WO2019015364A1

WO2019015364A1 - Method and device for executing structured query language (sql) instruction

Info

Publication number: WO2019015364A1
Application number: PCT/CN2018/083411
Authority: WO
Inventors: 陈志标; 陆军
Original assignee: 华为技术有限公司
Priority date: 2017-07-20
Filing date: 2018-04-17
Publication date: 2019-01-24
Also published as: CN110019287A; CN110019287B

Abstract

The present application discloses a method and a device for executing a structured query language (SQL) instruction, relating to the field of communications, used for supporting a standard SQL access object. The method comprises: receiving an SQL instruction, the SQL instruction being used for querying attribute data of an object from data of the object; parsing the SQL instruction to obtain the identifier of the object, the key value attribute of the object and a query condition; according to the identifier of the object and the mapping relationship between the identifier of the object and an object mode, obtaining the object mode and a relational data table corresponding to the object; according to the key value attribute of the object and the attribute for generating an index value defined in the object mode, determining the index value corresponding to the data of the object; and obtaining the data of the object according to the index value corresponding to the data of the object, and according to the query condition and the relational data table of the object, querying the data of the object to obtain attribute data. The embodiments of the present application can be applicable to object data access.

Description

Method and apparatus for executing structured query language SQL instructions

This application claims the priority of the Chinese Patent Application filed on Jan. 20, 2017, filed on Jan. 20, 2017, filed Jan. In this application.

Technical field

The present application relates to the field of communications, and in particular, to a method and apparatus for executing a Structured Query Language (SQL) instruction.

Background technique

Network Function Virtualization (NFV) architecture realizes the separation of business data and business processing. The business data is distributed storage, and object storage is the mainstream storage method in distributed storage.

Commonly used object access methods include SQL-like language operations and Object Relation Mapping (ORM). Among them, the SQL-like language is also called Object Query Language (OQL), its statement adopts the method of object expression, and its function is completely based on object-oriented entity query. Because OQL is a non-standard SQL interface, applications developed based on standard SQL interfaces cannot be used, and system openness is insufficient. In addition, different OQL languages need to be developed for different object differences, and it is difficult to implement function migration. And in terms of grammar, it can't be more succinct than SQL, and can't implement complex queries.

Summary of the invention

Embodiments of the present application provide a method and apparatus for executing SQL instructions for supporting standard SQL access objects.

To achieve the above objective, the embodiment of the present application adopts the following technical solutions:

In a first aspect, a method for executing a structured query language SQL instruction is provided, the method comprising: receiving a SQL instruction, the SQL instruction is used to query an attribute data of an object from an object data; and parsing the SQL instruction to obtain an object The identifier, the key value attribute of the object, and the query condition. The identifier of the object is used to indicate the type of the object, the key value attribute of the object is used to generate an index value, the index value corresponds to the data of the object, and the query condition is used in the data of the object. Query attribute data; according to the identifier of the object, and the mapping relationship between the object identifier and the object pattern, obtain the object pattern and the relational data table corresponding to the object, and the object pattern is used to define a tree data structure of a type of object, a tree shape The data structure includes at least one first node, the first node is used to define attributes of the attributes of such objects that can be used to generate index values, the relational data table is a two-dimensional table, and the two-dimensional table is in accordance with the node relationship indicated in the object pattern. Attribute data of the storage object; according to the key value attribute of the object and the attribute defined in the object pattern for generating the index value Determining an index value corresponding to the data object; the data object corresponding to the index value obtained data object, and query condition according to the object and relational data tables, a query attribute data obtained from the data object. In the embodiment of the present application, the method for executing the SQL instruction is used to parse the identifier of the object in the SQL instruction to obtain the corresponding object schema and the relational data table, and the key value attribute of the object in the SQL instruction is used to generate the data corresponding to the object. The index value, and the object data is obtained according to the index value, and the query condition in the parsing SQL instruction is used to query the attribute data of the object from the data of the object. The identifier of the object in the above SQL instruction, the key value attribute of the object, and the query condition are not the object class data, but the valid data such as variables and identifiers that can be passed by the standard SQL instruction, and the above valid data obtained by parsing the SQL instruction. The data of the mapping object and the attribute data of the object are implemented to support standard SQL access objects.

In a possible design, before receiving an SQL instruction, the method further includes: defining an object mode of the object and a relational data table; receiving a write request for the object, wherein the write request includes an identifier and an attribute of the object; The identifier of the object in the request, and the mapping relationship between the identifier of the object and the object pattern, obtaining the object schema and the relational data table of the object; converting the attribute in the write request into the object having the complete storage structure according to the object pattern of the object, And constructing an attribute in the write request as an index value according to an attribute for generating an index value defined in the object schema, wherein the object having the complete storage structure refers to an object stored in a storage space allocated according to all attributes of the object pattern, and writing The attributes in the request are stored in the storage space; the object with the complete storage structure is converted into the data of the object according to the relational data table of the object, and the data of the object is stored according to the index value. This design provides a way to store data for an object through an object interface.

