JPH0981426A - Record reading method and record retrieval method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform the retrieval of a record exactly and at high speed by newly setting the next time reading starting location according to the writing of the record. SOLUTION: A read starting pointer control part 4 stores the starting pointer of the next time reading (read) for every file 1 in a read starting pointer control table T 4. A starting pointer change processing part 5 is called when a record detection part 3 receives a write request with 'read starting pointer change' and a processing is executed. A decision is performed in accordance with the condition stored in a read starting pointer change condition storage area T 5-1 or the condition added to the write request. When the procedure stored in a read starting pointer change procedure storage area T 5-2 or the change procedure added to the write request exists if the decision result is true, the read starting pointer of the read starting pointer control part 4 is changed according to the procedure.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a record reading method and a record searching method, and more particularly to reading a filed record, managing a search starting position, reading, reading at the time of searching, reading from a search starting position,
The present invention relates to a record reading method and a record searching method for starting a search.

With the recent spread of OA systems, various business data are input to a computer system and temporarily stored in a large-capacity storage device, and then desired data (referred to as a record) is retrieved as appropriate and a predetermined data is stored in the record. In many cases, transaction processing is performed one after another and the final processing result is stored in the mass storage device for a certain period or longer.

Therefore, as the business is continued, the number of accumulated records increases, and it takes time to retrieve a desired record. However, even if the number of accumulated records increases in this way, it is necessary to secure the capacity of the computer system. In other words, there is a demand for a record reading / searching method that can shorten the search time even if the number of stored records increases and does not reduce the performance of the computer system.

[0004]

2. Description of the Related Art FIG. 15 shows an operation explanatory diagram of a conventional general file system during record processing. In a conventional file system, status information indicating the processing status is added to records and stored in a file in the order of occurrence, records are read from the file, transaction processing is performed, status information according to the processing is added, and the file is stored again in the file. It is said that. Here, the status information is information indicating the processing status of each record, and includes, for example, a status indicating a processing waiting status in the processing, a status indicating a processing, and a status indicating that the processing is completed.

In FIG. 15A, the record number. 1
-10 is status S3, 11-18 is S2, 19-2
7 is S1, and pointers of S2 and S1 are record N, respectively.
o. 10 and No. It shows when it is 18. Record No. Was previously searched as the record of S2, but the status has advanced to S3 as a result of the processing. At this time, when the transaction processing unit that performs a predetermined transaction process on the record makes a search request using S2 as the search condition (step S11-1), the search is performed with the record number.
Start from 10, but record N as a condition matching record
o. Notify 11. At the same time, as shown in FIG. 15B, the pointer of S2 is set to the record number. Change to 11.

Here, the record number. The transaction processing unit notified of the record No. 11 is the record number. When an update request for returning the status of S11 to S1 is made (step S11-2), the pointer of S1 is originally the record number as shown in FIG. The record number should be 11. It remains 18. In this state, when the transaction processing unit makes a search request with S1 as the search condition (step S11-3), the record N
o. Record No. 11 is the place where notification must be made. There were cases where 18 was notified.

FIG. 16 shows an explanatory view of another conventional example.
FIG. 16 shows a file system proposed by the present applicant in Japanese Patent Application No. 2-244610 (Japanese Patent Application Laid-Open No. 4-123239). In the file system shown in Japanese Patent Application No. 2-244610, record retrieval / processing is executed as follows. (1) The transaction processing unit 21 issues a search request (designated to be updated later). (2) Upon receiving the request, the record search unit 22 obtains the head record position of the group corresponding to the search condition from the search result management table 23a via the search result management unit 23. (3) The search is started from the position of the first record of the group, and the record matching the search condition is acquired. (4) The position of the acquired record is reflected in the search result management table 23a as the first record of the group. (5) The retrieved record is the transaction processing unit 1
1 is notified. (6) The record received as the search result is processed and the update request is made (including the change of the status). (7) The update request record is reflected in the file 24. (8) Receive the processing result of the update processing. (9) The pointer optimization processing unit 25 specifies the top record of the file 24 as the search start position and issues a search request. (10) The search is started from the position of the first record of the file 24, and the record matching the search condition is acquired. (11) The position of the acquired record is reflected in the search result management table 23a as the first record of the group (when the status has returned, the pointer also returns to that position). (12) The retrieved record is the pointer optimization processing unit 25.
Will be notified.

(1) to (8) are repeated each time the search condition is changed each time the search is required. (9) ~
(12) is repeated for all search conditions that form the status group at regular intervals.

FIG. 17 shows a change status of each record and pointer of the file when the pointer optimization processing described in FIG. 16 is performed. FIG. 17A shows the record number. 1
-10 is status S3, record No. 12 to 18 are status S2 and record No. 11, 19 to 27 are status S1, and the pointers of status S2 and status S1 are the record number. 11 and the record number. 1
When the pointer optimization processing unit 25 designates the head of the file as the search start position and makes a search request using S1 and S2 as search conditions (step S12-1), the search is performed. Record No. Start from 10, but record N for S1 as a condition matching record
o. 11 and S2, the record number. 12 is notified. At the same time, as shown in FIG. 17B, the pointer of S1 is set to the record number. 11, the pointer of S2 is set to the record No. Change to 11.

