thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-29 11:02:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

Better scheme cut

Browse Source
This commit is contained in:
alesapin 2019-06-25 12:04:35 +03:00
parent 48451b2b0f
commit fa88954e56
2 changed files with 9 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
dbms/src/Functions/URL/domain.h
View File

@ -32,23 +32,19 @@ inline StringRef getURLHost(const char * data, size_t size)
Pos end = data + size; Pos end = data + size;
Pos slash_pos = find_first_symbols<'/'>(pos, end); Pos slash_pos = find_first_symbols<'/'>(pos, end);
if (slash_pos != end) if (slash_pos < end - 1 && *(slash_pos + 1) == '/')
pos = slash_pos; pos = slash_pos + 2;
else else
pos = data; pos = data;
if (pos != data) if (pos != data)
{ {
StringRef scheme = getURLScheme(data, size); StringRef scheme = getURLScheme(data, end - pos);
Pos scheme_end = data + scheme.size; Pos scheme_end = data + scheme.size;
if (pos - scheme_end != 1 || *scheme_end != ':') if (scheme.size && (pos - scheme_end != 3 || *scheme_end != ':'))
return StringRef{}; return StringRef{};
} }
// Check with we still have // character from the scheme
if (end - pos > 2 && *(pos) == '/' && *(pos + 1) == '/')
pos += 2;
auto start_of_host = pos; auto start_of_host = pos;
Pos dot_pos = nullptr; Pos dot_pos = nullptr;
for (; pos < end; ++pos) for (; pos < end; ++pos)

10
dbms/tests/queries/1_stateful/00038_uniq_state_merge2.reference
View File

@ -1,16 +1,16 @@
ru 262911 69218 ru 262914 69218
92101 89421 92101 89421
com 63297 30285 com 63298 30285
ua 29037 17475 ua 29037 17475
html 25077 15037 html 25079 15039
tr 16770 11857 tr 16770 11857
net 16387 11686 net 16387 11686
php 14373 10307 php 14374 10307
yandsearch 12024 9484 yandsearch 12024 9484
by 8192 6915 by 8192 6915
yandex 7211 6124 yandex 7211 6124
org 4890 4514 org 4890 4514
kz 4677 4209 kz 4679 4211
tv 4400 3928 tv 4400 3928
su 2602 2396 su 2602 2396
phtml 2409 2226 phtml 2409 2226