In a possible design, the method further comprises: receiving a read request for the object, wherein the read request includes an identifier of the object and a key value attribute; according to the identifier of the object in the read request, and the identifier of the object and the object mode The mapping relationship is obtained, and the corresponding object pattern is obtained; according to the key value attribute of the object and the attribute that can be used to generate the index value defined in the object pattern, the index value corresponding to the data of the object is determined; and the data of the object is obtained according to the index value. This design provides a way to read the data of an object through the object interface.

In a possible design, the tree data structure of the object includes at least one parent node, and the parent node includes an identifier and a domain, wherein the identifier of the parent node is used to indicate the category of the object corresponding to the parent node, and the domain of the parent node is used for Contains child nodes. This design provides a way to structure a tree's data structure.

In one possible design, each node of the relational data table includes a parent table identifier, a root table identifier, and a foreign key, the parent table identifier is a two-dimensional table for the parent node of the child node, and the root table identifier is used to point A two-dimensional table of the root node of a tree-shaped data structure with a foreign key pointing to the primary key of the object schema. This design provides a way to structure a relational data table.

A second aspect provides an apparatus for executing a structured query language SQL instruction, the apparatus comprising: a SQL interface unit, a relational SQL access processing unit, an object relational conversion unit, a metadata management unit, and an object storage unit, wherein The SQL interface unit is configured to receive an SQL instruction, and the SQL instruction is used to query the attribute data of the object from the data of the object; the relational SQL access processing unit is configured to parse the SQL instruction received by the SQL interface unit to obtain the identifier of the object, The key value attribute of the object and the query condition. The identifier of the object is used to indicate the type of the object. The key value attribute of the object is used to generate an index value. The index value corresponds to the data of the object, and the query condition is used to query the attribute from the data of the object. Data; an object-relational conversion unit, configured to obtain an object identifier corresponding to an object according to a relationship between an identifier of an object obtained by the relational SQL access processing unit and an identifier of the object stored by the metadata management unit, and an object mode corresponding to the object Relational data table, object schema is used to define the tree data structure of a class of objects, the number of trees The structure includes at least one first node, the first node is used to define attributes of the attributes of such objects that can be used to generate index values, the relational data table is a two-dimensional table, and the two-dimensional table is stored according to the node relationship indicated in the object pattern. The attribute data of the object; the object relational conversion unit is further configured to determine an index value corresponding to the data of the object according to the key value attribute of the object and the attribute for generating the index value defined in the object pattern; the object relation type conversion unit further The data of the object is obtained according to the index value corresponding to the data of the object, and the attribute data is obtained from the data of the object stored in the object storage unit according to the query condition and the relational data table of the object. Based on the same inventive concept, the principle and the beneficial effects of the device can be referred to the first aspect and the possible method embodiments of the first aspect and the beneficial effects. Therefore, the implementation of the device can be referred to the first The aspects and implementations of the various possible methods of the first aspect are not repeated here.

A third aspect provides an apparatus for executing a structured query language SQL instruction, including: a processor, a memory, a bus, and a communication interface; the memory is configured to store a computer execution instruction, and the processor and the memory are connected through a bus when the system is running The processor executes the memory stored computer to execute the instructions to perform the method as in the first aspect and any of its possible designs.

In a fourth aspect, a computer storage medium is provided, comprising instructions that, when run on a computer, cause the computer to perform a method as in the first aspect and any of its possible designs.

In a fifth aspect, there is provided a computer program product comprising instructions which, when run on a computer, cause the computer to perform a method as in the first aspect and any of its possible designs.

It should be noted that the technical effects brought by the third aspect to the fifth aspect can be referred to the technical effects brought by the corresponding method, and details are not described herein again.

DRAWINGS

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings to be used in the embodiments or the prior art description will be briefly described below.

FIG. 1 is a schematic structural diagram of an apparatus for executing an SQL instruction according to an embodiment of the present application;

2 is a schematic flowchart of a method for writing an object by using an object interface unit according to an embodiment of the present application;

FIG. 3 is a schematic diagram of an object mode provided by an embodiment of the present application; FIG.

4 is a schematic diagram of an object storage structure provided by an embodiment of the present application;

FIG. 5 is a schematic flowchart of a method for executing an SQL instruction by using an SQL interface unit according to an embodiment of the present application;

FIG. 6 is a schematic diagram of a common SQL query operation required to perform a combined operation according to an embodiment of the present application;

FIG. 7 is a schematic diagram of a method for executing an SQL instruction according to an embodiment of the present application to reduce a combined operation;

FIG. 8 is a schematic flowchart of a method for reading an object by using an object interface unit according to an embodiment of the present application;

FIG. 9 is a schematic structural diagram of another apparatus for executing an SQL instruction according to an embodiment of the present application;

FIG. 10 is a schematic structural diagram of another apparatus for executing an SQL instruction according to an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be clearly and completely described in the following with reference to the accompanying drawings in the embodiments.

