[BugFix](functions) fix multi_search_all_positions #18682
This commit is contained in:
@ -416,70 +416,4 @@ struct LibCASCIICaseInsensitiveStringSearcher : public StringSearcherBase {
|
||||
return search(haystack, haystack + haystack_size);
|
||||
}
|
||||
};
|
||||
|
||||
template <typename StringSearcher>
|
||||
class MultiStringSearcherBase {
|
||||
private:
|
||||
/// needles
|
||||
const std::vector<StringRef>& needles;
|
||||
/// searchers
|
||||
std::vector<StringSearcher> searchers;
|
||||
/// last index of needles that was not processed
|
||||
size_t last;
|
||||
|
||||
public:
|
||||
explicit MultiStringSearcherBase(const std::vector<StringRef>& needles_)
|
||||
: needles {needles_}, last {0} {
|
||||
searchers.reserve(needles.size());
|
||||
|
||||
size_t size = needles.size();
|
||||
for (int i = 0; i < size; ++i) {
|
||||
const char* cur_needle_data = needles[i].data;
|
||||
const size_t cur_needle_size = needles[i].size;
|
||||
|
||||
searchers.emplace_back(cur_needle_data, cur_needle_size);
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* while (hasMoreToSearch())
|
||||
* {
|
||||
* search inside the haystack with the known needles
|
||||
* }
|
||||
*/
|
||||
bool hasMoreToSearch() {
|
||||
if (last >= needles.size()) {
|
||||
return false;
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
bool searchOne(const uint8_t* haystack, const uint8_t* haystack_end) {
|
||||
const size_t size = needles.size();
|
||||
if (last >= size) {
|
||||
return false;
|
||||
}
|
||||
|
||||
if (searchers[++last].search(haystack, haystack_end) != haystack_end) {
|
||||
return true;
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
template <typename CountCharsCallback, typename AnsType>
|
||||
void searchOneAll(const uint8_t* haystack, const uint8_t* haystack_end, AnsType* answer,
|
||||
const CountCharsCallback& count_chars) {
|
||||
const size_t size = needles.size();
|
||||
for (; last < size; ++last) {
|
||||
const uint8_t* ptr = searchers[last].search(haystack, haystack_end);
|
||||
if (ptr != haystack_end) {
|
||||
answer[last] = count_chars(haystack, ptr);
|
||||
}
|
||||
}
|
||||
}
|
||||
};
|
||||
|
||||
using MultiStringSearcher = MultiStringSearcherBase<ASCIICaseSensitiveStringSearcher>;
|
||||
|
||||
} // namespace doris
|
||||
|
||||
@ -18,6 +18,9 @@
|
||||
// https://github.com/ClickHouse/ClickHouse/blob/master/src/Functions/FunctionsMultiStringPosition.h
|
||||
// and modified by Doris
|
||||
|
||||
#include <cstdint>
|
||||
#include <iterator>
|
||||
|
||||
#include "function.h"
|
||||
#include "function_helpers.h"
|
||||
#include "vec/columns/column_array.h"
|
||||
@ -103,48 +106,53 @@ public:
|
||||
}
|
||||
};
|
||||
|
||||
template <typename Impl>
|
||||
struct FunctionMultiSearchAllPositionsImpl {
|
||||
public:
|
||||
using ResultType = Int32;
|
||||
|
||||
using SingleSearcher = ASCIICaseSensitiveStringSearcher;
|
||||
static constexpr auto name = "multi_search_all_positions";
|
||||
|
||||
static Status vector_constant(const ColumnString::Chars& haystack_data,
|
||||
const ColumnString::Offsets& haystack_offsets,
|
||||
const Array& needles_arr, PaddedPODArray<Int32>& vec_res,
|
||||
PaddedPODArray<UInt64>& offsets_res) {
|
||||
if (needles_arr.size() > std::numeric_limits<UInt8>::max())
|
||||
if (needles_arr.size() > std::numeric_limits<UInt8>::max()) {
|
||||
return Status::InvalidArgument(
|
||||
"number of arguments for function {} doesn't match: "
|
||||
"passed {}, should be at most 255",
|
||||
name, needles_arr.size());
|
||||
}
|
