Add fixed_list dataType

CMakeLists.txt

@@ -1,6 +1,6 @@
 cmake_minimum_required(VERSION 3.11)
 
-project(Kuzu VERSION 0.0.2 LANGUAGES CXX)
+project(Kuzu VERSION 0.0.3 LANGUAGES CXX)
 
 find_package(Threads REQUIRED)
 

dataset/tinysnb/eWorkAt.csv

@@ -1,3 +1,3 @@
-3,4,2015
-5,6,2010
-7,6,2015
+3,4,2015,"[3.8,2.5]"
+5,6,2010,"[2.1,4.4]"
+7,6,2015,"[9.2,3.1]"

dataset/tinysnb/schema.cypher

@@ -1,8 +1,8 @@
-create node table person (ID INt64, fName StRING, gender INT64, isStudent BoOLEAN, isWorker BOOLEAN, age INT64, eyeSight DOUBLE, birthdate DATE, registerTime TIMESTAMP, lastJobDuration interval, workedHours INT64[], usedNames STRING[], courseScoresPerTerm INT64[][], PRIMARY KEY (ID));
+create node table person (ID INt64, fName StRING, gender INT64, isStudent BoOLEAN, isWorker BOOLEAN, age INT64, eyeSight DOUBLE, birthdate DATE, registerTime TIMESTAMP, lastJobDuration interval, workedHours INT64[], usedNames STRING[], courseScoresPerTerm INT64[][], grades INT64[4],PRIMARY KEY (ID));
 create node table organisation (ID INT64, name STRING, orgCode INT64, mark DOUBLE, score INT64, history STRING, licenseValidInterval INTERVAL, rating DOUBLE, PRIMARY KEY (ID));
 create node table movies (name STRING, PRIMARY KEY (name));
 create rel table knows (FROM person TO person, date DATE, meetTime TIMESTAMP, validInterval INTERVAL, comments STRING[], MANY_MANY);
 create rel table studyAt (FROM person TO organisation, year INT64, places STRING[], MANY_ONE);
-create rel table workAt (FROM person TO organisation, year INT64, MANY_ONE);
+create rel table workAt (FROM person TO organisation, year INT64, grading DOUBLE[2], MANY_ONE);
 create rel table meets (FROM person TO person, MANY_ONE);
 create rel table marries (FROM person TO person, usedAddress STRING[], note STRING, ONE_ONE);

dataset/tinysnb/vPerson.csv

@@ -1,9 +1,9 @@
-id,fname,Gender,ISStudent,isWorker,age,eyeSight,birthdate,registerTime,lastJobDuration,workedHours,usedNames,courseScoresPerTerm
-0,Alice,1,true,false,35,5.0,1900-01-01,2011-08-20 11:25:30Z+00:00,3 years 2 days 13 hours 2 minutes,"[10,5]","[Aida]","[[10,8],[6,7,8]]"
-2,Bob,2,true,false,30,5.1,1900-01-01,2008-11-03 13:25:30.000526-02:00,10 years 5 months 13 hours 24 us,"[12,8]","[Bobby]","[[8,9],[9,10]]"
-3,Carol,1,false,true,45,5.0,1940-06-22,1911-08-20 02:32:21,48 hours 24 minutes 11 seconds,"[4,5]","[Carmen,Fred]","[[8,10]]"
-5,Dan,2,false,true,20,4.8,1950-7-23,2031-11-30 12:25:30Z,10 years 5 months 13 hours 24 us,"[1,9]","[Wolfeschlegelstein,Daniel]","[[7,4],[8,8],[9]]"
-7,Elizabeth,1,false,true,20,4.7,1980-10-26,1976-12-23 11:21:42,48 hours 24 minutes 11 seconds,"[2]","[Ein]","[[6],[7],[8]]"
-8,Farooq,2,true,false,25,4.5,1980-10-26,1972-07-31 13:22:30.678559,18 minutes 24 milliseconds,"[3,4,5,6,7]","[Fesdwe]","[[8]]"
-9,Greg,2,false,false,40,4.9,1980-10-26,1976-12-23 11:21:42Z+06:40,10 years 5 months 13 hours 24 us,"[1]","[Grad]","[[10]]"
-10,Hubert Blaine Wolfeschlegelsteinhausenbergerdorff,2,false,true,83,4.9,1990-11-27,2023-02-21 13:25:30,3 years 2 days 13 hours 2 minutes,"[10,11,12,3,4,5,6,7]","[Ad,De,Hi,Kye,Orlan]","[[7],[10],[6,7]]"
+id,fname,Gender,ISStudent,isWorker,age,eyeSight,birthdate,registerTime,lastJobDuration,workedHours,usedNames,courseScoresPerTerm,grades
+0,Alice,1,true,false,35,5.0,1900-01-01,2011-08-20 11:25:30Z+00:00,3 years 2 days 13 hours 2 minutes,"[10,5]","[Aida]","[[10,8],[6,7,8]]","[96,54,86,92]"
+2,Bob,2,true,false,30,5.1,1900-01-01,2008-11-03 13:25:30.000526-02:00,10 years 5 months 13 hours 24 us,"[12,8]","[Bobby]","[[8,9],[9,10]]","[98,42,93,88]"
+3,Carol,1,false,true,45,5.0,1940-06-22,1911-08-20 02:32:21,48 hours 24 minutes 11 seconds,"[4,5]","[Carmen,Fred]","[[8,10]]","[91,75,21,95]"
+5,Dan,2,false,true,20,4.8,1950-7-23,2031-11-30 12:25:30Z,10 years 5 months 13 hours 24 us,"[1,9]","[Wolfeschlegelstein,Daniel]","[[7,4],[8,8],[9]]","[76,88,99,89]"
+7,Elizabeth,1,false,true,20,4.7,1980-10-26,1976-12-23 11:21:42,48 hours 24 minutes 11 seconds,"[2]","[Ein]","[[6],[7],[8]]","[96,59,65,88]"
+8,Farooq,2,true,false,25,4.5,1980-10-26,1972-07-31 13:22:30.678559,18 minutes 24 milliseconds,"[3,4,5,6,7]","[Fesdwe]","[[8]]","[80,78,34,83]"
+9,Greg,2,false,false,40,4.9,1980-10-26,1976-12-23 11:21:42Z+06:40,10 years 5 months 13 hours 24 us,"[1]","[Grad]","[[10]]","[43,83,67,43]"
+10,Hubert Blaine Wolfeschlegelsteinhausenbergerdorff,2,false,true,83,4.9,1990-11-27,2023-02-21 13:25:30,3 years 2 days 13 hours 2 minutes,"[10,11,12,3,4,5,6,7]","[Ad,De,Hi,Kye,Orlan]","[[7],[10],[6,7]]","[77,64,100,54]"

