From 2aab4e03c3a2867abd4555dc776eebc8b0dba176 Mon Sep 17 00:00:00 2001 From: Maxim Filippov Date: Tue, 1 Jan 2019 23:26:40 +0300 Subject: [PATCH] Add OGP parser --- lib/pleroma/web/rich_media/data.ex | 3 +++ lib/pleroma/web/rich_media/parser.ex | 14 +++++++++++ lib/pleroma/web/rich_media/parsers/ogp.ex | 30 +++++++++++++++++++++++ mix.exs | 3 ++- mix.lock | 2 ++ test/fixtures/rich_media/ogp.html | 9 +++++++ test/web/rich_media/parser_test.exs | 26 ++++++++++++++++++++ 7 files changed, 86 insertions(+), 1 deletion(-) create mode 100644 lib/pleroma/web/rich_media/data.ex create mode 100644 lib/pleroma/web/rich_media/parser.ex create mode 100644 lib/pleroma/web/rich_media/parsers/ogp.ex create mode 100644 test/fixtures/rich_media/ogp.html create mode 100644 test/web/rich_media/parser_test.exs 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 diff --git a/mix.exs b/mix.exs index efdd4b7ed..0b8052d0a 100644 --- a/mix.exs +++ b/mix.exs @@ -75,7 +75,8 @@ defp deps do {:web_push_encryption, "~> 0.2.1"}, {:swoosh, "~> 0.20"}, {:gen_smtp, "~> 0.13"}, - {:websocket_client, git: "https://github.com/jeremyong/websocket_client.git", only: :test} + {:websocket_client, git: "https://github.com/jeremyong/websocket_client.git", only: :test}, + {:floki, "~> 0.20.0"} ] end diff --git a/mix.lock b/mix.lock index 323f55bb9..d6836231d 100644 --- a/mix.lock +++ b/mix.lock @@ -20,9 +20,11 @@ "ex_aws_s3": {:hex, :ex_aws_s3, "2.0.1", "9e09366e77f25d3d88c5393824e613344631be8db0d1839faca49686e99b6704", [:mix], [{:ex_aws, "~> 2.0", [hex: :ex_aws, repo: "hexpm", optional: false]}, {:sweet_xml, ">= 0.0.0", [hex: :sweet_xml, repo: "hexpm", optional: true]}], "hexpm"}, "ex_doc": {:hex, :ex_doc, "0.19.1", "519bb9c19526ca51d326c060cb1778d4a9056b190086a8c6c115828eaccea6cf", [:mix], [{:earmark, "~> 1.1", [hex: :earmark, repo: "hexpm", optional: false]}, {:makeup_elixir, "~> 0.7", [hex: :makeup_elixir, repo: "hexpm", optional: false]}], "hexpm"}, "ex_machina": {:hex, :ex_machina, "2.2.0", "fec496331e04fc2db2a1a24fe317c12c0c4a50d2beb8ebb3531ed1f0d84be0ed", [:mix], [{:ecto, "~> 2.1", [hex: :ecto, repo: "hexpm", optional: true]}], "hexpm"}, + "floki": {:hex, :floki, "0.20.4", "be42ac911fece24b4c72f3b5846774b6e61b83fe685c2fc9d62093277fb3bc86", [:mix], [{:html_entities, "~> 0.4.0", [hex: :html_entities, repo: "hexpm", optional: false]}, {:mochiweb, "~> 2.15", [hex: :mochiweb, repo: "hexpm", optional: false]}], "hexpm"}, "gen_smtp": {:hex, :gen_smtp, "0.13.0", "11f08504c4bdd831dc520b8f84a1dce5ce624474a797394e7aafd3c29f5dcd25", [:rebar3], [], "hexpm"}, "gettext": {:hex, :gettext, "0.15.0", "40a2b8ce33a80ced7727e36768499fc9286881c43ebafccae6bab731e2b2b8ce", [:mix], [], "hexpm"}, "hackney": {:hex, :hackney, "1.14.3", "b5f6f5dcc4f1fba340762738759209e21914516df6be440d85772542d4a5e412", [:rebar3], [{:certifi, "2.4.2", [hex: :certifi, repo: "hexpm", optional: false]}, {:idna, "6.0.0", [hex: :idna, repo: "hexpm", optional: false]}, {:metrics, "1.0.1", [hex: :metrics, repo: "hexpm", optional: false]}, {:mimerl, "1.0.2", [hex: :mimerl, repo: "hexpm", optional: false]}, {:ssl_verify_fun, "1.1.4", [hex: :ssl_verify_fun, repo: "hexpm", optional: false]}], "hexpm"}, + "html_entities": {:hex, :html_entities, "0.4.0", "f2fee876858cf6aaa9db608820a3209e45a087c5177332799592142b50e89a6b", [:mix], [], "hexpm"}, "html_sanitize_ex": {:hex, :html_sanitize_ex, "1.3.0", "f005ad692b717691203f940c686208aa3d8ffd9dd4bb3699240096a51fa9564e", [:mix], [{:mochiweb, "~> 2.15", [hex: :mochiweb, repo: "hexpm", optional: false]}], "hexpm"}, "httpoison": {:hex, :httpoison, "1.2.0", "2702ed3da5fd7a8130fc34b11965c8cfa21ade2f232c00b42d96d4967c39a3a3", [:mix], [{:hackney, "~> 1.8", [hex: :hackney, repo: "hexpm", optional: false]}], "hexpm"}, "idna": {:hex, :idna, "6.0.0", "689c46cbcdf3524c44d5f3dde8001f364cd7608a99556d8fbd8239a5798d4c10", [:rebar3], [{:unicode_util_compat, "0.4.1", [hex: :unicode_util_compat, repo: "hexpm", optional: false]}], "hexpm"}, diff --git a/test/fixtures/rich_media/ogp.html b/test/fixtures/rich_media/ogp.html new file mode 100644 index 000000000..c886b5871 --- /dev/null +++ b/test/fixtures/rich_media/ogp.html @@ -0,0 +1,9 @@ + + + The Rock (1996) + + + + + + diff --git a/test/web/rich_media/parser_test.exs b/test/web/rich_media/parser_test.exs new file mode 100644 index 000000000..bb0d663e9 --- /dev/null +++ b/test/web/rich_media/parser_test.exs @@ -0,0 +1,26 @@ +defmodule Pleroma.Web.RichMedia.ParserTest do + use ExUnit.Case, async: true + + setup do + Tesla.Mock.mock(fn + %{ + method: :get, + url: "http://example.com/ogp" + } -> + %Tesla.Env{status: 200, body: File.read!("test/fixtures/rich_media/ogp.html")} + end) + + :ok + end + + test "parses ogp" do + assert Pleroma.Web.RichMedia.Parser.parse("http://example.com/ogp") == + %Pleroma.Web.RichMedia.Data{ + description: nil, + image: "http://ia.media-imdb.com/images/rock.jpg", + title: "The Rock", + type: "video.movie", + url: "http://www.imdb.com/title/tt0117500/" + } + end +end