[0010]

However, in the conventional record search method, it is premised that the status in the record does not return to the previous status. Therefore, when the status returns to the original status, the pointer has already returned the status. There is a problem in that the record goes ahead of the record and the record may not be a search target as one record in the group to which the returned status belongs.

In order to solve the above problems, FIG.
Like the file system of Japanese Patent Application No. 2-244610 shown in No. 6, separately from the "transaction processing unit 21",
It is necessary to provide a "pointer optimization processing unit 25" that searches from the head of the file at a constant time interval according to the search condition for each group, and to change the pointer to the latest state by the pointer optimization processing unit 25. Japanese Patent Application No. 2-2446
The file system of No. 10 has a problem that the processing becomes complicated and the processing speed is lowered.

The present invention has been made in view of the above points, and it is an object of the present invention to provide a record reading method and a record searching method which enable accurate and high-speed record searching.

[0013]

According to a first aspect of the present invention, there is provided a record reading method for managing a read start position of a record filed in a file section and starting the read from the read start position when the file is read. A record write detection procedure for detecting writing of a record to a copy section and a new read start position for the next read when a record write to the file section is detected in the record write detection procedure It is characterized by having a read start position setting procedure.

According to the first aspect of the present invention, the read start position for the next read can be newly set according to the writing of the record. Therefore, when the contents of the record are changed, the record is deleted, etc., this is detected as the record writing. Then, a new read start position can be set, so that the record position where the content of the record has been changed or the record has been deleted can be read again, and the read can be performed without causing a drop in the read of the record.

According to a second aspect of the present invention, the record position arbitrarily designated in the read start position setting procedure is set as the read start position. According to the second aspect, the read start position can be set according to the change content of the record by setting the record position arbitrarily designated at the time of writing the record as the read start position. And, it can be implemented at high speed.

According to a third aspect, the read start position setting procedure is
It is characterized in that the preset record position is set as the read start position only when a certain condition is satisfied at the time of writing the record detected by the record writing detection procedure. According to the third aspect, by setting the designated record position as the read start position only when the instructed condition is satisfied, it is possible to set the read start position combined with various conditions. It is possible to set an appropriate read start position, and to perform the next read at a high speed because the record is not missed, the read is accurate, and unnecessary read is not performed.

According to a fourth aspect of the present invention, the read start position of the record is set according to a fixed procedure in the read start position setting procedure. According to the fourth aspect, the record read start position can be reset according to a certain procedure, so that the record read start position can be reliably set.

According to a fifth aspect of the present invention, when a certain condition is satisfied in the read start position setting procedure, the record read start position is set according to a predetermined procedure. According to the fifth aspect, when a certain condition is satisfied, the record read start position is set according to a certain procedure.
The read start position can be reset when the predetermined conditions are satisfied, and unnecessary reset of the read start position can be eliminated.

According to a sixth aspect of the present invention, when a record write is detected in the record write detection procedure in the read start position setting procedure, a desired condition or procedure is selected and selected from preset conditions and procedures. It is characterized in that the record read start position is set in accordance with a condition or procedure.

According to the sixth aspect, when the record writing is detected in the record writing detecting procedure, a desired condition or procedure is selected from preset conditions and procedures, and the record is recorded according to the selected condition or procedure. Since the reading start position can be set, the user can select the optimum condition or procedure depending on the record contents and the record input order.

According to a seventh aspect of the present invention, conditions and procedures for setting a record read start position when record write is detected in the record write detection procedure in the read start position setting procedure are set. . According to claim 7, since the condition and the procedure for setting the read start position of the record can be set when the record writing is detected in the record writing detection procedure, the condition and the procedure required for each record writing. Can be set, and processing can always be performed under optimal conditions.

According to an eighth aspect of the present invention, a record storage unit that stores status information indicating a processing state of the record and stores the records in the order of occurrence, a record search unit that searches for a desired record in the order of occurrence, and a record search unit In a record search method of a computer system provided with a transaction processing unit for performing predetermined transaction processing and updating status information of the record, a status group is set based on a transition characteristic of the status information,
When the first record of a predetermined group is searched by the record search instruction, and the correspondence between the group and the first record is specified, when the record search of the predetermined group is specified, the first record of the group is stored in the search result management unit. When the first record belonging to a predetermined group is searched by searching for the record belonging to the specified group by referring to the status information from the top record, the search result management unit determines that the specified group It is characterized in that the correspondence relation with the first record is updated, and the record achievement management unit updates the correspondence relation between the designated group and the first record during the record content updating process.

According to the eighth aspect, when the record content is updated, the search start position for the next search can be newly set according to the change in the status provided corresponding to the record. When a record is deleted, this can be detected as a status change and a new search start position can be set. Therefore, it is possible to search again for the record position where the record content has been changed or the record has been deleted. , Search can be performed without omission in search.

According to a ninth aspect of the present invention, the search result management section sets the record position arbitrarily designated according to the change of the status as the search start position when the record content is updated. According to the ninth aspect, since the record position arbitrarily designated according to the change of the status can be set as the search start position when updating the record contents, the process can be started from the optimum position according to the change of the status. Can be done quickly.