src/antlr4/Cypher.g4

@@ -99,7 +99,7 @@ kU_DataType
 
 kU_ListIdentifiers : kU_ListIdentifier ( kU_ListIdentifier )* ;
 
-kU_ListIdentifier : '[' ']' ;
+kU_ListIdentifier : '[' oC_IntegerLiteral? ']' ;
 
 oC_AnyCypherOption
     : oC_Explain

src/binder/bind/bind_ddl.cpp

@@ -79,7 +79,7 @@ std::unique_ptr<BoundStatement> Binder::bindAddProperty(const Statement& stateme
     validateTableExist(catalog, tableName);
     auto catalogContent = catalog.getReadOnlyVersion();
     auto tableID = catalogContent->getTableID(tableName);
-    auto dataType = Types::dataTypeFromString(addProperty.getDataType());
+    auto dataType = bindDataType(addProperty.getDataType());
     if (catalogContent->getTableSchema(tableID)->containProperty(addProperty.getPropertyName())) {
         throw BinderException("Property: " + addProperty.getPropertyName() + " already exists.");
     }
@@ -137,8 +137,8 @@ std::vector<PropertyNameDataType> Binder::bindPropertyNameDataTypes(
                     propertyNameDataType.first.c_str()));
         }
         StringUtils::toUpper(propertyNameDataType.second);
-        boundPropertyNameDataTypes.emplace_back(
-            propertyNameDataType.first, Types::dataTypeFromString(propertyNameDataType.second));
+        auto dataType = bindDataType(propertyNameDataType.second);
+        boundPropertyNameDataTypes.emplace_back(propertyNameDataType.first, dataType);
         boundPropertyNames.emplace(propertyNameDataType.first);
     }
     return boundPropertyNameDataTypes;
@@ -176,5 +176,30 @@ property_id_t Binder::bindPropertyName(TableSchema* tableSchema, const std::stri
         tableSchema->tableName + " table doesn't have property: " + propertyName + ".");
 }
 
