6 files changed, 57 insertions, 53 deletions

diff --git a/lib/pleroma/web/rich_media/helpers.ex b/lib/pleroma/web/rich_media/helpers.ex
index 1729141e9..6210f2c5a 100644
--- a/lib/pleroma/web/rich_media/helpers.ex
+++ b/lib/pleroma/web/rich_media/helpers.ex
@@ -9,12 +9,17 @@ defmodule Pleroma.Web.RichMedia.Helpers do
   alias Pleroma.Object
   alias Pleroma.Web.RichMedia.Parser
 
+  @rich_media_options [
+    pool: :media,
+    max_body: 2_000_000
+  ]
+
   @spec validate_page_url(URI.t() | binary()) :: :ok | :error
   defp validate_page_url(page_url) when is_binary(page_url) do
-    validate_tld = Application.get_env(:auto_linker, :opts)[:validate_tld]
+    validate_tld = Pleroma.Config.get([Pleroma.Formatter, :validate_tld])
 
     page_url
-    |> AutoLinker.Parser.url?(scheme: true, validate_tld: validate_tld)
+    |> Linkify.Parser.url?(validate_tld: validate_tld)
     |> parse_uri(page_url)
   end
 
@@ -49,11 +54,11 @@ defmodule Pleroma.Web.RichMedia.Helpers do
     |> hd
   end
 
-  def fetch_data_for_activity(%Activity{data: %{"type" => "Create"}} = activity) do
+  def fetch_data_for_object(object) do
     with true <- Config.get([:rich_media, :enabled]),
-         %Object{} = object <- Object.normalize(activity),
          false <- object.data["sensitive"] || false,
-         {:ok, page_url} <- HTML.extract_first_external_url(object, object.data["content"]),
+         {:ok, page_url} <-
+           HTML.extract_first_external_url(object, object.data["content"]),
          :ok <- validate_page_url(page_url),
          {:ok, rich_media} <- Parser.parse(page_url) do
       %{page_url: page_url, rich_media: rich_media}
@@ -62,10 +67,35 @@ defmodule Pleroma.Web.RichMedia.Helpers do
     end
   end
 
+  def fetch_data_for_activity(%Activity{data: %{"type" => "Create"}} = activity) do
+    with true <- Config.get([:rich_media, :enabled]),
+         %Object{} = object <- Object.normalize(activity) do
+      fetch_data_for_object(object)
+    else
+      _ -> %{}
+    end
+  end
+
   def fetch_data_for_activity(_), do: %{}
 
   def perform(:fetch, %Activity{} = activity) do
     fetch_data_for_activity(activity)
     :ok
   end
+
+  def rich_media_get(url) do
+    headers = [{"user-agent", Pleroma.Application.user_agent() <> "; Bot"}]
+
+    options =
+      if Application.get_env(:tesla, :adapter) == Tesla.Adapter.Hackney do
+        Keyword.merge(@rich_media_options,
+          recv_timeout: 2_000,
+          with_body: true
+        )
+      else
+        @rich_media_options
+      end
+
+    Pleroma.HTTP.get(url, headers, options)
+  end
 end
diff --git a/lib/pleroma/web/rich_media/parser.ex b/lib/pleroma/web/rich_media/parser.ex
index 7b45ecb9c..ca592833f 100644
--- a/lib/pleroma/web/rich_media/parser.ex
+++ b/lib/pleroma/web/rich_media/parser.ex
@@ -3,13 +3,6 @@
 # SPDX-License-Identifier: AGPL-3.0-only
 
 defmodule Pleroma.Web.RichMedia.Parser do
-  @hackney_options [
-    pool: :media,
-    recv_timeout: 2_000,
-    max_body: 2_000_000,
-    with_body: true
-  ]
-
   defp parsers do
     Pleroma.Config.get([:rich_media, :parsers])
   end
@@ -78,7 +71,7 @@ defmodule Pleroma.Web.RichMedia.Parser do
 
   defp parse_url(url) do
     try do
-      {:ok, %Tesla.Env{body: html}} = Pleroma.HTTP.get(url, [], adapter: @hackney_options)
+      {:ok, %Tesla.Env{body: html}} = Pleroma.Web.RichMedia.Helpers.rich_media_get(url)
 
       html
       |> parse_html()
@@ -97,8 +90,8 @@ defmodule Pleroma.Web.RichMedia.Parser do
   defp maybe_parse(html) do
     Enum.reduce_while(parsers(), %{}, fn parser, acc ->
       case parser.parse(html, acc) do
-        {:ok, data} -> {:halt, data}
-        {:error, _msg} -> {:cont, acc}
+        data when data != %{} -> {:halt, data}
+        _ -> {:cont, acc}
       end
     end)
   end
diff --git a/lib/pleroma/web/rich_media/parsers/meta_tags_parser.ex b/lib/pleroma/web/rich_media/parsers/meta_tags_parser.ex
index 2762b5902..3d577e254 100644
--- a/lib/pleroma/web/rich_media/parsers/meta_tags_parser.ex
+++ b/lib/pleroma/web/rich_media/parsers/meta_tags_parser.ex
@@ -3,22 +3,15 @@
 # SPDX-License-Identifier: AGPL-3.0-only
 