A class of objects described in the embodiments of the present application refers to an abstract description having a common data structure and attribute data. The data structure or attribute data of different types of objects are different. For example, the attribute data describing the scientist object and the customer object are different. For the scientist object, the age, the working age, the affiliation, and the like can be described; and for the customer object, The age, contact information, address, purchase history, etc. can be described, so the scientist object and the customer object belong to different classes of objects. Attribute data is data that defines its attributes for the convenience of distinguishing between different objects. It can be described from the perspective of a class of objects or for the data of a specific object. For example, for an object such as a scientist, age, length of service, affiliated institution, etc. is its attribute data, and for a specific scientist object, age=30, length of service=5, affiliated institution=“Chinese Academy of Sciences” Attribute data. The embodiment of the present application is described by the customer, and it can be understood that it is not intended to limit the specific object type.

The embodiment of the present application provides an apparatus for executing an SQL instruction, so that data is stored in units of objects, supports an efficient object access interface, and object-based distributed storage, and the device also supports standard SQL interface access. Referring to FIG. 1, the apparatus 11 includes: an object interface unit 1101, an object access processing unit 1102, a metadata management unit 1103, an object storage unit 1104, an object relation type conversion unit 1105, a relational SQL access processing unit 1106, and SQL. Interface unit 1107. The above-mentioned units may use the same or different processes or threads to perform the corresponding functions, and the embodiments of the present application are not limited herein.

The methods provided by the embodiments of the present application may be implemented by program code in a memory, or by a chip on a router or a server, or by a simulator. It can also be implemented in a chip, such as Field-Programmable Gate Array (FPGA), Digital Signal Processing (DSP), or by software in heterogeneous acceleration hardware such as tensor processing. Implemented on a Tensor Processing Unit (TPU), a Graphics Processing Unit (GPU), or implemented by software on a Central Processing Unit (CPU), or by a software/hardware simulator.

The object interface unit 1101 provides a transceiving function of different forms of objects in an application scenario. Different forms of objects include, but are not limited to, objects represented by Java Object Notation (JSON), objects represented by Extensible Mark-up Language (XML), which are related to the actual needs of the application scenario. Is a non-standard interface. The object interface unit 1101 has a function of predefining an object schema (Schema) and a relational data table (object relational schema). The object schema described in this paper is an abstract description of the common data structure and attribute data that a class of objects has, including the Field node and the Keys node. The Field node describes the attribute data of the object, and the key value. The (Keys) node describes the index of the lookup object. The attributes in the index key in the Keys node must come from the domain node. The relational data table describes the associations and constraints between different nodes in the class. These two modes can be collectively referred to as object patterns in this paper. Object schemas and relational data tables can also be defined in a variety of ways, such as JSON, XML, and so on. Only the data of the object defined in the object interface unit 1101 can be parsed. The object with complete structure in the device includes the data of a specific object and the mode of the object (object mode and relational data table), the data of the object is unique, and the mode of the object can be shared for a class of objects.

The object access processing unit 1102 provides a standardized conversion function of the non-standard object representation, making it an object that can be processed by the system, and processing the object information by the object mode and the relational data table provided by the metadata management unit 1103, so that the object The data is stored in the object storage unit 1104 in an agreed manner. The object mode is used to match the attribute data of the object, and is used to obtain a corresponding index key according to the definition of the key value Keys in the object mode and the attribute data of the object. The relational data table processes the association relationship between the parent and child nodes, and can not only guide the data of the object to be saved in the object storage unit 1104, but also provide support and optimization for the relational SQL operation.

The metadata management unit 1103 manages the object schema and the relational data table. The object access processing unit 1102 decomposes the transformation of the object, relying on the object schema and the relational data table in the metadata management unit 1103.

The object storage unit 1104 provides data of the object converted according to the mode, and is a result of serialization of the data of the object.

The object relation type conversion unit 1105 is responsible for the conversion between the relational data and the object type data, and depends on the object mode and the relational data table in the metadata management unit 1103. The object-relational conversion unit 1105 is a hub for the device to provide standard SQL access.

The relational SQL access processing unit 1106 is responsible for parsing the optimized SQL query language.

The SQL interface unit 1107 provides acceptance of the SQL query language and return of the result data.

Referring to FIG. 2, the method of writing an object by the object interface unit 1101 includes:

S101. Define an object mode of the object and a relational data table.

The object schema of the object and the relational data table can be defined by calling the corresponding function of the object interface unit 1101. The result is stored in the metadata management unit 1103.

The object schema is used to define a tree data structure of a class of objects. The tree data structure includes at least one node, which is used to define attributes of such objects that can be used to generate index values, and the relational data table is two-dimensional. The table, the two-dimensional table stores the attribute data of the object according to the node relationship indicated in the object schema.

Referring to FIG. 3, the left image is an example of a consumer object, and the right image is an example of an object mode in the form of a JSON whose mode name is "consumer".

