From 1e992d9193647a3bdff913fe0bc12feff288826d Mon Sep 17 00:00:00 2001
From: Bianca Nenciu <nbianca@users.noreply.github.com>
Date: Fri, 18 Jun 2021 18:55:24 +0300
Subject: [PATCH] FIX: Do not check for duplicate links in Onebox (#13345)

If a user posted a URL that appeared inside a Onebox, then the user
got a duplicate link notice. This was fixed by skipping those links in
Ruby.

If a user posted a URL that was Oneboxes and contained other links that
appeared in previous posts, then the user got a duplicate link notice.
This was fixed by skipping those links in JavaScript.
---
 .../discourse/app/controllers/composer.js        | 10 +++++++++-
 lib/pretty_text.rb                               |  7 +++++--
 spec/components/pretty_text_spec.rb              | 16 ++++++++++++++++
 3 files changed, 30 insertions(+), 3 deletions(-)

diff --git a/app/assets/javascripts/discourse/app/controllers/composer.js b/app/assets/javascripts/discourse/app/controllers/composer.js
index 0c4be8430c4..12c76c2547a 100644
--- a/app/assets/javascripts/discourse/app/controllers/composer.js
+++ b/app/assets/javascripts/discourse/app/controllers/composer.js
@@ -456,7 +456,7 @@ export default Controller.extend({
       $links.each((idx, l) => {
         const href = l.href;
         if (href && href.length) {
-          // skip links in quotes
+          // skip links in quotes and oneboxes
           for (let element = l; element; element = element.parentElement) {
             if (
               element.tagName === "DIV" &&
@@ -471,6 +471,14 @@ export default Controller.extend({
             ) {
               return true;
             }
+
+            if (
+              element.tagName === "ASIDE" &&
+              element.classList.contains("onebox") &&
+              href !== element.dataset["onebox-src"]
+            ) {
+              return true;
+            }
           }
 
           const [warn, info] = linkLookup.check(post, href);
diff --git a/lib/pretty_text.rb b/lib/pretty_text.rb
index cb897760c3c..a298ce4f80a 100644
--- a/lib/pretty_text.rb
+++ b/lib/pretty_text.rb
@@ -324,8 +324,11 @@ module PrettyText
     links = []
     doc = Nokogiri::HTML5.fragment(html)
 
-    # remove href inside quotes & elided part
-    doc.css("aside.quote a, .elided a").each { |a| a["href"] = "" }
+    # extract onebox links
+    doc.css("aside.onebox[data-onebox-src]").each { |onebox| links << DetectedLink.new(onebox["data-onebox-src"], false) }
+
+    # remove href inside quotes & oneboxes & elided part
+    doc.css("aside.quote a, aside.onebox a, .elided a").remove
 
     # extract all links
     doc.css("a").each do |a|
diff --git a/spec/components/pretty_text_spec.rb b/spec/components/pretty_text_spec.rb
index 3d31c1ae1e3..c2e5a611a54 100644
--- a/spec/components/pretty_text_spec.rb
+++ b/spec/components/pretty_text_spec.rb
@@ -779,6 +779,22 @@ describe PrettyText do
       ].sort)
     end
 
+    it "should not extract links inside oneboxes" do
+      onebox = <<~EOF
+        <aside class="onebox twitterstatus" data-onebox-src="https://twitter.com/EDBPostgres/status/1402528437441634306">
+          <header class="source">
+            <a href="https://twitter.com/EDBPostgres/status/1402528437441634306" target="_blank" rel="noopener">twitter.com</a>
+            <a href="https://twitter.com/EDBPostgres/status/1402528437441634306" target="_blank" rel="noopener">twitter.com</a>
+          </header>
+          <article class="onebox-body">
+            <div class="tweet">Example URL: <a target="_blank" href="https://example.com" rel="noopener">example.com</a></div>
+          </article>
+        </aside>
+      EOF
+
+      expect(PrettyText.extract_links(onebox).map(&:url)).to contain_exactly("https://twitter.com/EDBPostgres/status/1402528437441634306")
+    end
+
     it "should not preserve tags in code blocks" do
       expect(PrettyText.excerpt("<pre><code class='handlebars'>&lt;h3&gt;Hours&lt;/h3&gt;</code></pre>", 100)).to eq("&lt;h3&gt;Hours&lt;/h3&gt;")
     end