According to a tenth aspect of the present invention, the search result management section sets a record position according to the status of the status as a search start position when updating the record contents. According to the tenth aspect, the processing can be executed for each status by setting the search start position according to the status, so that the processing can be executed according to the status.

In the eleventh aspect of the present invention, the search result management unit sets the search start position to the instructed position when a certain condition is satisfied when updating the record contents. According to the eleventh aspect, by setting the search start position to the instructed position when a certain condition is satisfied, the search start position can be set when the predetermined condition is satisfied, which is unnecessary. It is possible to eliminate re-setting of a proper reading start position.

According to a twelfth aspect, when the search result management unit is instructed to change the search start position when updating the record contents,
It is characterized in that the search start position is set to the instructed position when the instructed position is before the current set position. According to the twelfth aspect, when the change of the search start position is instructed when updating the record contents, the search start position is set to the instructed position when the instructed position is before the current set position. It is possible to prevent omission of record processing between the designated position and the current set position.

A thirteenth aspect of the present invention is characterized in that the search result management section detects a change in the status at the time of updating the record content as one of the certain conditions. According to claim 13,
Since the change in status is detected and the search start position is set, the search start position can be set according to the status.
Processing can be executed for each status.

[0029] According to a fourteenth aspect, the search result management unit sets the search start position of the record according to a predetermined procedure. According to the fourteenth aspect, since the record search start position can be set according to a predetermined procedure, the record read start position can be surely set.

According to a fifteenth aspect of the present invention, the search result management section can select the fixed condition and the fixed procedure for setting the search start position of a record from preset conditions and procedures. According to the fifteenth aspect, since the certain condition and the certain procedure for setting the record search start position can be selected from preset conditions and procedures, the user can optimize the record content and the input order of the record. Conditions or procedures can be selected.

According to a sixteenth aspect of the present invention, the search result management section can newly set the certain condition and the certain procedure for setting the record search start position. According to claim 16, since the fixed condition and the fixed procedure for setting the record search start position can be newly set, the necessary condition and procedure can be set when necessary.
Processing can always be performed under optimal conditions.

[0032]

FIG. 1 is a block diagram showing a first embodiment of the present invention. This embodiment will explain a case where the record reading method of the present invention is applied to a general file system such as a UNIX file system and a DOS file system.

In FIG. 1, 1 indicates a file. File 1, record 1, record 2, ... for each file name
, Record n is stored, and the record length is not always fixed. Reference numeral 2 denotes a transaction processing unit that makes a file access request. The transaction processing unit 2 issues a record write / read request to the file 1.

Reference numeral 3 denotes a record detection unit. The record detection unit 3 controls access to the file 1 according to a read or write request to the file 1. Reference numeral 4 indicates a read start pointer management unit. The read start pointer management unit 4 stores the next read (read) start pointer for each file 1 in the read start pointer management table T4. Also, the start pointer is taught in response to a request from the record detection unit 3.

Reference numeral 5 denotes a read start pointer change processing unit for write. In the start pointer change processing unit 5, the record detection unit 3 writes “read start pointer change”
When the e request is received, the process is called and the process described below is executed. The determination is made according to the condition stored in the read start pointer changing condition storage area T5-1 or the condition added to the write request, and if the result of the determination is true, it is stored in the read start pointer changing procedure storage area T5-2. If there is a procedure or change procedure added to the write request,
According to the procedure, read start pointer management section rea
d Change the start pointer.

When the read start pointer change condition is set,
The designated condition is stored in the read start pointer change condition storage area T-1. When the read start pointer change procedure is set, the designated procedure is stored in the read start pointer change procedure storage area T5-2.

When the transaction processing unit 2 makes a write request with a change in the read start pointer, the record detection unit 3 recognizes that the pointer has been changed and re
The change processing unit 5 at the time of updating the ad start pointer is called. The pointer change processing unit 5 has a registered change condition in the read start pointer change condition registration table T5-1, or
It is checked whether there is a change condition added at the time of write request, and if there is a pointer change condition, condition comparison is performed. If the comparison result is true, it is checked whether or not there is a registered change procedure or a change procedure added at the time of write request in the read start pointer change procedure registration table T5-2, and if there is, a read start pointer according to the procedure. The read start pointer management unit 4 is requested to change.

When the transaction processing unit 2 issues a read start pointer change procedure setting request, the read start pointer change processing unit 5 at write time is called via the record detection unit 3 and the specified read start pointer change condition storage area is set. Store in T5-2.

FIG. 2 shows a processing flow chart of the transaction processing unit 2. First, the target data is read by the input device and the detection conditions are designated (steps S2-1, S2-
2). Next, a request for detection is made to the record detection unit 3 under the detection conditions designated in step S2-2 (step S2-
3).

Next, from the record detecting section 3 to step S2-
When the detection result for the detection request requested in 3 is received,
The contents of the record such as the status are changed according to the detection result (step S2-4). Next, the record detection unit 3 is requested to update the record according to the change of the record content such as the status changed in step S2-4 (step S2-5). When the record detection unit 3 is requested to update the record contents in step S2-5, the detection start pointer can be newly set.