A consumer may have no order or multiple orders, and each order includes at least one item, that is, the consumer object contains the order object, and the order object contains the item object. Therefore, in the tree-shaped object pattern formed, the consumer object is the root node, the order object is the leaf node of the consumer object, and the commodity object is the leaf node of the order object. In the embodiment of the present application, the consumer, the order, and the commodity are no longer separate objects, and they form a tree-shaped object pattern having a root node according to the nested relationship as a parent node.

In a relational database, an index is a storage structure that sorts the values of one or more columns in a database table by a unique value, which can be a collection of values in one or more columns of the table and a page pointing to the values in the table. A list of logical pointers to the table. The role of the index is equivalent to the directory of the book, you can quickly find the content you need based on the page number in the directory. In the traditional database of key-value pairs, the key and the value are in one-to-one correspondence, which limits the diversification of the data of the object to a certain extent. Therefore, we provide a variety of keys for the object through the object mode. Enhance data operability and provide support for query optimization.

The object mode is a tree data structure composed of nodes nested with each other, and the associated sub-objects correspond to respective nodes of the tree data structure, the tree data structure includes at least one parent node, and the object mode is described in the root node. Representation (that is, the schema name). Each node may include three attributes: a name, a type, and a field. The name and type are required attributes, and the identifier is used to indicate that the corresponding object belongs to the node. Kind, such as consumer, merchandise order number, etc.; type is used to indicate the data type of the node, such as integer, record, and the like. Fields appear only in the parent node of a one-to-many parent child node, similar to the form of an array. The fields of the parent node are used to contain child nodes (that is, the child nodes are nested in the parent node's domain (fields). )in).

The root node of the tree data structure also includes keys, and multiple index keys can be defined in the key value, but there is only one primary key, which is used to uniquely identify the object, each Index keys can consist of single or multiple domains in the same node. The primary key points to the storage location of the object, and the index key of the non-primary key points to the primary key.

The key value example in Figure 3 only includes one primary key. It can also create other index keys according to the actual needs of the business. Its function is similar to the index in the relational database. A reasonable number of index keys are beneficial for query optimization.

When the object mode of the object is defined, the metadata management unit 1103 automatically generates a relational data table based on the object mode and related information input by the user. The relational data table corresponds to the object pattern, and the relational data table can be automatically derived from the object pattern. The relational data table essentially decomposes each object node in the object schema into a two-dimensional table, including the parent table identifier, the root table identifier, and the foreign key in the two-dimensional table corresponding to the child node, and the parent table identifier is used to point A two-dimensional table of the child nodes of the child nodes, the root table identifies a two-dimensional table that points to the root node of the tree data structure, and the foreign key points to the primary key of the object schema. Relationship maintenance of nodes within an object is achieved by inheriting the identity fields of the parent and root tables and creating dependencies (such as foreign keys). Relational data representations such as the following:

Each node customer, orders, and items are broken down into a two-dimensional table. According to the definitions in the keys, the ID (customers.id) in the customer is the primary key of the root table and its own primary key. Orders are child nodes of the customer, inherit the root table (root_field integer inherit) primary key customers.id, inherit the parent table (parent_field integer inherit) primary key customers.id, foreign key dependencies (FOREIGN KEY) customers.id . The items are the children of the orders, inherit the root table's primary key customers.id, inherit the parent table's primary key oders.id, and the foreign keys depend on the customers.id. The other content is the attribute of each node.

It can be seen that the relational data table causes each node in the object to generate a table, and the relationship between the nodes is embodied by the child nodes (tables) inheriting the identification fields of the parent node (table) and the root node (table), and creating a dependency relationship. The foreign key guarantees strong dependency; the object mode key defines the attribute of the index key, and supports one object with multiple index keys. In addition, additional information can be added to the relational data table to optimize query access and improve query performance.

After the definition is completed, the object schema and the relational data table of all objects can be stored in the metadata management unit 1103.

S102. Receive a write request for an object, where the write request includes an identifier and an attribute of the object.

Specifically, it is received by the object interface unit 1101. Since the data representation in the write request may be JSON format or XML format, etc., the object interface unit 1101 parses the identifier and key value attribute of the object obj into a unified form that the system can process, and sends it to the object access processing unit. 1102 processing.

The identity of the object is used to indicate the type of the object. Exemplarily, referring to FIG. 4, "customer" in (1) is the identifier of the object, and the other content in (1) is the key value attribute of the object.

S103. Obtain an object mode and a relational data table of the object according to the identifier of the object in the write request, and the mapping relationship between the identifier of the object and the object mode.

Exemplarily, the object access processing unit 1102 can query the metadata management unit 1103 to obtain the name (name) shown in FIG. 2 as "consumer" according to the mode name "customer" shown in FIG. The object mode and the corresponding relational data table.

S104. Convert an attribute in the write request to an object having a complete storage structure according to an object mode of the object, and construct an attribute in the write request as an index value according to an attribute defined in the object pattern for generating an index value, where An object of a complete storage structure refers to an object stored in a storage space allocated according to all attributes of the object mode, and attributes in the write request are stored in the storage space.

