# Pleroma: A lightweight social networking server
-# Copyright © 2017-2019 Pleroma Authors <https://pleroma.social/>
+# Copyright © 2017-2020 Pleroma Authors <https://pleroma.social/>
# SPDX-License-Identifier: AGPL-3.0-only
defmodule Pleroma.Formatter do
- alias Pleroma.Emoji
alias Pleroma.HTML
alias Pleroma.User
- alias Pleroma.Web.MediaProxy
- @tag_regex ~r/((?<=[^&])|\A)(\#)(\w+)/u
+ @safe_mention_regex ~r/^(\s*(?<mentions>(@.+?\s+){1,})+)(?<rest>.*)/s
+ @link_regex ~r"((?:http(s)?:\/\/)?[\w.-]+(?:\.[\w\.-]+)+[\w\-\._~%:/?#[\]@!\$&'\(\)\*\+,;=.]+)|[0-9a-z+\-\.]+:[0-9a-z$-_.+!*'(),]+"ui
@markdown_characters_regex ~r/(`|\*|_|{|}|[|]|\(|\)|#|\+|-|\.|!)/
- # Modified from https://www.w3.org/TR/html5/forms.html#valid-e-mail-address
- @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
-
- def parse_tags(text, data \\ %{}) do
- Regex.scan(@tag_regex, text)
- |> Enum.map(fn ["#" <> tag = full_tag | _] -> {full_tag, String.downcase(tag)} end)
- |> (fn map ->
- if data["sensitive"] in [true, "True", "true", "1"],
- do: [{"#nsfw", "nsfw"}] ++ map,
- else: map
- end).()
+ @auto_linker_config hashtag: true,
+ hashtag_handler: &Pleroma.Formatter.hashtag_handler/4,
+ mention: true,
+ mention_handler: &Pleroma.Formatter.mention_handler/4,
+ scheme: true
+
+ def escape_mention_handler("@" <> nickname = mention, buffer, _, _) do
+ case User.get_cached_by_nickname(nickname) do
+ %User{} ->
+ # escape markdown characters with `\\`
+ # (we don't want something like @user__name to be parsed by markdown)
+ String.replace(mention, @markdown_characters_regex, "\\\\\\1")
+
+ _ ->
+ buffer
+ end
end
- @doc "Parses mentions text and returns list {nickname, user}."
- @spec parse_mentions(binary()) :: list({binary(), User.t()})
- def parse_mentions(text) do
- Regex.scan(@mentions_regex, text)
- |> List.flatten()
- |> Enum.uniq()
- |> Enum.map(fn nickname ->
- with nickname <- String.trim_leading(nickname, "@"),
- do: {"@" <> nickname, User.get_cached_by_nickname(nickname)}
- end)
- |> Enum.filter(fn {_match, user} -> user end)
+ def mention_handler("@" <> nickname, buffer, opts, acc) do
+ case User.get_cached_by_nickname(nickname) do
+ %User{id: id} = user ->
+ user_url = user.uri || user.ap_id
+ nickname_text = get_nickname_text(nickname, opts)
+
+ link =
+ Phoenix.HTML.Tag.content_tag(
+ :span,
+ Phoenix.HTML.Tag.content_tag(
+ :a,
+ ["@", Phoenix.HTML.Tag.content_tag(:span, nickname_text)],
+ "data-user": id,
+ class: "u-url mention",
+ href: user_url,
+ rel: "ugc"
+ ),
+ class: "h-card"
+ )
+ |> Phoenix.HTML.safe_to_string()
+
+ {link, %{acc | mentions: MapSet.put(acc.mentions, {"@" <> nickname, user})}}
+
+ _ ->
+ {buffer, acc}
+ end
end
- def emojify(text) do
- emojify(text, Emoji.get_all())
- end
