better formatting for titles

config/config.exs

@@ -282,8 +282,11 @@
 
 config :pleroma, :feed,
   post_title: %{
+    # Set max_length to 0 to suppress titles (Dave Winer suggestion)
     max_length: 100,
-    omission: "..."
+    omission: "...",
+    # New method to extract title
+    parse_source: false
   }
 
 config :pleroma, :markup,

config/description.exs

@@ -2919,6 +2919,7 @@
   %{
     group: :pleroma,
     key: :feed,
+    label: "RSS Feeds",
     type: :group,
     description: "Configure feed rendering",
     children: [
@@ -2938,7 +2939,13 @@
             type: :string,
             description: "Replacement which will be used after truncating string",
             suggestions: ["..."]
-          }
+          },
+          %{
+            key: :parse_source,
+            type: :boolean,
+            description: "Use content type-specific parsers to extract title (ignores max_length)",
+            suggestions: [true]
+          },
         ]
       }
     ]

lib/pleroma/formatter.ex

@@ -150,15 +150,19 @@ def html_escape(text, "text/plain") do
     |> Enum.join("")
   end
 
-  def truncate(text, max_length \\ 200, omission \\ "...") do
+  def truncate(text, max_length \\ 200, omission \\ "...") when max_length >= 0 do
     # Remove trailing whitespace
     text = Regex.replace(~r/([^ \t\r\n])([ \t]+$)/u, text, "\\g{1}")
 
-    if String.length(text) < max_length do
-      text
-    else
-      length_with_omission = max_length - String.length(omission)
-      String.slice(text, 0, length_with_omission) <> omission
+    length_with_omission = max_length - String.length(omission)
+
+    cond do
+      String.length(text) <= max_length ->
+        text
+      length_with_omission > 0 ->
+        String.slice(text, 0, length_with_omission) <> omission
+      true ->
+        String.slice(text, 0, max_length)
     end
   end
 

lib/pleroma/web/feed/feed_view.ex

@@ -12,6 +12,7 @@ defmodule Pleroma.Web.Feed.FeedView do
   alias Pleroma.Web.Gettext
   alias Pleroma.Web.MediaProxy
 
+  require Logger
   require Pleroma.Constants
 
   @spec pub_date(String.t() | DateTime.t()) :: String.t()
@@ -69,7 +70,24 @@ def logo(user) do
 
   def last_activity(activities), do: List.last(activities)
 
-  def activity_title(%{"content" => content}, opts \\ %{}) do
+  def maybe_activity_title(activity, opts \\ %{}) do
+    case activity_title(activity, opts) do
+      "" -> ""
+      title -> "<title>#{title}</title>"
+    end
+  end
+
+  def activity_title(activity, opts \\ %{})
+
+  def activity_title(
+        %{"source" => %{"mediaType" => content_type, "content" => content}},
+        %{parse_source: true} = opts
+      ) do
+    split_content(content, content_type, opts) |> elem(0)
+  end
+
+  # TODO: scrub_html should replace <p> with " "
+  def activity_title(%{"content" => content}, opts) do
     content
     |> Pleroma.Web.Metadata.Utils.scrub_html()
     |> Pleroma.Emoji.Formatter.demojify()
@@ -77,13 +95,36 @@ def activity_title(%{"content" => content}, opts \\ %{}) do
     |> escape()
   end
 
-  def activity_content(%{"content" => content}) do
-    content
-    |> String.replace(~r/[\n\r]/, "")
+  def activity_title(_, _), do: ""
+
+  def activity_content(activity, opts \\ %{})
+
+  def activity_content(
+        %{"source" => %{"mediaType" => content_type, "content" => content}},
+        %{parse_source: true} = opts
+      ) do
+    start = split_content(content, content_type, opts) |> elem(1)
+    length = String.length(content)
+
+    {text, _mentions, _tags} =
+      String.slice(content, start, length)
+      |> Pleroma.Web.CommonAPI.Utils.format_input(content_type)
+
+    text
+    |> String.replace(~r/(\r?\n)+/, " ")
+    |> String.trim()
     |> escape()
   end
 
-  def activity_content(_), do: ""
+  def activity_content(%{"content" => content}, _opts) do
+    # Replace 1 or more newlines with 1 space
+    content
+    |> String.replace(~r/(\r?\n)+/, " ")
+    |> String.trim()
+    |> escape()
+  end
+
+  def activity_content(_, _), do: ""
 
   def activity_context(activity), do: escape(activity.data["context"])
 