||||
|
||||
std::vector<StringRef> needles;
|
||||
needles.reserve(needles_arr.size());
|
||||
for (const auto& needle : needles_arr) needles.emplace_back(needle.get<StringRef>());
|
||||
|
||||
auto res_callback = [](const UInt8* start, const UInt8* end) -> Int32 {
|
||||
return 1 + Impl::count_chars(reinterpret_cast<const char*>(start),
|
||||
reinterpret_cast<const char*>(end));
|
||||
};
|
||||
|
||||
auto searcher = Impl::create_multi_searcher(needles);
|
||||
const size_t needles_size = needles_arr.size();
|
||||
std::vector<SingleSearcher> searchers;
|
||||
searchers.reserve(needles_size);
|
||||
for (const auto& needle : needles_arr) {
|
||||
searchers.emplace_back(needle.get<StringRef>().data, needle.get<StringRef>().size);
|
||||
}
|
||||
|
||||
const size_t haystack_size = haystack_offsets.size();
|
||||
const size_t needles_size = needles.size();
|
||||
|
||||
vec_res.resize(haystack_size * needles.size());
|
||||
vec_res.resize(haystack_size * needles_size);
|
||||
offsets_res.resize(haystack_size);
|
||||
|
||||
std::fill(vec_res.begin(), vec_res.end(), 0);
|
||||
|
||||
while (searcher.hasMoreToSearch()) {
|
||||
// we traverse to generator answer by Vector's slot of ColumnVector, not by Vector.
|
||||
// TODO: check if the order of loop is best. The large data may make us writing across the line which size out of L2 cache.
|
||||
for (size_t ans_slot_in_row = 0; ans_slot_in_row < searchers.size(); ans_slot_in_row++) {
|
||||
// is i.e. answer slot index in one Vector(row) of answer
|
||||
auto& searcher = searchers[ans_slot_in_row];
|
||||
size_t prev_haystack_offset = 0;
|
||||
for (size_t j = 0, from = 0; j < haystack_size; ++j, from += needles_size) {
|
||||
|
||||
for (size_t haystack_index = 0, res_index = ans_slot_in_row;
|
||||
haystack_index < haystack_size; ++haystack_index, res_index += needles_size) {
|
||||
const auto* haystack = &haystack_data[prev_haystack_offset];
|
||||
const auto* haystack_end = haystack + haystack_offsets[j] - prev_haystack_offset;
|
||||
searcher.searchOneAll(haystack, haystack_end, &vec_res[from], res_callback);
|
||||
prev_haystack_offset = haystack_offsets[j];
|
||||
const auto* haystack_end =
|
||||
haystack - prev_haystack_offset + haystack_offsets[haystack_index];
|
||||
|
||||
auto ans_now = searcher.search(haystack, haystack_end);
|
||||
vec_res[res_index] =
|
||||
ans_now >= haystack_end ? 0 : std::distance(haystack, ans_now) + 1;
|
||||
prev_haystack_offset = haystack_offsets[haystack_index];
|
||||
}
|
||||
}
|
||||
|
||||
@ -166,72 +174,72 @@ struct FunctionMultiSearchAllPositionsImpl {
|
||||
size_t prev_haystack_offset = 0;
|
||||
size_t prev_needles_offset = 0;
|
||||
|
||||
auto res_callback = [](const UInt8* start, const UInt8* end) -> Int32 {
|
||||
return 1 + Impl::count_chars(reinterpret_cast<const char*>(start),
|
||||
reinterpret_cast<const char*>(end));
|
||||
};
|
||||
|
||||
offsets_res.reserve(haystack_offsets.size());
|
||||
offsets_res.reserve(haystack_data.size());
|
||||
uint64_t offset_now = 0;
|
||||
|
||||
auto& nested_column =
|
||||
vectorized::check_and_get_column<vectorized::ColumnNullable>(needles_data)
|
||||
->get_nested_column();
|
||||
const ColumnString* needles_data_string = check_and_get_column<ColumnString>(nested_column);
|
||||
|
||||
std::vector<StringRef> needles;
|
||||
for (size_t i = 0; i < haystack_offsets.size(); ++i) {
|
||||
needles.reserve(needles_offsets[i] - prev_needles_offset);
|
||||
std::vector<StringRef> needles_for_row;
|
||||
// haystack first, row by row.
