JP2010165170A - Retrieval processing method, system and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress increase of time required for retrieval processing even if a data amount of a retrieval target increases. <P>SOLUTION: A retrieval target data management device 102 extracts a classification target item of a storage storing retrieval target data and a value of the item from a retrieval expression corresponding to a retrieval demand received by a retrieval controller 101, and specifies a retrieval target data storage 104 set as the target by the retrieval expression based thereon. The retrieval controller 101 bundles the respective retrieval expressions corresponding to one or more retrieval demands by specified storage 104, and issues a retrieval request including identification information of the storage 104 and information of the bundled retrieval expressions by storage 104. A retrieval processor 105 executes each retrieval processing corresponding to each retrieval expression included in the request to the retrieval target data inside the storage 104 corresponding to the identification information included in the retrieval request based on the retrieval request. The retrieval controller 101 distributes a retrieval result from the retrieval processor 105 to each retrieval expression, and transmits the retrieval result to a demand source of each retrieval demand corresponding to each retrieval expression. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The disclosed technique relates to a technique for retrieving target information from a large amount of accumulated data.

  In order to prove business results and deal with troubles, in addition to the results of corporate activities, it is an era of collecting and accumulating various data related to the business, and it is important to quickly extract the target information from a large amount of data Yes.

  As a first conventional technique for searching for target information, data is stored in a relational database, and a query (search expression) in which a search condition is described is input from a requesting terminal device. Techniques for retrieving information are well known.

  As a second conventional technique for searching for target information, a technique described in Patent Document 1 below is known. In this technology, search requests issued from a plurality of request source terminal devices are accumulated for a certain period of time, and then the search conditions included in those search requests are bundled to perform a batch search on the search target data. . The search results are distributed to the individual search results for each of the plurality of search requests, and the individual search results are returned to the requesting terminal device.

JP 2002-222194 A

  In the first prior art described above, in the case of a search condition that can use an index created in advance, the search result can be quickly retrieved. However, in the case of a search condition in which an index cannot be used, the search is performed after reading all the stored data, so that the search time varies depending on the size of the stored data. As a result, there is a problem that the search time becomes longer as the size of stored data becomes larger.

  Further, if a search request is issued from a plurality of terminals at the same time, the search processing is performed in a multiplex manner, and a load is applied to the search device or disk I / O competes. As a result, as shown in FIG. 30 (a), a search request is waited in the search device, so that it takes a long time to search and it becomes difficult to predict the search time. Was.

  On the other hand, in the second prior art described above, as shown in FIG. 30B, even when search requests are issued from a plurality of terminals at the same time, the search conditions included in each search request are constant by the search device. A batch search is performed by bundling every hour. As a result, the search device can return the search results to a plurality of terminals with substantially the same response time as a search request from one terminal. That is, it is possible to extract results for a plurality of search conditions only by reading all data once (solid licking) for a large amount of search target data. Therefore, as shown in FIG. 30B, the search time can be predicted even if the multiplicity of search requests increases.

  However, in the second prior art, when the amount of data to be searched increases, as shown in FIG. 31, the time 3101 required for one full data read by the search processing device becomes longer. As a result, there has been a problem that the response time of the search device for the search request becomes long.

  In response to such a problem, as shown by 3201 in FIG. 32, the search target data is divided into a plurality of files and the search processing is performed in parallel, so that the response time can be shortened. However, when the search target data is divided into a plurality of files, the following problems occur.

  When the search request source wants to search only data in a specific range, the search control device searches for the search target data by specifying the file name or database of the search target data as shown as 3301 in FIG. The range of target data can be narrowed down. However, in order to make this designation, the search request source needs to know the search target data that the user wants to search. Therefore, when the amount of data to be searched increases, it becomes difficult to specify information to be searched from the search request source.

  Furthermore, since the technique according to the second prior art is used, as shown in FIG. 34, when the search control device performs a search by combining search conditions from a plurality of search request sources, the search target data is bundled. Multiple search expressions must be specified. As a result, as shown as 3401 in FIG. 34, a search with respect to an unrelated search formula is performed even on data that is not a search target, and the search processing is wasted.

  The problem to be solved by the disclosed technology is that, even when the amount of search target data increases, the search target data can be narrowed down and the search conditions can be specified efficiently, and the time required for search processing increases. It is to suppress.

  In order to solve the above problems, the disclosed technology receives a search request, generates a search request based on a search expression corresponding to the search request, and classifies the search requests into a plurality of search target data storages based on the search request. The search processing method, system, or program is executed on the assumption that the search processing is executed on the search target data stored in this manner, and the search result of the search processing is notified to the request source of the search request.

  The search target data specifying processing step extracts a classification target item and a value of the item used when the search target data is classified and stored in a plurality of search target data storages from a search formula corresponding to the search request, Based on the value of the classification target item, the search target data storage that is targeted by the search formula is specified.

  In the search request processing step, each search expression corresponding to one or more search requests is bundled for each search target data storage specified by the search target data specifying processing step corresponding to each search expression, and for each search target data storage A search request including the identification information and information of each search expression bundled is generated and issued.

  In the search processing step, each search process corresponding to each search expression included in the search request is performed on the search target data in the search target data storage corresponding to the identification information included in the search request based on the search request. Execute.

  The search result notification processing step distributes the search results obtained in the search processing step for each search formula and transmits the search results to the corresponding request sources.

According to the disclosed technology, it is possible to automatically narrow down the search range of the search target data, and the search requester does not need to know the search range of the search target data. Therefore, even if the amount of search target data increases, the search range of the search target data can be appropriately specified, and control can be performed so that useless search processing does not occur.

  In the search process, only the search process corresponding to the specified search target data is started. For this reason, the number of search processes can be greatly reduced, and the response time to the search requester can be greatly shortened.

