[Feature](json) Support json_search function in 2.1 #41590
Conversation
|
Thank you for your contribution to Apache Doris. Since 2024-03-18, the Document has been moved to doris-website. |
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
| bool seek(const char* string, size_t length); | ||
|
|
||
| void add_leg_to_leg_vector(std::unique_ptr<leg_info> leg) { | ||
| leg_vector.emplace_back(leg.release()); |
There was a problem hiding this comment.
warning: use of undeclared identifier 'leg_vector' [clang-diagnostic-error]
leg_vector.emplace_back(leg.release());
^| leg_vector.emplace_back(leg.release()); | ||
| } | ||
|
|
||
| void pop_leg_from_leg_vector() { leg_vector.pop_back(); } |
There was a problem hiding this comment.
warning: use of undeclared identifier 'leg_vector' [clang-diagnostic-error]
void pop_leg_from_leg_vector() { leg_vector.pop_back(); }
^|
|
||
| bool to_string(std::string* res) const { | ||
| res->push_back(SCOPE); | ||
| for (const auto& leg : leg_vector) { |
There was a problem hiding this comment.
warning: use of undeclared identifier 'leg_vector' [clang-diagnostic-error]
for (const auto& leg : leg_vector) {
^| return true; | ||
| } | ||
|
|
||
| size_t get_leg_vector_size() { return leg_vector.size(); } |
There was a problem hiding this comment.
warning: use of undeclared identifier 'leg_vector' [clang-diagnostic-error]
size_t get_leg_vector_size() { return leg_vector.size(); }
^|
|
||
| size_t get_leg_vector_size() { return leg_vector.size(); } | ||
|
|
||
| leg_info* get_leg_from_leg_vector(size_t i) { return leg_vector[i].get(); } |
There was a problem hiding this comment.
warning: use of undeclared identifier 'leg_vector' [clang-diagnostic-error]
leg_info* get_leg_from_leg_vector(size_t i) { return leg_vector[i].get(); }
^| } else if (element.isArray()) { | ||
| const SimdJSONParser::Array& array = element.getArray(); | ||
| bool find = false; | ||
| for (size_t i = 0; i < array.size(); ++i) { |
There was a problem hiding this comment.
warning: use range-based for loop instead [modernize-loop-convert]
| for (size_t i = 0; i < array.size(); ++i) { | |
| for (auto child_element : array) { |
be/src/vec/functions/function_jsonb.cpp:1676:
- const SimdJSONParser::Element& child_element = array[i];
- // construct an array cell path leg.
+ // construct an array cell path leg.| return Status::OK(); | ||
| } | ||
|
|
||
| Status execute_impl(FunctionContext* context, Block& block, const ColumnNumbers& arguments, |
There was a problem hiding this comment.
warning: function 'execute_impl' exceeds recommended size/complexity thresholds [readability-function-size]
Status execute_impl(FunctionContext* context, Block& block, const ColumnNumbers& arguments,
^Additional context
be/src/vec/functions/function_jsonb.cpp:1827: 120 lines including whitespace and comments (threshold 80)
Status execute_impl(FunctionContext* context, Block& block, const ColumnNumbers& arguments,
^| } | ||
| } | ||
|
|
||
| Status FunctionLike::construct_like_const_state(FunctionContext* context, const StringRef& pattern, |
There was a problem hiding this comment.
warning: function 'construct_like_const_state' exceeds recommended size/complexity thresholds [readability-function-size]
Status FunctionLike::construct_like_const_state(FunctionContext* context, const StringRef& pattern,
^Additional context
be/src/vec/functions/like.cpp:818: 96 lines including whitespace and comments (threshold 80)
Status FunctionLike::construct_like_const_state(FunctionContext* context, const StringRef& pattern,
^
|
run buildall |
liutang123
force-pushed
the
json_search-2.1
branch
from
7ef7457 to
d144000
Compare
|
run buildall |
|
TeamCity be ut coverage result: |
Like mysql, json_search returns the path which point to a json string
witch match the pattern.
`SELECT JSON_SEARCH('["A",[{"B":"1"}],{"C":"AB"},{"D":"BC"}]', 'one',
'A_') as res;`
```
+----------+
| res |
+----------+
| "$[2].C" |
+----------+
```
liutang123
force-pushed
the
json_search-2.1
branch
from
d144000 to
36f9479
Compare
|
run buildall |
|
TeamCity be ut coverage result: |
2 participants
cherry-pick #40948
Like mysql, json_search returns the path which point to a json string witch match the pattern.
SELECT JSON_SEARCH('["A",[{"B":"1"}],{"C":"AB"},{"D":"BC"}]', 'one', 'A_') as res;Proposed changes
Issue Number: close #xxx