+ def hashtag_handler("#" <> tag = tag_text, _buffer, _opts, acc) do
+ tag = String.downcase(tag)
+ url = "#{Pleroma.Web.base_url()}/tag/#{tag}"
- def emojify(text, nil), do: text
-
- def emojify(text, emoji, strip \\ false) do
- Enum.reduce(emoji, text, fn {emoji, file}, text ->
- emoji = HTML.strip_tags(emoji)
- file = HTML.strip_tags(file)
-
- String.replace(
- text,
- ":#{emoji}:",
- if not strip do
- "<img height='32px' width='32px' alt='#{emoji}' title='#{emoji}' src='#{
- MediaProxy.url(file)
- }' />"
- else
- ""
- end
+ link =
+ Phoenix.HTML.Tag.content_tag(:a, tag_text,
+ class: "hashtag",
+ "data-tag": tag,
+ href: url,
+ rel: "tag ugc"
)
- |> HTML.filter_tags()
- end)
- end
+ |> Phoenix.HTML.safe_to_string()
- def demojify(text) do
- emojify(text, Emoji.get_all(), true)
+ {link, %{acc | tags: MapSet.put(acc.tags, {tag_text, tag})}}
end
- def demojify(text, nil), do: text
-
- def get_emoji(text) when is_binary(text) do
- Enum.filter(Emoji.get_all(), fn {emoji, _} -> String.contains?(text, ":#{emoji}:") end)
- end
+ @doc """
+ Parses a text and replace plain text links with HTML. Returns a tuple with a result text, mentions, and hashtags.
- def get_emoji(_), do: []
+ If the 'safe_mention' option is given, only consecutive mentions at the start the post are actually mentioned.
+ """
+ @spec linkify(String.t(), keyword()) ::
+ {String.t(), [{String.t(), User.t()}], [{String.t(), String.t()}]}
+ def linkify(text, options \\ []) do
+ options = options ++ @auto_linker_config
- @link_regex ~r/[0-9a-z+\-\.]+:[0-9a-z$-_.+!*'(),]+/ui
+ if options[:safe_mention] && Regex.named_captures(@safe_mention_regex, text) do
+ %{"mentions" => mentions, "rest" => rest} = Regex.named_captures(@safe_mention_regex, text)
+ acc = %{mentions: MapSet.new(), tags: MapSet.new()}
- @uri_schemes Application.get_env(:pleroma, :uri_schemes, [])
- @valid_schemes Keyword.get(@uri_schemes, :valid_schemes, [])
+ {text_mentions, %{mentions: mentions}} = AutoLinker.link_map(mentions, acc, options)
+ {text_rest, %{tags: tags}} = AutoLinker.link_map(rest, acc, options)
- # TODO: make it use something other than @link_regex
- def html_escape(text, "text/html") do
- HTML.filter_tags(text)
- end
+ {text_mentions <> text_rest, MapSet.to_list(mentions), MapSet.to_list(tags)}
+ else
+ acc = %{mentions: MapSet.new(), tags: MapSet.new()}
+ {text, %{mentions: mentions, tags: tags}} = AutoLinker.link_map(text, acc, options)
- def html_escape(text, "text/plain") do
- Regex.split(@link_regex, text, include_captures: true)
- |> Enum.map_every(2, fn chunk ->
- {:safe, part} = Phoenix.HTML.html_escape(chunk)
- part
- end)
- |> Enum.join("")
+ {text, MapSet.to_list(mentions), MapSet.to_list(tags)}
+ end
end
@doc """
Escapes a special characters in mention names.