The transaction processing unit 2 is step S2.
The record detection unit 3 obtains the update result according to the change request in -5.
When received from, the process ends. FIG. 3 shows a processing flowchart of the record detection unit 3. If the record detection unit 3 receives a processing request from the transaction processing unit 2, whether the processing request from the transaction processing unit 2 detects a record,
It is determined whether the record is updated (step S3-1).

In step S3-1, when the processing request from the transaction processing unit 2 is a record detection of the file 1, the record detection unit 3 reads the read start pointer management unit 4
The read start point is acquired (step S3-
2). The record detection unit 3 reads in step S3-2.
When the start point is acquired, detection is sequentially performed from the next acquired read start point (step S3-3).

The record detection unit 3 notifies the read start pointer management unit 4 of the detected record when the detection is performed in step S3-3, and the processing is terminated (step S3-).
4). If the processing request from the transaction processing unit 2 is the record update of the file 1 (record registration / addition / writing / writing / change / deletion, etc.) in step S3-1, the record detection unit 3 detects the record of the file 1 The requested record update processing is performed (step S3-
5).

Next, the record detection unit 3 judges whether or not there is a change start pointer change designation from the transaction processing unit 2, and if there is no read start pointer change designation, the process is terminated (step S3-6). Next, at step S3-6, if there is a change start pointer designation for updating processing or the like, it is then determined whether or not the pointer is changed at the time of status change (step S3).
3-7). If it is not designated to change the pointer at the time of status change in step S3-7, the read start pointer management unit 4 is notified so that the pointer is changed to the read start position designated by the read start pointer change designation ( Step S3-10).

If the pointer is designated to be changed when the status is changed in step S3-7, then it is determined whether or not the status indicating the record has changed due to the update processing (step S3-8). . Step S
If there is no change in the status in 3-8, it means that a process not instructed has been executed, so the process is terminated as it is.

If there is a change in the status of the record in step S3-8, it is determined whether or not it has become a status for changing the read start pointer (step S3-9). In step S3-9, if the status in which the pointer is changed is not reached in the update processing in step S3-5, the status is not the designated status, and the processing is terminated as it is. Also, in step S3-9, step S
In the update processing of 3-5, when the status to change the read start pointer is reached, the read start pointer management unit 4 moves the pointer to the status to be changed.
Is notified (step S3-10).

FIG. 4 shows a processing flowchart of the read start pointer management unit 4. At the time of the detection implementation management process, first, the content of the request process is determined (step S4-1). When the processing content requested in step S4-1 is a result acquisition request, the designated pointer is read from the read start pointer table T4, and the pointer position is notified to the record detection unit 3 (steps S4-2 and S4-3).

When the processing requested in step S4-1 is a pointer change request, it is determined whether the change designation pointer is at a position after the current pointer (step S).
4-4). If the change designation pointer is before the current pointer in step S4-4, the pointer is changed to the designated position on the read start pointer management table T4 (step S4-6).

If the change designation pointer is after the current pointer in step S4-4, then it is determined whether or not the pointers after the previous pointer are designated not to be changed (step S4-5). If it is after the current pointer in step S4-5, or if the pointer position is not changed, the process is terminated.

If it is specified in step S4-5 that the pointer position is to be changed if it is a pointer after the current pointer, then the pointer specified at the specified position on the read start pointer management table is next. The information is changed and the process ends (step S4-6).

FIG. 5 shows the read start pointer change processing unit 5
The processing flowchart of is shown. When the update processing is executed, the read start pointer change processing unit 5 determines whether or not the read start pointer change condition itself is designated (step S5-1). If the read start pointer change condition is not specified in step S5-1, next read
It is determined whether or not the use of the ad start pointer changing condition is designated (step S5-2). Step S5-
If the use of the read start pointer change condition is specified in 2, the re specified in table T5-1 is next.
It is determined whether there is an ad start pointer change condition (step S5-3).

In step S5-1, the user reads
When the start pointer change condition itself is specified, and the read start pointer change condition preset by the user in steps S5-2 and S5-3 is specified, the specified read start pointer change condition is set in the table T5-1.
If it exists, it is determined whether or not the condition of the next requested process matches the change condition (step S5-
4).

Also, the re specified in step S5-3
If there is no ad start pointer change condition, an error is notified to the transaction processing unit 2 and the r specified by the user
Notification is made that there is no ead start pointer changing condition (step S5-10).

If the change conditions do not match in step S5-4, the change process ends (step S5-4). If the change conditions match in step S5-4 and if it is determined in step S5-2 that the use of the read start pointer change condition is not specified, then the user starts the read start pointer change procedure itself. It is determined whether or not it is designated (step S5-5).

Read by the user in step S5-5
If the start pointer changing procedure is not specified, it is determined whether the user next specifies the read start pointer changing procedure registered in advance in the table T5-2 (step S5-6). Table T in step S5-6
When it is determined that the use of the read start pointer changing procedure registered in advance in 5-2 is not specified, the present read result is unconditionally notified to the read start pointer table T4 as a new point position, and the process ends. Yes (step S5-8).

