FIX: don't index urls to local files

This commit is contained in:
Régis Hanol 2018-09-13 18:53:53 +02:00 committed by Sam
parent 74eec1849d
commit 39a2d92417
2 changed files with 27 additions and 3 deletions

View File

@ -167,6 +167,8 @@ class SearchIndexer
class HtmlScrubber < Nokogiri::XML::SAX::Document
DIACRITICS ||= /([\u0300-\u036f]|[\u1AB0-\u1AFF]|[\u1DC0-\u1DFF]|[\u20D0-\u20FF])/
def self.strip_diacritics(str)
s = str.unicode_normalize(:nfkd)
s.gsub!(DIACRITICS, "")
@ -196,12 +198,12 @@ class SearchIndexer
attributes = Hash[*attributes.flatten]
ATTRIBUTES.each do |name|
characters(attributes[name]) if attributes[name].present?
if attributes[name].present?
characters(attributes[name]) unless name == "href" && UrlHelper.is_local(attributes[name])
end
end
end
DIACRITICS ||= /([\u0300-\u036f]|[\u1AB0-\u1AFF]|[\u1DC0-\u1DFF]|[\u20D0-\u20FF])/
def characters(str)
str = HtmlScrubber.strip_diacritics(str) if @strip_diacritics
scrubbed << " #{str} "

View File

@ -2,6 +2,7 @@ require 'rails_helper'
describe SearchIndexer do
let(:post_id) { 99 }
it 'correctly indexes chinese' do
SiteSetting.default_locale = 'zh_CN'
data = "你好世界"
@ -37,6 +38,27 @@ describe SearchIndexer do
expect(scrubbed).to eq(" HELLO Heterogeneite Здравствуите هتاف للترحيب 你好 ")
end
it "doesn't index local files" do
html = <<~HTML
<p><img src="https://www.discourse.org/logo.png" alt="Discourse"></p>
<p><img src="#{Discourse.base_url_no_prefix}/uploads/episodeinteractive/original/3X/0/f/0f40b818356bdc1d80acfa905034e95cfd112a3a.png" alt="51%20PM" width="289" height="398"></p>
<div class="lightbox-wrapper">
<a class="lightbox" href="#{Discourse.base_url_no_prefix}/uploads/episodeinteractive/original/3X/1/6/16790095df3baf318fb2eb1d7e5d7860dc45d48b.jpg" data-download-href="#{Discourse.base_url_no_prefix}/uploads/episodeinteractive/16790095df3baf318fb2eb1d7e5d7860dc45d48b" title="Untitled design (21).jpg" rel="nofollow noopener">
<img src="#{Discourse.base_url_no_prefix}/uploads/episodeinteractive/optimized/3X/1/6/16790095df3baf318fb2eb1d7e5d7860dc45d48b_1_563x500.jpg" alt="Untitled%20design%20(21)" width="563" height="500">
<div class="meta">
<span class="filename">Untitled design (21).jpg</span>
<span class="informations">1280x1136 472 KB</span>
<span class="expand"></span>
</div>
</a>
</div>
HTML
scrubbed = SearchIndexer::HtmlScrubber.scrub(html).gsub(/\s+/, " ")
expect(scrubbed).to eq(" Discourse 51%20PM Untitled design (21).jpg Untitled%20design%20(21) Untitled design (21).jpg 1280x1136 472 KB ")
end
it 'correctly indexes a post according to version' do
# Preparing so that they can be indexed to right version
SearchIndexer.update_posts_index(post_id, "dummy", "", nil, nil)