@@ -99,6 +140,117 @@ def attachment_type(attachment) do
     |> Map.get("mediaType")
   end
 
+  def source_content(%{"source" => %{"mediaType" => _, "content" => content}}) do
+    xml_escape(content)
+  end
+
+  def parse_title(activity, opts \\ %{})
+
+  def parse_title(%{"source" => %{"mediaType" => content_type, "content" => content}}, opts) do
+    split_content(content, content_type, opts) |> elem(0)
+  end
+
+  def parse_title(_activity, _opts), do: 0
+
+  def parse_description_offset(activity, opts \\ %{})
+
+  def parse_description_offset(
+        %{"source" => %{"mediaType" => content_type, "content" => content}},
+        opts
+      ) do
+    split_content(content, content_type, opts) |> elem(1)
+  end
+
+  def parse_description_offset(_activity, _opts), do: 0
+
+  @spec split_content(binary(), binary(), any()) :: {binary(), non_neg_integer()}
+  def split_content(content, "text/html", _opts) do
+    case Regex.named_captures(
+           ~r/^[ \t]*<(?<tag>h[12])([ \t][^>]+)?>(?<title_inner>[^<]+)<\/h[12](?<title_end>[ \t]*>)/im,
+           content,
+           return: :index
+         ) do
+      %{
+        "title_inner" => {title_start, title_length},
+        "title_end" => {title_end_start, title_end_length}
+      } ->
+        title = String.slice(content, title_start, title_length) |> String.trim()
+        {title, title_end_start + title_end_length}
+
+      _ ->
+        Logger.error("No H1/H2 match")
+        {"", 0}
+    end
+  end
+
+  def split_content(content, "text/plain", opts) do
+    content
+    |> split_text_lines(opts)
+    |> case do
+      {_str, 0, 0} ->
+        {"", 0}
+
+      {str, te, ds} ->
+        {str
+         |> String.split_at(te)
+         |> elem(0)
+         |> String.trim(), ds}
+    end
+  end
+
+  def split_content(content, "text/bbcode", opts) do
+    content
+    |> split_text_lines(opts)
+    |> match_title(~r/\[b\](?<title>[^\[]+)\[\/b\]/)
+  end
+
+  def split_content(content, "text/markdown", opts) do
+    content
+    |> split_text_lines(opts)
+    |> match_title(~r/^[#]{1,2}[ \t](?<title>.+)/)
+  end
+
+  def split_content(content, "text/x.misskeymarkdown", opts) do
+    content
+    |> split_text_lines(opts)
+    |> match_title(~r/^\*\*(?<title>.+)\*\*/)
+  end
+
+  def split_content(_, _, _), do: {0, 0}
+
+  def split_text_lines(str, _opts) do
+    case String.split(str, ~r/\n/) do
+      [] ->
+        {str, 0, 0}
+
+      [_first, _rest] ->
+        {str, 0, 0}
+
+      [first | [second | _rest]] ->
+        title_end = String.trim_trailing(first) |> String.length()
+        description_start = String.length(first) + String.length(second)
+        {str, title_end, description_start}
+    end
+  end
+
+  def match_title({_str, 0, 0}, _regex), do: {"", 0}
+
+  def match_title({str, te, ds}, regex) do
+    maybe_title =
+      str
+      |> String.split_at(te)
+      |> elem(0)
+      |> String.trim()
+
+    case Regex.named_captures(regex, maybe_title) do
+      %{"title" => title} ->
+        {String.trim(title), ds}
+
+      _ ->
+        {"", 0}
+    end
+  end
+
   def get_href(id) do
     with %Object{data: %{"external_url" => external_url}} <- Object.get_cached_by_ap_id(id) do
       external_url
@@ -112,4 +264,37 @@ def escape(html) do
     |> html_escape()
     |> safe_to_string()
   end
+
+  def cdata(str), do: "<![CDATA[" <> str <> "]]>"
+
+  def xml_escape(nil), do: ""
+
+  def xml_escape(str) when is_binary(str) do
+    str
+    |> xml_escape_string()
+    |> to_string()
+  end
+
+  defp xml_escape_string(""), do: ""
+  defp xml_escape_string(<<"&"::utf8, rest::binary>>), do: xml_escape_entity(rest)
+  defp xml_escape_string(<<"<"::utf8, rest::binary>>), do: ["&lt;" | xml_escape_string(rest)]
+  defp xml_escape_string(<<">"::utf8, rest::binary>>), do: ["&gt;" | xml_escape_string(rest)]
+  defp xml_escape_string(<<"\t"::utf8, rest::binary>>), do: ["&#9;" | xml_escape_string(rest)]
+  defp xml_escape_string(<<"\n"::utf8, rest::binary>>), do: ["&#10;" | xml_escape_string(rest)]
+  defp xml_escape_string(<<"\r"::utf8, rest::binary>>), do: ["&#13;" | xml_escape_string(rest)]
+  defp xml_escape_string(<<"\""::utf8, rest::binary>>), do: ["&#22;" | xml_escape_string(rest)]
+  defp xml_escape_string(<<"\'"::utf8, rest::binary>>), do: ["&#39;" | xml_escape_string(rest)]
+  defp xml_escape_string(<<c::utf8, rest::binary>>), do: [c | xml_escape_string(rest)]
+
+  defp xml_escape_entity(<<"amp;"::utf8, rest::binary>>), do: ["&amp;" | xml_escape_string(rest)]
+  defp xml_escape_entity(<<"lt;"::utf8, rest::binary>>), do: ["&lt;" | xml_escape_string(rest)]
+  defp xml_escape_entity(<<"gt;"::utf8, rest::binary>>), do: ["&gt;" | xml_escape_string(rest)]
+
+  defp xml_escape_entity(<<"quot;"::utf8, rest::binary>>),
+    do: ["&quot;" | xml_escape_string(rest)]
+
+  defp xml_escape_entity(<<"apos;"::utf8, rest::binary>>),
+    do: ["&apos;" | xml_escape_string(rest)]
+
+  defp xml_escape_entity(rest), do: ["&amp;" | xml_escape_string(rest)]
 end

