JPH04114276A - Condition retrieving system to file having index - Google Patents

Abstract

PURPOSE:To execute a retrieval processing at high speed by providing an intermediate index entry extraction part to extract the intermediate index entry of a least-significant extracted index block in the order designated in advance. CONSTITUTION:This system is equipped with a storage device 1, index block extraction part 2, index block hierarchy judgement part 3, top index entry extraction part 4, intermediate index entry extraction part 5, final index entry extraction part 6 and data comparison part 7. In the case of a conditional expression is '<' and '<=' after completing the processing to one index block, a preceding index block 12 shown by a PP (Prior Pointer) is read and when the conditional expression is the other one, a following index block 14 shown by an NP (Next Pointer) is read in so as to execute a comparison processing, which is completed up to the moment, to a newly read index block 14. Thus, since an objective result can be obtained without comparing all the index entries in the index block, time for condition retrieval is shortened.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a conditional search method for files having an index, and particularly to a conditional search method that increases the processing speed of the search.

[Conventional technology]

Conventionally, in the search process for files with indexes,
The index part stored in the storage device is traced from the highest index block to the lowest index block based on the specified value in the search condition (consisting of a conditional expression and a specified value), and the lowest index block is reached. Afterwards, the item values of all index entries in the lowest index block (the combination of the item value serving as the index in the data and the address indicating the location where the data is stored) are compared with the search conditions, and the items that meet the conditions are compared. It was generating a subset of index entries.

[Problem to be solved by the invention]

In the conventional conditional search method described above, the index block is traced from the highest level to the lowest level, and after locating the index entry with the same value as the specified value in the lowest index block, the conditional expression (=, ≦ , ≧.

<, >), read forward or backward from that index entry one by one, and compare the read values one by one with the specified value. The disadvantage is that it takes time to compare.

SUMMARY OF THE INVENTION An object of the present invention is to provide a conditional search method for files with indexes that allows high-speed search processing without comparing all index entries.

[Means to solve the problem]

The conditional search method for files with the index of the present invention is as follows:
An index block extraction unit that extracts an index block, which is a collection of index entries including item values and addresses, from the index unit of the storage device in which the file with the index exists; and an index block extraction unit that inspects the ID unit of the extracted index block and extracts the If it is not an index, instruct the index block extraction unit to extract the index block in the next lower hierarchy indicated by the index entry with the item value that is equal to or immediately higher (or immediately lower) than the specified value in the search condition. An index block hierarchy determination unit that repeatedly sends out a block until it is extracted, a first index entry extraction unit that extracts the index entry at the beginning of the extracted lowest index block, and an index entry at the end of the extracted lowest index block. a trailing index entry extraction section that extracts the index entries in the middle of the extracted lowest index block in a prespecified order, and an intermediate index entry extraction section that extracts the intermediate index entries of the extracted lowest index block in a prespecified order, and A data comparison section that compares the specified value with a specified value, a specified comparison condition judgment section that judges the comparison result and the conditional expression in the search condition to determine whether the search condition is satisfied, and an annotation first index entry extraction section. Yamabe, an intermediate index entry extraction section, a trailing index entry extraction section, a data comparison section, and a subset generation section that controls the specified comparison condition determination section and generates a subset of index entries satisfying the search condition. ing.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention, in which 1 is a storage device in which a file with an index section exists, 2 is an index block extraction section that extracts an index block from the index section, and 3 is an index block extractor that extracts an index block from the index section. The ID section is inspected, and if it is not the lowest index, the index block extraction unit 2 is instructed to extract the index block in the next lower hierarchy indicated by the index entry with the item value that is equal to the specified value in the search condition or has the immediately higher item value. 4 is an index block hierarchy determination unit that repeatedly sends out the index block until it extracts the lowest index block; 4 is a first index entry extraction unit that extracts the index entry at the beginning of the extracted lowest index block; 5 is an index block in the middle of the lowest index block that has been extracted; an intermediate index entry extraction unit that extracts index entries in a prespecified order;
Reference numeral 6 denotes a trailing index entry extraction unit that extracts the index entry at the end of the extracted index block; 7 a data comparison unit that compares the item value of the extracted index entry with the specified value specified in the search condition; 8 9 is a search condition storage unit that records specified search conditions; 9 is a specified comparison condition determination unit that determines whether the search condition is satisfied by comparing the conditional expression in the search condition with the comparison result of the data comparison unit 7; 10 is a specified comparison condition determination unit that determines whether the search condition is satisfied Top index entry extraction unit 4. Intermediate index entry extraction unit5. Trailing index entry extraction unit6. This is a subset generation unit that controls the data comparison unit 7 and the specified condition determination unit 9 to generate a subset of index entries that meet the search conditions.

Next, the operation of this embodiment will be explained using the index section shown in FIG. 2 as an example.

When a search condition is specified and a conditional search is instructed, the index block extracting section 2 is first activated, and the topmost index block 11 shown in FIG. 2 is extracted from the index section of the storage device.

Now, it is assumed that the search condition ``index entry item value≦specified value (40)J'' is specified and stored in the search condition storage unit 8.

The extracted index block 11 is checked by the index block hierarchy determination unit 3 for the ID part at the top, and it is determined whether it is the lowest index block or not. In this case, since the index block 11 is at the top level, among the index entries in the index block 11 (the item value and its address of the last index entry in the index block at the next lower level), it is equal to the specified value or The index entry with the most recent item value greater than, in this case, the second r40. An instruction is issued to the index block extraction unit 2 to extract the index block 13 at the address indicated by ****J.

