2021-11-23 03:32:12 +08:00
|
|
|
# frozen_string_literal: true
|
|
|
|
|
|
|
|
module CookedProcessorMixin
|
|
|
|
def post_process_oneboxes
|
2022-03-23 23:36:08 +08:00
|
|
|
limit = SiteSetting.max_oneboxes_per_post - @doc.css("aside.onebox, a.inline-onebox").size
|
2021-11-23 03:32:12 +08:00
|
|
|
oneboxes = {}
|
|
|
|
inlineOneboxes = {}
|
|
|
|
|
|
|
|
Oneboxer.apply(@doc, extra_paths: [".inline-onebox-loading"]) do |url, element|
|
|
|
|
is_onebox = element["class"] == Oneboxer::ONEBOX_CSS_CLASS
|
|
|
|
map = is_onebox ? oneboxes : inlineOneboxes
|
|
|
|
skip_onebox = limit <= 0 && !map[url]
|
|
|
|
|
|
|
|
if skip_onebox
|
|
|
|
if is_onebox
|
|
|
|
element.remove_class("onebox")
|
|
|
|
else
|
|
|
|
remove_inline_onebox_loading_class(element)
|
|
|
|
end
|
|
|
|
|
|
|
|
next
|
|
|
|
end
|
|
|
|
|
|
|
|
limit -= 1
|
|
|
|
map[url] = true
|
|
|
|
|
|
|
|
if is_onebox
|
|
|
|
onebox =
|
|
|
|
Oneboxer.onebox(
|
|
|
|
url,
|
|
|
|
invalidate_oneboxes: !!@opts[:invalidate_oneboxes],
|
|
|
|
user_id: @model&.user_id,
|
|
|
|
category_id: @category_id,
|
|
|
|
)
|
|
|
|
|
|
|
|
@has_oneboxes = true if onebox.present?
|
|
|
|
onebox
|
|
|
|
else
|
|
|
|
process_inline_onebox(element)
|
|
|
|
false
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
FEATURE: Allow hotlinked media to be blocked (#16940)
This commit introduces a new site setting: `block_hotlinked_media`. When enabled, all attempts to hotlink media (images, videos, and audio) will fail, and be replaced with a linked placeholder. Exceptions to the rule can be added via `block_hotlinked_media_exceptions`.
`download_remote_image_to_local` can be used alongside this feature. In that case, hotlinked images will be blocked immediately when the post is created, but will then be replaced with the downloaded version a few seconds later.
This implementation is purely server-side, and does not impact the composer preview.
Technically, there are two stages to this feature:
1. `PrettyText.sanitize_hotlinked_media` is called during `PrettyText.cook`, and whenever new images are introduced by Onebox. It will iterate over all src/srcset attributes in the post HTML and check if they're allowed. If not, the attributes will be removed and replaced with a `data-blocked-hotlinked-src(set)` attribute
2. In the `CookedPostProcessor`, we iterate over all `data-blocked-hotlinked-src(set)` attributes and check whether we have a downloaded version of the media. If yes, we update the src to use the downloaded version. If not, the entire media element is replaced with a placeholder. The placeholder is labelled 'external media', and is a link to the offsite media.
2022-06-07 22:23:04 +08:00
|
|
|
PrettyText.sanitize_hotlinked_media(@doc)
|
|
|
|
|
2021-11-23 03:32:12 +08:00
|
|
|
oneboxed_images.each do |img|
|
|
|
|
next if img["src"].blank?