It is a block diagram of embodiment. It is operation | movement explanatory drawing of the search object data registration process in a search object data management apparatus. It is a figure which shows the operation example of the search object data registration process in a search object data management apparatus. It is operation | movement explanatory drawing of the search object data specific process in a search object data management apparatus. It is a figure which shows the operation example of the search object data specific process in a search object data management apparatus. It is operation | movement explanatory drawing of the search process of the multi file by a search control apparatus. It is an operation | movement flowchart which shows the search object data registration process in a search object data management apparatus. It is a figure which shows the data structural example of search object data. It is a figure which shows the example of a data structure of a search object data location management book (when a search object data storage is a file). It is a figure which shows the data registration example of a search object data location management book (when a search object data storage is a file). It is a figure which shows the example of the search object data registration process in a search object data management apparatus (when search object data storage is a CSV file). It is a figure which shows the example of the search object data registration process in a search object data management apparatus (when search object data storage is an XML file). It is an operation | movement flowchart which shows the search object data specific process in a search object data management apparatus. It is a figure (example 1 when a search object data storage is a file) which shows the example of a query to a search object data location management book. It is a figure (example 2 when a search object data storage is a file) which shows the example of a query to a search object data location management book. It is a figure which shows the example of the search object data specific process in a search object data management apparatus (when search object data storage is a CSV file). It is a figure which shows the example of the search object data specific process in a search object data management apparatus (when search object data storage is an XML file). It is an operation | movement flowchart which shows the registration process of the search expression for every search object file in a search control apparatus. It is operation | movement explanatory drawing of the search formula bundling process in a search control apparatus. It is a figure which shows the example of a data structure used for the search expression bundling process in a search control apparatus. It is operation | movement explanatory drawing of the search request process with respect to a search processing apparatus. It is an operation | movement flowchart which shows the merge process of the output file of the search result in a search control apparatus. It is operation | movement explanatory drawing of the notification process of the search result to a search request source. It is a figure which shows the example of a data structure of a search object data location management book (when a search object data storage is a database). It is a figure which shows the data registration example of a search object data location management book (when a search object data storage is a database). It is a figure which shows the example of the search object data registration process in a search object data management apparatus (when search object data storage is a database). It is a figure which shows the example of a query to a search object data location management book (when a search object data storage is a database). It is a figure which shows the example of the search object data specific process in a search object data management apparatus (when a search object data storage is a database). It is a figure which shows the hardware constitutions of the computer system which can run the program based on embodiment. It is explanatory drawing of a prior art. It is explanatory drawing (the 1) of the problem of a prior art. It is explanatory drawing (the 2) of the problem of a prior art. It is explanatory drawing (the 3) of the problem of a prior art. It is explanatory drawing (the 4) of the problem of a prior art.

Hereinafter, embodiments will be described in detail.
FIG. 1 is a configuration diagram of the embodiment.
The present embodiment includes a search control device 101, a search target data management device 102, a search target data location management list 103, a search target data storage 104, and a search processing device 105.

First, the operation of the search target data management apparatus 102 will be described.
The device 102 executes the following two processes.
・ Search target data registration process ・ Search target data identification process

First, search target data registration processing executed in the search target data management apparatus 102 will be described with reference to an operation explanatory diagram shown in FIG.
(1) Classification target items and classification conditions are defined in advance as classification condition definitions for classifying data. A plurality of classification condition definitions can be defined.
(2) Thereafter, registration target data to be registered is analyzed, and it is confirmed whether the classification target item in the classification condition definition is included in the registration target data.

(3) When the classification target item is included in the registration target data, the value of the classification target item is acquired from the registration target data. Based on the classification condition in the classification condition definition including the classification target item and the acquired value, the search target data storage 104 for storing the registration target data is determined, and the registration target data is stored in the storage. Stored. In the example of FIG. 2, registration target data is stored in C of the search target data storage 104. If the search target data storage 104 corresponding to the classification target item and the classification condition does not exist, the new storage is created and the registration target data is stored therein. Here, the search target data storage 104 may be a classified data storage unit that manages a plurality of files such as a classified table in a database, a folder, or a directory, or may be a file. In the case of a file, the CSV format or the XML format can be adopted as the file format, and registration target data is additionally written in each classified file.
(4) When the registration target data does not include any classification target item in any classification condition definition, the registration target data is stored in the search target data storage 104 corresponding to the absence of the classification target.

  In the search target data registration process described above, refer to the search target data location management list 103 for the analysis process of the registration target data in the process (2) and the determination process of the search target data storage 104 in the process (3). Can do. This will be described later.

  FIG. 3 is a diagram illustrating an operation example of search target data registration processing. First, registration target data can be described in, for example, an XML document format in which data values are delimited by tags. A plurality of classification condition definitions such as classification condition definition 1 and classification condition definition 2 can be defined. In the classification condition definition 1, the classification target item is “/ document / area”, and it is defined as “classified into specified value units” as the classification condition. In the classification condition definition 2, the classification target item is “/ document / date”, and it is defined that “classify by month” as the classification condition. For such a classification condition definition, as the search target data storage 104, the classification target item “/ document / area” is classified for each designated value representing a region, and the classification target item “/ document / date” is A plurality of hangars categorized for each month may be created. Each hangar may be a CSV or XML file, or a database classified table. Now, for example, it is assumed that registration target data in which the specified value of the classification target item “/ document / area” is “Kanto” and the specified value of the classification target item “/ document / date” is “20080202” is registered. In this case, the classification target item “/ document / area” is classified as “Kanto”, and the classification target item “/ document / date” is classified as “February 2008”. Data is registered.

Next, search target data specifying processing executed in the search target data management apparatus 102 will be described with reference to an operation explanatory diagram shown in FIG.
First, in the search request from the search requester (request source terminal device), only the search formula is described and the search request is issued. In FIG. 4, the search request is described as “condition 1)”.

Upon receiving the search request, the search control apparatus 101 designates a search formula and requests the search target data management apparatus 102 to specify the search target file.
Upon receiving the request, the search target data management apparatus 102 executes the following search target data specifying process.
(1) A search expression is received from the search control apparatus 101. In FIG. 4, for example, condition 1) is received.
(2) The syntax of the received search expression is analyzed, and it is confirmed whether the search expression includes a classification target item in the classification condition definition defined in advance prior to the search target data registration process described above. .

  (3) When the classification target item is included in the search formula, the value of the classification target item is acquired from the search formula. Then, based on the classification condition in the classification condition definition including the classification target item and the acquired value, the search target data storage 104 to which the value belongs is specified. The identification information for specifying the search target data storage 104 may be a file name, a table name in the database, a folder name, or a directory name. In FIG. 4, for example, among the search target data storages 104 of A to Z, the search target data storages 104 of A, B, and C are specified.

(4) The identification information of the specified search target data storage 104 is returned to the search control apparatus 101. In FIG. 4, for example, the identification information is A, B, and C.
(5) When the classification target item in any classification condition definition is not included in the search formula, information indicating all search target data as a search target is returned to the search control apparatus 101.

  In the search target data specifying process described above, the search target data location management list 103 may be referred to for the search expression analysis process in the process (2) and the search target data storage 104 specifying process in the process (3). it can. This will be described later.

  In response to the search target data specifying process by the search target data management apparatus 102, the search control apparatus 101 executes the following process. That is, the search control apparatus 101 performs a search request for searching only the search target data storage 104 corresponding to the identification information A, B, C returned from the search target data management apparatus 102 as a search process for the condition 1). Issued in parallel to the search processing device 105 corresponding to the hangar. At this time, as will be described later, the search control apparatus 101 executes a process of requesting a batch search by bundling search formulas from a plurality of search requesters. As a result, the search control apparatus 101 collects the search results corresponding to the condition 1) among the search results returned from each search processing apparatus 105, generates a result 1), and returns it to the search requester.

