com.alibaba.innodb.java.reader.service.impl

类 IndexServiceImpl

java.lang.Object

com.alibaba.innodb.java.reader.service.impl.IndexServiceImpl

所有已实现的接口:

IndexService

public class IndexServiceImpl
extends Object
implements IndexService

Innodb index page service, providing read-only query operations.

作者:

xu.zx

字段概要

从接口继承的字段 com.alibaba.innodb.java.reader.service.IndexService

NOP_PROJECTION

构造器概要

构造器

构造器和说明
IndexServiceImpl(StorageService storageService, TableDef tableDef, KeyComparator keyComparator)

方法概要

限定符和类型	方法和说明
Iterator<GenericRecord>	getQueryAllIterator(Optional<List<String>> recordProjection, boolean ascOrder) Return an iterator to query all records of a tablespace.
Iterator<GenericRecord>	getQueryIteratorBySk(String skName, List<Object> lower, ComparisonOperator lowerOperator, List<Object> upper, ComparisonOperator upperOperator, Optional<List<String>> recordProjection, boolean ascOrder) Return record iterator by secondary key (SK) in a tablespace.
Iterator<GenericRecord>	getRangeQueryIterator(List<Object> lower, ComparisonOperator lowerOperator, List<Object> upper, ComparisonOperator upperOperator, Optional<List<String>> recordProjection, boolean ascOrder) The implementation is different from the way queryAll(Optional, Optional) works.
Iterator<GenericRecord>	getSkRangeQueryIterator(TableDef skTableDef, long skRootPageNumber, List<Object> lower, ComparisonOperator lowerOperator, List<Object> upper, ComparisonOperator upperOperator, boolean ascOrder) Return secondary key (SK) record iterator in a tablespace.
Index	loadIndexPage(long pageNumber) Load index page by page number.
List<GenericRecord>	queryAll(Optional<java.util.function.Predicate<GenericRecord>> recordPredicate, Optional<List<String>> recordProjection) Query all records in a tablespace.
List<GenericRecord>	queryByPageNumber(int pageNumber) Query all records by single index page.
List<GenericRecord>	queryByPageNumber(long pageNumber) Query all records by single index page.
GenericRecord	queryByPrimaryKey(List<Object> key, Optional<List<String>> recordProjection) Query record by primary key in a tablespace with projection list.
List<GenericRecord>	rangeQueryByPrimaryKey(List<Object> lower, ComparisonOperator lowerOperator, List<Object> upper, ComparisonOperator upperOperator, Optional<java.util.function.Predicate<GenericRecord>> recordPredicate, Optional<List<String>> recordProjection) Range query records by primary key in a tablespace.

从类继承的方法 java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

构造器详细资料

IndexServiceImpl

public IndexServiceImpl(StorageService storageService,
                        TableDef tableDef,
                        KeyComparator keyComparator)

方法详细资料

queryByPageNumber

public List<GenericRecord> queryByPageNumber(int pageNumber)

Query all records by single index page.

指定者:

queryByPageNumber 在接口中 IndexService

参数:

pageNumber - page number (int type), can be leaf or non-leaf page

返回:

list of records

queryByPageNumber

public List<GenericRecord> queryByPageNumber(long pageNumber)

Query all records by single index page.

指定者:

queryByPageNumber 在接口中 IndexService

参数:

pageNumber - page number (long type), can be leaf or non-leaf page

返回:

list of records

queryAll

public List<GenericRecord> queryAll(Optional<java.util.function.Predicate<GenericRecord>> recordPredicate,
                                    Optional<List<String>> recordProjection)

Query all records in a tablespace.

Note this will cause out-of-memory if the table size is too big. Make sure fields are included in projection for predicate to use.

指定者:

queryAll 在接口中 IndexService

参数:

recordPredicate - optional filtering, if predicate returns true upon record, then it will be added to result set

recordProjection - optional projection of selected column names, if no present, all fields will be included

返回:

all records

queryByPrimaryKey

public GenericRecord queryByPrimaryKey(List<Object> key,
                                       Optional<List<String>> recordProjection)

Query record by primary key in a tablespace with projection list.

For single key, the the list size should be one, for composite key the size will be more than one.

指定者:

queryByPrimaryKey 在接口中 IndexService

参数:

key - primary key, single key or a composite key

recordProjection - optional projection of selected column names, if no present, all fields will be included

返回:

record

getQueryAllIterator

public Iterator<GenericRecord> getQueryAllIterator(Optional<List<String>> recordProjection,
                                                   boolean ascOrder)

Return an iterator to query all records of a tablespace.