|
|
|
|
|
|
|
|
parent = img.parent
|
|
|
|
|
2022-05-13 21:11:45 +08:00
|
|
|
if respond_to?(:process_hotlinked_image, true)
|
|
|
|
still_an_image = process_hotlinked_image(img)
|
|
|
|
next if !still_an_image
|
2021-11-23 03:32:12 +08:00
|
|
|
end
|
|
|
|
|
|
|
|
# make sure we grab dimensions for oneboxed images
|
|
|
|
# and wrap in a div
|
|
|
|
limit_size!(img)
|
|
|
|
|
|
|
|
next if img["class"]&.include?("onebox-avatar")
|
|
|
|
|
|
|
|
parent = parent&.parent if parent&.name == "a"
|
|
|
|
parent_class = parent && parent["class"]
|
|
|
|
width = img["width"].to_i
|
|
|
|
height = img["height"].to_i
|
|
|
|
|
|
|
|
if parent_class&.include?("onebox-body") && width > 0 && height > 0
|
|
|
|
# special instruction for width == height, assume we are dealing with an avatar
|
|
|
|
if (img["width"].to_i == img["height"].to_i)
|
|
|
|
found = false
|
|
|
|
parent = img
|
|
|
|
while parent = parent.parent
|
2022-11-24 23:28:21 +08:00
|
|
|
if parent["class"] && parent["class"].match?(/\b(allowlistedgeneric|discoursetopic)\b/)
|
2021-11-23 03:32:12 +08:00
|
|
|
found = true
|
|
|
|
break
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
if found
|
|
|
|
img["class"] = img["class"].to_s + " onebox-avatar"
|
|
|
|
next
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
if width < 64 && height < 64
|
|
|
|
img["class"] = img["class"].to_s + " onebox-full-image"
|
|
|
|
else
|
|
|
|
img.delete("width")
|
|
|
|
img.delete("height")
|
|
|
|
new_parent =
|
|
|
|
img.add_next_sibling(
|
|
|
|
"<div class='aspect-image' style='--aspect-ratio:#{width}/#{height};'/>",
|
|
|
|
)
|
|
|
|
new_parent.first.add_child(img)
|
|
|
|
end
|
|
|
|
elsif (
|
|
|
|
parent_class&.include?("instagram-images") || parent_class&.include?("tweet-images") ||
|
|
|
|
parent_class&.include?("scale-images")
|
|
|
|
) && width > 0 && height > 0
|
|
|
|
img.remove_attribute("width")
|
|
|
|
img.remove_attribute("height")
|
|
|
|
parent["class"] = "aspect-image-full-size"
|
|
|
|
parent["style"] = "--aspect-ratio:#{width}/#{height};"
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
if @omit_nofollow || !SiteSetting.add_rel_nofollow_to_user_content
|
|
|
|
@doc
|
|
|
|
.css(".onebox-body a[rel], .onebox a[rel]")
|
|
|
|
.each do |a|
|
|
|
|
rel_values = a["rel"].split(" ").map(&:downcase)
|
|
|
|
rel_values.delete("nofollow")
|
|
|
|
rel_values.delete("ugc")
|
|
|
|
if rel_values.blank?
|
|
|
|
a.remove_attribute("rel")
|
|
|
|
else
|
|
|
|
a["rel"] = rel_values.join(" ")
|
2023-01-09 20:10:19 +08:00
|
|
|
end
|
2021-11-23 03:32:12 +08:00
|
|
|
end
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
def limit_size!(img)
|
|
|
|
# retrieve the size from
|
|
|
|
# 1) the width/height attributes
|
|
|
|
# 2) the dimension from the preview (image_sizes)
|
|
|
|
# 3) the dimension of the original image (HTTP request)
|
|
|
|
w, h =
|
|
|
|
get_size_from_attributes(img) || get_size_from_image_sizes(img["src"], @opts[:image_sizes]) ||
|
|
|
|
get_size(img["src"])
|
|
|
|
|
|
|
|
# limit the size of the thumbnail
|
|
|
|
img["width"], img["height"] = ImageSizer.resize(w, h)
|
|
|
|
end
|
|
|
|
|
|
|
|
def get_size_from_attributes(img)
|
|
|
|
w, h = img["width"].to_i, img["height"].to_i
|
2023-02-16 17:40:11 +08:00
|
|
|
return w, h if w > 0 && h > 0
|
2021-11-23 03:32:12 +08:00
|
|
|
# if only width or height are specified attempt to scale image
|
|
|
|
if w > 0 || h > 0
|
|
|
|
w = w.to_f
|
|
|
|
h = h.to_f
|
|
|
|
|
|
|
|
return unless original_image_size = get_size(img["src"])
|
|
|
|
original_width, original_height = original_image_size.map(&:to_f)
|
|
|
|
|
|
|
|
if w > 0
|
|
|
|
ratio = w / original_width
|
|
|
|
[w.floor, (original_height * ratio).floor]
|
|
|
|
else
|
|
|
|
ratio = h / original_height
|
|
|
|
[(original_width * ratio).floor, h.floor]
|
|
|
|
end
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
def get_size_from_image_sizes(src, image_sizes)
|
2024-05-27 18:27:13 +08:00
|
|
|
return if image_sizes.blank?