  Here, the search target data storage 104 to be searched is divided into a plurality of files, database tables, folders, or directories. Thereby, even if the data amount of search object data increases, the increase in the response time accompanying data reading can be suppressed.

  In addition, the search control apparatus 101 automatically determines the search range of the search target data by receiving from the search target data management apparatus 102 identification information of the search target data to which the item value in the search expression to be issued a search request belongs. Can be narrowed down. In the example of FIG. 4, in the search process for condition 1), the search range 401 of only A, B, and C is specified among the search target data of A to Z. In this case, the search requester does not need to know the search range 401 of the search target data. Therefore, even if the data amount of the search target data increases, the search range of the search target data can be appropriately specified, and useless search processing does not occur.

  Further, the search control device 101 does not need to start the search processing device 105 corresponding to all the search target data, and in the case of the example of FIG. 4, the search processing device 105 corresponding to the search target data A, B, C. Just start up. Therefore, the number of search processes executed by one search processing device 105 in one transaction can be greatly reduced, and the response time from the search control device 101 to the search requester can be greatly shortened.

  FIG. 5 is a diagram illustrating an operation example of the search target data specifying process. Assume that the search formula shown in FIG. 5 is issued from the search requester. This search formula has a date (/ document / date) before “January 1, 2008” to “March 1, 2008”, and the region (/ document / area) is “Kanto”. A search for data to be searched whose store number (/ document / shopid) is “12345” is requested. This search formula is compiled at regular intervals by the search control apparatus 101 and notified to the search target data management apparatus 102. The search target data management apparatus 102 executes the search target data specifying process described above. As a result, the classification target item “/ document / area” is classified as “Kanto”, and the classification target item “/ document / date” is classified as “January 2008” and “February 2008”. Identification information of the hangar 104 is obtained. That is, in FIG. 5, identification information “kanto — 200801” and “kanto — 200802” are obtained. Such identification information is returned from the search target data management apparatus 102 to the search control apparatus 101. As a result, the search control apparatus 101 need only specify the search processing apparatus 105 for the two search target data storages 104 corresponding to the identification information for the search formula.

FIG. 6 is an operation explanatory diagram showing multi-file search processing executed by the search control apparatus 101 of FIG.
The search control apparatus 101 is realized as a resident program on a server computer, for example, and executes the following series of processes.

  (1) First, the search control apparatus 101 receives a search request from a search requester. The search control apparatus 101 can accept search requests from a plurality of search requesters. In the example of FIG. 6, the search control apparatus 101 specifies a search request specifying condition 1) from search requester 1, a search request specifying condition 2) from search requester 2, and a condition 3) from search requester 3 Each search request is accepted. In the search request from each search requester, one search expression may be requested, or a plurality of search expressions describing completely different search conditions may be requested in a lump.

  (2) Next, the search control device 101 requests the search target data management device 102 to perform search target data specifying processing for each search formula included in each received search request. As a result, in the example of FIG. 6, the search target data ranges of A and B for the search formula of condition 1), B and C for the search formula of condition 2), and Z for the search formula of condition 3) are obtained. Each piece of identification information shown is returned from the search target data management device 102 to the search control device 101.

  (3) The search control apparatus 101 accumulates each search request from the search requester and identification information corresponding to each search request for a certain period of time. In the example of FIG. 6, each search request corresponding to the conditions 1, 2, and 3 is accumulated. Here, there may be no waiting time (0 seconds) for a certain period of time. The waiting time can be configured to be changed in the search control apparatus 101.

  (4) The search control apparatus 101 bundles the search formulas of the search requests that have been received so far after a certain period of time for each search target file identified by the identification information (integration of search requests). In the example of FIG. 6, condition 1) for search target data A, conditions 1) and 2) for search target data B, condition 2) for search target data C, and condition 3 for search target data Z are shown. ) Are bundled together. Here, in search expression bundling processing (integration processing of search requests), a plurality of search expressions are not combined into a single search expression, but a plurality of search expressions are held as they are and searched. A search request is issued to the processing device 105.

  (5) The search control device 101 causes each search processing device 105 to search for the search target file name, a plurality of search requests (search formulas) bundled for the file, and the search requests. A search request including the output file name to output the result is issued. In the example of FIG. 6, one search processing device 105 is assigned to each of the search target data A, B, C, and Z. However, when the number is insufficient, one search processing device 105 is allocated. A plurality of requests may be sequentially issued to 105.

(6) The search control apparatus 101 waits for a response to the search result from each search processing apparatus 105.
(7) The search control apparatus 101 receives search results for a plurality of search requests bundled from each search processing apparatus 105, and distributes the search results for each search request for each search request.

(8) When the search request is for a plurality of search target data (for example, files), the search control apparatus 101 waits until search results from all the search target data are obtained. Then, the search control apparatus 101 notifies the search requester of the search results when the search results are prepared for each search request. In the example of FIG. 6, the search results 1), 2), and 3) are returned for the search requests of the conditions 1), 2), and 3).

Next, the operation of the search processing device 105 in FIG. 1 will be described.
Only one search processing apparatus 105 may operate for the search control apparatus 101, or a plurality of search processing apparatuses 105 may operate. Further, the server computer on which the search processing apparatus 105 operates may be the same server computer as the search control apparatus 101 or may be different.

The search processing device 105 executes the following series of processes.
(1) The search processing device 105 receives a search request from the search control device 101.
(2) The search processing device 105 is based on search request information (a search target file name, one or more search requests (search formulas), and an output file name to which a search result is output corresponding to each search request). The search process is executed on the search target data.
(3) When the search processing is completed, the search processing device 105 transmits the output file name that outputs each search result for a plurality of search requests to the search control device 101.
(4) The search processing device 105 waits for the next search request from the search control device 101.

  FIG. 7 shows a case where the search target data registration process by the search target data management apparatus 102 of FIG. 1 described above with reference to FIGS. 2 and 3 is realized as a process in which the server computer executes the search target data registration process program. It is an operation | movement flowchart which shows operation | movement. FIGS. 8 to 10 are diagrams showing an example of the configuration of each data used in this operation flowchart. Further, FIG. 11 and FIG. 12 are diagrams showing processing examples according to this operation flowchart. Hereinafter, the details of the search object data registration process will be described with reference to these drawings.

  First, FIG. 8 is a diagram illustrating a data configuration example of data to be registered as search target data. FIG. 8A shows an example of an XML format data configuration, and FIG. 8B shows an example of a CSV format data configuration. In the embodiment, both data configurations can be registered.

