 'co\\.uk'];
 
-const urlRegexBase = `|(?:[-a-zA-Z0-9@:%._\\+~#=]{1,64}\\.))[-a-zA-Z0-9@:%._\\+~#=]{2,256}\\.[a-zA-Z][a-zA-Z0-9]{1,62}|(?:[-a-zA-Z0-9@:%._\\+~#=]{2,256}\\.(?:${tldWhitelist.join('|')})\\b))(?:\\/[^\\s.,'")}\\]]*[?#.,)}\\]'"]?[^\\s.,'")}\\]]+|(?:[^!(){}.,\\[\\]'"\\s]+)|\\/[^\\s.,'")}\\]]*[^\\s?#.,'")}\\]]*)?`;
+const urlRegexBase = `|(?:www.))[-a-zA-Z0-9@:%._\\+~#=]{2,256}\\.[a-zA-Z][a-zA-Z0-9]{1,62}|(?:[-a-zA-Z0-9@:%._\\+~#=]{2,256}\\.(?:${tldWhitelist.join('|')})\\b))(?:\\/[^\\s.,'")}\\]]*[?#.,)}\\]'"]?[^\\s.,'")}\\]]+|(?:[^!(){}.,\\[\\]'"\\s]+)|\\/[^\\s.,'")}\\]]*[^\\s?#.,'")}\\]]*)?`;
 const httpCapturedRegex= `(https?:\\/\\/)`;
 
 export const URL_REGEX = new RegExp(`((?:(?:${httpCapturedRegex}${urlRegexBase})`, 'i');
diff --git a/addons/web_editor/static/src/js/editor/odoo-editor/test/spec/link.test.js b/addons/web_editor/static/src/js/editor/odoo-editor/test/spec/link.test.js
index 92a8423f45b..257ff164f2b 100644
--- a/addons/web_editor/static/src/js/editor/odoo-editor/test/spec/link.test.js
+++ b/addons/web_editor/static/src/js/editor/odoo-editor/test/spec/link.test.js
@@ -87,7 +87,7 @@ describe('Link', () => {
         testUrlRegex('https://www.google.com/');
         testNotUrlRegex('google.shop/');
         testUrlRegex('http://google.com/foo#test');
-        testUrlRegex('a.bcd.ef');
+        testNotUrlRegex('a.bcd.ef');
         testUrlRegex('a.bc.de');
         testNotUrlRegex('a.bc.d');
         testNotUrlRegex('a.b.bc');
diff --git a/addons/web_editor/static/src/js/editor/odoo-editor/test/spec/urlRegex.test.js b/addons/web_editor/static/src/js/editor/odoo-editor/test/spec/urlRegex.test.js
index 265f901e83f..f812e87013c 100644
--- a/addons/web_editor/static/src/js/editor/odoo-editor/test/spec/urlRegex.test.js
+++ b/addons/web_editor/static/src/js/editor/odoo-editor/test/spec/urlRegex.test.js
@@ -15,8 +15,8 @@ describe('urlRegex', () => {
         const match = text.match(URL_REGEX);
         chai.expect(match).to.be.equal(null);
     });
-    it('should match abc.abc.abc', () => {
-        const url = 'abc.abc.abc';
+    it('should match www.abc.abc', () => {
+        const url = 'www.abc.abc';
         const text = `abc ${url} abc`;
         const match = text.match(URL_REGEX);
         chai.expect(match[0]).to.be.equal(url);