Specifically, the object access processing unit 1102 encapsulates the attribute data into an object of the complete storage structure of the object pattern according to the object mode of the object, and refers to the size of the opened storage space as the size of the entire object mode, and the write request for the received object. The included attribute data is filled in the storage space according to the object mode, and the attribute data not included in the write request of the received object is null (NULL). Exemplarily, referring to FIG. 4, if the attribute data of the object received in (1) does not include birthday, the birthday is null in the object of the complete structure in (2) (NULL ).

Constructing an index value according to the attribute data and the definition of the key value in the object schema means that one of the key values shown in FIG. 2 is a consumer ID, and therefore, the consumer in the attribute data of the object received in FIG. ID (Customer.id) 2 calculates a hash value as an index value (first key value). When the query is performed in this way, the object can be queried according to the first key value.

It should be noted that, although only one key value is listed in the figure, the embodiment of the present application does not limit the number of key values and the index relationship, for example, the first key value is a primary key, and the second key value is an indirect key. The second key value can index the first key value and so on.

S105. Convert an object having a complete storage structure into data of the object according to the relational data table of the object, and store the data of the object according to the index value.

Specifically, the object access processing unit 1102 stores the above data to the object storage unit 1104.

Exemplarily, referring to FIG. 4, (3) is the data of the object obtained by splitting the object having the complete storage structure in (2) according to the relational data table, from which it can be seen that each node Separate storage, parent-child relationship between nodes is maintained by a pair of bidirectional pointers Ptr.

S106. The feedback write status identifier.

Specifically, the object storage unit 1104 feeds back the write status identifier to the object interface unit 1101 through the object access processing unit 1102, and the identifier is used to indicate whether the write is successful.

Referring to FIG. 5, the method of executing the SQL instruction by the SQL interface unit 1107 includes:

S201. Receive an SQL instruction, where the SQL instruction is used to query attribute data of the object from data of an object.

Specifically, this step can receive standard SQL operations for the object through the SQL interface unit 1107. For example, the SQL operation can be an SQL query operation for querying attribute data in a corresponding object. For example, the SQL query statement of the object querying items.product_id=1 can be written as "select*from customers where items.product_id=1;", which obtains all the attributes of the root node that satisfy the query condition from the result set of the SQL. Or can be written as "select*from customers.items where product_id=1;", which gets all the attributes of the root node that match the query condition from the SQL result set.

S202. Parse the SQL instruction to obtain an identifier of the object to be queried, a key value attribute of the object, and a query condition. The identifier of the object is used to indicate the type of the object, and the key value attribute of the object is used to generate an index value, the index value and the storage. The data of the object corresponds to the query condition for querying the corresponding attribute data from the data of the object.

The object's identity is used to query the object's two-dimensional table information and can be mapped to a unique object pattern. The key-value attribute refers to an attribute that can be used to generate the index value according to the attribute in which the object is defined in the mapped object pattern, and the key value (index value) is obtained by hash calculation. For example, suppose an object has ten attributes named F0~F9. In the object mode of the object, key indexes (Keys) define two indexes, Key1 and Key2, where Key1 is the primary key, Key2 is the secondary index, and Key2 points. The value of Key1. Key1 is obtained by hashing the values of the objects F0 and F1 of the object, and Key2 is obtained by hashing the values of the attributes F2 and F3 of the object. Then F0, F1, F2, and F3 are the key value attributes of the object, and other attributes F4 to F9 are non-key value attributes. When reading an object, you can construct Key1 by passing the values of F0 and F1 according to the definition of the key values in the object mode, or you can construct Key2 by passing in the values of F2 and F3, thus obtaining object storage. s position. The query condition includes other attribute conditions than the key value attribute.

Assume that the user passes the SQL query statement of "select*from customers.items where product_id=1 and price>30;", and the SQL interface unit 1107 receives the SQL query operation and passes it to the relational SQL query access processing unit 1106. The relational SQL access processing unit 1106 analyzes the SQL operation to identify the relationship operators select, from, where, and identifies valid data other than the relationship operator (node information customers.items, key attribute product_id=1, and query condition price) >30), it is sent to the object-relational conversion unit 1105.

S203. Obtain an object mode and a relational data table corresponding to the object according to the identifier of the object and the mapping relationship between the identifier of the object and the object mode.

Specifically, the object-relational conversion unit 1105 maps to the mode name customers according to the identifier of the object, and then searches for the corresponding object schema and the relational data table from the metadata management unit 1103.

S204. Determine an index value corresponding to the data of the object according to the key value attribute of the object and the attribute used to generate the index value defined in the object mode.

Illustratively, the object-relational conversion unit 1105 looks up the corresponding node in the object schema by customers.items in the valid data, and uses product_id as the key-value attribute in the node.

For example, the key value can be defined by some sort of hashing of the product_id to get the index value.