|
||||
for (size_t haystack_index = 0; haystack_index < haystack_offsets.size();
|
||||
++haystack_index) {
|
||||
// get haystack for this row.
|
||||
const auto* haystack = &haystack_data[prev_haystack_offset];
|
||||
const auto* haystack_end =
|
||||
haystack - prev_haystack_offset + haystack_offsets[haystack_index];
|
||||
|
||||
for (size_t j = prev_needles_offset; j < needles_offsets[i]; ++j) {
|
||||
needles.emplace_back(needles_data_string->get_data_at(j));
|
||||
// build needles for this row.
|
||||
needles_for_row.reserve(needles_offsets[haystack_index] - prev_needles_offset);
|
||||
for (size_t j = prev_needles_offset; j < needles_offsets[haystack_index]; ++j) {
|
||||
needles_for_row.emplace_back(needles_data_string->get_data_at(j));
|
||||
}
|
||||
|
||||
const size_t needles_size = needles.size();
|
||||
if (needles_size > std::numeric_limits<UInt8>::max())
|
||||
const size_t needles_row_size = needles_for_row.size();
|
||||
if (needles_row_size > std::numeric_limits<UInt8>::max()) {
|
||||
return Status::InvalidArgument(
|
||||
"number of arguments for function {} doesn't match: "
|
||||
"passed {}, should be at most 255",
|
||||
name, needles_size);
|
||||
|
||||
vec_res.resize(vec_res.size() + needles_size);
|
||||
|
||||
auto searcher = Impl::create_multi_searcher(needles);
|
||||
|
||||
std::fill(vec_res.begin() + vec_res.size() - needles_size, vec_res.end(), 0);
|
||||
|
||||
while (searcher.hasMoreToSearch()) {
|
||||
const auto* haystack = &haystack_data[prev_haystack_offset];
|
||||
const auto* haystack_end = haystack + haystack_offsets[i] - prev_haystack_offset;
|
||||
searcher.searchOneAll(haystack, haystack_end,
|
||||
&vec_res[vec_res.size() - needles_size], res_callback);
|
||||
name, needles_row_size);
|
||||
}
|
||||
|
||||
if (offsets_res.empty())
|
||||
offsets_res.push_back(needles_size);
|
||||
else
|
||||
offsets_res.push_back(offsets_res.back() + needles_size);
|
||||
// each searcher search for one needle.
|
||||
std::vector<SingleSearcher> searchers;
|
||||
searchers.clear();
|
||||
searchers.reserve(needles_row_size);
|
||||
for (auto needle : needles_for_row) {
|
||||
searchers.emplace_back(needle.data, needle.size);
|
||||
}
|
||||
|
||||
prev_haystack_offset = haystack_offsets[i];
|
||||
prev_needles_offset = needles_offsets[i];
|
||||
needles.clear();
|
||||
// search for first so that the ans's size is constant for each row.
|
||||
auto ans_row_begin = vec_res.size();
|
||||
vec_res.resize(vec_res.size() + needles_row_size);
|
||||
offset_now += searchers.size();
|
||||
offsets_res.emplace_back(offset_now);
|
||||
|
||||
//for now haystack, apply needle to search, generator answer by order.