|
2021-11-23 03:32:12 +08:00
|
|
|
image_sizes.each do |image_size|
|
|
|
|
url, size = image_size[0], image_size[1]
|
2023-01-04 02:27:05 +08:00
|
|
|
if url && src && url.include?(src) && size && size["width"].to_i > 0 &&
|
2021-11-23 03:32:12 +08:00
|
|
|
size["height"].to_i > 0
|
|
|
|
return size["width"], size["height"]
|
|
|
|
end
|
|
|
|
end
|
|
|
|
nil
|
|
|
|
end
|
|
|
|
|
|
|
|
def add_to_size_cache(url, w, h)
|
|
|
|
@size_cache[url] = [w, h]
|
|
|
|
end
|
|
|
|
|
|
|
|
def get_size(url)
|
|
|
|
return @size_cache[url] if @size_cache.has_key?(url)
|
|
|
|
|
|
|
|
absolute_url = url
|
2023-01-21 02:52:49 +08:00
|
|
|
absolute_url = Discourse.base_url_no_prefix + absolute_url if absolute_url =~ %r{\A/[^/]}
|
2021-11-23 03:32:12 +08:00
|
|
|
|
|
|
|
return unless absolute_url
|
|
|
|
|
|
|
|
# FastImage fails when there's no scheme
|
|
|
|
absolute_url = SiteSetting.scheme + ":" + absolute_url if absolute_url.start_with?("//")
|
|
|
|
|
2022-09-29 07:24:33 +08:00
|
|
|
# we can't direct FastImage to our secure-uploads url because it bounces
|
2021-11-23 03:32:12 +08:00
|
|
|
# anonymous requests with a 404 error
|
2022-09-29 07:24:33 +08:00
|
|
|
if url && Upload.secure_uploads_url?(url)
|
|
|
|
absolute_url = Upload.signed_url_from_secure_uploads_url(absolute_url)
|
2021-11-23 03:32:12 +08:00
|
|
|
end
|
|
|
|
|
|
|
|
return unless is_valid_image_url?(absolute_url)
|
|
|
|
|
2022-11-25 09:40:31 +08:00
|
|
|
upload = Upload.get_from_url(absolute_url)
|
|
|
|
if upload && upload.width && upload.width > 0
|
|
|
|
@size_cache[url] = [upload.width, upload.height]
|
|
|
|
else
|
2024-05-31 10:52:53 +08:00
|
|
|
@size_cache[url] = FastImage.size(absolute_url)
|
2022-11-25 09:40:31 +08:00
|
|
|
end
|
2021-11-23 03:32:12 +08:00
|
|
|
rescue Zlib::BufError, URI::Error, OpenSSL::SSL::SSLError
|
|
|
|
# FastImage.size raises BufError for some gifs, leave it.
|
|
|
|
end
|
|
|
|
|
|
|
|
def is_valid_image_url?(url)
|
|
|
|
uri = URI.parse(url)
|
|
|
|
%w[http https].include? uri.scheme
|
|
|
|
rescue URI::Error
|
|
|
|
end
|
|
|
|
|
|
|
|
def add_large_image_placeholder!(img)
|
|
|
|
url = img["src"]
|
|
|
|
|
|
|
|
is_hyperlinked = is_a_hyperlink?(img)
|
|
|
|
|
|
|
|
placeholder = create_node("div", "large-image-placeholder")
|
|
|
|
img.add_next_sibling(placeholder)
|
|
|
|
placeholder.add_child(img)
|
|
|
|
|
|
|
|
a = create_link_node(nil, url, true)
|
|
|
|
img.add_next_sibling(a)
|
|
|
|
|
|
|
|
span = create_span_node("url", url)
|
|
|
|
a.add_child(span)
|
|
|
|
span.add_previous_sibling(create_icon_node("far-image"))
|
|
|
|
span.add_next_sibling(
|
|
|
|
create_span_node(
|
|
|
|
"help",
|
|
|
|
I18n.t(
|
|
|
|
"upload.placeholders.too_large_humanized",
|
|
|
|
max_size:
|
|
|
|
ActiveSupport::NumberHelper.number_to_human_size(
|
|
|
|
SiteSetting.max_image_size_kb.kilobytes,
|
|
|
|
),
|
|
|
|
),
|
2023-01-09 20:10:19 +08:00
|
|
|
),
|
2021-11-23 03:32:12 +08:00
|
|
|
)
|
|
|
|
|
|
|
|
# Only if the image is already linked
|
|
|
|
if is_hyperlinked
|
|
|
|
parent = placeholder.parent
|
|
|
|
parent.add_next_sibling(placeholder)
|
|
|
|
|
|
|
|
if parent.name == "a" && parent["href"].present?