  Next, FIG. 9A is a diagram illustrating a data configuration example of the classification condition definition defined in advance, which is included as a part of the search target data location management book 103 (see FIG. 1). This data configuration example is an example in which the search target data storage 104 is configured as a file. As described above, a plurality of classification condition definitions can be defined. The data configuration example shown in FIG. 9A corresponds to one classification condition definition.

In FIG. 9A, the “classification condition definition name” is set with the name of the classification condition definition.
In the “data format”, the data format of the search target data to be registered is set. Here, either “CSV” or “XML” format can be set.

  In the “type of storage”, the data type of the search target data storage 104 is set. In the example of FIG. 9A, “FILE” is set because the search target data storage 104 is realized as a file. As in the case of FIG. 24 described later, when the search target data storage 104 is realized as a database table, “DB” is set.

A plurality of sets of “classification condition item name”, “classification target item”, and “classification condition” can be designated.
In “classification condition item name”, the name of the classification condition is set.
In “classification target item”, an item to be classified in the registered data is set. Here, when “Data Format” is “CSV”, the item name is set, and “Data Format” is “X”.
In the case of “ML”, an XML path is set.

  In the “classification condition”, a condition for determining a storage location where registered data is stored is specified. For example, “yearly”, “monthly”, “dayly” or the like can be set as a classification condition for specifying a period by date data. In addition, as a classification condition for specifying a range, for example, a range can be set with a maximum value and a minimum value. When “value specification” is set as the classification condition, the classification is performed based on the value of the classification item.

  Next, based on the classification condition definition as shown in FIG. 9A, the classification target item management table included as a part of the search target data location management list 103 (see FIG. 1) is shown in FIG. It is created with the data structure as shown in In practice, this table is defined as a database table.

  In FIG. 9B, items corresponding to the “classification target item” and “classification condition” in FIG. 9A are set as 901 and 902, for example. For example, when “classification condition” is “value specification”, a column for storing a value is defined as 901. In 901 of FIG. 9B, a column “region” is defined. When the “classification condition” is a period designation or a range designation, each column for storing a minimum value and a maximum value indicating the range is defined as 902. In FIG. 9B 902, columns of “date_MIN” and “date_MAX” are defined. Furthermore, as “903”, one “search target file” column for storing the search target file name is always defined.

  Using the search target data location management list 103 (classification condition definition and classification target item management table) created as described above, search target data registration processing is executed based on the operation flowchart of FIG.

  First, one record of registered data is read (step S701 in FIG. 7). The data to be registered is, for example, data in the CSV format as shown in FIG. 10, and a plurality of registration records can be included in one file.

Next, the record of the registration data is analyzed (step S702 in FIG. 7).
As a result of the above analysis, it is determined whether or not a classification target item (see FIG. 9A) in the classification condition definition exists in the data record (step S703 in FIG. 7).

  If a classification target item exists in the data record and the determination in step S703 is YES, the value of the classification target item is acquired from the data record. Then, based on the classification condition in the classification condition definition including the classification target item and the acquired value, to the classification target item management table (see FIG. 9B) in the search target data location management book 103. Is made (step S704 in FIG. 7).

  Then, as a result of this inquiry, it is determined whether or not a record can be searched on the classification target item management table, that is, whether or not the search target data storage 104 corresponding to the value of the classification target item in the data record exists. (Step S705 in FIG. 7).

  The registration contents of the classification target item management table are as shown in FIG. 10, for example, as search target data registration proceeds. The registration contents of this table include “Hokkaido”, “Kanto”, “Chubu”, and “January 1, 2008 to January 31, 2008”, “February 01, 2008 to February 29, 2008”. ”Indicates that the search target files that are the search target data storages 104 are created and classified for each month. As for the registration contents of this table, there is no registration record in the initial state, but as the search target data to be registered progresses, the corresponding records are sequentially registered as the search target file is created. Go.

  Now, for example, as data to be registered in the CSV file format, a data record “20080101”, “Hokkaido”, “00126”, “Otaru store”,... To do. In step S704 in FIG. 7, “Hokkaido” is extracted as the value of the classification target item “area” and “20080101” is extracted as the value of the classification target item “date” for the registration target data. Next, the classification condition definition in FIG. As a result, in step S704 of FIG. 7, for example, an inquiry is executed on the classification target item management table shown in FIG. 10 under the following conditions.

The value of the “region” column is equal to the value “Hokkaido” of the classification target item “area”. The value of the “date_MIN” column is less than or equal to the value “2008011” of the classification target item “date”.
In addition, the value of the “date_MAX” column is greater than or equal to the value “20080101” of the classification target item “date”.

  As a result, for example, the first record on the classification target item management table of FIG. 10 is extracted, and “hokaido_200801.csv” is set as the search target file name of the search target data storage 104 from the “search target file” column of the record. can get. Accordingly, the determination in step S705 of FIG. 7 is YES, and the registration target data is added to the search target file (search target data storage 104) corresponding to the search target file name (step S708 of FIG. 7). . If the search target file that is the search target data storage 104 is in the CSV format, the registration target data is added to the search target file “hokaido_200801.csv” in the format shown in FIG. 11, for example.

  The registration target data and the search target file (= search target data storage 104) may be in the XML format. In this case, for example, the registration target data in the XML format is in the XML format as shown in FIG. Is added to the search target file.

  On the other hand, if no record is extracted on the classification target item management table as a result of the inquiry in step S704 in FIG. 7, the determination in step S705 in FIG. 7 is NO. As a result, a search target file that is the search target data storage 104 corresponding to the value of the categorized target item that has been queried is newly created (step S706 in FIG. 7).

  Subsequently, as shown in FIG. 10, a record corresponding to the value of the categorized target item for which the inquiry has been made is registered in the categorized target item management table. The file name of the search target file that is the newly created search target data storage 104 is registered (step S707 in FIG. 7).

Then, registration target data is added to the newly created search target file (step S708 in FIG. 7).
In the determination in step S703 of FIG. 7 described above, if the classification target item does not exist in the data record and the determination is NO, all search target files corresponding to no classification target created in advance ( The registration target data is stored in (all search target data storage) (step S709 in FIG. 7).

The above-described series of processing from step S701 to S709 in FIG. 7 is repeatedly executed by the number of data records of registration target data (step S710 in FIG. 7).
FIG. 13 shows a case where the search target data specifying process by the search target data management apparatus 102 of FIG. 1 described above with reference to FIGS. 4 and 5 is realized as a process in which the server computer executes the search target data specifying process program. It is an operation | movement flowchart which shows operation | movement. FIG. 14 and FIG. 15 are diagrams showing an example of the configuration of each data used in this operation flowchart. Further, FIGS. 16 and 17 are diagrams showing a processing example according to this operation flowchart. The details of the search target data specifying process will be described below with reference to these drawings.

