VectorScan patterns can grow large (up to multiple MBs) and queries
involving different VectorScan patterns lead to an ever increasing
pattern cache size.
With this commit, the unbounded cache for vectorscan patterns is
replaced by a bounded cache with "CacheTable" eviction strategy, similar
to what we have for re2 patterns. The cache size is currently hard-coded
to 500 entries.
Fixes#19869
After making function multi[Fuzzy]Match(Any|AnyIndex|AllIndices)() work
with non-const needles, 12 more functions started to fail in test
"00233_position_function_family":
multiSearchAny()
multiSearchAnyCaseInsensitive()
multiSearchAnyUTF8
multiSearchAnyCaseInsensitiveUTF8()
multiSearchFirstPosition()
multiSearchFirstPositionCaseInsensitive()
multiSearchFirstPositionUTF8()
multiSearchFirstPositionCaseInsensitiveUTF8()
multiSearchFirstIndex()
multiSearchFirstIndexCaseInsensitive()
multiSearchFirstIndexUTF8()
multiSearchFirstIndexCaseInsensitiveUTF8()
Failing queries take the form
select 0 = multiSearchAny('\0', CAST([], 'Array(String)'));
- With this, invalid combinations of the FindAny/FindAnyIndex bools are
no longer possible and we can remove the corresponding check
- Also makes the instantiations more readable.
- The patterns are not used in hashing, there should not be a performance
impact when we use stuff from the standard library instead.
- added forgotten .reserve() in FunctionsMultiStringPosition.h
This commit migrates ClickHouse to Vectorscan. The first 10 min of
[0] explain the reasons for it.
(*) Addresses (but does not resolve) #38046
(*) Config parameter names (e.g. "max_hyperscan_regexp_length") are
preserved for compatibility. Likewise, error codes (e.g.
"ErrorCodes::HYPERSCAN_CANNOT_SCAN_TEXT") and function/class names (e.g.
"HyperscanDeleter") are preserved as vectorscan aims to be a drop-in
replacement.
[0] https://www.youtube.com/watch?v=KlZWmmflW6M
- introduced class MatchTraits with enums that replace bool template
parameters
- (minor: made negation the last template parameters because negation
executes last during evaluation)