|
||||
for (size_t ans_slot_in_row = 0; ans_slot_in_row < searchers.size();
|
||||
ans_slot_in_row++) {
|
||||
// is i.e. answer slot index in one Vector(row) of answer
|
||||
auto& searcher = searchers[ans_slot_in_row];
|
||||
|
||||
auto ans_now = searcher.search(haystack, haystack_end);
|
||||
vec_res[ans_row_begin + ans_slot_in_row] =
|
||||
ans_now >= haystack_end ? 0 : std::distance(haystack, ans_now) + 1;
|
||||
}
|
||||
|
||||
prev_haystack_offset = haystack_offsets[haystack_index];
|
||||
prev_needles_offset = needles_offsets[haystack_index];
|
||||
needles_for_row.clear();
|
||||
}
|
||||
|
||||
return Status::OK();
|
||||
}
|
||||
};
|
||||
|
||||
struct MultiSearcherImpl {
|
||||
using MultiSearcher = MultiStringSearcher;
|
||||
|
||||
static MultiSearcher create_multi_searcher(const std::vector<StringRef>& needles) {
|
||||
return MultiSearcher(needles);
|
||||
}
|
||||
|
||||
static size_t count_chars(const char* begin, const char* end) { return end - begin; }
|
||||
};
|
||||
|
||||
using FunctionMultiSearchAllPositions =
|
||||
FunctionMultiStringPosition<FunctionMultiSearchAllPositionsImpl<MultiSearcherImpl>>;
|
||||
FunctionMultiStringPosition<FunctionMultiSearchAllPositionsImpl>;
|
||||
|
||||
void register_function_multi_string_position(SimpleFunctionFactory& factory) {
|
||||
factory.register_function<FunctionMultiSearchAllPositions>();
|
||||
|
||||
@ -30,8 +30,7 @@ under the License.
|
||||
|
||||
`ARRAY<INT> multi_search_all_positions(VARCHAR haystack, ARRAY<VARCHAR> needles)`
|
||||
|
||||
|
||||
Searches for the substrings `needles` in the string `haystack`, and returns array of positions of the found corresponding substrings in the string. Positions are indexed starting from 1.
|
||||
Returns an `ARRAY` where the `i`-th element is the position of the `i`-th element in `needles`(i.e. `needle`)'s **first** occurrence in the string `haystack`. Positions are counted from 1, with 0 meaning the element was not found. **Case-sensitive**.
|
||||
|
||||
### example
|
||||
|
||||
@ -43,12 +42,13 @@ mysql> select multi_search_all_positions('Hello, World!', ['hello', '!', 'world'
|
||||
| [0,13,0] |
|
||||
+----------------------------------------------------------------------+
|
||||
|
||||
mysql> select multi_search_all_positions('abc', ['a', 'bc', 'd']);
|
||||
+-----------------------------------------------------+
|
||||
| multi_search_all_positions('abc', ['a', 'bc', 'd']) |
|
||||
+-----------------------------------------------------+
|
||||
| [1,2,0] |
|
||||
+-----------------------------------------------------+
|
||||
select multi_search_all_positions("Hello, World!", ['hello', '!', 'world', 'Hello', 'World']);
|
||||
+---------------------------------------------------------------------------------------------+
|
||||
| multi_search_all_positions('Hello, World!', ARRAY('hello', '!', 'world', 'Hello', 'World')) |
|
||||
+---------------------------------------------------------------------------------------------+
|
||||
| [0, 13, 0, 1, 8] |
|
||||
+---------------------------------------------------------------------------------------------+
|
||||
```
|
||||
|
||||
### keywords
|
||||
MULTI_SEARCH,SEARCH,POSITIONS
|
||||
|
||||
@ -30,8 +30,7 @@ under the License.
