logo

pleroma

My custom branche(s) on git.pleroma.social/pleroma/pleroma
commit: de1da7b3d1c855df87262b8d24d6cdb6a6a518bc
parent: fa0392e49cd46e4d2845840232a4b56e52729369
Author: kaniini <nenolod@gmail.com>
Date:   Mon, 14 Jan 2019 07:28:04 +0000

Merge branch 'feature/add-oembed-parser' into 'develop'

Add OEmbed parser

See merge request pleroma/pleroma!661

Diffstat:

Mlib/pleroma/web/rich_media/parser.ex6+++++-
Alib/pleroma/web/rich_media/parsers/oembed_parser.ex27+++++++++++++++++++++++++++
Atest/fixtures/rich_media/oembed.html3+++
Atest/fixtures/rich_media/oembed.json1+
Mtest/web/rich_media/parser_test.exs41+++++++++++++++++++++++++++++++++++++++++
5 files changed, 77 insertions(+), 1 deletion(-)

diff --git a/lib/pleroma/web/rich_media/parser.ex b/lib/pleroma/web/rich_media/parser.ex @@ -1,5 +1,9 @@ defmodule Pleroma.Web.RichMedia.Parser do - @parsers [Pleroma.Web.RichMedia.Parsers.OGP, Pleroma.Web.RichMedia.Parsers.TwitterCard] + @parsers [ + Pleroma.Web.RichMedia.Parsers.OGP, + Pleroma.Web.RichMedia.Parsers.TwitterCard, + Pleroma.Web.RichMedia.Parsers.OEmbed + ] if Mix.env() == :test do def parse(url), do: parse_url(url) diff --git a/lib/pleroma/web/rich_media/parsers/oembed_parser.ex b/lib/pleroma/web/rich_media/parsers/oembed_parser.ex @@ -0,0 +1,27 @@ +defmodule Pleroma.Web.RichMedia.Parsers.OEmbed do + def parse(html, _data) do + with elements = [_ | _] <- get_discovery_data(html), + {:ok, oembed_url} <- get_oembed_url(elements), + {:ok, oembed_data} <- get_oembed_data(oembed_url) do + {:ok, oembed_data} + else + _e -> {:error, "No OEmbed data found"} + end + end + + defp get_discovery_data(html) do + html |> Floki.find("link[type='application/json+oembed']") + end + + defp get_oembed_url(nodes) do + {"link", attributes, _children} = nodes |> hd() + + {:ok, Enum.into(attributes, %{})["href"]} + end + + defp get_oembed_data(url) do + {:ok, %Tesla.Env{body: json}} = Pleroma.HTTP.get(url) + + {:ok, Poison.decode!(json)} + end +end diff --git a/test/fixtures/rich_media/oembed.html b/test/fixtures/rich_media/oembed.html @@ -0,0 +1,3 @@ +<link rel="alternate" type="application/json+oembed" + href="http://example.com/oembed.json" + title="Bacon Lollys oEmbed Profile" /> diff --git a/test/fixtures/rich_media/oembed.json b/test/fixtures/rich_media/oembed.json @@ -0,0 +1 @@ +{"type":"photo","flickr_type":"photo","title":"Bacon Lollys","author_name":"\u202e\u202d\u202cbees\u202c","author_url":"https:\/\/www.flickr.com\/photos\/bees\/","width":"1024","height":"768","url":"https:\/\/farm4.staticflickr.com\/3040\/2362225867_4a87ab8baf_b.jpg","web_page":"https:\/\/www.flickr.com\/photos\/bees\/2362225867\/","thumbnail_url":"https:\/\/farm4.staticflickr.com\/3040\/2362225867_4a87ab8baf_q.jpg","thumbnail_width":150,"thumbnail_height":150,"web_page_short_url":"https:\/\/flic.kr\/p\/4AK2sc","license":"All Rights Reserved","license_id":0,"html":"<a data-flickr-embed=\"true\" href=\"https:\/\/www.flickr.com\/photos\/bees\/2362225867\/\" title=\"Bacon Lollys by \u202e\u202d\u202cbees\u202c, on Flickr\"><img src=\"https:\/\/farm4.staticflickr.com\/3040\/2362225867_4a87ab8baf_b.jpg\" width=\"1024\" height=\"768\" alt=\"Bacon Lollys\"><\/a><script async src=\"https:\/\/embedr.flickr.com\/assets\/client-code.js\" charset=\"utf-8\"><\/script>","version":"1.0","cache_age":3600,"provider_name":"Flickr","provider_url":"https:\/\/www.flickr.com\/"} diff --git a/test/web/rich_media/parser_test.exs b/test/web/rich_media/parser_test.exs @@ -15,6 +15,18 @@ defmodule Pleroma.Web.RichMedia.ParserTest do } -> %Tesla.Env{status: 200, body: File.read!("test/fixtures/rich_media/twitter_card.html")} + %{ + method: :get, + url: "http://example.com/oembed" + } -> + %Tesla.Env{status: 200, body: File.read!("test/fixtures/rich_media/oembed.html")} + + %{ + method: :get, + url: "http://example.com/oembed.json" + } -> + %Tesla.Env{status: 200, body: File.read!("test/fixtures/rich_media/oembed.json")} + %{method: :get, url: "http://example.com/empty"} -> %Tesla.Env{status: 200, body: "hello"} end) @@ -48,4 +60,33 @@ defmodule Pleroma.Web.RichMedia.ParserTest do description: "View the album on Flickr." }} end + + test "parses OEmbed" do + assert Pleroma.Web.RichMedia.Parser.parse("http://example.com/oembed") == + {:ok, + %{ + "author_name" => "‮‭‬bees‬", + "author_url" => "https://www.flickr.com/photos/bees/", + "cache_age" => 3600, + "flickr_type" => "photo", + "height" => "768", + "html" => + "<a data-flickr-embed=\"true\" href=\"https://www.flickr.com/photos/bees/2362225867/\" title=\"Bacon Lollys by ‮‭‬bees‬, on Flickr\"><img src=\"https://farm4.staticflickr.com/3040/2362225867_4a87ab8baf_b.jpg\" width=\"1024\" height=\"768\" alt=\"Bacon Lollys\"></a><script async src=\"https://embedr.flickr.com/assets/client-code.js\" charset=\"utf-8\"></script>", + "license" => "All Rights Reserved", + "license_id" => 0, + "provider_name" => "Flickr", + "provider_url" => "https://www.flickr.com/", + "thumbnail_height" => 150, + "thumbnail_url" => + "https://farm4.staticflickr.com/3040/2362225867_4a87ab8baf_q.jpg", + "thumbnail_width" => 150, + "title" => "Bacon Lollys", + "type" => "photo", + "url" => "https://farm4.staticflickr.com/3040/2362225867_4a87ab8baf_b.jpg", + "version" => "1.0", + "web_page" => "https://www.flickr.com/photos/bees/2362225867/", + "web_page_short_url" => "https://flic.kr/p/4AK2sc", + "width" => "1024" + }} + end end