In step S5-6, the table T5-
If the use of the read start pointer changing procedure registered in advance in step 2 is specified, then the table T5-2 is used.
It is determined whether or not the read start pointer changing procedure corresponding to the designation is registered (step S5-7).

If there is a read start pointer change procedure specified in step S5-7, and step S5
When the user himself sets the read start pointer changing procedure in -5, the pointer position of the table T4 is changed according to the designated and set read start pointer changing procedure (step S5-9).

FIG. 6 is an operation explanatory diagram when a read request of 5 bytes of data is made from the file 1. When the transaction processing unit 2 requests the record detection unit 3 to read 5 bytes of data, the record of the file 1 has the structure shown in FIG.
Assuming that the start pointer P ₀ is at “H”, FIG.
As shown in (B), 5-byte records'H ',' I ',' J ',' K ', and'L' are read from'H 'and notified to the transaction processing unit 2.

The pointer P ₀ is set in FIG.
It automatically moves to the'M 'position as shown in (C).
Next, as shown in FIG. 6D, the transaction processing unit 2 causes the 5-byte data “H”, “I”, “J”,
Instead of'K ',' L ',' ☆ ',' ★ ',' ○ ',
When a request for writing the data of “●” and “◎” is made and the position of the pointer P ₁ is designated to be set at the 17th byte at this time, 8 to ₁ as shown in FIG.
2-byte data '☆', '★', '○', '●',
'◎' pointer P ₁ together with written is moved to the 17th byte.

FIG. 7 shows a block diagram of an application example of the first embodiment of the present invention. This application example is applied to a server-client type information processing system. (1) Send a write request to the transaction processing unit 2. At this time, "read start pointer change designation" is added as necessary. (2) wri with "read start pointer change designation"
Upon receiving the te request, the record detection unit 3 first writes the requested data to the file 1. (3) After writing to the file 1, the record detection unit 3
Requests the read start pointer change processing unit during write. (4) The read start pointer change processing unit during write has already registered the read start pointer change condition, or
If a change condition is added at the time of write processing request, the condition is judged according to the change condition, and if true, read
If the start pointer change procedure has been registered or if the change procedure has been added when the write processing request is made, the read start position management unit is requested to change the read start pointer management table according to the change procedure. After the read start position management unit changes the read start pointer management table, the process is returned to the record detection unit 3. (5) read Start position management part read when write
Read based on a request from the start pointer change processing unit
Changed start pointer management table. (6) The transaction processing unit receives the writ processing result.

(1) 'and (6)' are (1),
An example in which another transaction processing unit (on the same WS or another WS) performs the same process as (6) is shown.
Here, the update process refers to all the change processes for the record of the file. All the record changing processes such as record registration / addition, record writing, record content change, record deletion, etc. are targeted. A record is a collection of data having a specific length, and does not particularly define the length from 1 byte to n bytes.

As described above, according to the present embodiment, it is possible to freely set the read position of a record at the time of detection, freely respond to changes and additions of records, and the operability is good. Becomes FIG. 8 shows a block diagram of the second embodiment of the present invention. This example is Japanese Patent Publication No. 4-123.
No. 239 is proposed, and an example applied to a system configured to add a status information indicating a processing status to a record at the time of a file and to perform a file will be described.

In FIG. 8, 11 indicates a record storage section. The record storage unit 11 stores records 1, record 2, ..., Record n with status information Si (i = 1, 2, ...) Indicating processing status in the order of occurrence. 1
Reference numeral 2 denotes a transaction processing unit that makes a file access request. The transaction processing unit 12 issues a search request, performs a predetermined transaction process on the searched record, further changes the status information of the record, and stores it in the record storage unit.

FIG. 9 shows a processing flow chart of the transaction processing unit 12. First, the target data is read by the input device and search conditions are designated (steps S6-1 and S6-
2).

Next, the record search unit 13 is requested to perform a search under the search conditions specified in step S6-2 (step S6-3). Next, from the record search unit 13 to step S6
When the search result corresponding to the search request requested in -3 is received, the status and the like are changed according to the search result (step S6-4).

Next, the record retrieving unit 13 is requested to update the record according to the change of the record contents such as the status changed in step S6-4 (step S6-5). When the record search unit 13 is requested to update the record contents in step S6-5, a new search start pointer can be set.

The transaction processing unit 12 executes step S
When the update result according to the change request in 6-5 is received from the record search unit 13, the process ends. Reference numeral 13 indicates a record search unit. The record search unit 13 detects a desired record for which a record search request is issued.

FIG. 10 shows a processing flowchart of the record search unit 13. Upon receiving a processing request from the transaction processing unit 12, the record search unit 13 determines whether the processing request from the transaction processing unit 12 is a record search or a record update (step S7-1).

When the processing request from the transaction processing unit 12 is the record search of the file 11 in step S7-1, the record search unit 13 acquires the read start point from the read start pointer management unit 14 (step S7-2). . The record search unit 13 is step S7-2.
When you get the read start point in
Search is performed in order from the ad start point (step S7).
-3).