|
||||
|
||||
`ARRAY<INT> multi_search_all_positions(VARCHAR haystack, ARRAY<VARCHAR> needles)`
|
||||
|
||||
|
||||
搜索字符串 `haystack` 中的子字符串 `needles`,并返回在 `haystack` 中找到的相应 `needles` 的位置数组。位置的下标从 1 开始。
|
||||
返回一个 `ARRAY`,其中第 `i` 个元素为 `needles` 中第 `i` 个元素 `needle`,在字符串 `haystack` 中**首次**出现的位置。位置从1开始计数,0代表未找到该元素。**大小写敏感**。
|
||||
|
||||
### example
|
||||
|
||||
@ -43,12 +42,13 @@ mysql> select multi_search_all_positions('Hello, World!', ['hello', '!', 'world'
|
||||
| [0,13,0] |
|
||||
+----------------------------------------------------------------------+
|
||||
|
||||
mysql> select multi_search_all_positions('abc', ['a', 'bc', 'd']);
|
||||
+-----------------------------------------------------+
|
||||
| multi_search_all_positions('abc', ['a', 'bc', 'd']) |
|
||||
+-----------------------------------------------------+
|
||||
| [1,2,0] |
|
||||
+-----------------------------------------------------+
|
||||
select multi_search_all_positions("Hello, World!", ['hello', '!', 'world', 'Hello', 'World']);
|
||||
+---------------------------------------------------------------------------------------------+
|
||||
| multi_search_all_positions('Hello, World!', ARRAY('hello', '!', 'world', 'Hello', 'World')) |
|
||||
+---------------------------------------------------------------------------------------------+
|
||||
| [0, 13, 0, 1, 8] |
|
||||
+---------------------------------------------------------------------------------------------+
|
||||
```
|
||||
|
||||
### keywords
|
||||
MULTI_SEARCH,SEARCH,POSITIONS
|
||||
|
||||
@ -1,25 +1,30 @@
|
||||
-- This file is automatically generated. You should know what you did if you want to edit this
|
||||
-- !select --
|
||||
-- !table_select1 --
|
||||
[0, 13, 0, 1, 8]
|
||||
[0, 13, 0, 1, 8]
|
||||
[1, 13, 8, 0, 0]
|
||||
[1, 13, 8, 0, 0]
|
||||
[0, 6, 0, 0, 0]
|
||||
|
||||
-- !table_select2 --
|
||||
[0, 0]
|
||||
[0, 0, 1, 13]
|
||||
[0, 8]
|
||||
[1, 8, 0, 13]
|
||||
[1, 1, 4, 0]
|
||||
|
||||
-- !select1 --
|
||||
[4, 1, 1, 2, 6, 1, 1, 0, 4, 1, 14, 0, 10, 0, 16, 6]
|
||||
-- !select --
|
||||
|
||||
-- !select2 --
|
||||
[0, 0, 0, 2, 3, 0, 1, 0, 5, 0, 0, 0, 11, 10, 6, 7]
|
||||
-- !select --
|
||||
|
||||
-- !select3 --
|
||||
[1, 1, 0, 0, 0, 1, 1, 1, 4, 0, 6, 6, 0, 10, 1, 5]
|
||||
-- !select --
|
||||
|
||||
-- !select4 --
|
||||
[1, 0, 0, 8, 6, 0, 7, 1, 3, 0, 0, 0, 0, 12]
|
||||
-- !select --
|
||||
|
||||
-- !select5 --
|
||||
[7, 0, 0, 8, 0, 2, 0, 0, 6, 0, 2, 0, 3, 1]
|
||||
-- !select --
|
||||
[0, 0, 0, 19, 14, 22, 10, 0, 0, 13, 0, 8]
|
||||
-- !select --
|
||||
[6, 1, 1, 0, 0, 5, 1, 0, 8, 0, 5, 0, 2, 12, 0, 15, 0, 0]
|
||||
-- !select --
|
||||
[0, 0, 5, 1, 1, 0, 15, 1, 5, 10, 4, 0, 1, 0, 3, 0, 0, 0]
|
||||
-- !select --
|
||||
[0, 1, 11, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4, 1]
|
||||
-- !select --
|
||||
[0, 0, 0, 3, 0, 15, 0, 0, 12, 7, 0, 0, 0, 0, 5, 0]
|
||||
-- !select --
|
||||
[0, 0, 12, 4, 4, 0, 13, 23, 0, 1, 0, 2, 0, 0, 0, 3, 0, 0]
|
||||
-- !select --
|
||||
[6, 8, 1, 4, 0, 10, 0, 1, 14, 0, 1, 0, 5, 0, 0, 0, 0, 15, 0, 1]
|
||||
|
||||
|
||||
@ -16,16 +16,39 @@
|
||||
// under the License.
