thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-24 00:22:29 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix topLevelDomain() for IDN hosts

Browse Source
This commit is contained in:
Azat Khuzhin 2021-06-09 02:30:29 +03:00
parent 38ac83dff9
commit e0c1780370
4 changed files with 12 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
src/Functions/URL/topLevelDomain.cpp
View File

@ -28,7 +28,10 @@ struct ExtractTopLevelDomain
return; return;
/// For IPv4 addresses select nothing. /// For IPv4 addresses select nothing.
if (last_dot[1] <= '9') ///
/// NOTE: it is safe to access last_dot[1]
/// since getURLHost() will not return a host if there is symbol after dot.
if (isNumericASCII(last_dot[1]))
return; return;
res_data = last_dot + 1; res_data = last_dot + 1;

3
tests/queries/0_stateless/00398_url_functions.reference
View File

@ -35,6 +35,9 @@ ru
com com
com com
com com
рф
====PATH==== ====PATH====
П П
%D%9 %D%9

3
tests/queries/0_stateless/00398_url_functions.sql
View File

@ -38,6 +38,9 @@ SELECT topLevelDomain('svn+ssh://example.ru.?q=hello%20world') AS Domain;
SELECT topLevelDomain('//www.example.com') AS Domain; SELECT topLevelDomain('//www.example.com') AS Domain;
SELECT topLevelDomain('www.example.com') as Domain; SELECT topLevelDomain('www.example.com') as Domain;
SELECT topLevelDomain('example.com') as Domain; SELECT topLevelDomain('example.com') as Domain;
SELECT topLevelDomain('example.рф') as Domain;
SELECT topLevelDomain('example.') as Domain;
SELECT topLevelDomain('example') as Domain;
SELECT '====PATH===='; SELECT '====PATH====';
SELECT decodeURLComponent('%D0%9F'); SELECT decodeURLComponent('%D0%9F');

4
tests/queries/1_stateful/00038_uniq_state_merge2.reference
View File

@ -1,5 +1,5 @@
ru 262914 69218 ru 262914 69218
92101 89421 91872 89417
com 63298 30285 com 63298 30285
ua 29037 17475 ua 29037 17475
html 25079 15039 html 25079 15039
@ -53,6 +53,7 @@ eu 237 234
liveinteria 218 218 liveinteria 218 218
to 215 213 to 215 213
mamba 214 214 mamba 214 214
рф 209 204
auto-supers 208 208 auto-supers 208 208
sberbank 207 207 sberbank 207 207
tj 205 205 tj 205 205
@ -97,4 +98,3 @@ loveplaceOfSearchplus 111 111
nl 111 111 nl 111 111
bstatistika 107 107 bstatistika 107 107
br 102 102 br 102 102
sport 99 99