 defmodule Pleroma.Web.RichMedia.Parsers.MetaTagsParser do
-  def parse(html, data, prefix, error_message, key_name, value_name \\ "content") do
-    meta_data =
-      html
-      |> get_elements(key_name, prefix)
-      |> Enum.reduce(data, fn el, acc ->
-        attributes = normalize_attributes(el, prefix, key_name, value_name)
-
-        Map.merge(acc, attributes)
-      end)
-      |> maybe_put_title(html)
-
-    if Enum.empty?(meta_data) do
-      {:error, error_message}
-    else
-      {:ok, meta_data}
-    end
+  def parse(data, html, prefix, key_name, value_name \\ "content") do
+    html
+    |> get_elements(key_name, prefix)
+    |> Enum.reduce(data, fn el, acc ->
+      attributes = normalize_attributes(el, prefix, key_name, value_name)
+
+      Map.merge(acc, attributes)
+    end)
+    |> maybe_put_title(html)
   end
 
   defp get_elements(html, key_name, prefix) do
diff --git a/lib/pleroma/web/rich_media/parsers/oembed_parser.ex b/lib/pleroma/web/rich_media/parsers/oembed_parser.ex
index db8ccf15d..1fe6729c3 100644
--- a/lib/pleroma/web/rich_media/parsers/oembed_parser.ex
+++ b/lib/pleroma/web/rich_media/parsers/oembed_parser.ex
@@ -7,9 +7,9 @@ defmodule Pleroma.Web.RichMedia.Parsers.OEmbed do
     with elements = [_ | _] <- get_discovery_data(html),
          oembed_url when is_binary(oembed_url) <- get_oembed_url(elements),
          {:ok, oembed_data} <- get_oembed_data(oembed_url) do
-      {:ok, oembed_data}
+      oembed_data
     else
-      _e -> {:error, "No OEmbed data found"}
+      _e -> %{}
     end
   end
 
@@ -22,7 +22,7 @@ defmodule Pleroma.Web.RichMedia.Parsers.OEmbed do
   end
 
   defp get_oembed_data(url) do
-    with {:ok, %Tesla.Env{body: json}} <- Pleroma.HTTP.get(url, [], adapter: [pool: :media]) do
+    with {:ok, %Tesla.Env{body: json}} <- Pleroma.Web.RichMedia.Helpers.rich_media_get(url) do
       Jason.decode(json)
     end
   end
diff --git a/lib/pleroma/web/rich_media/parsers/ogp.ex b/lib/pleroma/web/rich_media/parsers/ogp.ex
index 3e9012588..b3b3b059c 100644
--- a/lib/pleroma/web/rich_media/parsers/ogp.ex
+++ b/lib/pleroma/web/rich_media/parsers/ogp.ex
@@ -3,13 +3,8 @@
 # SPDX-License-Identifier: AGPL-3.0-only
 
 defmodule Pleroma.Web.RichMedia.Parsers.OGP do
-  def parse(html, data) do
-    Pleroma.Web.RichMedia.Parsers.MetaTagsParser.parse(
-      html,
-      data,
-      "og",
-      "No OGP metadata found",
-      "property"
-    )
+  @deprecated "OGP parser is deprecated. Use TwitterCard instead."
+  def parse(_html, _data) do
+    %{}
   end
 end
diff --git a/lib/pleroma/web/rich_media/parsers/twitter_card.ex b/lib/pleroma/web/rich_media/parsers/twitter_card.ex
index 09d4b526e..4a04865d2 100644
--- a/lib/pleroma/web/rich_media/parsers/twitter_card.ex
+++ b/lib/pleroma/web/rich_media/parsers/twitter_card.ex
@@ -5,18 +5,11 @@
 defmodule Pleroma.Web.RichMedia.Parsers.TwitterCard do
   alias Pleroma.Web.RichMedia.Parsers.MetaTagsParser
 
-  @spec parse(String.t(), map()) :: {:ok, map()} | {:error, String.t()}
+  @spec parse(list(), map()) :: map()
   def parse(html, data) do
     data
-    |> parse_name_attrs(html)
-    |> parse_property_attrs(html)
-  end
-
-  defp parse_name_attrs(data, html) do
-    MetaTagsParser.parse(html, data, "twitter", %{}, "name")
-  end
-
-  defp parse_property_attrs({_, data}, html) do
-    MetaTagsParser.parse(html, data, "twitter", "No twitter card metadata found", "property")
+    |> MetaTagsParser.parse(html, "og", "property")
+    |> MetaTagsParser.parse(html, "twitter", "name")
+    |> MetaTagsParser.parse(html, "twitter", "property")
   end
 end