lib/pleroma/web/metadata/utils.ex

@@ -42,8 +42,12 @@ def scrub_html(content) when is_binary(content) do
     content
     # html content comes from DB already encoded, decode first and scrub after
     |> HtmlEntities.decode()
-    |> String.replace(~r/<br\s?\/?>/, " ")
+    |> String.replace(~r/<(br|p)[^>]*>/, "\\0&nbsp;")
+    |> String.replace(~r/<\/p\s*\/?>/, "&nbsp;\\0")
     |> HTML.strip_tags()
+    # strip_tags will convert &nbsp; to U+00A0, adding /u will match these to " "
+    |> String.replace(~r/\s+/u, " ")
+    |> String.trim()
   end
 
   def scrub_html(content), do: content

lib/pleroma/web/templates/feed/feed/_activity.atom.eex

@@ -2,8 +2,10 @@
   <activity:object-type>http://activitystrea.ms/schema/1.0/note</activity:object-type>
   <activity:verb>http://activitystrea.ms/schema/1.0/post</activity:verb>
   <id><%= @data["id"] %></id>
-  <title><%= activity_title(@data, Keyword.get(@feed_config, :post_title, %{})) %></title>
-  <content type="html"><%= activity_content(@data) %></content>
+
+  <%= maybe_activity_title(@data, Keyword.get(@feed_config, :post_title, %{})) %>
+
+  <content type="html"><%= activity_content(@data, Keyword.get(@feed_config, :post_title, %{})) %></content>
   <published><%= @activity.data["published"] %></published>
   <updated><%= @activity.data["published"] %></updated>
   <ostatus:conversation ref="<%= activity_context(@activity) %>">
@@ -15,6 +17,13 @@
     <summary><%= escape(@data["summary"]) %></summary>
   <% end %>
 
+  <%= if !is_nil(get_in(@data, ["source", "mediaType"])) do %>
+    <source:contentType><%= get_in(@data, ["source", "mediaType"]) %></source:contentType>
+    <source:content><%= source_content(@data) %></source:content>
+    <source:title><%= parse_title(@data, Keyword.get(@feed_config, :post_title, %{})) %></source:titleEnd>
+    <source:descriptionOffset><%= parse_description_offset(@data, Keyword.get(@feed_config, :post_title, %{})) %></source:descriptionStart>
+  <% end %>
+
   <%= if @activity.local do %>
     <link type="application/atom+xml" href='<%= @data["id"] %>' rel="self"/>
     <link type="text/html" href='<%= @data["id"] %>' rel="alternate"/>

lib/pleroma/web/templates/feed/feed/_activity.rss.eex

@@ -2,8 +2,10 @@
   <activity:object-type>http://activitystrea.ms/schema/1.0/note</activity:object-type>
   <activity:verb>http://activitystrea.ms/schema/1.0/post</activity:verb>
   <guid><%= @data["id"] %></guid>
-  <title><%= activity_title(@data, Keyword.get(@feed_config, :post_title, %{})) %></title>
-  <description><%= activity_content(@data) %></description>
+
+  <%= maybe_activity_title(@data, Keyword.get(@feed_config, :post_title, %{})) %>
+
+  <description><%= activity_content(@data, Keyword.get(@feed_config, :post_title, %{})) %></description>
   <pubDate><%= @activity.data["published"] %></pubDate>
   <updated><%= @activity.data["published"] %></updated>
   <ostatus:conversation ref="<%= activity_context(@activity) %>">
@@ -11,7 +13,14 @@
   </ostatus:conversation>
 
