Implement array to string function (#3320)

src/function/CMakeLists.txt

@@ -3,6 +3,7 @@ add_subdirectory(arithmetic)
 add_subdirectory(cast)
 add_subdirectory(pattern)
 add_subdirectory(table)
+add_subdirectory(list)
 
 add_library(kuzu_function
         OBJECT

src/function/function_collection.cpp

@@ -110,7 +110,7 @@ FunctionCollection* FunctionCollection::getFunctions() {
         SCALAR_FUNCTION(ListSumFunction), SCALAR_FUNCTION(ListProductFunction),
         SCALAR_FUNCTION(ListDistinctFunction), SCALAR_FUNCTION(ListUniqueFunction),
         SCALAR_FUNCTION(ListAnyValueFunction), SCALAR_FUNCTION(ListReverseFunction),
-        SCALAR_FUNCTION(SizeFunction),
+        SCALAR_FUNCTION(SizeFunction), SCALAR_FUNCTION(ListToStringFunction),
 
         // Cast functions
         SCALAR_FUNCTION(CastToDateFunction), SCALAR_FUNCTION_ALIAS(CastToDateFunction),

src/function/list/CMakeLists.txt

@@ -0,0 +1,7 @@
+add_library(kuzu_list_function
+        OBJECT
+        list_to_string_function.cpp)
+
+set(ALL_OBJECT_FILES
+        ${ALL_OBJECT_FILES} $<TARGET_OBJECTS:kuzu_list_function>
+        PARENT_SCOPE)

src/function/list/list_to_string_function.cpp

@@ -0,0 +1,25 @@
+#include "function/list/functions/list_to_string_function.h"
+
+#include "common/type_utils.h"
+
+namespace kuzu {
+namespace function {
+
+void ListToString::operation(common::list_entry_t& input, common::ku_string_t& delim,
+    common::ku_string_t& result, common::ValueVector& inputVector,
+    common::ValueVector& /*delimVector*/, common::ValueVector& resultVector) {
+    std::string resultStr = "";
+    auto dataVector = common::ListVector::getDataVector(&inputVector);
+    for (auto i = 0u; i < input.size - 1; i++) {
+        resultStr += common::TypeUtils::entryToString(dataVector->dataType,
+            common::ListVector::getListValuesWithOffset(&inputVector, input, i), dataVector);
+        resultStr += delim.getAsString();
+    }
+    resultStr += common::TypeUtils::entryToString(dataVector->dataType,
+        common::ListVector::getListValuesWithOffset(&inputVector, input, input.size - 1),
+        dataVector);
+    common::StringVector::addString(&resultVector, result, resultStr);
+}
+
+} // namespace function
+} // namespace kuzu

src/function/vector_list_functions.cpp

@@ -20,6 +20,7 @@
 #include "function/list/functions/list_slice_function.h"
 #include "function/list/functions/list_sort_function.h"
 #include "function/list/functions/list_sum_function.h"
+#include "function/list/functions/list_to_string_function.h"
 #include "function/list/functions/list_unique_function.h"
 #include "function/scalar_function.h"
 
@@ -924,5 +925,15 @@ function_set ListReverseFunction::getFunctionSet() {
     return result;
 }
 
+function_set ListToStringFunction::getFunctionSet() {
+    function_set result;
+    result.push_back(std::make_unique<ScalarFunction>(name,
+        std::vector<LogicalTypeID>{LogicalTypeID::LIST, LogicalTypeID::STRING},
+        LogicalTypeID::STRING,
+        ScalarFunction::BinaryExecListStructFunction<list_entry_t, ku_string_t, ku_string_t,
+            ListToString>));
+    return result;
+}
+
 } // namespace function
 } // namespace kuzu

src/include/function/list/functions/list_to_string_function.h

@@ -0,0 +1,16 @@
+#pragma once
+
+#include "common/types/ku_string.h"
+#include "common/vector/value_vector.h"
+
+namespace kuzu {
+namespace function {
+
+struct ListToString {
+    static void operation(common::list_entry_t& input, common::ku_string_t& delim,
+        common::ku_string_t& result, common::ValueVector& inputVector,
+        common::ValueVector& /*delimVector*/, common::ValueVector& resultVector);
+};
+
+} // namespace function
+} // namespace kuzu

src/include/function/list/functions/list_unique_function.h

@@ -17,18 +17,23 @@ struct ValueEquality {
 
 using ValueSet = std::unordered_set<common::Value, ValueHashFunction, ValueEquality>;
 
-using duplicateValueHandler = std::function<void(const std::string&)>;
-using uniqueValueHandler = std::function<void(common::ValueVector& dataVector, uint64_t pos)>;
+using duplicate_value_handler = std::function<void(const std::string&)>;
+using unique_value_handler = std::function<void(common::ValueVector& dataVector, uint64_t pos)>;
+using null_value_handler = std::function<void()>;
 
 struct ListUnique {
     static uint64_t appendListElementsToValueSet(common::list_entry_t& input,
-        common::ValueVector& inputVector, duplicateValueHandler duplicateValHandler = nullptr,
-        uniqueValueHandler uniqueValueHandler = nullptr) {
+        common::ValueVector& inputVector, duplicate_value_handler duplicateValHandler = nullptr,
+        unique_value_handler uniqueValueHandler = nullptr,
+        null_value_handler nullValueHandler = nullptr) {
         ValueSet uniqueKeys;
         auto dataVector = common::ListVector::getDataVector(&inputVector);
         auto val = common::Value::createDefaultValue(dataVector->dataType);
         for (auto i = 0u; i < input.size; i++) {
             if (dataVector->isNull(input.offset + i)) {
+                if (nullValueHandler != nullptr) {
+                    nullValueHandler();
+                }
                 continue;
             }
             auto entryVal = common::ListVector::getListValuesWithOffset(&inputVector, input, i);

src/include/function/list/vector_list_functions.h

@@ -129,5 +129,11 @@ struct ListReverseFunction {
     static function_set getFunctionSet();
 };
 