|
||||
|
||||
suite("test_multi_string_position") {
|
||||
qt_select "select multi_search_all_positions('jmdqwjbrxlbatqeixknricfk', ['qwjbrxlba', 'jmd', '', 'mdqwjbrxlbatqe', 'jbrxlbatqeixknric', 'jmdqwjbrxlbatqeixknri', '', 'fdtmnwtts', 'qwjbrxlba', '', 'qeixknricfk', 'hzjjgrnoilfkvzxaemzhf', 'lb', 'kamz', 'ixknr', 'jbrxlbatq'])"
|
||||
qt_select "select multi_search_all_positions('coxcctuehmzkbrsmodfvx', ['bkhnp', 'nlypjvriuk', 'rkslxwfqjjivcwdexrdtvjdtvuu', 'oxcctuehm', 'xcctuehmzkbrsm', 'kfrieuocovykjmkwxbdlkgwctwvcuh', 'coxc', 'lbwvetgxyndxjqqwthtkgasbafii', 'ctuehmzkbrsmodfvx', 'obzldxjldxowk', 'ngfikgigeyll', 'wdaejjukowgvzijnw', 'zkbr', 'mzkb', 'tuehm', 'ue'])"
|
||||
qt_select "select multi_search_all_positions('mpswgtljbbrmivkcglamemayfn', ['', 'm', 'saejhpnfgfq', 'rzanrkdssmmkanqjpfi', 'oputeneprgoowg', 'mp', '', '', 'wgtljbbrmivkcglamemay', 'cbpthtrgrmgfypizi', 'tl', 'tlj', 'xuhs', 'brmivkcglamemayfn', '', 'gtljb'])"
|
||||
qt_select "select multi_search_all_positions('arbphzbbecypbzsqsljurtddve', ['arbphzb', 'mnrboimjfijnti', 'cikcrd', 'becypbz', 'z', 'uocmqgnczhdcrvtqrnaxdxjjlhakoszuwc', 'bbe', '', 'bp', 'yhltnexlpdijkdzt', 'jkwjmrckvgmccmmrolqvy', 'vdxmicjmfbtsbqqmqcgtnrvdgaucsgspwg', 'witlfqwvhmmyjrnrzttrikhhsrd', 'pbzsqsljurt'])"
|
||||
qt_select "select multi_search_all_positions('aizovxqpzcbbxuhwtiaaqhdqjdei', ['qpzcbbxuhw', 'jugrpglqbm', 'dspwhzpyjohhtizegrnswhjfpdz', 'pzcbbxuh', 'vayzeszlycke', 'i', 'gvrontcpqavsjxtjwzgwxugiyhkhmhq', 'gyzmeroxztgaurmrqwtmsxcqnxaezuoapatvu', 'xqpzc', 'mjiswsvlvlpqrhhptqq', 'iz', 'hmzjxxfjsvcvdpqwtrdrp', 'zovxqpzcbbxuhwtia', 'ai'])"
|
||||
qt_select "select multi_search_all_positions('ydfgiluhyxwqdfiwtzobwzscyxhuov', ['srsoubrgghleyheujsbwwwykerzlqphgejpxvog', 'axchkyleddjwkvbuyhmekpbbbztxdlm', 'zqodzvlkmfe', 'obwz', 'fi', 'zsc', 'xwq', 'pvmurvrd', 'uulcdtexckmrsokmgdpkstlkoavyrmxeaacvydxf', 'dfi', 'mxcngttujzgtlssrmluaflmjuv', 'hyxwqdfiwtzobwzscyxhu'])"
|
||||
qt_select "select multi_search_all_positions('pyepgwainvmwekwhhqxxvzdjw', ['w', '', '', 'gvvkllofjnxvcu', 'kmwwhboplctvzazcyfpxhwtaddfnhekei', 'gwainv', 'pyepgwain', 'ekpnogkzzmbpfynsunwqp', 'invmwe', 'hrxpiplfplqjsstuybksuteoz', 'gwa', 'akfpyduqrwosxcbdemtxrxvundrgse', 'yepgwainvmw', 'wekwhhqxxvzdjw', 'fyimzvedmyriubgoznmcav', 'whhq', 'ozxowbwdqfisuupyzaqynoprgsjhkwlum', 'vpoufrofekajksdp'])"