When the record search unit 13 performs the search in step S7-3, the record search unit 13 notifies the read start pointer management unit 14 of the searched record and ends the process (step S7).
-4). If the processing request from the transaction processing unit 12 is a record update (record registration / addition / writing / writing / change / deletion, etc.) of the file 11 in step S7-1, the record search unit 13 sends the record of the file 11 to the record. The requested record update processing is performed (step S7-5).

Next, the record search unit 13 determines whether or not there is a change start pointer change designation from the transaction processing unit 2, and if there is no read start pointer change designation, the process is terminated (step S7-6). Next, at step S7-6, if there is a change designation of the read start pointer at the time of the update processing, it is judged whether or not it is designated to change the pointer at the next status change (step S7).
7-7). If it is not designated to change the pointer when the status is changed in step S7-7, the read start pointer management unit 14 is notified so that the pointer is changed to the read start position designated by the read start pointer change designation ( Step S7-10).

If the pointer is designated to be changed when the status is changed in step S7-7, it is then determined whether or not the status indicating the state of the record is changed by the update processing (step S7-8). Step S7
At -8, if there is no change in the status, it means that a process not instructed has been executed, so the process is terminated as it is.

When the status of the record is changed in step S7-8, it is determined whether or not the status is such that the read start pointer is changed (step S7-9). If the status in which the pointer is changed is not obtained in the update processing in step S3-5 in step S7-9, the status is not the designated status, and the processing is terminated as it is. Also, in step S7-9, step S7
In the update process of -5, when the status to change the read start pointer is reached, the read start pointer management unit 14 moves the pointer to the status to be changed.
Is notified (step S7-10).

Reference numeral 14 denotes a search result management unit. When the number of status groups G1 to Gn is set based on the transition characteristics of the statuses S1 to Sn, the search result management unit 14 sets a relationship between the group and its head record every time the head record of the predetermined group is searched. The correspondence is stored in the search result management table T4.

FIG. 11 shows a processing flowchart of the search result management section 14. At the time of the search result management process, first, the content of the request process is determined (step S8-1). Step S8
When the processing content requested in -1 is a result acquisition request, r
The designated pointer 4 is read from the ead start pointer table T4, and the pointer position is notified to the record search unit 14 (steps S8-2, S8-3).

When the processing requested in step S8-1 is a pointer change request, it is determined whether the change designation pointer is at a position after the current pointer (step S).
8-4). If the change designation pointer is before the current pointer in step S8-4, the pointer is changed to the designated position on the read start pointer management table T4 (step S8-6).

If the change designation pointer is after the current pointer in step S8-4, then it is determined whether or not the pointer after the previous pointer is designated not to be changed (step S8-5). If it is specified in step S8-5 that the pointer position is not changed if it is after the current pointer, the process is terminated.

If it is specified in step S8-5 that the pointer position is to be changed after the current pointer, the pointer specified at the specified position on the read start pointer management table is next read. The information is changed, and the process ends (step S8-6).

Reference numeral 15 shows a change process at the time of updating the search start pointer. The search start pointer update time change processing unit 15 is called when the record search unit 13 receives an update request with "change search start pointer", and executes the following processing. FIG. 12 shows a processing flowchart of the search start pointer change processing unit 15.

Search start pointer change condition storage area T5-1
The result is judged according to the condition stored in the step S1 or the condition added to the update request, and if the result is true, the procedure stored in the search start pointer change procedure storage area T5-2 or the procedure added to the update request. If there is, the search start pointer of the search result management unit is changed according to the procedure (step S9).
-1 to S9-5, S9-9).

FIG. 13 shows an operation explanatory diagram of the second embodiment of the present invention. Here, when the search start pointer change condition is set,
Search for specified conditions Search start pointer change condition storage area T5
-1, and when the search start pointer changing procedure is set, the designated procedure is stored in the search start pointer changing procedure storage area T5-.
Store in 2.

At this time, the status is as shown in FIG.
It is assumed that the state is as shown in (a). Here, when the transaction processing unit 12 makes a search request with S2 as the search condition, as shown in FIG.
Is searched as a search target. Further, when the transaction processing unit 12 specifies the change of the search start pointer at the time of the update request, the record search unit 13 recognizes that the pointer is changed and calls the update time search start pointer change processing unit 15. . For example, when the status of the record 11 in FIG. 13B is returned to S1 and an update request is made, it is simultaneously notified that the status of the record 11 is returned to S1 as shown in FIG. 13C, and transaction processing is performed. The unit 12 makes a search request using S1 as a search condition.

At this time, the update-time pointer change processing unit 15
Checks whether there is a registered change condition in the search start pointer change condition registration table T5-1 or a change condition added at the time of the update request, and if there is a pointer change condition, the condition comparison is performed. If the comparison result is true, the search start pointer change procedure registration table T5-2 registered change procedure,
Alternatively, it is checked whether or not the change procedure added at the time of the update request exists, and if there is, the search result management unit 14 is requested to change the search start pointer in the search result management table T4 according to the procedure.

When the transaction processing unit 12 issues a search start pointer change procedure setting request, the update-time search start pointer change processing unit 15 is called via the record search unit 13 to specify the specified search start pointer change condition. Store in storage area T5-2.

