1
0
Fork 0
mirror of https://framagit.org/framasoft/mobilizon.git synced 2024-12-23 00:14:48 +00:00
mobilizon/lib/service/formatter/html.ex
Thomas Citharel 8e6aa8c85c
Decode HTML entities when sanitized
Signed-off-by: Thomas Citharel <tcit@tcit.fr>
2021-03-30 09:12:53 +02:00

42 lines
1.2 KiB
Elixir

# Portions of this file are derived from Pleroma:
# Copyright © 2017-2019 Pleroma Authors <https://pleroma.social>
# SPDX-License-Identifier: AGPL-3.0-only
# Upstream: https://git.pleroma.social/pleroma/pleroma/blob/develop/lib/pleroma/html.ex
defmodule Mobilizon.Service.Formatter.HTML do
@moduledoc """
Service to filter tags out of HTML content.
"""
alias FastSanitize.Sanitizer
alias Mobilizon.Service.Formatter.{DefaultScrubbler, OEmbed}
def filter_tags(html), do: Sanitizer.scrub(html, DefaultScrubbler)
def strip_tags(html) do
case FastSanitize.strip_tags(html) do
{:ok, html} ->
HtmlEntities.decode(html)
_ ->
raise "Failed to filter tags"
end
end
@doc """
Inserts a space before tags closing so that words are not attached once tags stripped
`<h1>test</h1>next` thing becomes `test next` instead of `testnext`
"""
@spec strip_tags_and_insert_spaces(String.t()) :: String.t()
def strip_tags_and_insert_spaces(html) when is_binary(html) do
html
|> String.replace("><", "> <")
|> strip_tags()
end
def strip_tags_and_insert_spaces(html), do: html
def filter_tags_for_oembed(html), do: Sanitizer.scrub(html, OEmbed)
end