database/doris
2
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

[optimize](string) optimize char_length function by SIMD (#18925)

Browse Source 
Optimize char_length function by SIMD
(1) optimize utf8_len compute
(2) 840% up
This commit is contained in:
ZhangYu0123
2023-04-28 17:22:35 +08:00
committed by GitHub
parent aef9355cd3
commit 6626f26506
3 changed files with 51 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
be/src/vec/functions/function_string.cpp
View File

@ -95,7 +95,7 @@ struct StringUtf8LengthImpl {
for (int i = 0; i < size; ++i) {
const char* raw_str = reinterpret_cast<const char*>(&data[offsets[i - 1]]);
int str_size = offsets[i] - offsets[i - 1];
res[i] = get_char_len(StringRef(raw_str, str_size), str_size);
res[i] = simd::VStringFunctions::get_char_len(raw_str, str_size);
}
return Status::OK();
}
@ -223,7 +223,8 @@ struct StringInStrImpl {
// Hive returns positions starting from 1.
int loc = search.search(&lstr_ref);
if (loc > 0) {
loc = get_char_len(lstr_ref, loc);
size_t len = std::min(lstr_ref.size, (size_t)loc);
loc = simd::VStringFunctions::get_char_len(lstr_ref.data, len);
}
res[i] = loc + 1;
}
@ -263,7 +264,8 @@ struct StringInStrImpl {
// Hive returns positions starting from 1.
int loc = search.search(&strl);
if (loc > 0) {
loc = get_char_len(strl, loc);
size_t len = std::min((size_t)loc, strl.size);
loc = simd::VStringFunctions::get_char_len(strl.data, len);
}
return loc + 1;