JPH1055373A - Index generating method and information retrieval device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable fast retrieval of an attribute value beyond the limit of individual tables by grouping information into tables which are collection of records consisting of a plurality of attributes, and referring to all indexes generated in common as to corresponding attributes of respective tables in the group and retrieving a target record. SOLUTION: A retrieval means knows an index number Iij regarding an attribute ANij to be retrieved with type definition information, compares the attribute value 81 in the index with an attribute value given an index condition and finds a record pointer 82 corresponding to the attribute value 81 when finding the attribute value 81 meeting the condition, and takes target records #T1 :R1 and #T2 :R2 out of a secondary storage device with the record pointer 82. Consequently, the records #T1 :R1 and #T2 :R2 can be searched for beyond the limit of the tables.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information retrieval apparatus generically referred to as a database.

[0002]

2. Description of the Related Art In a conventional information retrieval apparatus, each table is an independent management unit, there is no treatment of grouping tables, and an index for retrieval is usually created in table units. .

[0003] The relational database has a concept of a view, and it is possible to create a virtual table that does not actually exist. However, the view is fixed at the time of definition, and each time a new table is created. There are no dynamic features to be incorporated into.

[0004] Even in the world of a simple language that does not reach a database, some tables manage a plurality of tables in units called binders, and can search not only individual tables but also all tables in the binder.

[0005]

However, when it becomes necessary to search a plurality of tables of the same type beyond the table frame, the index created for each attribute is independent. In a search for an attribute value included in any of the attributes, a search condition expression is considerably complicated, and the search speed is significantly reduced.

Further, the binder is also a binder to the extent that it gets tired, and the management of the tables in the binder and the contents of the tables are free to the user. The management of the group is left to the user and the format is limited to the binder name. .

In view of the above, the present invention automatically creates a common index for the corresponding attribute of a plurality of tables, or deletes all indexes of a grouped table even if a common index is not created. It is an object of the present invention to provide an information retrieval apparatus that automatically retrieves information and enables high-speed retrieval beyond the boundaries of individual tables.

[0008]

According to the present invention, there is provided an information management means for managing information by grouping a table which is a set of records including a plurality of attributes, and a common index for a corresponding attribute of each table in the group. The information retrieval apparatus is provided with an index creating means for creating a target record and a search means for searching for a target record by referring to all indexes created in common for corresponding attributes of each table in the group.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention has the above-described structure,
The indexing means automatically creates a common index for corresponding attributes with respect to indexing in the grouped tables managed by the information managing means, and the searching means comprises a grouped index managed by the information managing means. By automatically retrieving all indices for the corresponding attributes for retrieval within a table, the retrieval of attribute values across a plurality of tables can be performed at high speed.

FIG. 1 is a block diagram showing an information retrieval apparatus according to an embodiment of the present invention. In FIG. 1, 1 is an information management unit, 2 is an index creation unit, 3 is a search unit, and 4 is an index.

The information management means 1 manages information consisting of a collection of a plurality of attributes in units of records, and manages a table, which is a set of records, in a main storage or a secondary storage.
Now, assuming that records are fixed length and all records are stored in the secondary storage, the storage form is as shown in FIG. In FIG. 2, 5 is a secondary storage, 51 to 5b
Is a table, and 511 to 51a are records. All records are given a record serial number, and the storage position of each record in the table is (record serial number) x
(Record length). Hereinafter, a table number is considered as a table identifier and a record serial number is considered as a record identifier, and a record pointer is represented by # (table number) :( record serial number). A record is composed of a plurality of attributes, and the format of the record, that is, the type of the table is managed by type definition information. An example is shown in FIG. In FIG. 3, 6 is type definition information, 61 is a type number, 62 is an attribute name, 63 is an attribute type (character string, numeric value, etc.), 64 is an attribute length (width),
65 is an index number. The table is created in one of the types registered in the type definition information 6, and the correspondence is managed by the table management information. An example is shown in FIG. 4, where 7 is table management information, 71 is a table number, 72 is a model number, 73 is a table index common flag, and 74 is a table group number.

When the index creating means 2 is instructed to create an index for a certain attribute AN _ij in the table T of type i, first, the value F _T of the index common flag 73 in the table management information 7 (1: application of the index common, 0: not applicable) examined, F _T
If = 1, the attribute AN in the record is determined by the type definition information 6.
The position and length of _ij are known, and an index, which is information on the correspondence between attribute values and record pointers, is created in the index indicated by index number _Iij (FIG. 5). If F _T = 0, index commonalization is not applied, so that it is similarly created in the index TI _ij dedicated to the table T. In FIG. 5, 8 is an index I _ij , 81 is an attribute value, 82
Is the record pointer of the record having that attribute value.

The search means 3 searches for an attribute AN _ij in a group of tables grouped by type,
Know the index number I _ij about attributes AN _ij of the type definition information 6 to be searched, comparing the attribute values with the given attribute value 81 in that index as a search condition (matching, mismatch,
When an attribute value 81 that meets the condition is found, a record pointer 42 corresponding to the attribute value is obtained, and the target record is retrieved from the secondary storage 5 by the record pointer 82. Also, a table that is not grouped by index commonality (the value of index commonality flag 73 =
0), the same operation is performed on the index TI _ij relating to the attribute AN _ij to be searched.