|
||||
qt_select "select multi_search_all_positions('lqwahffxurkbhhzytequotkfk', ['rwjqudpuaiufle', 'livwgbnflvy', 'hffxurkbhh', '', '', 'xcajwbqbttzfzfowjubmmgnmssat', 'zytequ', 'lq', 'h', 'rkbhh', 'a', 'immejthwgdr', '', 'llhhnlhcvnxxorzzjt', 'w', 'cvjynqxcivmmmvc', 'wexjomdcmursppjtsweybheyxzleuz', 'fzronsnddfxwlkkzidiknhpjipyrcrzel'])"
|
||||
qt_select "select multi_search_all_positions('nkddriylnakicwgdwrfxpodqea', ['izwdpgrgpmjlwkanjrffgela', '', 'kicw', 'hltmfymgmrjckdiylkzjlvvyuleksikdjrg', 'yigveskrbidknjxigwilmkgyizewikh', 'xyvzhsnqmuec', 'odcgzlavzrwesjks', 'oilvfgliktoujukpgzvhmokdgkssqgqot', 'llsfsurvimbahwqtbqbp', 'nxj', 'pimydixeobdxmdkvhcyzcgnbhzsydx', 'couzmvxedobuohibgxwoxvmpote', 'driylnakicwgdwrf', 'nkddr'])"
|
||||
qt_select "select multi_search_all_positions('jnckhtjqwycyihuejibqmddrdxe', ['tajzx', 'vuddoylclxatcjvinusdwt', 'spxkhxvzsljkmnzpeubszjnhqczavgtqopxn', 'ckhtjqwycyi', 'xlbfzdxspldoes', 'u', 'czosfebeznt', 'gzhabdsuyreisxvyfrfrkq', 'yihuejibqmd', 'jqwycyihuejibqm', 'cfbvprgzx', 'hxu', 'vxbhrfpzacgd', 'afoaij', 'htjqwycyihu', 'httzbskqd'])"
|
||||
qt_select "select multi_search_all_positions('dzejajvpoojdkqbnayahygidyrjmb', ['khwxxvtnqhobbvwgwkpusjlhlzifiuclycml', 'nzvuhtwdaivo', 'dkqbnayahygidyr', 'jajvpoo', 'j', 'wdtbvwmeqgyvetu', 'kqbn', 'idyrjmb', 'tsnxuxevsxrxpgpfdgrkhwqpkse', '', 'efsdgzuefhdzkmquxu', 'zejajvpoojdkqbnayahyg', 'ugwfuighbygrxyctop', 'fcbxzbdugc', 'dxmzzrcplob', 'ejaj', 'wmmupyxrylvawsyfccluiiene', 'ohzmsqhpzbafvbzqwzftbvftei'])"
|
||||
qt_select "select multi_search_all_positions('ffaujlverosspbzaqefjzql', ['lvero', 'erossp', 'f', 'ujlverosspbz', 'btfimgklzzxlbkbuqyrmnud', 'osspb', 'muqexvtjuaar', 'f', 'bzaq', 'lprihswhwkdhqciqhfaowarn', 'ffaujlve', 'uhbbjrqjb', 'jlver', 'umucyhbbu', 'pjthtzmgxhvpbdphesnnztuu', 'xfqhfdfsbbazactpastzvzqudgk', 'lvovjfoatc', 'z', 'givejzhoqsd', ''])"
|
||||
def table_name = "strings"
|
||||
|
||||
sql """ DROP TABLE IF EXISTS ${table_name} """
|
||||
sql """ CREATE TABLE IF NOT EXISTS ${table_name}
|
||||
(
|
||||
`col1` INT NOT NULL,
|
||||
`content` TEXT NOT NULL,
|
||||
`mode` ARRAY<TEXT> NOT NULL
|
||||
) ENGINE=OLAP
|
||||
DUPLICATE KEY(`col1`)
|
||||
COMMENT 'OLAP'
|
||||
DISTRIBUTED BY HASH(`col1`) BUCKETS 3
|
||||
PROPERTIES (
|
||||
"replication_allocation" = "tag.location.default: 1",
|
||||