In this process, the search target data location management list 103 (classification condition definition and classification target item management table) created in the search target data registration process described above is used.
First, the search expression delivered from the search control device is analyzed (step S1301 in FIG. 13) (see FIG. 16).

As a result of the above analysis, it is determined whether or not a classification target item (see FIG. 9A) in the classification condition definition exists in the search formula (step S1302 in FIG. 13).
If there is a classification target item in the search formula and the determination in step S1302 is YES, the value of one classification target item is selected from the search formula. Then, based on the classification condition in the classification condition definition (see FIG. 9A) including the classification target item and the acquired value, the classification target item management table (in the search target data location management book 103) ( Inquiry conditions for making an inquiry to (see FIG. 9B) are determined (step S1303 in FIG. 13).

  In accordance with the determined inquiry condition, an inquiry is executed on the classification object item management table constituting the search object data location management list 103 (step S1304 in FIG. 13) (see FIG. 16).

As a result, a hit flag is set corresponding to the record extracted from the classification target item management table (step S1305 in FIG. 13).
Now, for example, when a search expression as shown in FIG. 14A is processed, the values of the classification target items extracted from the search expression and the query conditions are as shown in FIG. 14B. Further, as shown in FIG. 14C, the value “20080221” of one classification target item “date” is selected from the search formula, and the query condition “date_MAX is“ 20080221 ”or more” is correspondingly selected. It is determined. Based on this query condition, a query is executed on the classification target item management table as shown in FIG. As a result, as shown as the determination result A in FIG. 14C, a hit flag is set for the record extracted from the classification target item management table. A record with a “◯” mark in the determination result A is a record for which a hit flag is set. The hit flag can be set as array data on the on-memory, for example.

The series of processing from step S1303 to S1305 in FIG. 13 is repeatedly executed for the number of classification target items selected from the search formula.
In the example of FIG. 14C, the same hit flag setting process as in the case of the determination result A is executed in the order of the determination result B and the determination result C.

  After the setting of the hit flag for all the classification target items in the search formula is completed, the hit flag for each classification target item is determined based on the logical operator of the search formula by a series of processing in steps S1307 to S1312 in FIG. Is done. Thereby, the process for specifying the search object file which is the search object data storage 104 targeted by the search expression is executed.

First, determination logic between hit flags for each classification target item is generated based on logical operators of a plurality of classification target items in the search formula (step S1307 in FIG. 13). For example, in the example of FIG. 14, hit flags are all set in the determination results A, B, and C shown in FIG. 14C based on the logical operators in the search expression shown in FIG. Decision logic is generated with the record as a specific record. That is, as shown in FIG. 14 (c), in the determination results A, B, and C, there are records for which “◯” marks are all set (records for which “◯” marks are set in the “Result” column). Will be selected as a specific record.

After the determination logic is generated, one record in which at least one hit flag is set is extracted (step S1308 in FIG. 13).
Next, the determination logic generated in step S1307 is performed on each hit flag set in the extracted record (step S1309 in FIG. 13).

As a result, it is determined whether or not the extracted record is a record to be specified (step S1310 in FIG. 13).
If the determination in step S1310 is YES, the extracted record is stored as a specific record (step S1311 in FIG. 13).
If the determination in step S1310 is NO, the extracted record is not a specific record.

  The series of processing from step S1308 to S1311 is executed for each record while sequentially extracting records in which at least one hit flag is set (step S1312 in FIG. 13).

  As a result of the above processing, each search target file name registered in the “search target file” column of each record finally stored as a specific record is retrieved from the search target data management apparatus 102 of FIG. 101 is notified, and the search target data specifying process ends (step S1313 in FIG. 13) (see FIG. 16).

  In the example of FIG. 14C, two search target files “kanto_200802.csv” and “kanto_200803.csv” are finally identified and notified to the search control apparatus 101 of FIG.

  FIG. 15 is a diagram illustrating an example of a query for a more complicated search expression. In this example, with respect to the search formula shown in FIG. 15A, the values of the classification target items extracted from the search formula and the query conditions are as shown in FIG. 15B. Thus, as shown in FIG. 15C, hit flags are set for the respective determination results A to E. The determination logic of the hit flag based on the search expression is that E of all the records with the A and B hit flags (◯) or the records with all the C and D hit flags (◯). A record with a hit flag (◯) is a specific record. Therefore, “kanto — 200801.csv” and “kanto — 200803.csv” are notified as search target file names from these specific records.

  FIG. 16 shows the flow of the above process when the search target file in the search target data storage 104 is configured as a CSV file. On the other hand, when the search target file is an XML file, as shown in FIG. 17, only the description method of the classification target item is described by the XML method, and the basic processing flow is that of the CSV file. Same as the case.

Note that, as a result of the search expression analysis processing in step 1301 of FIG. 13, it is determined that the search expression does not include the classification target item in the classification condition definition, and the determination in step S1302 is NO. The following processing is executed. That is, the search control apparatus 101 is notified of the names of files to be searched, and the search target data specifying process ends (step S1314 in FIG. 13).

  FIG. 18 is an operation flowchart showing an operation when a part of the multi-file search process by the search control apparatus 101 of FIG. 1 described above with reference to FIG. 6 is realized as a process in which the server computer executes the search process program. It is. FIG. 19 is a diagram showing a configuration example of each data used in this operation flowchart. Further, FIG. 18 is an operation explanatory diagram of this operation flowchart. The details of the search process will be described below with reference to these drawings.

  First, although not particularly illustrated, as described above with reference to FIG. 6, the search control apparatus 101 can sequentially receive search requests from a plurality of search requesters. Then, the search control apparatus 101 requests the search target data management apparatus 102 to perform search target data specifying processing for each search formula included in each received search request. As a result, the search target file name to be searched is returned from the search target data management apparatus 102 to the search control apparatus 101.

  The search control apparatus 101 holds a request source information management table shown in FIG. 20A for each search request. This table includes, for each search request, a request source ID for specifying a request source (terminal device), a pointer to a storage address of a search expression, and the search request returned from the search target data management device 102. A pointer to the storage address of the file name to be searched is held. The request source information management table can be realized as structure data, for example.

  As a result, as shown in the upper half of FIG. 19, the search control apparatus 101 uses the request source information management table for each search request source and one or more returned from the search target data management apparatus 102. The search target file name is managed.

While executing the above management, the search control apparatus 101 further executes the program of the operation flowchart shown in FIG.
As described above, the search control apparatus 101 accumulates the search requests from each search request source in the format of the request source information management table every certain time.

Therefore, it is first determined whether or not a search request has been accumulated for a certain period of time, particularly using means such as a timer (not shown) (step S1801 in FIG. 18).
When a certain time has elapsed, a search request search expression for accessing each search target file is bundled for each search target file by a series of processes in steps S1802 to S1808 in FIG.