The operation of the information retrieval apparatus of the present embodiment configured as described above will be described below.

FIG. 6 shows a first management mode when the information management means 1 according to the present embodiment is grouped by sharing an index. 6, reference numeral 10 denotes table management information A, which is 1 to M, (M + 1) to (M + N).
Is a table. Here, (M + 1) to (M + N) are groups of the first table having the same type i and having a common index,
F _{M + 1 to} F _{M + N} are all set to 1.

Different Tables T₁, T_Two((M + 1) ≦ T_p≤
Two records #T belonging to (M + N), p = 1, 2)
₁: R₁, #T_Two: R_TwoFor the attribute AN_ijIndex for
Is instructed to create the index, the index creating means 2
Is the record pointer #T₁: R₁Table number T₁To
Then, it is known from the table management information A10 that the type is i.
Next, the configuration of the attribute of type i is known from the type definition information 6.
Attribute AN_ijIndex number I for_ijKnow the above function
Record #T₁: R₁Attribute AN_ijAttribute value K₁When
Record pointer #T₁: R₁And index I_ijTo
register. Record #T_Two: R_TwoFollow the same procedure for
From the attribute value K_TwoAnd record pointer #T_Two: R _TwoPair with
Index I_ij(FIG. 7).

The search means 3 has a common index by type.
When retrieving a table, refer to the type definition information 6.
And the attribute AN in type i_ijIs the index I_ijKnow that
Index I_ijGoing beyond the table by searching for
Record #T₁: R₁, #T _Two: R_TwoCan be found
It will work.

FIG. 8 shows a second management mode when the information management means 1 in the present embodiment is grouped by sharing the index. In FIG. 8, reference numeral 20 denotes table management information B, 1 to M, (M + 1) to (M + N).
Is a table. Here, (M + 1) to (M + N) are tables belonging to the second group S having the same type but having no common index, and F _{M + 1 to} F _{M + N} are all set to 0.

Different tables T ₃ , T ₄ ((M + 1) ≦ T _p ≦
Two records #T belonging to (M + N), p = 3, 4)
₃ : R ₃ , #T ₄ : When an instruction to create an index for the attribute AN _ij is given to R ₄ , the index creating means 2
The record pointer #T _3: from R _3, to give the table number T _3, the table management information B20 know that the table T ₃ has its own index, attribute values K ₃ in the index TI _K3 for the attribute AN _K
And record pointer #T ₃ : R ₃ are registered in association with each other. Record #T ₄ : For R ₄ ,
The attribute value K _{4 is associated} with the record pointer #T ₄ : R ₄ and registered in the index TI _K4 (FIG. 9).

The retrieval means 3 knows that the tables T ₃ and T ₄ belong to the same group S by referring to the table management information B20, and finds the indexes TI _K3 and TI _K4 for the attribute AN _K and all other corresponding indexes. by searching the index TI _KS, record #T ₃ beyond the framework of Table: R _3, # T
_4: it is possible to locate the R _4.

As described above, according to the present embodiment, a grouping related to index creation or a grouping related to search is defined for a plurality of tables in a database, so that a search beyond a table frame is performed. Can be performed at high speed.

The method of creating the index is described in ISA
There are several methods, such as M, B-TREE, and HASH, which are not limited to the present embodiment. It goes without saying that grouping related to index creation and grouping related to search can be defined independently of each other.

[0023]

As described above, according to the present invention,
Searching for attribute values over a plurality of tables having the same format can be expected to be performed at high speed without complicating the search condition expression, and the practical effect is large.

[Brief description of the drawings]

FIG. 1 is a block diagram of an information retrieval apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a storage form of tables and records in a secondary storage;

FIG. 3 is a structural diagram of type definition information.

FIG. 4 is a structural diagram of table management information.

FIG. 5 is a structural diagram of an index.

FIG. 6 is a diagram showing a first table management form according to the embodiment;

FIG. 7 is an explanatory diagram of registration in an index in the first table management mode.

FIG. 8 is a diagram showing a second table management form according to the embodiment;

FIG. 9 is an explanatory diagram of registration to an index in a second table management mode.

[Description of Signs] 1 Information management means 2 Index creation means 3 Search means 4 Index 5 Secondary storage 51-5b Tables 511-51a Records 6 Type definition information 7 Table management information 8 Index 10, 20 Table management information

Claims (4)

[Claims] 1. An index creation method, wherein an index is created by associating an attribute value stored in a record constituting a table with a value (record pointer) capable of specifying the record beyond the table frame. . 2. The index creation method according to claim 1, wherein the record pointer is a value generated based on at least a table identifier and a record identifier. 3. An index created by the index creation method according to claim 1 is searched by comparing it with an attribute value indicated in a search condition, and a record pointer associated with an attribute value matching the search condition is obtained. Information retrieval device. 4. An index created by the index creation method according to claim 2 is searched for by comparing with an attribute value indicated in the search condition, and a record pointer associated with an attribute value that matches the search condition is searched for. An information retrieval device for obtaining at least a table identifier and a record identifier.