IntenseWebs/discourse
3
0
Fork 0
mirror of https://github.com/discourse/discourse.git synced 2025-02-25 18:55:32 -06:00
Code Issues Packages Projects Releases Wiki Activity

FIX: Improve regexp for matching version lexeme.

Browse Source 
Follow up to b70f1084f7
This commit is contained in:
Guo Xiang Tan 2020-07-27 15:17:49 +08:00
parent de5ef5d895
commit 0f53ad58c2
No known key found for this signature in database
GPG Key ID: FBD110179AAC1F20
2 changed files with 10 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
app/services/search_indexer.rb
View File

@ -59,7 +59,7 @@ class SearchIndexer
tsvector.scan(/'(([a-zA-Z0-9]+\.)+[a-zA-Z0-9]+)'\:([\w+,]+)/).reduce(additional_lexemes) do |array, (lexeme, _, positions)| tsvector.scan(/'(([a-zA-Z0-9]+\.)+[a-zA-Z0-9]+)'\:([\w+,]+)/).reduce(additional_lexemes) do |array, (lexeme, _, positions)|
count = 0 count = 0
if lexeme !~ /^(\d+\.)?(\d+\.)?(\*|\d+)$/ if lexeme !~ /^(\d+\.)?(\d+\.)*(\*|\d+)$/
loop do loop do
count += 1 count += 1
break if count >= 10 # Safeguard here to prevent infinite loop when a term has many dots break if count >= 10 # Safeguard here to prevent infinite loop when a term has many dots

11
spec/services/search_indexer_spec.rb
View File

@ -143,10 +143,17 @@ describe SearchIndexer do
it 'should not tokenize versions' do it 'should not tokenize versions' do
post.topic.update!(title: "this is a title that I am testing") post.topic.update!(title: "this is a title that I am testing")
post.update!(raw: '1.2.2') post.update!(raw: '123.223')
expect(post.post_search_data.search_data).to eq( expect(post.post_search_data.search_data).to eq(
"'1.2.2':10 'test':8A 'titl':4A 'uncategor':9B" "'123.223':10 'test':8A 'titl':4A 'uncategor':9B"
)
post.update!(raw: '15.2.231.423')
post.reload
expect(post.post_search_data.search_data).to eq(
"'15.2.231.423':10 'test':8A 'titl':4A 'uncategor':9B"
) )
end end