"in_memory" = "false",
|
||||
"storage_format" = "V2"
|
||||
);
|
||||
"""
|
||||
|
||||
sql """ INSERT INTO ${table_name} (col1, content, mode) VALUES
|
||||
(1, 'Hello, World!', ['hello', 'world'] ),
|
||||
(2, 'Hello, World!', ['hello', 'world', 'Hello', '!'] ),
|
||||
(3, 'hello, world!', ['Hello', 'world'] ),
|
||||
(4, 'hello, world!', ['hello', 'world', 'Hello', '!'] ),
|
||||
(5, 'HHHHW!', ['H', 'HHHH', 'HW', 'WH'] );
|
||||
"""
|
||||
|
||||
qt_table_select1 "select multi_search_all_positions(content, ['hello', '!', 'world', 'Hello', 'World']) from ${table_name} order by col1"
|
||||
qt_table_select2 "select multi_search_all_positions(content, mode) from ${table_name} order by col1"
|
||||
|
||||
qt_select1 "select multi_search_all_positions('jmdqwjbrxlbatqeixknricfk', ['qwjbrxlba', 'jmd', '', 'mdqwjbrxlbatqe', 'jbrxlbatqeixknric', 'jmdqwjbrxlbatqeixknri', '', 'fdtmnwtts', 'qwjbrxlba', '', 'qeixknricfk', 'hzjjgrnoilfkvzxaemzhf', 'lb', 'kamz', 'ixknr', 'jbrxlbatq'])"
|
||||
qt_select2 "select multi_search_all_positions('coxcctuehmzkbrsmodfvx', ['bkhnp', 'nlypjvriuk', 'rkslxwfqjjivcwdexrdtvjdtvuu', 'oxcctuehm', 'xcctuehmzkbrsm', 'kfrieuocovykjmkwxbdlkgwctwvcuh', 'coxc', 'lbwvetgxyndxjqqwthtkgasbafii', 'ctuehmzkbrsmodfvx', 'obzldxjldxowk', 'ngfikgigeyll', 'wdaejjukowgvzijnw', 'zkbr', 'mzkb', 'tuehm', 'ue'])"
|
||||
qt_select3 "select multi_search_all_positions('mpswgtljbbrmivkcglamemayfn', ['', 'm', 'saejhpnfgfq', 'rzanrkdssmmkanqjpfi', 'oputeneprgoowg', 'mp', '', '', 'wgtljbbrmivkcglamemay', 'cbpthtrgrmgfypizi', 'tl', 'tlj', 'xuhs', 'brmivkcglamemayfn', '', 'gtljb'])"
|
||||
qt_select4 "select multi_search_all_positions('arbphzbbecypbzsqsljurtddve', ['arbphzb', 'mnrboimjfijnti', 'cikcrd', 'becypbz', 'z', 'uocmqgnczhdcrvtqrnaxdxjjlhakoszuwc', 'bbe', '', 'bp', 'yhltnexlpdijkdzt', 'jkwjmrckvgmccmmrolqvy', 'vdxmicjmfbtsbqqmqcgtnrvdgaucsgspwg', 'witlfqwvhmmyjrnrzttrikhhsrd', 'pbzsqsljurt'])"
|
||||
qt_select5 "select multi_search_all_positions('aizovxqpzcbbxuhwtiaaqhdqjdei', ['qpzcbbxuhw', 'jugrpglqbm', 'dspwhzpyjohhtizegrnswhjfpdz', 'pzcbbxuh', 'vayzeszlycke', 'i', 'gvrontcpqavsjxtjwzgwxugiyhkhmhq', 'gyzmeroxztgaurmrqwtmsxcqnxaezuoapatvu', 'xqpzc', 'mjiswsvlvlpqrhhptqq', 'iz', 'hmzjxxfjsvcvdpqwtrdrp', 'zovxqpzcbbxuhwtia', 'ai'])"
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user