database/mariadb-columnstore-engine
1
0
Fork 0
mirror of https://github.com/mariadb-corporation/mariadb-columnstore-engine.git synced 2025-07-30 19:23:07 +03:00
Code Activity

Part#1 MCOL-495 Make string comparison not case sensitive

Browse Source 
Fixing field='str' for long (Dict) string data types.
This commit is contained in:
Alexander Barkov
2020-12-01 10:48:59 +04:00
parent a53f93df43
commit 0ff6a6ec20
7 changed files with 96 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
primitives/linux-port/dictionary.cpp
View File

@ -186,11 +186,11 @@ void PrimitiveProcessor::p_TokenByScan(const TokenByScanRequestHeader* h,
if (eqFilter)
{
// MCOL-1246 Trim whitespace before match
// TODO MCOL-3536 use CHARSET_INFO* cs for collation
// cs->hash_sort(hash_sort(const uchar *key, size_t len, ulong *nr1, ulong *nr2))
if (cs != & eqFilter->getCharset())
{
//throw runtime_error("Collations mismatch: TokenByScanRequestHeader and DicEqualityFilter");
}
string strData(sig, siglen);
boost::trim_right_if(strData, boost::is_any_of(" "));
bool gotIt = eqFilter->find(strData) != eqFilter->end();
if ((h->COP1 == COMPARE_EQ && gotIt) || (h->COP1 == COMPARE_NE &&

23
primitives/linux-port/primitiveprocessor.h
View File

@ -102,7 +102,28 @@ public:
typedef std::tr1::unordered_set<int64_t, pcfHasher, pcfEqual> prestored_set_t;
typedef std::tr1::unordered_set<int128_t, pcfHasher128, pcfEqual128> prestored_set_t_128;
typedef std::tr1::unordered_set<std::string, utils::Hasher> DictEqualityFilter;
class DictEqualityFilter: public std::tr1::unordered_set<std::string,
datatypes::CollationAwareHasher,
datatypes::CollationAwareComparator>
{
public:
DictEqualityFilter(const datatypes::Charset &cs)
:std::tr1::unordered_set<std::string,
datatypes::CollationAwareHasher,
datatypes::CollationAwareComparator>
(10,
datatypes::CollationAwareHasher(cs),
datatypes::CollationAwareComparator(cs))
{ }
CHARSET_INFO & getCharset() const
{
idbassert(& _M_h1.getCharset() == & _M_eq.getCharset());
return _M_h1.getCharset();
}
};
struct idb_regex_t
{

2
primitives/primproc/dictstep.cpp
View File

@ -86,7 +86,7 @@ void DictStep::createCommand(ByteStream& bs)
{
string strTmp;
eqFilter.reset(new primitives::DictEqualityFilter());
eqFilter.reset(new primitives::DictEqualityFilter(my_charset_latin1));
bs >> eqOp;
//cout << "saw the eqfilter count=" << filterCount << endl;

9
primitives/primproc/primitiveserver.cpp
View File

@ -1804,12 +1804,15 @@ public:
private:
void createEqualityFilter()
{
uint32_t uniqueID, count, i;
uint32_t uniqueID, count, i, charsetNumber;
string str;
boost::shared_ptr<DictEqualityFilter> filter(new DictEqualityFilter());
bs->advance(sizeof(ISMPacketHeader));
*bs >> uniqueID;
*bs >> charsetNumber;
datatypes::Charset cs(charsetNumber);
boost::shared_ptr<DictEqualityFilter> filter(new DictEqualityFilter(cs));
*bs >> count;
for (i = 0; i < count; i++)