Fix case-insensetive search in UTF8 strings

[tavplubix]

Changelog category (leave one):

• Bug Fix (user-visible misbehaviour in official stable or prestable release)

Changelog entry (a user-readable short description of the changes that goes to CHANGELOG.md):
Functions for case-insensitive search in UTF8 strings like positionCaseInsensitiveUTF8 and countSubstringsCaseInsensitiveUTF8 might find substrings that actually does not match, it's fixed.

[alexey-milovidov]

Do I understand correctly that we just perform full comparison after found a candidate match, and before this PR it was something less trivial?

PS. Do we have perf test for it?

[tavplubix]

we just perform full comparison after found a candidate match?

Yep

before this PR it was something less trivial

Not precisely, before this PR we just skipped first cache_valid_len bytes of needle assuming that these bytes match (this assumption was incorrect) and then performed full comparison of the rest bytes. The only difference is that now we do not skip needle prefix and compare it too.

PS. Do we have perf test for it?

ClickHouse/tests/performance/cpu_synthetic.xml

Lines 38 to 40 in 990f8a2

	<!-- Sophisticated substring search. -->
	<query>SELECT count() FROM hits_10m_single WHERE positionCaseInsensitiveUTF8(URL, 'новости') != 0 SETTINGS max_threads = 1</query>
	<query>SELECT count() FROM hits_100m_single WHERE positionCaseInsensitiveUTF8(URL, 'новости') != 0</query>

[alexey-milovidov]

Alright! I expect there will be no visible performance degradation.

[alexey-milovidov]

Integration tests were temporarily broken for a moment - ok.

[alexey-milovidov]

No difference in peft test.