git.squeep.com Git - akkoma/blob - lib/pleroma/formatter.ex

   1 # Pleroma: A lightweight social networking server
   2 # Copyright © 2017-2019 Pleroma Authors <https://pleroma.social/>
   3 # SPDX-License-Identifier: AGPL-3.0-only
   4
   5 defmodule Pleroma.Formatter do
   6   alias Pleroma.User
   7   alias Pleroma.Web.MediaProxy
   8   alias Pleroma.HTML
   9   alias Pleroma.Emoji
  10
  11   @tag_regex ~r/((?<=[^&])|\A)(\#)(\w+)/u
  12   @markdown_characters_regex ~r/(`|\*|_|{|}|[|]|\(|\)|#|\+|-|\.|!)/
  13
  14   # Modified from https://www.w3.org/TR/html5/forms.html#valid-e-mail-address
  15   @mentions_regex ~r/@[a-zA-Z0-9.!#$%&'*+\/=?^_`{|}~-]*@?[a-zA-Z0-9_-](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\.[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)*/u
  16
  17   def parse_tags(text, data \\ %{}) do
  18     Regex.scan(@tag_regex, text)
  19     |> Enum.map(fn ["#" <> tag = full_tag | _] -> {full_tag, String.downcase(tag)} end)
  20     |> (fn map ->
  21           if data["sensitive"] in [true, "True", "true", "1"],
  22             do: [{"#nsfw", "nsfw"}] ++ map,
  23             else: map
  24         end).()
  25   end
  26
  27   @doc "Parses mentions text and returns list {nickname, user}."
  28   @spec parse_mentions(binary()) :: list({binary(), User.t()})
  29   def parse_mentions(text) do
  30     Regex.scan(@mentions_regex, text)
  31     |> List.flatten()
  32     |> Enum.uniq()
  33     |> Enum.map(fn nickname ->
  34       with nickname <- String.trim_leading(nickname, "@"),
  35            do: {"@" <> nickname, User.get_cached_by_nickname(nickname)}
  36     end)
  37     |> Enum.filter(fn {_match, user} -> user end)
  38   end
  39
  40   def emojify(text) do
  41     emojify(text, Emoji.get_all())
  42   end
  43
  44   def emojify(text, nil), do: text
  45
  46   def emojify(text, emoji, strip \\ false) do
  47     Enum.reduce(emoji, text, fn {emoji, file}, text ->
  48       emoji = HTML.strip_tags(emoji)
  49       file = HTML.strip_tags(file)
  50
  51       String.replace(
  52         text,
  53         ":#{emoji}:",
  54         if not strip do
  55           "<img height='32px' width='32px' alt='#{emoji}' title='#{emoji}' src='#{
  56             MediaProxy.url(file)
  57           }' />"
  58         else
  59           ""
  60         end
  61       )
  62       |> HTML.filter_tags()
  63     end)
  64   end
  65
  66   def demojify(text) do
  67     emojify(text, Emoji.get_all(), true)
  68   end
  69
  70   def demojify(text, nil), do: text
  71
  72   def get_emoji(text) when is_binary(text) do
  73     Enum.filter(Emoji.get_all(), fn {emoji, _} -> String.contains?(text, ":#{emoji}:") end)
  74   end
  75
  76   def get_emoji(_), do: []
  77
  78   @link_regex ~r/[0-9a-z+\-\.]+:[0-9a-z$-_.+!*'(),]+/ui
  79
  80   @uri_schemes Application.get_env(:pleroma, :uri_schemes, [])
  81   @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, [])
  82
  83   # TODO: make it use something other than @link_regex
  84   def html_escape(text, "text/html") do
  85     HTML.filter_tags(text)
  86   end
  87
  88   def html_escape(text, "text/plain") do
  89     Regex.split(@link_regex, text, include_captures: true)
  90     |> Enum.map_every(2, fn chunk ->
  91       {:safe, part} = Phoenix.HTML.html_escape(chunk)
  92       part
  93     end)
  94     |> Enum.join("")
  95   end
  96
  97   @doc """