S205. Obtain data of the object according to an index value corresponding to the data of the object, and query the attribute data from the data of the object according to the query condition and the relational data table of the object.

The object-relational conversion unit 1105 simultaneously generates an execution plan according to the operation of the SQL. If the attribute product_id included in the key-value attribute is a key-value attribute, and the index value can be constructed, the object-relational conversion unit 1105 can directly pass the index value. The corresponding object is indexed from the object storage unit 1104.

It should be noted that if the product_id is a non-key value attribute, the object relational type conversion unit 1105 traverses all the object queries to obtain a corresponding object.

When the object is accessed through SQL, the data of the object is parsed and optimized according to the relational data table. When the relational operator needs to access the object, the object-relational conversion unit 1105 converts the data of the optimized object into a relationship according to the relational operator. The type data is returned to the location where the request is initiated by the relational SQL access processing unit 1106 and the SQL interface unit 1107, and the relationship operator can be processed without modification. For example, when the SQL operation is a SQL query operation, to query the customer, order id, and order time for which a certain product_id is purchased, the traditional SQL query operation is:

Select customers.name, orders.id, orders.order_date from customers, orders, items where customers.id=orders.customer_id and orders.order_id=items.order_id and items.product_id in(x,y,z);where,x , y, z refers to the specific product_id value.

This operation performs an Object-Oriented Language (OML) operation within the object as shown in FIG. 6. It can be seen that, according to the query conditions, the data items of items.product belonging to x, y, z are queried in the items table by Filter items_filter where items.product_id in(x, y, z); In the orders table, the data set of orders.id=items.order_id is queried by Filter order_filter where orders_id=items.order_id; all customers are queried in the consumer (custormers) table by Filter customer_filter where customer.id=orders.customer_id. The data set of id=orders.customer_id. Then the three sets of data sets are joined in the order of the SQL query, and finally the qualified query result set is returned to the relational operator.

Referring to FIG. 7, the SQL access object method of the present application is used to query the items.product belonging to x, y, z through the Filter items_filter where items.product_id in(x, y, z) only during the initial OML operation. data set. Then you can rely on the parent key of the parent node inherited in the relational data table to query the parent node information. According to this method, the entire object information can be queried, which greatly reduces the number of join operations.

S206. Return the SQL query result in the form of a result set conforming to the SQL standard.

The object storage unit 1104 returns to the SQL interface unit 1107 via the relational SQL access processing unit 1106.

In the embodiment of the present application, the method for executing the SQL instruction is used to parse the identifier of the object in the SQL instruction to obtain the corresponding object schema and the relational data table, and the key value attribute of the object in the SQL instruction is used to generate the data corresponding to the object. The index value, and the object data is obtained according to the index value, and the query condition in the parsing SQL instruction is used to query the attribute data from the object data. The identifier of the object in the above SQL instruction, the key value attribute of the object, and the query condition are not the object class data, but the valid data such as variables and identifiers that can be passed and parsed by the standard SQL interface, and the above validated by the SQL command parsing. The data of the data mapping object and the attribute data of the object implement the standard SQL access object.

Referring to FIG. 8, the method of reading an object by the object interface unit 1101 includes:

S301. Receive a read request for an object, where the read request includes an identifier of the object and a key value attribute.

This step differs from S202 in that it is not necessary to remove the relational operator in the SQL operation, and the read request explicitly includes the mode name and the key value attribute.

Similar to step S102, since the data representation in the read request may be a JSON format, an XML format, or the like, the object interface unit 1101 parses the object's schema name and key value attribute into a unified form that the system can handle, and The processing is sent to the object access processing unit 1102.

S302. Obtain a corresponding object mode according to the identifier of the object in the read request and the mapping relationship between the identifier of the object and the object mode.

This step is different from step S203 in that it is not necessary to query to obtain a relational data table.

S303. Determine an index value corresponding to the data of the object according to the key value attribute of the object and the attribute that can be used to generate the index value defined in the object mode.

This step is similar to step S204, and details are not described herein again.

S304. Query the data of the object according to the index value.

This step is similar to S205, except that the query gets the data of the entire object, not the attribute data in the data of the object.

S305. Return data of the object obtained by the query.

The object storage unit 1104 returns the data of the object to the object interface unit 1101 through the object access processing unit 1102.

In summary, the foregoing method and apparatus provided by the embodiments of the present application first solve the problem that a system using object storage cannot provide standard relational SQL access; the reason is that the device simultaneously provides an object interface and a SQL interface, and the object is Pattern and relational data tables enable objects to be stored in associations in the device. Not only do the object storage attributes with stored values of keys have relational attributes, they can provide standard relational SQL access in the system where the objects are stored. Secondly, through automatic conversion, it is possible to optimize and execute related technologies using relational SQL that is already very mature; the reason is that relational databases have many excellent features, which not only maintain data consistency, but also rely on standardization. Data update has a small overhead and can perform complex queries such as joins. These technologies can be fully utilized by automatic conversion on storage devices that provide standard SQL interfaces. In addition to this, the object schema definition itself is optimized for SQL queries. Thirdly, the distributed characteristics of the system are preserved, and the programmability and portability of the system are improved, which has wider application prospects. The reason is that the device can be used in a stand-alone environment or in a distributed environment. in. The standard SQL interface makes it easy to deploy the device in an existing standard SQL-based application environment, and the variety of interfaces increases the developer's programming choices. Under the general trend of big data distribution, it has broad application prospects.