+DataType Binder::bindDataType(const std::string& dataType) {
+    auto boundType = Types::dataTypeFromString(dataType);
+    if (boundType.typeID == common::FIXED_LIST) {
+        auto validNumericTypes = common::DataType::getNumericalTypeIDs();
+        if (find(validNumericTypes.begin(), validNumericTypes.end(), boundType.childType->typeID) ==
+            validNumericTypes.end()) {
+            throw common::BinderException(
+                "The child type of a fixed list must be a numeric type. Given: " +
+                common::Types::dataTypeToString(*boundType.childType) + ".");
+        }
+        if (boundType.fixedNumElementsInList == 0) {
+            // Note: the parser already guarantees that the number of elements is a non-negative
+            // number. However, we still need to check whether the number of elements is 0.
+            throw common::BinderException(
+                "The number of elements in a fixed list must be greater than 0. Given: " +
+                std::to_string(boundType.fixedNumElementsInList) + ".");
+        }
+        if (Types::getDataTypeSize(boundType) > common::BufferPoolConstants::DEFAULT_PAGE_SIZE) {
+            throw common::BinderException("The size of fixed list is larger than a "
+                                          "DEFAULT_PAGE_SIZE, which is not supported yet.");
+        }
+    }
+    return boundType;
+}
+
 } // namespace binder
 } // namespace kuzu

src/binder/bind/bind_reading_clause.cpp

@@ -50,7 +50,7 @@ std::unique_ptr<BoundReadingClause> Binder::bindMatchClause(const ReadingClause&
 std::unique_ptr<BoundReadingClause> Binder::bindUnwindClause(const ReadingClause& readingClause) {
     auto& unwindClause = (UnwindClause&)readingClause;
     auto boundExpression = expressionBinder.bindExpression(*unwindClause.getExpression());
-    boundExpression = ExpressionBinder::implicitCastIfNecessary(boundExpression, LIST);
+    boundExpression = ExpressionBinder::implicitCastIfNecessary(boundExpression, VAR_LIST);
     auto aliasExpression =
         createVariable(unwindClause.getAlias(), *boundExpression->dataType.childType);
     return make_unique<BoundUnwindClause>(std::move(boundExpression), std::move(aliasExpression));

src/binder/bind_expression/bind_function_expression.cpp

@@ -180,8 +180,9 @@ std::shared_ptr<Expression> ExpressionBinder::bindNodeLabelFunction(const Expres
     auto nodeTableIDs = catalogContent->getNodeTableIDs();
     expression_vector children;
     children.push_back(node.getInternalIDProperty());
-    auto labelsValue = std::make_unique<Value>(DataType(LIST, std::make_unique<DataType>(STRING)),
-        populateLabelValues(nodeTableIDs, *catalogContent));
+    auto labelsValue =
+        std::make_unique<Value>(DataType(VAR_LIST, std::make_unique<DataType>(STRING)),
+            populateLabelValues(nodeTableIDs, *catalogContent));
     children.push_back(createLiteralExpression(std::move(labelsValue)));
     auto execFunc = function::LabelVectorOperation::execFunction;
     auto uniqueExpressionName = ScalarFunctionExpression::getUniqueName(LABEL_FUNC_NAME, children);
@@ -199,8 +200,9 @@ std::shared_ptr<Expression> ExpressionBinder::bindRelLabelFunction(const Express
     auto relTableIDs = catalogContent->getRelTableIDs();
     expression_vector children;
     children.push_back(rel.getInternalIDProperty());
-    auto labelsValue = std::make_unique<Value>(DataType(LIST, std::make_unique<DataType>(STRING)),
-        populateLabelValues(relTableIDs, *catalogContent));
+    auto labelsValue =
+        std::make_unique<Value>(DataType(VAR_LIST, std::make_unique<DataType>(STRING)),
+            populateLabelValues(relTableIDs, *catalogContent));
     children.push_back(createLiteralExpression(std::move(labelsValue)));
     auto execFunc = function::LabelVectorOperation::execFunction;
     auto uniqueExpressionName = ScalarFunctionExpression::getUniqueName(LABEL_FUNC_NAME, children);

src/binder/expression_binder.cpp

@@ -67,15 +67,15 @@ std::shared_ptr<Expression> ExpressionBinder::implicitCastIfNecessary(
         return expression;
     }
     if (expression->dataType.typeID == ANY) {
-        if (targetTypeID == LIST) {
+        if (targetTypeID == VAR_LIST) {
             // e.g. len($1) we cannot infer the child type for $1.
             throw BinderException("Cannot resolve recursive data type for expression " +
                                   expression->getRawName() + ".");
         }
         resolveAnyDataType(*expression, DataType(targetTypeID));
         return expression;
     }
-    assert(targetTypeID != LIST);
+    assert(targetTypeID != VAR_LIST);
     return implicitCast(expression, DataType(targetTypeID));
 }
 