  In order to realize this, the search control apparatus 101 holds a search processing management table shown in FIG. In this table, for each search target file, a pointer to information (search request information) related to a search request that uses the search target file as a search target is held. The search request information is further indicated by pointers to the same address as the storage address of the search formula referenced from the request source information management table of FIG. 20A ("circled number 1" to "circled number 3" in the figure). ), Holds the output file name where the search results are output. The search processing management table and the search request information can be realized as structure data, for example.

  With this search processing management table, as shown in the lower half of FIG. 19, a search request search expression that uses the search target file as a search target is linked and bundled for each search target file.

Hereinafter, a series of processes in steps S1802 to S1808 in FIG. 18 for realizing this bundling process will be described.
First, by referring to the request source information management table (FIG. 20A) corresponding to one search request that has been accumulated for a certain period of time, one search target file name corresponding to the search request is acquired. (Step S1802 in FIG. 18).

Next, the search processing management table (FIG. 20B) is searched based on the acquired search target file name (step S1803 in FIG. 18).
Then, it is determined whether or not a search processing management table corresponding to the search target file name exists (step S1804 in FIG. 18).

  If no search processing management table corresponding to the search target file name exists and the determination in step S1804 is NO, a search processing management table corresponding to the search target file name is created (step S1805 in FIG. 18). ).

  Subsequently, in step S1806 of FIG. 18, search request information is newly created and combined with the search processing management table that has been successfully searched in step S1803 or the search processing management table that has been newly created in step S1805. (See FIG. 20 (b)). In this search request information, a pointer to the storage address of the search expression of the search request currently being processed and the output file name to which the search result is output are set. The pointer can be obtained from the request source information management table (FIG. 20A) of the search request currently being processed.

With the above processing, for one search target file selected in the search request currently being processed, the search expression of the search request currently being processed is bundled in the search target file.
Next, returning to the process of step S1802 in FIG. 18, the request source information management table (FIG. 20A) corresponding to the search request currently being processed is referred again, so that another request corresponding to the search request is returned. The search target file name is further acquired. Then, the process of bundling the search expressions of the search request currently being processed is executed in the same manner as described above for the newly selected search target file. Such a process is repeatedly executed until all the search target files are selected on the request source information management table (FIG. 20A) corresponding to the search request currently being processed (step S1807 in FIG. 18). ).

  In this way, when the process of bundling the search requests for each search target file used for one search request is completed, the process returns to the process of step S1802 in FIG. The same processing as described above is executed for the search request. Such a process is repeatedly executed until it is processed for all the search requests accumulated for a certain time (step S1808 in FIG. 18).

When the processing is completed for all the search requests accumulated for a certain period of time, the process returns to the waiting process for a certain period of time after step S1801 in FIG.
With the above processing, for example, in FIG. 19, only the search expression of the request source 1 is bundled in the search target file “D: \ data \ kanto — 200801.csv”. In addition, only the search expression of the request source 2 is bundled in the search target file “D: \ data \ kanto — 200803.csv” or the like. On the other hand, in the search target file “D: \ data \ kanto — 200805.csv”, the search expressions of the request source 2 and the request source 3 are bundled.

  As described above, when the bundling process for each search target file is completed in the search control apparatus 101 of FIG. 1, the search control apparatus 101 sequentially refers to the search process management table of FIG. As a result, the search control apparatus 101 sequentially receives a search request including a search target file name, one or more search expressions bundled for the file, and an output file name that outputs a search result for each search expression. And sequentially issue them to the appropriate search processing device 105 (FIG. 1).

  When the search control device 101 and the search processing device 105 are executed on the same server computer, the search formula is notified in the search request by notifying pointer information to the storage address of each search formula. It may be realized as follows.

  As a result, for example, as shown in FIG. 21, for the search processing device 105 of # 1, the search target file name “D: \ data \ kanto — 200801.csv”, the search formula of the request source 1, and the request source A search request including the output file name for 1 is issued. In response to this, the search processing device 105 # 1 searches for the search target file “D: \ data \ kanto — 200801.csv”, which is the search target data storage 104 (FIG. 1), using the search formula of the request source 1. Execute. Then, the search processing apparatus 105 of # 1 outputs the search result by the search formula of the request source 1 to the search result file having the output file name designated for the request source 1.

  Similarly, for the search processing device 105 of # 2, the search target file name “D: \ data \ kanto — 200802.csv”, the search formula of the request source 1, the search formula of the request source 2, and the request source 1 and A search request including each output file name for 2 is issued. In response to this, the search processing device 105 # 2 performs a search process based on the search formula of the request source 1 and a search process based on the search formula of the request source 2 on the search target file “D: \ data \ kanto — 200802.csv”. Run each one. Then, the search processing device 105 of # 2 outputs the results of the respective searches by the respective search expressions of the request sources 1 and 2 to the search result files of the respective output file names respectively designated for the request sources 1 and 2. To do.

  Here, in the example of FIG. 21, four search requests (gray portions in FIG. 21) are currently issued to the four search processing devices 105 from # 1 to # 4. Consider a case where a search request is waiting to be issued. In this case, when a new search request (request sources 4 and 5 in FIG. 21) is generated for the search target files corresponding to the fifth and subsequent search requests, those new search requests are added to the search target file. It is possible to further bundle the search formula for the request. Thereby, the waiting time for the search process can be further reduced.

When the search is completed normally, the search processing device 105 returns the file name of the search result file to the search control device 101 that issued the search request.
As described above, the search control apparatus 101 receives search results for a plurality of search requests bundled from each search processing apparatus 105, and distributes the search results for each search request for each search request. At this time, if the search request is for a plurality of search target files, the search control apparatus 101 waits until search results from all the search processing apparatuses 105 are obtained. Then, the search control apparatus 101 notifies the search requester of the search results when the search results are prepared for each search request.

  In this case, the request source terminal device or the search control device 101 responds to a single search request with information (output file name, etc.) on each search result file storing each search result for a plurality of search target files. Can be specified to be notified to the requesting terminal device as it is. Alternatively, the request source terminal device or the search control device 101 can specify that each search result file is merged into one output file and the merged output file name is notified.

FIG. 22 is an operation flowchart showing the operation of the search control apparatus 101 that realizes the above-described output file merging process. FIG. 23 is a diagram for explaining the operation.
First, for each search request, the search control apparatus 101 repeatedly executes the operation of receiving the search result file name from the search processing apparatus 105 for the number of search target files targeted by the search request (step S2201 in FIG. 22). And S2202).

  Next, the search control apparatus 101 determines whether or not to merge the search result files (to one file) (step S2203 in FIG. 22). For example, the requesting terminal device can specify whether or not to specify merge at the time of a search request, or the server administrator can set whether or not to specify merge as a default setting.