In the case of employing an integrated unit, FIG. 9 shows a possible structural diagram of an apparatus for executing an SQL instruction involved in the above embodiment. The apparatus 11 that executes the SQL instruction includes a processing module 1122 and a communication module 1123. The processing module 1122 is configured to control and manage the action of the apparatus 11 that executes the SQL instruction. For example, the processing module 1122 is configured to support the apparatus 11 that executes the SQL instruction to execute the processes S101-S106 in FIG. 2, the process S201-S206 in FIG. , the process S301-S305 in FIG. Communication module 1123 is used to support communication of devices executing SQL instructions with other entities. The apparatus 11 executing the SQL instructions may further include a storage module 1121 for storing program codes and data of the apparatus that executes the SQL instructions.

When the processing module 1122 is a processor, the communication module 1123 is a network port, and the storage module 1121 is a memory, the apparatus for executing the SQL instruction according to the embodiment of the present application may be a device for executing the SQL instruction as described below.

Referring to FIG. 10, the apparatus 11 for executing an SQL command includes a processor 1132, a network port 1133, a memory 1131, and a bus 1134. The network port 1133, the processor 1132, and the memory 1131 are connected to each other through a bus 1134. The bus 1134 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. Wait. The bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is shown in the figure, but it does not mean that there is only one bus or one type of bus.

It should be understood that, in the various embodiments of the present application, the size of the sequence numbers of the foregoing processes does not mean the order of execution sequence, and the order of execution of each process should be determined by its function and internal logic, and should not be applied to the embodiment of the present application. The implementation process constitutes any limitation.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the solution. A person skilled in the art can use different methods to implement the described functions for each particular application, but such implementation should not be considered to be beyond the scope of the present application.

A person skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the system, the device and the unit described above can refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.

In the several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be electrical, mechanical or otherwise.

The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using a software program, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in accordance with embodiments of the present application are generated in whole or in part. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer readable storage medium or transferred from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be from a website site, computer, server or data center Transmission to another website site, computer, server, or data center by wire (eg, coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (eg, infrared, wireless, microwave, etc.). The computer readable storage medium can be any available media that can be accessed by a computer or a data storage device that includes one or more servers, data centers, etc. that can be integrated with the media. The usable medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium (such as a Solid State Disk (SSD)) or the like.

The foregoing is only a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application. It should be covered by the scope of protection of this application. Therefore, the scope of protection of the present application should be determined by the scope of the claims.

Claims

A method for executing a structured query language SQL instruction, comprising:

Receiving an SQL instruction, the SQL instruction is used to query attribute data of the object from data of an object;

Parsing the SQL instruction to obtain an identifier of the object, a key value attribute of the object, and a query condition, where the identifier of the object is used to indicate a type of the object, and the key value attribute of the object is used to generate an index value The index value corresponds to data of the object, and the query condition is used to query the attribute data from data of the object;

Obtaining an object schema and a relational data table corresponding to the object according to the identifier of the object, and a mapping relationship between the identifier of the object and the object schema, where the object schema is used to define a tree data structure of a type of object, The tree data structure includes at least one first node, the first node is used to define an attribute of an attribute of such an object that can be used to generate the index value, and the relational data table is a two-dimensional table. The two-dimensional table stores attribute data of the object according to a node relationship indicated in the object mode;

Determining an index value corresponding to the data of the object according to the key value attribute of the object and the attribute for generating an index value defined in the object mode;

Obtaining data of the object according to an index value corresponding to the data of the object, and querying the attribute data from the data of the object according to the query condition and the relational data table of the object.
The method of claim 1, wherein before the receiving an SQL instruction, the method further comprises:

Defining an object schema of the object and a relational data table;

Receiving a write request for the object, where the write request includes an identifier and an attribute of the object;

Obtaining an object mode and a relational data table of the object according to the identifier of the object in the write request, and a mapping relationship between the identifier of the object and the object mode;

Converting an attribute in the write request into an object having a complete storage structure according to an object schema of the object, and constructing an attribute in the write request as an attribute according to an attribute defined in the object pattern for generating an index value An index value, wherein the object having a complete storage structure refers to an object stored in a storage space allocated according to all attributes of the object mode, and attributes in the write request are stored in the storage space;

Converting the object having the complete storage structure into data of the object according to the relational data table of the object, and storing the data of the object according to the index value.
The method of claim 1 further comprising:

Receiving a read request for an object, the read request including an identifier and a key value attribute of the object;