|
|
|
|
if url == parent["href"]
|
|
|
|
parent.remove
|
|
|
|
else
|
|
|
|
parent["class"] = "link"
|
|
|
|
a.add_previous_sibling(parent)
|
|
|
|
|
|
|
|
lspan = create_span_node("url", parent["href"])
|
|
|
|
parent.add_child(lspan)
|
|
|
|
lspan.add_previous_sibling(create_icon_node("link"))
|
|
|
|
end
|
|
|
|
end
|
|
|
|
end
|
|
|
|
|
|
|
|
img.remove
|
|
|
|
true
|
|
|
|
end
|
|
|
|
|
|
|
|
def add_broken_image_placeholder!(img)
|
|
|
|
img.name = "span"
|
|
|
|
img.set_attribute("class", "broken-image")
|
|
|
|
img.set_attribute("title", I18n.t("post.image_placeholder.broken"))
|
2024-11-07 02:35:07 +08:00
|
|
|
img << "<svg class=\"fa d-icon d-icon-link-slash svg-icon\" aria-hidden=\"true\"><use href=\"#link-slash\"></use></svg>"
|
2021-11-23 03:32:12 +08:00
|
|
|
img.remove_attribute("src")
|
|
|
|
img.remove_attribute("width")
|
|
|
|
img.remove_attribute("height")
|
|
|
|
true
|
|
|
|
end
|
|
|
|
|
FEATURE: Allow hotlinked media to be blocked (#16940)
This commit introduces a new site setting: `block_hotlinked_media`. When enabled, all attempts to hotlink media (images, videos, and audio) will fail, and be replaced with a linked placeholder. Exceptions to the rule can be added via `block_hotlinked_media_exceptions`.
`download_remote_image_to_local` can be used alongside this feature. In that case, hotlinked images will be blocked immediately when the post is created, but will then be replaced with the downloaded version a few seconds later.
This implementation is purely server-side, and does not impact the composer preview.
Technically, there are two stages to this feature:
1. `PrettyText.sanitize_hotlinked_media` is called during `PrettyText.cook`, and whenever new images are introduced by Onebox. It will iterate over all src/srcset attributes in the post HTML and check if they're allowed. If not, the attributes will be removed and replaced with a `data-blocked-hotlinked-src(set)` attribute
2. In the `CookedPostProcessor`, we iterate over all `data-blocked-hotlinked-src(set)` attributes and check whether we have a downloaded version of the media. If yes, we update the src to use the downloaded version. If not, the entire media element is replaced with a placeholder. The placeholder is labelled 'external media', and is a link to the offsite media.
2022-06-07 22:23:04 +08:00
|
|
|
def add_blocked_hotlinked_image_placeholder!(el)
|
|
|
|
el.name = "a"
|
|
|
|
el.set_attribute("href", el[PrettyText::BLOCKED_HOTLINKED_SRC_ATTR])
|
|
|
|
el.set_attribute("class", "blocked-hotlinked-placeholder")
|
|
|
|
el.set_attribute("title", I18n.t("post.image_placeholder.blocked_hotlinked_title"))
|
|
|
|
el << "<svg class=\"fa d-icon d-icon-link svg-icon\" aria-hidden=\"true\"><use href=\"#link\"></use></svg>"
|
|
|
|
el << "<span class=\"notice\">#{CGI.escapeHTML(I18n.t("post.image_placeholder.blocked_hotlinked"))}</span>"
|
|
|
|
|
|
|
|
true
|
|
|
|
end
|
|
|
|
|
|
|
|
def add_blocked_hotlinked_media_placeholder!(el, src)
|
|
|
|
placeholder = Nokogiri::XML::Node.new("a", el.document)