  If the merge specification is not made and the determination in step S2203 is NO, the search control apparatus 101 responds to the search request with search result file names created for the number of search target files targeted by the search request. The request is transmitted as it is (step S2208 in FIG. 22). The requesting terminal device accesses the search result file stored on the network folder or the like based on the notified search result file name.

  In the example shown in FIG. 23, in response to the search request from the request source 1, the search result device 105 from the search processor # 1 sends a search result file name “result_request source 1_processing device 1. csv "is returned (FIG. 23-1)). Similarly, “result_request source 1_processing device 2.csv” is returned as the search result file name from the search processing device 105 of # 2. If the merge specification is not made for the request source 1, the search control apparatus 101 returns the two search result file names as search results for the search request from the request source 1 (FIG. 23-2)).

  On the other hand, if merge is specified and the determination in step S2203 is YES, the search control apparatus 101 first creates a new merge file in response to the search request (step S2204 in FIG. 22).

Next, the search control apparatus 101 repeatedly executes the operation of outputting (additional writing) information of each search result file to the merge file for the number of search target files targeted by the search request (step in FIG. 22). S2205 and S2206).
Then, after deleting each search result file, the search control apparatus 101 transmits the file name of the merge file to the request source (step S2207 in FIG. 22).

In the example shown in FIG. 23, in response to the search request from the request source 2, the search processing device 105 from # 2 sends a search result file name “result_request source 2_processing device 2. csv "is returned. Similarly, “result_request source 2_processing device 3.csv” is returned as the search result file name from the search processing device 105 # 3. If the merge specification is made for the request source 2, the search control apparatus 101 executes the merge process on the two search result files (3 in FIG. 23). Then, the search control apparatus 101 generates a merge file “result_request source 2.csv” and returns the file name as a search result for the search request from the request source 2 (FIG. 23-4)).
As described above, the search result can be processed and returned according to the request of the requester or the operator.

  24 shows the data structure of the classification condition definition (FIG. (A)) and the classification target item management table (FIG. (B)) when the search target data storage 104 of FIG. 1 is configured as a database table. It is a figure which shows an example.

The basic configuration is the same as the case of FIGS. 9A and 9B in the case where the search target data storage 104 is configured by a file. The difference is that the type of hangar in the classification condition definition is not “FILE” but “DB” representing a database, and the classification target item management table has a “search target table” column instead of a “search target file” column. This is the point to be set.

  The operation of the search target data registration process shown in FIGS. 25 and 26 is the same as the case of FIGS. 10 and 11 except that the search target data storage 104 is not a file but a database table. The operation is based on the operation flowchart of FIG.

  Further, the operation of the search target data specifying process shown in FIGS. 27 and 28 is the same as the case of FIGS. 14 and 16 except that the search target data storage 104 is not a file but a database table. The operation is based on the operation flowchart of FIG.

FIG. 29 is a diagram illustrating an example of a hardware configuration of a computer that can implement the search system according to the above-described embodiment.
The computer shown in FIG. 29 includes a CPU 2901, a memory 2902, an input device 2903, an output device 2904, an external storage device 2905, a portable recording medium driving device 2906 into which a portable recording medium 2909 is inserted, and a network connection device 2907. These have a configuration in which they are connected to each other by a bus 2908. The configuration shown in the figure is an example of a computer that can implement the above system, and such a computer is not limited to this configuration.

  The CPU 2901 controls the entire computer. The memory 2902 is a memory such as a RAM that temporarily stores a program or data stored in the external storage device 2905 (or portable recording medium 2909) when executing a program, updating data, or the like. The CPU 2901 performs overall control by reading the program into the memory 2902 and executing it.

  The input device 2903 includes, for example, a keyboard, a mouse, etc. and their interface control devices. The input device 2903 detects an input operation by a user using a keyboard, a mouse, or the like, and notifies the CPU 2901 of the detection result.

  The output device 2904 includes a display device, a printing device, etc. and their interface control devices. The output device 2904 outputs data sent under the control of the CPU 2901 to a display device or a printing device.

The external storage device 2905 is, for example, a hard disk storage device. Mainly used for storing various data and programs.
The portable recording medium driving device 2906 accommodates a portable recording medium 2909 such as an optical disc, SDRAM, or Compact Flash (registered trademark), and has an auxiliary role for the external storage device 2905.

The network connection device 2907 is a device for connecting, for example, a LAN (local area network) or a WAN (wide area network) communication line.
The system according to the present embodiment is realized by the CPU 2901 executing a program equipped with processing functions realized by the operation flowcharts of FIGS. 7, 13, 18, and 22 described above. For example, the program may be recorded and distributed in the external storage device 2905 or the portable recording medium 2909, or may be acquired from the network by the network connection device 2907.