Leverage getRangeQueryIterator(List, ComparisonOperator, List, ComparisonOperator, Optional, boolean).

This is friendly to memory since only one page is loaded per batch.

指定者:

getQueryAllIterator 在接口中 IndexService

参数:

recordProjection - optional projection of selected column names, if no present, all fields will be included

ascOrder - if set result records in ascending order

返回:

all records

getRangeQueryIterator

public Iterator<GenericRecord> getRangeQueryIterator(List<Object> lower,
                                                     ComparisonOperator lowerOperator,
                                                     List<Object> upper,
                                                     ComparisonOperator upperOperator,
                                                     Optional<List<String>> recordProjection,
                                                     boolean ascOrder)

The implementation is different from the way queryAll(Optional, Optional) works.

This method will do point query to search the nearest lower and upper bound record, then visit the leaf page, go through all the level 0 pages by the double-linked pages. While queryAll(Optional, Optional) traverses b+ tree in a depth-first way.

指定者:

getRangeQueryIterator 在接口中 IndexService

参数:

lower - lower bound

lowerOperator - comparison operator for lower

upper - upper bound

upperOperator - comparison operator for upper

recordProjection - optional projection of selected column names, if no present, all fields will be included

ascOrder - if set result records in ascending order

返回:

record iterator

getQueryIteratorBySk

public Iterator<GenericRecord> getQueryIteratorBySk(String skName,
                                                    List<Object> lower,
                                                    ComparisonOperator lowerOperator,
                                                    List<Object> upper,
                                                    ComparisonOperator upperOperator,
                                                    Optional<List<String>> recordProjection,
                                                    boolean ascOrder)

Return record iterator by secondary key (SK) in a tablespace. This is first go through all secondary keys and look up record back to clustered index.

This is friendly to memory since only one page is loaded per batch.

For single key the lower or upper list size should be one, for composite key the size will be more than one.

指定者:

getQueryIteratorBySk 在接口中 IndexService

参数:

skName - secondary key name in SHOW CREATE TABLE command

lower - lower bound

lowerOperator - comparison operator for lower

upper - upper bound

upperOperator - comparison operator for upper

ascOrder - if set result records in ascending order

recordProjection - optional projection of selected column names, if no present, all fields will be included

返回:

record iterator, record is composed by secondary key columns and primary key columns

getSkRangeQueryIterator

public Iterator<GenericRecord> getSkRangeQueryIterator(TableDef skTableDef,
                                                       long skRootPageNumber,
                                                       List<Object> lower,
                                                       ComparisonOperator lowerOperator,
                                                       List<Object> upper,
                                                       ComparisonOperator upperOperator,
                                                       boolean ascOrder)

Return secondary key (SK) record iterator in a tablespace.

参数:

skTableDef - secondary key virtual table definition

skRootPageNumber - secondary key root page number

lower - lower bound

lowerOperator - comparison operator for lower

upper - upper bound

upperOperator - comparison operator for upper

ascOrder - if set result records in ascending order

返回:

record iterator, record is composed by secondary key columns and primary key columns

rangeQueryByPrimaryKey

public List<GenericRecord> rangeQueryByPrimaryKey(List<Object> lower,
                                                  ComparisonOperator lowerOperator,
                                                  List<Object> upper,
                                                  ComparisonOperator upperOperator,
                                                  Optional<java.util.function.Predicate<GenericRecord>> recordPredicate,
                                                  Optional<List<String>> recordProjection)

Range query records by primary key in a tablespace.

For single key the lower or upper list size should be one, for composite key the size will be more than one.

Leverage getRangeQueryIterator(List, ComparisonOperator, List, ComparisonOperator, Optional, boolean) if range is specified, or else fallback to queryAll(Optional, Optional).

指定者:

rangeQueryByPrimaryKey 在接口中 IndexService

参数:

lower - if rangeQuery is true, then this is the lower bound

lowerOperator - if rangeQuery is true, then this is the comparison operator for lower

upper - if rangeQuery is true, then this is the upper bound

upperOperator - if rangeQuery is true, then this is the comparison operator for upper

recordPredicate - optional. evaluating record, if true then it will be added to result set, else skip it

recordProjection - optional projection of selected column names, if no present, all fields will be included

返回:

list of records

loadIndexPage

public Index loadIndexPage(long pageNumber)

从接口复制的说明: IndexService

Load index page by page number.

指定者:

loadIndexPage 在接口中 IndexService

参数:

pageNumber - page number

返回:

index page