FIG. 14 shows a configuration diagram of an application example of the second embodiment of the present invention. This application example shows an example applied to a server-client type information processing system. First, (1) a search request is made in a mode in which transaction processing is “updated later”. (2) The record search unit 13 obtains the head record position information from the search result management table by the search result management unit from the record corresponding to the request of (1). (3) The record is read from the position from the top record information obtained in (2), and the record corresponding to the request is read. (4) The search result management unit 14 is notified of the record position found in (3). The search result management unit 14 stores the notified information as new head record position information in the search result management table T4, and returns the process to the record search unit 13. The record search unit 13 notifies the transaction processing unit 12 of the search result. (5) The transaction processing unit 12 receives the search result. (6) The transaction processing unit 12 processes the record received in (5), and then issues an update request. If necessary, such as when the status is changed during record processing,
Make an update request with "Search start pointer change designation". (7) Upon receiving the update request with “Search start pointer change designation”, the record detection unit first writes the requested record in the file 11. (8) After writing to the file 11, the record search unit 13 requests the update-time search start pointer change processing unit 15 to perform processing. (9) The update-time search start pointer change processing unit 15 determines the condition according to the change condition when the search start pointer change condition is already registered or when the change condition is added when the update process is requested, and when it is true, the search is performed. If the start pointer changing procedure has been registered, or if the changing procedure has been added when the update processing is requested, the search result management table T
The search result management unit 14 is requested to change 4. After the search result management unit 14 changes the search start pointer management table, the process returns to the record search unit 13. (10) The search start position management unit changes the search result management table T4 based on a request from the update search start pointer change processing unit 15. (11) The transaction processing unit 12 receives the update processing result.

(1) ', (5)', (6) ', (11)'
Respectively perform the same processing as (1), (5), (6), and (11) in another transaction processing unit 12 (on the same WS, another W).
S)) shows an example of the case. As described above, according to the present embodiment, the position of the read start pointer can be arbitrarily set in response to the status change.
The position can be set freely. Therefore, the read position can be changed even when a record is changed or added, and it is possible to prevent omission of records to be searched at the time of searching and it is possible to accurately search for records.

[0087]

As described above, according to claim 1 of the present invention,
Since the new read start position for the next read can be set according to the write of the record, if the record content is changed or the record is deleted, this can be detected as record write and the new read start position can be set. Therefore, the record position where the content of the record is changed, the record is deleted, and the like can be read again, and the record can be read without any omission or the like.

According to claim 2, according to claim 2,
By setting a record position arbitrarily designated when writing a record as the read start position, the read start position can be set according to the change contents of the record, so that the next search can be performed at high speed without missing records. It has features such as

According to the third aspect, by setting the designated record position as the read start position only when the instructed condition is satisfied, it is possible to set the read start position combined with various conditions. It has the features that the optimum read start position can be set according to it, the next read can be performed accurately without skipping a record, and unnecessary read can be performed at high speed.

According to the fourth aspect, the read start position of the record can be reset according to a fixed procedure, so that there is a feature that the read start position of the record can be surely set. According to the fifth aspect, the read start position of the record is set according to a certain procedure when a certain condition is satisfied, so that the read start position can be reset when the predetermined condition is satisfied, It has features such as unnecessary resetting of the read start position can be eliminated.

According to the sixth aspect, when the record writing is detected in the record writing detection procedure, a desired condition or procedure is selected from preset conditions and procedures, and the record is recorded according to the selected condition or procedure. Since the read start position can be set, the user can select the optimum condition or procedure according to the record contents and the record input order.

According to the seventh aspect, the condition and the procedure for setting the read start position of the record can be set when the record writing is detected in the record writing detecting procedure, and therefore it is necessary for each record writing. It has features that conditions and procedures can be set and processing can always be performed under optimal conditions.

According to the eighth aspect, the search start position for the next search can be newly set according to the change in the status provided corresponding to the record when the record content is updated.
When a record content change or record deletion occurs, this can be detected as a status change and a new search start position can be set. Therefore, the record position where the record content change, record deletion, etc., occurs again It has features such as being able to perform a search, being able to perform a search without causing omissions in the search, and being able to perform an accurate search.

According to the ninth aspect, since the record position arbitrarily designated according to the status change can be set as the search start position at the time of updating the record contents, the processing is started from the optimum position according to the status change. It has the feature that it can be processed quickly.

According to the tenth aspect, since the processing can be executed for each status by setting the search start position according to the status of the status, there is a feature that the processing according to the status can be executed. According to the eleventh aspect, by setting the search start position to the instructed position when a certain condition is satisfied, the search start position can be set when the predetermined condition is satisfied, which is unnecessary. It has features such as eliminating the need to re-set various search start positions.

According to the twelfth aspect, when the change of the search start position is instructed at the time of updating the record content, the search start position is set to the instructed position when the instructed position is before the current set position. Therefore, it is possible to prevent omission of record processing between the instructed position and the current set position.

According to the thirteenth aspect, since the change of the status is detected and the search start position is set, the search start position can be set according to the status, and the processing can be executed for each status. . According to claim 14,
Since the record search start position can be set according to a predetermined procedure, the record read start position can be set with certainty.