+struct ListToStringFunction {
+    static constexpr const char* name = "LIST_TO_STRING";
+
+    static function_set getFunctionSet();
+};
+
 } // namespace function
 } // namespace kuzu

src/include/function/map/functions/map_creation_function.h

@@ -9,11 +9,17 @@
 namespace kuzu {
 namespace function {
 
+static void duplicateValueHandler(const std::string& key) {
+    throw common::RuntimeException{common::stringFormat("Found duplicate key: {} in map.", key)};
+}
+
+static void nullValueHandler() {
+    throw common::RuntimeException("Null value key is not allowed in map.");
+}
+
 static void validateKeys(common::list_entry_t& keyEntry, common::ValueVector& keyVector) {
-    ListUnique::appendListElementsToValueSet(keyEntry, keyVector, [](const std::string& key) {
-        throw common::RuntimeException{
-            common::stringFormat("Found duplicate key: {} in map.", key)};
-    });
+    ListUnique::appendListElementsToValueSet(keyEntry, keyVector, duplicateValueHandler,
+        nullptr /* uniqueValueHandler */, nullValueHandler);
 }
 
 struct MapCreation {

test/test_files/tinysnb/function/list.test

@@ -2130,3 +2130,55 @@ Ad
 -STATEMENT Return list_product(["ok"]);
 ---- error
 Binder exception: Unsupported inner data type for LIST_PRODUCT: STRING
+
+-LOG ListToStringInt128
+-STATEMENT RETURN LIST_TO_STRING([CAST(5, 'INT128'), CAST(7, 'INT128'), CAST(10, 'INT128')], '--')
+---- 1
+5--7--10
+
+-LOG ListToStringInt64
+-STATEMENT RETURN LIST_TO_STRING([CAST(255, 'INT64'), CAST(-51, 'INT64'), CAST(3132, 'INT64')], '--')
+---- 1
+255---51--3132
+
+-LOG ListToStringBOOL
+-STATEMENT RETURN LIST_TO_STRING([true, false, false, true], '/')
+---- 1
+True/False/False/True
+
+-LOG ListToStringUUID
+-STATEMENT MATCH (p:person) return list_to_string(collect(p.u), '*')
+---- 1
+a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a11*a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a12*a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a13*a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a14*a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a15*a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a16*a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a17*a0eebc99-9c0b-4ef8-bb6d-6bb9bd380a18
+
+-LOG ListToStringDate
+-STATEMENT MATCH (p:person) return list_to_string(collect(p.birthdate), '+')
+---- 1
+1900-01-01+1900-01-01+1940-06-22+1950-07-23+1980-10-26+1980-10-26+1980-10-26+1990-11-27
+
+-LOG ListToStringTimestamp
+-STATEMENT MATCH (p:person) return list_to_string(collect(p.registerTime), '==')
+---- 1
+2011-08-20 11:25:30==2008-11-03 15:25:30.000526==1911-08-20 02:32:21==2031-11-30 12:25:30==1976-12-23 11:21:42==1972-07-31 13:22:30.678559==1976-12-23 04:41:42==2023-02-21 13:25:30
+
+-LOG ListToStringList
+-STATEMENT MATCH (p:person) return list_to_string(p.workedHours, ',')
+---- 8
+10,5
+12,8
+4,5
+1,9
+2
+3,4,5,6,7
+1
+10,11,12,3,4,5,6,7
+
+-LOG ListToStringStruct
+-STATEMENT MATCH (o:organisation) return list_to_string(collect(o.state), '|')
+---- 1
+{revenue: 138, location: ['toronto','montr,eal'], stock: {price: [96,56], volume: 1000}}|{revenue: 152, location: ["vanco,uver north area"], stock: {price: [15,78,671], volume: 432}}|{revenue: 558, location: ['very long city name','new york'], stock: {price: [22], volume: 99}}
+
+-LOG ListToStringArray
+-STATEMENT MATCH (p:person) return list_to_string(collect(p.grades), '/')
+---- 1
+[96,54,86,92]/[98,42,93,88]/[91,75,21,95]/[76,88,99,89]/[96,59,65,88]/[80,78,34,83]/[43,83,67,43]/[77,64,100,54]

test/test_files/tinysnb/projection/single_label.test

@@ -667,6 +667,11 @@ Runtime exception: Found duplicate key: {a: 5, b: 3} in map.
 ---- error
 Runtime exception: Found duplicate key: 0 in map.
 
+-LOG MapWithNullKey
+-STATEMENT RETURN map([cast(5, 'int128'), NULL, cast(5, 'int128')], [20, 34, 50]);
+---- error
+Runtime exception: Null value key is not allowed in map.
+
 -LOG ReturnMapLiteralWithProp
 -STATEMENT MATCH (p:person) RETURN map([p.ID, p.age], [p.fName, p.fName]);
 ---- 8