   <%= if @data["summary"] do %>
-    <description><%= escape(@data["summary"]) %></description>
+    <masto:summary><%= escape(@data["summary"]) %></masto:summary>
+  <% end %>
+
+  <%= if !is_nil(get_in(@data, ["source", "mediaType"])) do %>
+    <source:contentType><%= get_in(@data, ["source", "mediaType"]) %></source:contentType>
+    <source:content><%= source_content(@data) %></source:content>
+    <source:title><%= parse_title(@data, Keyword.get(@feed_config, :post_title, %{})) %></source:titleEnd>
+    <source:descriptionOffset><%= parse_description_offset(@data, Keyword.get(@feed_config, :post_title, %{})) %></source:descriptionStart>
   <% end %>
 
   <%= if @activity.local do %>

lib/pleroma/web/templates/feed/feed/_tag_activity.atom.eex

@@ -5,8 +5,10 @@
     <%= render @view_module, "_tag_author.atom", assigns %>
 
     <id><%= @data["id"] %></id>
-    <title><%= activity_title(@data, Keyword.get(@feed_config, :post_title, %{})) %></title>
-    <content type="html"><%= activity_content(@data) %></content>
+
+    <%= maybe_activity_title(@data, Keyword.get(@feed_config, :post_title, %{})) %>
+
+    <content type="html"><%= activity_content(@data, Keyword.get(@feed_config, :post_title, %{})) %></content>
 
   <%= if @activity.local do %>
     <link type="application/atom+xml" href='<%= @data["id"] %>' rel="self"/>

lib/pleroma/web/templates/feed/feed/_tag_activity.xml.eex

@@ -1,12 +1,11 @@
 <item>
-  <title><%= activity_title(@data, Keyword.get(@feed_config, :post_title, %{})) %></title>
-
-
   <guid isPermalink="true"><%= activity_context(@activity) %></guid>
   <link><%= activity_context(@activity) %></link>
   <pubDate><%= pub_date(@activity.data["published"]) %></pubDate>
 
-  <description><%= activity_content(@data) %></description>
+  <%= maybe_activity_title(@data, Keyword.get(@feed_config, :post_title, %{})) %>
+
+  <description><%= activity_content(@data, Keyword.get(@feed_config, :post_title, %{})) %></description>
   <%= for attachment <- @data["attachment"] || [] do %>
     <enclosure url="<%= attachment_href(attachment) %>" type="<%= attachment_type(attachment) %>"/>
   <% end %>

lib/pleroma/web/templates/feed/feed/tag.atom.eex

@@ -7,7 +7,8 @@
       xmlns:media="http://purl.org/syndication/atommedia"
       xmlns:poco="http://portablecontacts.net/spec/1.0"
       xmlns:ostatus="http://ostatus.org/schema/1.0"
-      xmlns:statusnet="http://status.net/schema/api/1/">
+      xmlns:statusnet="http://status.net/schema/api/1/"
+      xmlns:source="http://source.scripting.com/">
 
     <id><%= '#{Routes.tag_feed_url(@conn, :feed, @tag)}.rss' %></id>
     <title>#<%= @tag %></title>

lib/pleroma/web/templates/feed/feed/user.atom.eex

@@ -4,7 +4,8 @@
   xmlns:thr="http://purl.org/syndication/thread/1.0"
   xmlns:activity="http://activitystrea.ms/spec/1.0/"
   xmlns:poco="http://portablecontacts.net/spec/1.0"
-  xmlns:ostatus="http://ostatus.org/schema/1.0">
+  xmlns:ostatus="http://ostatus.org/schema/1.0"
+  xmlns:source="http://source.scripting.com/">
 
   <id><%= Routes.user_feed_url(@conn, :feed, @user.nickname) <> ".atom" %></id>
   <title><%= @user.nickname <> "'s timeline" %></title>