Leatherface/mastodon
1
0
Fork 0
mirror of https://github.com/glitch-soc/mastodon.git synced 2025-12-15 16:59:41 +00:00
Code Issues Packages Projects Releases Wiki Activity

Improve handling of encoding problems when creating link previews (#30929)

Browse Source
This commit is contained in:
David Roetzel
2024-07-05 13:54:38 +02:00
committed by GitHub
parent 05f0d51005
commit 016c1e4e78
5 changed files with 76 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
spec/fixtures/requests/latin1_posing_as_utf8_broken.txt vendored Normal file
View File

@@ -0,0 +1,17 @@
HTTP/1.1 200 OK
server: nginx
date: Thu, 13 Jun 2024 14:33:13 GMT
content-type: text/html; charset=utf-8
content-length: 158
accept-ranges: bytes
<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Tofu <20> l'orange</title>
</head>
<body>
<h2>Tofu <20> l'orange</h2>
</body>
</html>

17
spec/fixtures/requests/latin1_posing_as_utf8_recoverable.txt vendored Normal file
View File

@@ -0,0 +1,17 @@
HTTP/1.1 200 OK
server: nginx
date: Thu, 13 Jun 2024 14:33:13 GMT
content-type: text/html; charset=utf-8
content-length: 158
accept-ranges: bytes
<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Tofu with orange sauce</title>
</head>
<body>
<h2>Tofu <20> l'orange</h2>
</body>
</html>

28
spec/services/fetch_link_card_service_spec.rb
View File

@@ -27,6 +27,8 @@ RSpec.describe FetchLinkCardService do
stub_request(:get, 'http://example.com/koi8-r').to_return(request_fixture('koi8-r.txt'))
stub_request(:get, 'http://example.com/windows-1251').to_return(request_fixture('windows-1251.txt'))
stub_request(:get, 'http://example.com/low_confidence_latin1').to_return(request_fixture('low_confidence_latin1.txt'))
stub_request(:get, 'http://example.com/latin1_posing_as_utf8_broken').to_return(request_fixture('latin1_posing_as_utf8_broken.txt'))
stub_request(:get, 'http://example.com/latin1_posing_as_utf8_recoverable').to_return(request_fixture('latin1_posing_as_utf8_recoverable.txt'))
stub_request(:get, 'http://example.com/aergerliche-umlaute').to_return(request_fixture('redirect_with_utf8_url.txt'))
Rails.cache.write('oembed_endpoint:example.com', oembed_cache) if oembed_cache
@@ -159,10 +161,30 @@ RSpec.describe FetchLinkCardService do
end
context 'with a URL of a page in ISO-8859-1 encoding, that charlock_holmes cannot detect' do
let(:status) { Fabricate(:status, text: 'Check out http://example.com/low_confidence_latin1') }
context 'when encoding in http header is correct' do
let(:status) { Fabricate(:status, text: 'Check out http://example.com/low_confidence_latin1') }
it 'decodes the HTML' do
expect(status.preview_card.title).to eq("Tofu á l'orange")
it 'decodes the HTML' do
expect(status.preview_card.title).to eq("Tofu á l'orange")
end
end
context 'when encoding in http header is incorrect' do
context 'when encoding problems appear in unrelated tags' do
let(:status) { Fabricate(:status, text: 'Check out http://example.com/latin1_posing_as_utf8_recoverable') }
it 'decodes the HTML' do
expect(status.preview_card.title).to eq('Tofu with orange sauce')
end
end
context 'when encoding problems appear in title tag' do
let(:status) { Fabricate(:status, text: 'Check out http://example.com/latin1_posing_as_utf8_broken') }
it 'does not create a preview card' do
expect(status.preview_card).to be_nil
end
end
end
end