Fix Rich Media Previews for updated activities

The Rich Media Previews were not regenerated when a post was updated due to a cache invalidation issue. They are now cached by the activity id so they can be evicted with the other activity cache objects in the :scrubber_cache.

4 files changed, 63 insertions, 1 deletions

diff --git a/test/fixtures/rich_media/google.html b/test/fixtures/rich_media/google.html
new file mode 100644
index 000000000..c068397a5
--- /dev/null
+++ b/test/fixtures/rich_media/google.html
@@ -0,0 +1,12 @@
+<meta property="og:url" content="https://google.com">
+<meta property="og:type" content="website">
+<meta property="og:title" content="Google">
+<meta property="og:description" content="Search the world's information, including webpages, images, videos and more. Google has many special features to help you find exactly what you're looking for.">
+<meta property="og:image" content="">
+
+<meta name="twitter:card" content="summary_large_image">
+<meta property="twitter:domain" content="google.com">
+<meta property="twitter:url" content="https://google.com">
+<meta name="twitter:title" content="Google">
+<meta name="twitter:description" content="Search the world's information, including webpages, images, videos and more. Google has many special features to help you find exactly what you're looking for.">
+<meta name="twitter:image" content="">
diff --git a/test/fixtures/rich_media/yahoo.html b/test/fixtures/rich_media/yahoo.html
new file mode 100644
index 000000000..41d8c5cd9
--- /dev/null
+++ b/test/fixtures/rich_media/yahoo.html
@@ -0,0 +1,12 @@
+<meta property="og:url" content="https://yahoo.com">
+<meta property="og:type" content="website">
+<meta property="og:title" content="Yahoo | Mail, Weather, Search, Politics, News, Finance, Sports & Videos">
+<meta property="og:description" content="Latest news coverage, email, free stock quotes, live scores and video are just the beginning. Discover more every day at Yahoo!">
+<meta property="og:image" content="https://s.yimg.com/cv/apiv2/social/images/yahoo_default_logo.png">
+
+<meta name="twitter:card" content="summary_large_image">
+<meta property="twitter:domain" content="yahoo.com">
+<meta property="twitter:url" content="https://yahoo.com">
+<meta name="twitter:title" content="Yahoo | Mail, Weather, Search, Politics, News, Finance, Sports & Videos">
+<meta name="twitter:description" content="Latest news coverage, email, free stock quotes, live scores and video are just the beginning. Discover more every day at Yahoo!">
+<meta name="twitter:image" content="https://s.yimg.com/cv/apiv2/social/images/yahoo_default_logo.png">
diff --git a/test/pleroma/web/rich_media/helpers_test.exs b/test/pleroma/web/rich_media/helpers_test.exs
index 3ef5705ce..8f6713ef8 100644
--- a/test/pleroma/web/rich_media/helpers_test.exs
+++ b/test/pleroma/web/rich_media/helpers_test.exs
@@ -83,6 +83,34 @@ defmodule Pleroma.Web.RichMedia.HelpersTest do
              Pleroma.Web.RichMedia.Helpers.fetch_data_for_activity(activity)
   end
 
+  test "recrawls URLs on updates" do
+    original_url = "https://google.com/"
+    updated_url = "https://yahoo.com/"
+
+    Pleroma.StaticStubbedConfigMock
+    |> stub(:get, fn
+      [:rich_media, :enabled] -> true
+      path -> Pleroma.Test.StaticConfig.get(path)
+    end)
+
+    user = insert(:user)
+    {:ok, activity} = CommonAPI.post(user, %{status: "I like this site #{original_url}"})
+
+    assert match?(
+             %{page_url: ^original_url, rich_media: _},
+             Pleroma.Web.RichMedia.Helpers.fetch_data_for_activity(activity)
+           )
+
+    {:ok, _} = CommonAPI.update(user, activity, %{status: "I like this site #{updated_url}"})
+
+    activity = Pleroma.Activity.get_by_id(activity.id)
+
+    assert match?(
+             %{page_url: ^updated_url, rich_media: _},
+             Pleroma.Web.RichMedia.Helpers.fetch_data_for_activity(activity)
+           )
+  end
+
   # This does not seem to work. The urls are being fetched.
   @tag skip: true
   test "refuses to crawl URLs of private network from posts" do
diff --git a/test/support/http_request_mock.ex b/test/support/http_request_mock.ex
index f76128312..b220fd051 100644
--- a/test/support/http_request_mock.ex
+++ b/test/support/http_request_mock.ex
@@ -1464,6 +1464,14 @@ defmodule HttpRequestMock do
      }}
   end
 
+  def get("https://google.com/", _, _, _) do
+    {:ok, %Tesla.Env{status: 200, body: File.read!("test/fixtures/rich_media/google.html")}}
+  end
+
+  def get("https://yahoo.com/", _, _, _) do
+    {:ok, %Tesla.Env{status: 200, body: File.read!("test/fixtures/rich_media/yahoo.html")}}
+  end
+
   def get(url, query, body, headers) do
     {:error,
      "Mock response not implemented for GET #{inspect(url)}, #{query}, #{inspect(body)}, #{inspect(headers)}"}
@@ -1539,7 +1547,9 @@ defmodule HttpRequestMock do
   @rich_media_mocks [
     "https://example.com/ogp",
     "https://example.com/ogp-missing-data",
-    "https://example.com/twitter-card"
+    "https://example.com/twitter-card",
+    "https://google.com/",
+    "https://yahoo.com/"
   ]
   def head(url, _query, _body, _headers) when url in @rich_media_mocks do
     {:ok, %Tesla.Env{status: 404, body: ""}}