If the index has several hierarchies, this index block extraction and extracted index hierarchy judgment are repeated until the lowest index proc is reached.

After reaching the lowest index block, the index entry at the head of the extracted index block 13 is extracted by the head index entry extraction unit 4.

The extracted index entry is compared with the specified value "40" stored in the search condition storage section 8 by the data comparison section 7, and the comparison result is [J, which has the larger specified value.

Next, in the specified comparison condition determination section 9, the search condition storage section 8
It is compared with the conditional expression "≦" stored in . As a result, since the first index entry of the index block 13 satisfies the search conditions, this information is notified to the subset generation unit IO.

The trailing index entry extraction unit 6 is activated in response to an instruction from the subset generation unit 10 that received this notification, and the index block 1
The index entry at the end of 3 is extracted. The extracted trailing index entry is compared with the designated value in the same way as the leading index entry in the data comparison unit 7, and the result is "equal to the designated value."

Furthermore, the specified comparison condition determination unit 9 determines the conditional expression in the same manner as in the case of the first index entry, and the trailing index entry of the index block 13 is determined to satisfy the search condition.
The information is notified to the subset generation unit 10.

As a result of the above, the index block 13 is [first index entry
Specified value = trailing index entry. In a normal index section organized in ascending order, the relationship between the first index entry and the last index entry is always "first index entry ≦ last index entry", so all the index entries stored in this index block 13 satisfy the search conditions. This means that Therefore, in this case, there is no need to compare item values of intermediate index entries, so a subset of all index entries in the index block 13 is generated.

Next, a case where the comparison results and search conditions are different from those described above will be explained.

If the comparison result that takes into account the conditional expression of the first index entry satisfies the search condition and the conditional expression is ">" or "≧", all index entries in the index block satisfy the search condition, so the intermediate index entry is A subset of all index entries in the index block is generated without comparing the item values of l to ri.

When the conditional expression is "<J "≦" or "-", if the comparison result including the conditional expression of the trailing index entry satisfies the search condition, then all index entries in the index block satisfy the specified condition. Generate a subset of all index entries in an index block.

In addition, if the comparison result taking into account the conditional expression of the trailing index entry does not satisfy the search condition, the intermediate index entry extraction unit 5 extracts the index entries one by one, performs a comparison taking into account the conditional expression, and then satisfies the search condition. Generate a subset with satisfied index entries.

When the comparison result taking into account the conditional expression of the first index entry does not satisfy the search condition and the conditional expression is ">J "≧J or "-", a comparison taking into account the conditional expression of the trailing index entry is performed,
If the search condition is satisfied, the intermediate index entry extraction unit 5 extracts the index entries one by one, performs a comparison taking into account the conditional expression, and generates a subset of index entries that satisfy the search condition.

When the conditional expression is other than ">", "≧J" or "-J", all index entries in the index block do not satisfy the specified condition, so all index entries in the index block are not added to the subset.

After the processing for one index block described above is completed, if the conditional expression is "<" or "≦", P P (Pr
1 or Po1nter), and if the conditional expression is other than the above, N P
The next index block 14 indicated by (NexLPointer) is read, and the comparison processing that has been performed so far is performed on the newly read index block.

Although the above embodiment has been described for the case where the index is in ascending order of item values, the same process can be performed even when the index is in descending order. In this case, the index block hierarchy determination unit may instruct extraction of an index block corresponding to an item value that is equal to or immediately below the specified value.

Note that the intermediate index entry extraction section extracts index entries one by one, but the extraction method is not particularly limited and may be extracted sequentially from the beginning or end. Furthermore, extraction and comparison are performed when the boundaries of the search conditions are discovered. It is also possible to cancel it.

[Effects of the Invention] As described in detail above, the present invention has the effect of shortening the condition search time because the desired result can be obtained without comparing all index entries in the index block.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, and FIG. 2 is an explanatory diagram showing the structure of an index section. ■... Storage device, 2... Index block extraction section, 3... Index block hierarchy determination section, 4.
--- Leading index entry extraction section, 5 --- Intermediate index entry extraction section, 6 --- Trailing index entry extraction section, 7 --- Data comparison section, 8 ---・・・・・・
Search condition storage unit, 9...Specified comparison condition determination unit,
10...---Subset generation unit, 11-14...
...index block.

Claims (1)

[Claims] An index block extraction unit that extracts an index block, which is a collection of index entries including item values and addresses, from the index unit of the storage device in which the file with the index exists; and an index block extraction unit that inspects the ID unit of the extracted index block and extracts the If it is not an index, instruct the index block extraction unit to extract the index block in the next lower hierarchy indicated by the index entry with the item value that is equal to or immediately higher (or immediately lower) than the specified value in the search condition. An index block hierarchy determination unit that repeatedly sends out a block until it is extracted, a first index entry extraction unit that extracts the index entry at the beginning of the extracted lowest index block, and an index entry at the end of the extracted lowest index block. a trailing index entry extraction section that extracts the index entries in the middle of the extracted lowest index block in a prespecified order, and an intermediate index entry extraction section that extracts the intermediate index entries of the extracted lowest index block in a prespecified order, and a data comparison unit that compares the specified value with a specified value; a designated comparison condition determination unit that determines whether the search condition is satisfied by determining the comparison result and a conditional expression in the search condition; and the first index entry extraction unit. , a subset generation unit that controls the intermediate index entry extraction unit, the trailing index entry extraction unit, the data comparison unit, and the specified comparison condition determination unit to generate a subset of index entries that satisfy the search conditions. A conditional search method for files with indexes.