Obtaining a corresponding object mode according to the identifier of the object in the read request, and the mapping relationship between the identifier of the object and the object mode;

Determining an index value corresponding to the data of the object according to the key value attribute of the object and an attribute that can be used to generate an index value defined in the object mode;

The data of the object is obtained according to the index value query.
The method according to any one of claims 1 to 3, wherein at least one parent node is included in the tree data structure of the object, the parent node includes an identifier and a domain, wherein the parent node The identifier is used to indicate the category to which the parent node corresponds, and the domain of the parent node is used to include the child node.
The method according to claim 4, wherein each node of the relational data table includes a parent table identifier, a root table identifier, and a foreign key, and the parent table identifier is used to point to a parent node of the child node A two-dimensional table identifying a two-dimensional table for pointing to a root node of the tree data structure, the foreign key pointing to a primary key of the object schema.
An apparatus for executing a structured query language SQL instruction, comprising: a SQL interface unit, a relational SQL access processing unit, an object relational conversion unit, a metadata management unit, and an object storage unit, wherein

The SQL interface unit is configured to receive an SQL instruction, where the SQL instruction is used to query attribute data of the object from data of an object;

The relational SQL access processing unit is configured to parse the SQL instruction received by the SQL interface unit to obtain an identifier of the object, a key value attribute of the object, and a query condition, where the identifier of the object is used to indicate a type of the object, the key value attribute of the object is used to generate an index value, the index value corresponding to data of the object, the query condition is used to query the attribute data from data of the object ;

The object-relational conversion unit is configured to obtain, according to the identifier of the object obtained by the relational SQL access processing unit, and a mapping relationship between an identifier of an object stored by the metadata management unit and an object mode, An object schema and a relational data table corresponding to the object, the object schema is used to define a tree data structure of a class of objects, the tree data structure includes at least one first node, and the first node is used to define An attribute of such an object that can be used to generate an attribute of the index value, the relational data table being a two-dimensional table, the two-dimensional table storing attribute data of the object according to a node relationship indicated in the object pattern ;

The object-relational conversion unit is further configured to determine an index value corresponding to the data of the object according to the key value attribute of the object and the attribute for generating an index value defined in the object mode;

The object-relational conversion unit is further configured to obtain data of the object according to an index value corresponding to the data of the object, and obtain the data from the object storage unit according to the query condition and the relational data table of the object. The attribute data is obtained by querying the data of the object stored in the object.
The device according to claim 6, wherein the device further comprises an object interface unit, an object access processing unit,

The object interface unit is configured to define an object mode and a relational data table of the object before the SQL interface unit receives an SQL instruction;

The object interface unit is further configured to receive a write request for the object, where the write request includes an identifier and an attribute of the object;

The object access processing unit is configured to obtain an object of the object according to an identifier of the object in the write request, and a mapping relationship between an identifier of an object stored by the metadata management unit and an object mode. Pattern and relational data tables;

The object access processing unit is further configured to convert an attribute in the write request into an object having a complete storage structure according to an object mode of the object, and according to an attribute defined in the object mode for generating an index value, An attribute in the write request is constructed as the index value, wherein the object having a complete storage structure refers to an object stored in a storage space allocated according to all attributes of the object pattern, and the attribute in the write request is stored in the In the storage space;

The object access processing unit is further configured to convert the object having the complete storage structure into data of the object according to the relational data table of the object, and store the data of the object according to the index value. To the object storage unit.
The device according to claim 6, wherein the device further comprises an object interface unit, an object access processing unit,

The object interface unit is configured to receive a read request for an object, where the read request includes an identifier and a key attribute of the object;

The object access processing unit is configured to obtain a corresponding object mode according to the identifier of the object in the read request and the mapping relationship between the identifier of the object stored in the metadata management unit and the object mode;

The object access processing unit is further configured to determine an index value corresponding to the data of the object according to the key value attribute of the object and an attribute that can be used to generate an index value defined in the object mode;

The object access processing unit is further configured to query, according to the index value, the data of the object from the object storage unit.
The apparatus according to any one of claims 6-8, wherein the tree data structure of the object includes at least one parent node, the parent node includes an identifier and a domain, wherein the parent node The identifier is used to indicate the category to which the parent node corresponds, and the domain of the parent node is used to include the child node.
The apparatus according to claim 9, wherein each node of said relational data table includes a parent table identifier, a root table identifier, and a foreign key, said parent table identifier being used to point to a parent node of said child node A two-dimensional table identifying a two-dimensional table for pointing to a root node of the tree data structure, the foreign key pointing to a primary key of the object schema.
An apparatus for executing a structured query language SQL instruction, comprising: a processor, a memory, a bus, and a communication interface; the memory is configured to store a computer to execute an instruction, and the processor and the memory pass through the bus Connected, when the device is in operation, the processor executes computer stored instructions stored in the memory to cause the device to perform the method of any of claims 1-5.
A computer storage medium, comprising instructions that, when executed on a computer, cause the computer to perform the method of any of claims 1-5.