diff --git a/config/locales/server.en.yml b/config/locales/server.en.yml
index 3d7fbc45c24..4e7d572a746 100644
--- a/config/locales/server.en.yml
+++ b/config/locales/server.en.yml
@@ -986,6 +986,7 @@ en:
post_onebox_maxlength: "Maximum length of a oneboxed Discourse post in characters."
onebox_domains_blacklist: "A list of domains that will never be oneboxed."
inline_onebox_domains_whitelist: "A list of domains that will be oneboxed in miniature form if linked without a title"
+ enable_inline_onebox_on_all_domains: "Ignore inline_onebox_domain_whitelist site setting and allow inline onebox on all domains."
max_oneboxes_per_post: "Maximum number of oneboxes in a post."
logo_url: "The logo image at the top left of your site, should be a wide rectangle shape. If left blank site title text will be shown."
diff --git a/config/site_settings.yml b/config/site_settings.yml
index b4c95b37e75..8fe5daf3fde 100644
--- a/config/site_settings.yml
+++ b/config/site_settings.yml
@@ -949,6 +949,8 @@ onebox:
inline_onebox_domains_whitelist:
default: ''
type: list
+ enable_inline_onebox_on_all_domains:
+ default: false
spam:
add_rel_nofollow_to_user_content: true
diff --git a/lib/inline_oneboxer.rb b/lib/inline_oneboxer.rb
index f345bef4d38..4c71cba888d 100644
--- a/lib/inline_oneboxer.rb
+++ b/lib/inline_oneboxer.rb
@@ -36,13 +36,15 @@ class InlineOneboxer
end
end
- if whitelist = SiteSetting.inline_onebox_domains_whitelist
+ always_allow = SiteSetting.enable_inline_onebox_on_all_domains
+ domains = SiteSetting.inline_onebox_domains_whitelist&.split('|') unless always_allow
+
+ if always_allow || domains
uri = URI(url) rescue nil
- domains = whitelist.split('|')
if uri.present? &&
uri.hostname.present? &&
- domains.include?(uri.hostname) &&
+ (always_allow || domains.include?(uri.hostname)) &&
title = RetrieveTitle.crawl(url)
return onebox_for(url, title, opts)
end
diff --git a/lib/retrieve_title.rb b/lib/retrieve_title.rb
index 62d5498c157..e68809a09f0 100644
--- a/lib/retrieve_title.rb
+++ b/lib/retrieve_title.rb
@@ -13,11 +13,11 @@ module RetrieveTitle
title = nil
if doc = Nokogiri::HTML(html)
- if node = doc.at('meta[property="og:title"]')
+ title = doc.at('title')&.inner_text
+
+ if !title && node = doc.at('meta[property="og:title"]')
title = node['content']
end
-
- title ||= doc.at('title')&.inner_text
end
if title.present?
@@ -42,9 +42,6 @@ module RetrieveTitle
# Fetch the beginning of a HTML document at a url
def self.fetch_beginning(url)
- # Never crawl in test mode
- return if Rails.env.test?
-
fd = FinalDestination.new(url)
uri = fd.resolve
return "" unless uri
diff --git a/spec/components/inline_oneboxer_spec.rb b/spec/components/inline_oneboxer_spec.rb
index d58a0d3a7e3..491f322a0cd 100644
--- a/spec/components/inline_oneboxer_spec.rb
+++ b/spec/components/inline_oneboxer_spec.rb
@@ -68,6 +68,25 @@ describe InlineOneboxer do
expect(onebox).to be_blank
end
+ it "will crawl anything if allowed to" do
+ SiteSetting.enable_inline_onebox_on_all_domains = true
+
+ # Final destination does a HEAD and a GET
+ stub_request(:head, "https://eviltrout.com/some-path").to_return(status: 200)
+
+ stub_request(:get, "https://eviltrout.com/some-path").
+ to_return(status: 200, body: "a blog", headers: {})
+
+ onebox = InlineOneboxer.lookup(
+ "https://eviltrout.com/some-path",
+ skip_cache: true
+ )
+
+ expect(onebox).to be_present
+ expect(onebox[:url]).to eq("https://eviltrout.com/some-path")
+ expect(onebox[:title]).to eq("a blog")
+ end
+
it "will lookup whitelisted domains" do
SiteSetting.inline_onebox_domains_whitelist = "eviltrout.com"
RetrieveTitle.stubs(:crawl).returns("Evil Trout's Blog")
diff --git a/spec/components/retrieve_title_spec.rb b/spec/components/retrieve_title_spec.rb
index 351a28acfab..a65be0b34d7 100644
--- a/spec/components/retrieve_title_spec.rb
+++ b/spec/components/retrieve_title_spec.rb
@@ -21,11 +21,22 @@ describe RetrieveTitle do
expect(title).to eq("Another Title")
end
+ it "will pick og:title if title is missing" do
+ title = RetrieveTitle.extract_title(<<~HTML
+
+
+ HTML
+ )
+
+ expect(title).to eq("Good Title")
+ end
+
it "will prefer the title from an opengraph tag" do
title = RetrieveTitle.extract_title(<<~HTML
- Bad Title
-
+ Good Title
+
HTML
)