|
|
|
|
placeholder.name = "a"
|
|
|
|
placeholder.set_attribute("href", src)
|
|
|
|
placeholder.set_attribute("class", "blocked-hotlinked-placeholder")
|
|
|
|
placeholder.set_attribute("title", I18n.t("post.media_placeholder.blocked_hotlinked_title"))
|
|
|
|
placeholder << "<svg class=\"fa d-icon d-icon-link svg-icon\" aria-hidden=\"true\"><use href=\"#link\"></use></svg>"
|
|
|
|
placeholder << "<span class=\"notice\">#{CGI.escapeHTML(I18n.t("post.media_placeholder.blocked_hotlinked"))}</span>"
|
|
|
|
|
|
|
|
el.replace(placeholder)
|
|
|
|
|
|
|
|
true
|
|
|
|
end
|
|
|
|
|
2021-11-23 03:32:12 +08:00
|
|
|
def oneboxed_images
|
|
|
|
@doc.css(".onebox-body img, .onebox img, img.onebox")
|
|
|
|
end
|
|
|
|
|
|
|
|
def is_a_hyperlink?(img)
|
|
|
|
parent = img.parent
|
|
|
|
while parent
|
|
|
|
return true if parent.name == "a"
|
|
|
|
parent = parent.parent if parent.respond_to?(:parent)
|
|
|
|
end
|
|
|
|
false
|
|
|
|
end
|
|
|
|
|
|
|
|
def process_inline_onebox(element)
|
|
|
|
inline_onebox =
|
|
|
|
InlineOneboxer.lookup(
|
|
|
|
element.attributes["href"].value,
|
|
|
|
invalidate: !!@opts[:invalidate_oneboxes],
|
|
|
|
user_id: @model&.user_id,
|
|
|
|
category_id: @category_id,
|
|
|
|
)
|
|
|
|
|
|
|
|
if title = inline_onebox&.dig(:title)
|
|
|
|
element.children = CGI.escapeHTML(title)
|
|
|
|
element.add_class("inline-onebox")
|
|
|
|
end
|
|
|
|
|
|
|
|
remove_inline_onebox_loading_class(element)
|
|
|
|
end
|
|
|
|
|
|
|
|
def remove_inline_onebox_loading_class(element)
|
|
|
|
element.remove_class("inline-onebox-loading")
|
|
|
|
end
|
|
|
|
|
|
|
|
def dirty?
|
|
|
|
@previous_cooked != html
|
|
|
|
end
|
|
|
|
|
|
|
|
def html
|
|
|
|
@doc.try(:to_html)
|
|
|
|
end
|
|
|
|
|
|
|
|
def create_link_node(klass, url, external = false)
|
|
|
|
a = create_node("a", klass)
|
|
|
|
a["href"] = url
|
|
|
|
if external
|
|
|
|
a["target"] = "_blank"
|
|
|
|
a["rel"] = "nofollow noopener"
|
|
|
|
end
|
|
|
|
a
|
|
|
|
end
|
|
|
|
|
|
|
|
def create_icon_node(klass)
|
|
|
|
icon = create_node("svg", "fa d-icon d-icon-#{klass} svg-icon")
|
|
|
|
icon.set_attribute("aria-hidden", "true")
|
2021-11-25 12:22:43 +08:00
|
|
|
icon << "<use href=\"##{klass}\"></use>"
|
2021-11-23 03:32:12 +08:00
|
|
|
end
|
|
|
|
|
|
|
|
def create_node(tag_name, klass)
|
2022-02-26 10:52:11 +08:00
|
|
|
node = @doc.document.create_element(tag_name)
|
2021-11-23 03:32:12 +08:00
|
|
|
node["class"] = klass if klass.present?
|
2022-02-26 10:52:11 +08:00
|
|
|
@doc.add_child(node)
|
2021-11-23 03:32:12 +08:00
|
|
|
node
|
|
|
|
end
|
|
|
|
|
|
|
|
def create_span_node(klass, content = nil)
|
|
|
|
span = create_node("span", klass)
|
|
|
|
span.content = content if content
|
|
|
|
span
|
|
|
|
end
|
2023-12-06 14:59:18 +08:00
|
|
|
|
|
|
|
def each_responsive_ratio
|
|
|
|
SiteSetting
|
|
|
|
.responsive_post_image_sizes
|
|
|
|
.split("|")
|
|
|
|
.map(&:to_f)
|
|
|
|
.sort
|
|
|
|
.each { |r| yield r if r > 1 }
|
|
|
|
end
|
2021-11-23 03:32:12 +08:00
|
|
|
end
|