"""
- @spec mentions_escape(String.t(), list({String.t(), any()})) :: String.t()
- def mentions_escape(text, mentions) do
- mentions
- |> Enum.reduce(text, fn {name, _}, acc ->
- escape_name = String.replace(name, @markdown_characters_regex, "\\\\\\1")
- String.replace(acc, name, escape_name)
- end)
- end
+ def mentions_escape(text, options \\ []) do
+ options =
+ Keyword.merge(options,
+ mention: true,
+ url: false,
+ mention_handler: &Pleroma.Formatter.escape_mention_handler/4
+ )
- @doc "changes scheme:... urls to html links"
- def add_links({subs, text}) do
- links =
- text
- |> String.split([" ", "\t", "<br>"])
- |> Enum.filter(fn word -> String.starts_with?(word, @valid_schemes) end)
- |> Enum.filter(fn word -> Regex.match?(@link_regex, word) end)
- |> Enum.map(fn url -> {Ecto.UUID.generate(), url} end)
- |> Enum.sort_by(fn {_, url} -> -String.length(url) end)
-
- uuid_text =
- links
- |> Enum.reduce(text, fn {uuid, url}, acc -> String.replace(acc, url, uuid) end)
-
- subs =
- subs ++
- Enum.map(links, fn {uuid, url} ->
- {uuid, "<a href=\"#{url}\">#{url}</a>"}
- end)
-
- {subs, uuid_text}
+ if options[:safe_mention] && Regex.named_captures(@safe_mention_regex, text) do
+ %{"mentions" => mentions, "rest" => rest} = Regex.named_captures(@safe_mention_regex, text)
+ AutoLinker.link(mentions, options) <> AutoLinker.link(rest, options)
+ else
+ AutoLinker.link(text, options)
+ end
end
- @doc "Adds the links to mentioned users"
- def add_user_links({subs, text}, mentions, options \\ []) do
- mentions =
- mentions
- |> Enum.sort_by(fn {name, _} -> -String.length(name) end)
- |> Enum.map(fn {name, user} -> {name, user, Ecto.UUID.generate()} end)
-
- uuid_text =
- mentions
- |> Enum.reduce(text, fn {match, _user, uuid}, text ->
- String.replace(text, match, uuid)
- end)
-
- subs =
- subs ++
- Enum.map(mentions, fn {match, %User{id: id, ap_id: ap_id, info: info}, uuid} ->
- ap_id =
- if is_binary(info.source_data["url"]) do
- info.source_data["url"]
- else
- ap_id
- end
-
- nickname =
- if options[:format] == :full do
- User.full_nickname(match)
- else
- User.local_nickname(match)
- end
-
- {uuid,
- "<span class='h-card'><a data-user='#{id}' class='u-url mention' href='#{ap_id}'>" <>
- "@<span>#{nickname}</span></a></span>"}
- end)
-
- {subs, uuid_text}
+ def html_escape({text, mentions, hashtags}, type) do
+ {html_escape(text, type), mentions, hashtags}
end
- @doc "Adds the hashtag links"
- def add_hashtag_links({subs, text}, tags) do
- tags =
- tags
- |> Enum.sort_by(fn {name, _} -> -String.length(name) end)
- |> Enum.map(fn {name, short} -> {name, short, Ecto.UUID.generate()} end)
-
- uuid_text =
- tags
- |> Enum.reduce(text, fn {match, _short, uuid}, text ->
- String.replace(text, ~r/((?<=[^&])|(\A))#{match}/, uuid)
- end)
-
- subs =
- subs ++
- Enum.map(tags, fn {tag_text, tag, uuid} ->
- url =
- "<a class='hashtag' data-tag='#{tag}' href='#{Pleroma.Web.base_url()}/tag/#{tag}' rel='tag'>#{
- tag_text
- }</a>"
-
- {uuid, url}
- end)
-
- {subs, uuid_text}
+ def html_escape(text, "text/html") do
+ HTML.filter_tags(text)
end
- def finalize({subs, text}) do
- Enum.reduce(subs, text, fn {uuid, replacement}, result_text ->
- String.replace(result_text, uuid, replacement)
+ def html_escape(text, "text/plain") do
+ Regex.split(@link_regex, text, include_captures: true)
+ |> Enum.map_every(2, fn chunk ->
+ {:safe, part} = Phoenix.HTML.html_escape(chunk)
+ part
end)
+ |> Enum.join("")
end
def truncate(text, max_length \\ 200, omission \\ "...") do
String.slice(text, 0, length_with_omission) <> omission
end
end
+
+ defp get_nickname_text(nickname, %{mentions_format: :full}), do: User.full_nickname(nickname)
+ defp get_nickname_text(nickname, _), do: User.local_nickname(nickname)
end
projects / akkoma / blobdiff