akkoma/lib/pleroma/formatter.ex

209 lines
5.9 KiB
Elixir
Raw Normal View History

# Pleroma: A lightweight social networking server
2018-12-31 15:41:47 +00:00
# Copyright © 2017-2019 Pleroma Authors <https://pleroma.social/>
# SPDX-License-Identifier: AGPL-3.0-only
2017-05-17 16:00:09 +00:00
defmodule Pleroma.Formatter do
alias Pleroma.User
alias Pleroma.Web.MediaProxy
alias Pleroma.HTML
alias Pleroma.Emoji
2017-05-17 16:00:09 +00:00
2018-12-18 18:38:15 +00:00
@tag_regex ~r/((?<=[^&])|\A)(\#)(\w+)/u
2018-12-14 18:21:37 +00:00
@markdown_characters_regex ~r/(`|\*|_|{|}|[|]|\(|\)|#|\+|-|\.|!)/
2018-12-14 09:41:55 +00:00
2018-12-20 09:35:01 +00:00
# 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
2017-05-17 16:00:09 +00:00
Regex.scan(@tag_regex, text)
2018-12-18 18:38:15 +00:00
|> Enum.map(fn ["#" <> tag = full_tag | _] -> {full_tag, String.downcase(tag)} end)
2018-03-30 13:01:53 +00:00
|> (fn map ->
if data["sensitive"] in [true, "True", "true", "1"],
do: [{"#nsfw", "nsfw"}] ++ map,
else: map
end).()
2017-05-17 16:00:09 +00:00
end
2018-12-20 09:35:01 +00:00
@doc "Parses mentions text and returns list {nickname, user}."
@spec parse_mentions(binary()) :: list({binary(), User.t()})
def parse_mentions(text) do
2018-12-20 09:35:01 +00:00
Regex.scan(@mentions_regex, text)
2018-03-30 13:01:53 +00:00
|> List.flatten()
|> Enum.uniq()
2018-12-20 09:35:01 +00:00
|> Enum.map(fn nickname ->
with nickname <- String.trim_leading(nickname, "@"),
do: {"@" <> nickname, User.get_cached_by_nickname(nickname)}
2018-03-30 13:01:53 +00:00
end)
|> Enum.filter(fn {_match, user} -> user end)
end
2017-06-19 15:51:43 +00:00
def emojify(text) do
emojify(text, Emoji.get_all())
end
2017-09-16 13:47:45 +00:00
def emojify(text, nil), do: text
2018-03-30 13:01:53 +00:00
def emojify(text, emoji, strip \\ false) do
2018-03-30 13:01:53 +00:00
Enum.reduce(emoji, text, fn {emoji, file}, text ->
emoji = HTML.strip_tags(emoji)
file = HTML.strip_tags(file)
2018-03-30 13:01:53 +00:00
String.replace(
text,
":#{emoji}:",
if not strip do
"<img height='32px' width='32px' alt='#{emoji}' title='#{emoji}' src='#{
MediaProxy.url(file)
}' />"
else
""
end
2018-03-30 13:01:53 +00:00
)
|> HTML.filter_tags()
2017-06-19 15:51:43 +00:00
end)
end
2017-09-16 14:14:23 +00:00
def demojify(text) do
emojify(text, Emoji.get_all(), true)
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)
2017-09-16 14:14:23 +00:00
end
2017-10-19 19:51:56 +00:00
def get_emoji(_), do: []
@link_regex ~r/[0-9a-z+\-\.]+:[0-9a-z$-_.+!*'(),]+/ui
@uri_schemes Application.get_env(:pleroma, :uri_schemes, [])
@valid_schemes Keyword.get(@uri_schemes, :valid_schemes, [])
2018-03-24 10:03:10 +00:00
# TODO: make it use something other than @link_regex
def html_escape(text, "text/html") do
HTML.filter_tags(text)
end
def html_escape(text, "text/plain") do
2018-03-24 10:03:10 +00:00
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
2018-12-14 09:41:55 +00:00
@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 ->
2018-12-14 18:21:37 +00:00
escape_name = String.replace(name, @markdown_characters_regex, "\\\\\\1")
2018-12-14 09:41:55 +00:00
String.replace(acc, name, escape_name)
end)
end
@doc "changes scheme:... urls to html links"
2018-03-24 10:03:10 +00:00
def add_links({subs, text}) do
2018-03-30 13:01:53 +00:00
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)
2018-03-24 10:03:10 +00:00
2018-03-30 13:01:53 +00:00
uuid_text =
links
|> Enum.reduce(text, fn {uuid, url}, acc -> String.replace(acc, url, uuid) end)
2018-03-24 10:03:10 +00:00
2018-03-30 13:01:53 +00:00
subs =
subs ++
Enum.map(links, fn {uuid, url} ->
{uuid, "<a href=\"#{url}\">#{url}</a>"}
2018-03-30 13:01:53 +00:00
end)
2018-03-24 10:03:10 +00:00
{subs, uuid_text}
end
@doc "Adds the links to mentioned users"
def add_user_links({subs, text}, mentions) do
2018-03-30 13:01:53 +00:00
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 ++
2018-12-05 19:35:41 +00:00
Enum.map(mentions, fn {match, %User{id: id, ap_id: ap_id, info: info}, uuid} ->
2018-10-17 19:15:20 +00:00
ap_id =
2018-11-18 20:40:52 +00:00
if is_binary(info.source_data["url"]) do
info.source_data["url"]
2018-10-17 19:15:20 +00:00
else
ap_id
end
2018-04-03 16:32:46 +00:00
2018-03-30 13:01:53 +00:00
short_match = String.split(match, "@") |> tl() |> hd()
2018-05-27 13:06:38 +00:00
{uuid,
2018-12-05 19:35:41 +00:00
"<span><a data-user='#{id}' class='mention' href='#{ap_id}'>@<span>#{short_match}</span></a></span>"}
2018-03-30 13:01:53 +00:00
end)
2018-03-24 10:03:10 +00:00
{subs, uuid_text}
end
@doc "Adds the hashtag links"
def add_hashtag_links({subs, text}, tags) do
2018-03-30 13:01:53 +00:00
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 ->
2018-12-18 18:38:15 +00:00
String.replace(text, ~r/((?<=[^&])|(\A))#{match}/, uuid)
2018-03-30 13:01:53 +00:00
end)
subs =
subs ++
Enum.map(tags, fn {tag_text, tag, uuid} ->
2018-12-05 19:35:41 +00:00
url =
"<a data-tag='#{tag}' href='#{Pleroma.Web.base_url()}/tag/#{tag}' rel='tag'>#{
tag_text
}</a>"
2018-03-30 13:01:53 +00:00
{uuid, url}
end)
2018-03-24 10:03:10 +00:00
{subs, uuid_text}
end
def finalize({subs, text}) do
2018-03-30 13:01:53 +00:00
Enum.reduce(subs, text, fn {uuid, replacement}, result_text ->
2018-03-24 10:03:10 +00:00
String.replace(result_text, uuid, replacement)
end)
end
def truncate(text, max_length \\ 200, omission \\ "...") 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
end
end
2017-05-17 16:00:09 +00:00
end