src/catalog/catalog.cpp

@@ -38,9 +38,9 @@ template<>
 uint64_t SerDeser::serializeValue<DataType>(
     const DataType& value, FileInfo* fileInfo, uint64_t offset) {
     offset = SerDeser::serializeValue<DataTypeID>(value.typeID, fileInfo, offset);
-    if (value.childType) {
-        assert(value.typeID == LIST);
-        return SerDeser::serializeValue<DataType>(*value.childType, fileInfo, offset);
+    offset = SerDeser::serializeValue<uint64_t>(value.fixedNumElementsInList, fileInfo, offset);
+    if (value.typeID == VAR_LIST || value.typeID == FIXED_LIST) {
+        offset = SerDeser::serializeValue<DataType>(*value.childType, fileInfo, offset);
     }
     return offset;
 }
@@ -49,11 +49,11 @@ template<>
 uint64_t SerDeser::deserializeValue<DataType>(
     DataType& value, FileInfo* fileInfo, uint64_t offset) {
     offset = SerDeser::deserializeValue<DataTypeID>(value.typeID, fileInfo, offset);
-    if (value.typeID == LIST) {
+    offset = SerDeser::deserializeValue<uint64_t>(value.fixedNumElementsInList, fileInfo, offset);
+    if (value.typeID == VAR_LIST || value.typeID == FIXED_LIST) {
         auto childDataType = std::make_unique<DataType>();
         offset = SerDeser::deserializeValue<DataType>(*childDataType, fileInfo, offset);
         value.childType = std::move(childDataType);
-        return offset;
     }
     return offset;
 }

src/common/arrow/arrow_converter.cpp

@@ -79,7 +79,7 @@ void ArrowConverter::setArrowFormat(
     case DataTypeID::STRING: {
         child.format = "u";
     } break;
-    case LIST: {
+    case VAR_LIST: {
         child.format = "+l";
         child.n_children = 1;
         rootHolder.nestedChildren.emplace_back();

src/common/arrow/arrow_row_batch.cpp

@@ -1,7 +1,6 @@
 #include "common/arrow/arrow_row_batch.h"
 
 #include "common/types/value.h"
-#include "processor/result/flat_tuple.h"
 
 namespace kuzu {
 namespace common {
@@ -41,7 +40,7 @@ void ArrowRowBatch::templateInitializeVector<STRING>(
 }
 
 template<>
-void ArrowRowBatch::templateInitializeVector<LIST>(
+void ArrowRowBatch::templateInitializeVector<VAR_LIST>(
     ArrowVector* vector, const main::DataTypeInfo& typeInfo, std::int64_t capacity) {
     initializeNullBits(vector->validity, capacity);
     assert(typeInfo.childrenTypesInfo.size() == 1);
@@ -105,8 +104,8 @@ std::unique_ptr<ArrowVector> ArrowRowBatch::createVector(
     case STRING: {
         templateInitializeVector<STRING>(result.get(), typeInfo, capacity);
     } break;
-    case LIST: {
-        templateInitializeVector<LIST>(result.get(), typeInfo, capacity);
+    case VAR_LIST: {
+        templateInitializeVector<VAR_LIST>(result.get(), typeInfo, capacity);
     } break;
     case INTERNAL_ID: {
         templateInitializeVector<INTERNAL_ID>(result.get(), typeInfo, capacity);
@@ -180,7 +179,7 @@ void ArrowRowBatch::templateCopyNonNullValue<STRING>(
 }
 
 template<>
-void ArrowRowBatch::templateCopyNonNullValue<LIST>(
+void ArrowRowBatch::templateCopyNonNullValue<VAR_LIST>(
     ArrowVector* vector, const main::DataTypeInfo& typeInfo, Value* value, std::int64_t pos) {
     vector->data.resize((pos + 2) * sizeof(std::uint32_t));
     auto offsets = (std::uint32_t*)vector->data.data();
@@ -194,7 +193,7 @@ void ArrowRowBatch::templateCopyNonNullValue<LIST>(
     for (auto i = currentNumBytesForChildValidity; i < numBytesForChildValidity; i++) {
         vector->childData[0]->validity.data()[i] = 0xFF; // Init each value to be valid (as 1).
     }
-    if (typeInfo.childrenTypesInfo[0]->typeID != LIST) {
+    if (typeInfo.childrenTypesInfo[0]->typeID != VAR_LIST) {
         vector->childData[0]->data.resize(
             numChildElements * Types::getDataTypeSize(typeInfo.childrenTypesInfo[0]->typeID));
     }
@@ -268,8 +267,8 @@ void ArrowRowBatch::copyNonNullValue(
     case STRING: {
         templateCopyNonNullValue<STRING>(vector, typeInfo, value, pos);
     } break;
-    case LIST: {
-        templateCopyNonNullValue<LIST>(vector, typeInfo, value, pos);
+    case VAR_LIST: {
+        templateCopyNonNullValue<VAR_LIST>(vector, typeInfo, value, pos);
     } break;
     case INTERNAL_ID: {
         templateCopyNonNullValue<INTERNAL_ID>(vector, typeInfo, value, pos);
@@ -303,7 +302,7 @@ void ArrowRowBatch::templateCopyNullValue<STRING>(ArrowVector* vector, std::int6
 }
 