According to the fifteenth aspect, the constant condition and the constant procedure for setting the record search start position can be selected from preset conditions and procedures.
It has a feature that the user can select the optimum condition or procedure according to the record contents or the record input order.

According to the sixteenth aspect, since the fixed condition and the fixed procedure for setting the record search start position can be newly set, the necessary condition and procedure can be set when necessary, and the optimum condition is always obtained. It has features such as processing under various conditions.

[Brief description of drawings]

FIG. 1 is a block diagram of a first embodiment of the present invention.

FIG. 2 is a flowchart of a transaction processing unit according to the first embodiment of this invention.

FIG. 3 is a flowchart of a record detection unit according to the first embodiment of this invention.

FIG. 4 is a flowchart of a read start pointer management unit according to the first embodiment of this invention.

FIG. 5 is a flowchart of a read start pointer change processing unit according to the first embodiment of this invention.

FIG. 6 is an operation explanatory diagram of the first embodiment of the present invention.

FIG. 7 is a configuration diagram of an application example of the first embodiment of the present invention.

FIG. 8 is a block diagram of a second embodiment of the present invention.

FIG. 9 is a flowchart of a transaction processing unit according to the second embodiment of the present invention.

FIG. 10 is a flow chart of a record detection unit according to the second embodiment of the present invention.

FIG. 11 is a flowchart of a search result management unit according to the second embodiment of this invention.

FIG. 12 is a flowchart of a search start pointer change processing unit according to the second embodiment of this invention.

FIG. 13 is an operation explanatory diagram of the second embodiment of the present invention.

FIG. 14 is a configuration diagram of an application example of the second embodiment of the present invention.

FIG. 15 is a diagram illustrating an operation of an example of the related art.

FIG. 16 is an explanatory diagram of another example of the related art.

FIG. 17 is an operation explanatory diagram of another example of the related art.

[Explanation of symbols]

 1, 11 File 2, 12 Transaction Processing Section 3, 13 Record Detection Section 4, 14 Start Pointer Management Section 5, 15 Start Pointer Change Processing Section

Claims (16)

[Claims] 1. A record reading method for managing a read start position of a record filed in a file part and starting the read from the read start position at the time of read, the record write detecting detecting writing of the record to the file part. And a read start position setting procedure for newly setting a read start position for the next read when a record write to the file unit is detected in the record write detection process. Record read method. 2. The record read method according to claim 1, wherein the read start position setting procedure sets an arbitrarily designated record position as the read start position. 3. The read start position setting procedure sets a preset record position as a read start position only when a certain condition is satisfied at the time of writing the record detected by the record write detection procedure. The record reading method according to claim 1 or 2. 4. The record read method according to claim 1, wherein in the read start position setting procedure, the read start position of the record is reset according to a fixed procedure. 5. The record read method according to claim 1, wherein the read start position setting procedure sets a record read start position when a certain condition is satisfied. . 6. The read start position setting procedure, when a record write is detected in the record write detection procedure,
6. The record read according to claim 1, wherein a desired condition or procedure is selected from preset conditions and procedures, and a record read start position is set according to the selected condition or procedure. Method. 7. The read start position setting procedure can set a condition and a procedure for setting a record read start position when a record write is detected in the record write detection procedure. 6. The record reading method according to any one of 1 to 5. 8. A record storage unit that stores status information indicating a processing state of the record in order of occurrence, a record search unit that searches for a desired record in order of occurrence, and a predetermined transaction for the searched record. In a record search method of a computer system including a transaction processing unit that performs processing and updates the status information of the record, a status group is set based on the transition characteristic of the status information, and a head of a predetermined group is set by a record search instruction. When a record is searched, a search result management unit that stores the correspondence relationship between the group and the first record thereof is provided, and when the record search of a predetermined group is designated, the first record of the group is searched as the search result management unit. The first record obtained from the correspondence stored in the copy Then, the record belonging to the designated group is searched by referring to the status information, and when the first record belonging to the designated group is searched, the search result management unit updates the correspondence between the designated group and the first record, The record search method, wherein the search result management unit updates the correspondence between the designated group and the first record during the content update process. 9. The record according to claim 8, wherein the search result management unit sets a record position arbitrarily designated according to the change of the status as the search start position when updating the record content. retrieval method. 10. The record search method according to claim 8, wherein the search result management unit sets a record position according to the status of the status as a search start position when the record content is updated. 11. The search result management unit sets the search start position to a designated position when a certain condition is satisfied when updating the record contents.
10 to the record search method. 12. The search result management unit is instructed to perform a search start position when a change in the search start position is instructed when updating the record content and when the instructed position is before the current set position. 12. The record search method according to claim 8, wherein the record search method is set at a position. 13. The search result management unit detects a change in the status when updating the record content as one of the predetermined conditions.
The record search method described in 2. 14. The record search method according to claim 8, wherein the search result management unit sets the search start position of the record according to a predetermined procedure. 15. The record according to claim 14, wherein the search result management unit can select the fixed condition and the fixed procedure for setting a search start position of a record from preset conditions and procedures. retrieval method. 16. The record search method according to claim 14, wherein the search result management unit can newly set the certain condition and the certain procedure for setting a record search start position.