Add OGP parser

3 files changed, 47 insertions, 0 deletions

diff --git a/lib/pleroma/web/rich_media/data.ex b/lib/pleroma/web/rich_media/data.ex
new file mode 100644
index 000000000..403d1d341
--- /dev/null
+++ b/lib/pleroma/web/rich_media/data.ex
@@ -0,0 +1,3 @@
+defmodule Pleroma.Web.RichMedia.Data do
+  defstruct [:title, :type, :image, :url, :description]
+end
diff --git a/lib/pleroma/web/rich_media/parser.ex b/lib/pleroma/web/rich_media/parser.ex
new file mode 100644
index 000000000..d9c1684d5
--- /dev/null
+++ b/lib/pleroma/web/rich_media/parser.ex
@@ -0,0 +1,14 @@
+defmodule Pleroma.Web.RichMedia.Parser do
+  @parsers [Pleroma.Web.RichMedia.Parsers.OGP]
+
+  def parse(url) do
+    {:ok, %Tesla.Env{body: html}} = Pleroma.HTTP.get(url)
+
+    Enum.reduce_while(@parsers, %Pleroma.Web.RichMedia.Data{}, fn parser, acc ->
+      case parser.parse(html, acc) do
+        {:ok, data} -> {:halt, data}
+        {:error, _msg} -> {:cont, acc}
+      end
+    end)
+  end
+end
diff --git a/lib/pleroma/web/rich_media/parsers/ogp.ex b/lib/pleroma/web/rich_media/parsers/ogp.ex
new file mode 100644
index 000000000..75084c7ee
--- /dev/null
+++ b/lib/pleroma/web/rich_media/parsers/ogp.ex
@@ -0,0 +1,30 @@
+defmodule Pleroma.Web.RichMedia.Parsers.OGP do
+  def parse(html, data) do
+    with elements = [_ | _] <- get_elements(html),
+         ogp_data =
+           Enum.reduce(elements, data, fn el, acc ->
+             attributes = normalize_attributes(el)
+
+             Map.merge(acc, attributes)
+           end) do
+      {:ok, ogp_data}
+    else
+      _e -> {:error, "No OGP metadata found"}
+    end
+  end
+
+  defp get_elements(html) do
+    html |> Floki.find("meta[property^='og:']")
+  end
+
+  defp normalize_attributes(tuple) do
+    {_tag, attributes, _children} = tuple
+
+    data =
+      Enum.into(attributes, %{}, fn {name, value} ->
+        {name, String.trim_leading(value, "og:")}
+      end)
+
+    %{String.to_atom(data["property"]) => data["content"]}
+  end
+end