  98   Escapes a special characters in mention names.
  99   """
 100   @spec mentions_escape(String.t(), list({String.t(), any()})) :: String.t()
 101   def mentions_escape(text, mentions) do
 102     mentions
 103     |> Enum.reduce(text, fn {name, _}, acc ->
 104       escape_name = String.replace(name, @markdown_characters_regex, "\\\\\\1")
 105       String.replace(acc, name, escape_name)
 106     end)
 107   end
 108
 109   @doc "changes scheme:... urls to html links"
 110   def add_links({subs, text}) do
 111     links =
 112       text
 113       |> String.split([" ", "\t", "<br>"])
 114       |> Enum.filter(fn word -> String.starts_with?(word, @valid_schemes) end)
 115       |> Enum.filter(fn word -> Regex.match?(@link_regex, word) end)
 116       |> Enum.map(fn url -> {Ecto.UUID.generate(), url} end)
 117       |> Enum.sort_by(fn {_, url} -> -String.length(url) end)
 118
 119     uuid_text =
 120       links
 121       |> Enum.reduce(text, fn {uuid, url}, acc -> String.replace(acc, url, uuid) end)
 122
 123     subs =
 124       subs ++
 125         Enum.map(links, fn {uuid, url} ->
 126           {uuid, "<a href=\"#{url}\">#{url}</a>"}
 127         end)
 128
 129     {subs, uuid_text}
 130   end
 131
 132   @doc "Adds the links to mentioned users"
 133   def add_user_links({subs, text}, mentions, options \\ []) do
 134     mentions =
 135       mentions
 136       |> Enum.sort_by(fn {name, _} -> -String.length(name) end)
 137       |> Enum.map(fn {name, user} -> {name, user, Ecto.UUID.generate()} end)
 138
 139     uuid_text =
 140       mentions
 141       |> Enum.reduce(text, fn {match, _user, uuid}, text ->
 142         String.replace(text, match, uuid)
 143       end)
 144
 145     subs =
 146       subs ++
 147         Enum.map(mentions, fn {match, %User{id: id, ap_id: ap_id, info: info}, uuid} ->
 148           ap_id =
 149             if is_binary(info.source_data["url"]) do
 150               info.source_data["url"]
 151             else
 152               ap_id
 153             end
 154
 155           nickname =
 156             if options[:format] == :full do
 157               User.full_nickname(match)
 158             else
 159               User.local_nickname(match)
 160             end
 161
 162           {uuid,
 163            "<span class='h-card'><a data-user='#{id}' class='u-url mention' href='#{ap_id}'>" <>
 164              "@<span>#{nickname}</span></a></span>"}
 165         end)
 166
 167     {subs, uuid_text}
 168   end
 169
 170   @doc "Adds the hashtag links"
 171   def add_hashtag_links({subs, text}, tags) do
 172     tags =
 173       tags
 174       |> Enum.sort_by(fn {name, _} -> -String.length(name) end)
 175       |> Enum.map(fn {name, short} -> {name, short, Ecto.UUID.generate()} end)
 176
 177     uuid_text =
 178       tags
 179       |> Enum.reduce(text, fn {match, _short, uuid}, text ->
 180         String.replace(text, ~r/((?<=[^&])|(\A))#{match}/, uuid)
 181       end)
 182
 183     subs =
 184       subs ++
 185         Enum.map(tags, fn {tag_text, tag, uuid} ->
 186           url =
 187             "<a class='hashtag' data-tag='#{tag}' href='#{Pleroma.Web.base_url()}/tag/#{tag}' rel='tag'>#{
 188               tag_text
 189             }</a>"
 190
 191           {uuid, url}
 192         end)
 193
 194     {subs, uuid_text}
 195   end
 196
 197   def finalize({subs, text}) do
 198     Enum.reduce(subs, text, fn {uuid, replacement}, result_text ->
 199       String.replace(result_text, uuid, replacement)
 200     end)
 201   end
 202
 203   def truncate(text, max_length \\ 200, omission \\ "...") do
 204     # Remove trailing whitespace
 205     text = Regex.replace(~r/([^ \t\r\n])([ \t]+$)/u, text, "\\g{1}")
 206
 207     if String.length(text) < max_length do
 208       text
 209     else
 210       length_with_omission = max_length - String.length(omission)
 211       String.slice(text, 0, length_with_omission) <> omission
 212     end
 213   end
 214 end