Regarding the above embodiment, the following additional notes are disclosed.
(Appendix 1)
A search request is received, a search request is generated based on a search expression corresponding to the search request, and a search is performed on the search target data classified and stored in a plurality of search target data storages based on the search request. In a search processing method for executing a process and notifying a search request of the search request of a search result of the search process,
From the search formula corresponding to the search request, extract the classification target item and the value of the item used when the search target data is classified and stored in the plurality of search target data storages, and the classification target item A search target data specifying processing step for specifying the search target data storage targeted by the search formula based on a value;
Each search expression corresponding to one or more of the search requests is bundled for each search target data storage specified by the search target data specifying processing step corresponding to each search expression, and each search expression of the storage is specified for each search target data storage. A search request processing step for generating and issuing a search request including identification information and information of each of the bundled search expressions;
Based on the search request, each search process corresponding to each search formula included in the search request for the search target data in the search target data storage corresponding to the identification information included in the search request A search processing step for executing
A search result notification processing step of distributing the search results obtained by the search processing step for each search formula and transmitting the search results to a request source of each search request corresponding to the search formula;
A search processing method characterized by comprising:
(Appendix 2)
In the search target data specifying processing step, based on the value of the classification target item, in which storage among the plurality of search target data storages the search target data is stored for each value of the classification target item. The search target data storage is specified by referring to the search target data location management list shown in FIG.
The search processing method according to supplementary note 1, wherein:
(Appendix 3)
The search target data storage is a search target file in which the search target data is written.
The search method according to any one of appendices 1 and 2, characterized in that:
(Appendix 4)
The search target data storage is a search target table of a database storing the search target data.
The search processing method according to any one of Supplementary Note 1 or 2, wherein:
(Appendix 5)
A search request is received, a search request is generated based on a search expression corresponding to the search request, and a search is performed on the search target data classified and stored in a plurality of search target data storages based on the search request. In a search processing system that executes a process and notifies the search request source of the search result of the search process,
From the search expression corresponding to the search request, extract the classification target item and the value of the item used when the search target data is classified and stored in the search target data storage, and the value of the classification target item is extracted. A search target data identification processing unit that identifies the search target data storage that is targeted by the search formula;
Each search expression corresponding to one or more of the search requests is bundled for each search target data storage specified by the search target data specifying processing unit corresponding to each search expression, and each search expression of the storage is stored for each search target data storage. A search request processing unit that generates and issues a search request including identification information and information on each of the bundled search expressions;
Based on the search request, each search process corresponding to each search formula included in the search request for the search target data in the search target data storage corresponding to the identification information included in the search request A search processing unit for executing
A search result notification processing unit that distributes the search results obtained by the search processing unit for each search formula and transmits the search results to a request source of each search request corresponding to each search formula;
A search processing system comprising:
(Appendix 6)
Based on the value of the classification target item, the search target data identification processing unit determines in which storage of the plurality of search target data storages the search target data is stored for each value of the classification target item. The search target data storage is specified by referring to the search target data location management list shown in FIG.
The search processing system according to appendix 5, characterized in that:
(Appendix 7)
The search target data storage is a search target file in which the search target data is written.
The search processing system according to any one of appendices 5 and 6, characterized in that:
(Appendix 8)
The search target data storage is a search target table of a database storing the search target data.
The search processing system according to any one of appendices 5 and 6, characterized in that:
(Appendix 9)
A search request is received, a search request is generated based on a search expression corresponding to the search request, and executed on search target data classified and stored in a plurality of search target data storages based on the search request A computer that executes a search process that receives a search result of the search process and notifies the requester of the search request;
From the search formula corresponding to the search request, extract the classification target item and the value of the item used when the search target data is classified and stored in the search target data storage, and the value of the classification target item is extracted. A search target data specifying processing step for specifying the search target data storage targeted by the search formula,
Each search formula corresponding to one or more of the search requests is bundled for each search target data storage specified by the search target data specifying processing step corresponding to each search formula, and for each search target data storage, A search request processing step for generating and issuing a search request including identification information and information of each of the bundled search expressions;
Based on the search request, corresponding to the search formulas included in the search request executed for the search target data in the search target data storage corresponding to the identification information included in the search request. A search result notification processing step of receiving each search result obtained by each search process, distributing each search result for each search formula, and transmitting it to the request source of each search request corresponding to each search formula;
Search processing program for executing
(Appendix 10)
In the search target data specifying processing step, based on the value of the classification target item, in which storage among the plurality of search target data storages the search target data is stored for each value of the classification target item. The search target data storage is identified by referring to the search target data location management list shown in FIG.
The search processing program according to supplementary note 9, wherein

  The disclosed technology can be used in a system or program that performs data operations such as profiling, cleansing, log analysis, and log trails on data stored in a general file or database.

DESCRIPTION OF SYMBOLS 101 Search control apparatus 102 Search object data management apparatus 103 Search object data location management book 104 Search object data storage 105 Search processing apparatus

Claims (7)

A search request is received, a search request is generated based on a search expression corresponding to the search request, and a search is performed on the search target data classified and stored in a plurality of search target data storages based on the search request. In a search processing method for executing a process and notifying a search request of the search request of a search result of the search process,
From the search formula corresponding to the search request, extract the classification target item and the value of the item used when the search target data is classified and stored in the plurality of search target data storages, and the classification target item A search target data specifying processing step for specifying the search target data storage targeted by the search formula based on a value;
Each search expression corresponding to one or more of the search requests is bundled for each search target data storage specified by the search target data specifying processing step corresponding to each search expression, and each search expression of the storage is specified for each search target data storage. A search request processing step for generating and issuing a search request including identification information and information of each of the bundled search expressions;
Based on the search request, each search process corresponding to each search formula included in the search request for the search target data in the search target data storage corresponding to the identification information included in the search request A search processing step for executing
A search result notification processing step of distributing the search results obtained by the search processing step for each search formula and transmitting the search results to a request source of each search request corresponding to the search formula;
A search processing method characterized by comprising: In the search target data specifying processing step, based on the value of the classification target item, in which storage among the plurality of search target data storages the search target data is stored for each value of the classification target item. The search target data storage is specified by referring to the search target data location management list shown in FIG.
The search processing method according to claim 1, wherein: The search target data storage is a search target file in which the search target data is written.
The search method according to any one of claims 1 and 2. The search target data storage is a search target table of a database storing the search target data.
The search processing method according to any one of claims 1 and 2. A search request is received, a search request is generated based on a search expression corresponding to the search request, and a search is performed on the search target data classified and stored in a plurality of search target data storages based on the search request. In a search processing system that executes a process and notifies the search request source of the search result of the search process,
From the search expression corresponding to the search request, extract the classification target item and the value of the item used when the search target data is classified and stored in the search target data storage, and the value of the classification target item is extracted. A search target data identification processing unit that identifies the search target data storage that is targeted by the search formula;
Each search expression corresponding to one or more of the search requests is bundled for each search target data storage specified by the search target data specifying processing unit corresponding to each search expression, and each search expression of the storage is stored for each search target data storage. A search request processing unit that generates and issues a search request including identification information and information on each of the bundled search expressions;
Based on the search request, each search process corresponding to each search formula included in the search request for the search target data in the search target data storage corresponding to the identification information included in the search request A search processing unit for executing
A search result notification processing unit that distributes the search results obtained by the search processing unit for each search formula and transmits the search results to a request source of each search request corresponding to each search formula;
A search processing system comprising: A search request is received, a search request is generated based on a search expression corresponding to the search request, and executed on search target data classified and stored in a plurality of search target data storages based on the search request A computer that executes a search process that receives a search result of the search process and notifies the requester of the search request;
From the search expression corresponding to the search request, extract the classification target item and the value of the item used when the search target data is classified and stored in the search target data storage, and the value of the classification target item is extracted. A search target data specifying processing step for specifying the search target data storage targeted by the search formula,
Each search expression corresponding to one or more of the search requests is bundled for each search target data storage specified by the search target data specifying processing step corresponding to each search expression, and each search expression of the storage is specified for each search target data storage. A search request processing step for generating and issuing a search request including identification information and information of each of the bundled search expressions;
Based on the search request, corresponding to each search expression included in the search request executed for the search target data in the search target data storage corresponding to the identification information included in the search request. A search result notification processing step of receiving each search result obtained by each search process, distributing each search result for each search formula, and transmitting it to the request source of each search request corresponding to each search formula;
Search processing program to execute. In the search target data specifying processing step, based on the value of the classification target item, in which storage among the plurality of search target data storages the search target data is stored for each value of the classification target item. The search target data storage is specified by referring to the search target data location management list shown in FIG.
The search processing program according to claim 6.