 template<>
-void ArrowRowBatch::templateCopyNullValue<LIST>(ArrowVector* vector, std::int64_t pos) {
+void ArrowRowBatch::templateCopyNullValue<VAR_LIST>(ArrowVector* vector, std::int64_t pos) {
     auto offsets = (std::uint32_t*)vector->data.data();
     offsets[pos + 1] = offsets[pos];
     setBitToZero(vector->validity.data(), pos);
@@ -333,8 +332,8 @@ void ArrowRowBatch::copyNullValue(ArrowVector* vector, Value* value, std::int64_
     case STRING: {
         templateCopyNullValue<STRING>(vector, pos);
     } break;
-    case LIST: {
-        templateCopyNullValue<LIST>(vector, pos);
+    case VAR_LIST: {
+        templateCopyNullValue<VAR_LIST>(vector, pos);
     } break;
     case INTERNAL_ID: {
         templateCopyNullValue<INTERNAL_ID>(vector, pos);
@@ -396,7 +395,7 @@ ArrowArray* ArrowRowBatch::templateCreateArray<STRING>(
 }
 
 template<>
-ArrowArray* ArrowRowBatch::templateCreateArray<LIST>(
+ArrowArray* ArrowRowBatch::templateCreateArray<VAR_LIST>(
     ArrowVector& vector, const main::DataTypeInfo& typeInfo) {
     auto result = createArrayFromVector(vector);
     vector.childPointers.resize(1);
@@ -465,8 +464,8 @@ ArrowArray* ArrowRowBatch::convertVectorToArray(
     case STRING: {
         return templateCreateArray<STRING>(vector, typeInfo);
     }
-    case LIST: {
-        return templateCreateArray<LIST>(vector, typeInfo);
+    case VAR_LIST: {
+        return templateCreateArray<VAR_LIST>(vector, typeInfo);
     }
     case INTERNAL_ID: {
         return templateCreateArray<INTERNAL_ID>(vector, typeInfo);

src/common/csv_reader/csv_reader.cpp

@@ -297,7 +297,7 @@ std::unique_ptr<Value> CSVReader::getList(const DataType& dataType) {
             case INTERVAL: {
                 val = std::make_unique<Value>(listCSVReader.getInterval());
             } break;
-            case LIST: {
+            case VAR_LIST: {
                 val = listCSVReader.getList(*dataType.childType);
             } break;
             default:
@@ -315,7 +315,7 @@ std::unique_ptr<Value> CSVReader::getList(const DataType& dataType) {
             BufferPoolConstants::DEFAULT_PAGE_SIZE, numBytesOfOverflow));
     }
     return std::make_unique<Value>(
-        DataType(LIST, std::make_unique<DataType>(dataType)), std::move(listVal));
+        DataType(VAR_LIST, std::make_unique<DataType>(dataType)), std::move(listVal));
 }
 
 void CSVReader::setNextTokenIsProcessed() {

src/common/in_mem_overflow_buffer_utils.cpp

@@ -42,7 +42,7 @@ void InMemOverflowBufferUtils::copyListRecursiveIfNested(const ku_list_t& src, k
                 ((ku_string_t*)dest.overflowPtr)[i], inMemOverflowBuffer);
         }
     }
-    if (dataType.childType->typeID == LIST) {
+    if (dataType.childType->typeID == VAR_LIST) {
         for (auto i = 0u; i < dest.size; i++) {
             InMemOverflowBufferUtils::copyListRecursiveIfNested(
                 ((ku_list_t*)src.overflowPtr)[i + srcStartIdx], ((ku_list_t*)dest.overflowPtr)[i],

src/common/type_utils.cpp

@@ -73,7 +73,7 @@ std::string TypeUtils::elementToString(
         return TypeUtils::toString(((interval_t*)overflowPtr)[pos]);
     case STRING:
         return TypeUtils::toString(((ku_string_t*)overflowPtr)[pos]);
-    case LIST:
+    case VAR_LIST:
         return TypeUtils::toString(((ku_list_t*)overflowPtr)[pos], dataType);
     default:
         throw RuntimeException("Invalid data type " + Types::dataTypeToString(dataType) +