CN1271549C - Information searching device for searching text to retrieve a certain character stream matching a certain keyword - Google Patents

Information searching device for searching text to retrieve a certain character stream matching a certain keyword Download PDF

Info

Publication number
CN1271549C
CN1271549C CN 02132175 CN02132175A CN1271549C CN 1271549 C CN1271549 C CN 1271549C CN 02132175 CN02132175 CN 02132175 CN 02132175 A CN02132175 A CN 02132175A CN 1271549 C CN1271549 C CN 1271549C
Authority
CN
China
Prior art keywords
real
character
time processing
processing data
time
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN 02132175
Other languages
Chinese (zh)
Other versions
CN1492360A (en
Inventor
木下哲也
小山隆正
菊池忠一
櫌本典行
篠木裕文
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Panasonic Holdings Corp
Original Assignee
Matsushita Electric Industrial Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Matsushita Electric Industrial Co Ltd filed Critical Matsushita Electric Industrial Co Ltd
Publication of CN1492360A publication Critical patent/CN1492360A/en
Application granted granted Critical
Publication of CN1271549C publication Critical patent/CN1271549C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Landscapes

  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

A plurality of characters arranged in a text are numbered in order of occurrence for each character type, and an occurrence number table in which one or more occurrence numbers are listed for each character type is prepared in an occurrence number preparing unit. Also, a set of two character types is defined as a character pattern, and a set of two occurrence numbers of two adjacent characters of which types correspond to one character pattern is defined as an occurrence number pattern. One or more occurrence number patterns respectively corresponding to one character pattern are arranged for each character pattern, and a list of occurrence number patterns is prepared as an index file in an index file preparing unit. Thereafter, a series of specified character patterns indicated by a key word is prepared, and a series of particular occurrence number patterns corresponding to the series of specified character patterns is selected on condition that a series of particular characters indicated by the series of particular occurrence number patterns agrees with the particular character stream. Therefore, because the characters of the text is converted to the occurrence number patterns and all occurrence number patterns are not searched, a searching operation can be performed at a high speed.

Description

Searching text meets the information searching apparatus of the character stream of a certain keyword with retrieval
The application is that application number is 97102954.7, the applying date be February 27 in 1997 day, be called the dividing an application of application of " the search text meets the information searching apparatus of the character stream of a certain keyword with retrieval ".
Technical field
The present invention generally speaking relates to information searching apparatus, and this device is searched a database for consistent with the character stream (keyword) of appointment respectively one or more character strings, thus searching character stream.The invention particularly relates to by search database with the information retrieval device of high speed retrieval, adding, renewal or the deletion one section character stream consistent with the character stream of appointment.
Background technology
In the information search system in modern age, when the user specifies a certain character stream, and when this character stream is input to common information searching apparatus as keyword, the database that one or more snippets character data search consistent with the character stream of appointment disposed for common information searching apparatus, and searching character data automatically.More particularly, made and be extensive use of a kind of related data base management system.
For example, first kind of common information searching apparatus seen through announcing but the content of announcement among uncensored Japanese patent application S63-155324 (1988), H1-282635 (1989), H2-30869 (1990) and the H3-113626 (1991).In patented claim S63-155324 (information indexing device), in each circulation, when the keyword of depositing by speech when moving, with the keyword deposited with by the search key of operator's appointment relatively, and carry out the information retrieval operation.In patented claim H1-282635 (system kept in index), keep a master index and a less subindex, as the index that is used for retrieving the data segment of depositing, the index record of depositing data that increases is added on the subindex, and to deposit renewal, increase and the deletion of data at a high speed.In patented claim H2-302869 (document-editing system), in order to remove zone useless in the storer, and specify the data of storing in the storer to be processed apace, form a table that is used for reserve storage free area pointer (pointer), and when relating to this table, just specify the character stream data of being stored in the storer to be processed.In patented claim H3-113626 (system of controller buffer and device),, carry out the retrieval of data in the impact damper concurrently and from the data transmission of auxilary unit in order to improve the handling capacity of a certain task.
In above-mentioned four first kind of common information searching apparatus, data field and Data Update district are arranged in the memory block in the recording mechanism unit, some data segments are stored in the Data Update district during Data Update operation, and data field and Data Update district mutual alternative (patented claim S63-155324, H1-282635 and H3-113626).Simultaneously, the increase of data, renewal and deletion are (the patented claim H2-302869) that the string data corresponding tables of being stored in the record with retrieving files and character stream data is carried out.
The data of carrying out in first kind of general information search device are described below deposit operation, data change operation and data deletion operation.
Fig. 1 is the process flow diagram flow chart that carries out in first kind of general information search device according to patented claim H2-302869.Fig. 2 illustrates as the raw data table T1 of first kind of common information searching apparatus and the tables of data T2 of change, has described the data stream that writes down among the table T1.Among the table T1, be a file to be processed with three recording setting.In A type record, write down the character stream of forming by two characters " 12 " or " AI " and another character stream of forming by two characters " 34 " or " UE ", form the data structure of A type record.In the Type B record, write down the character stream of forming by two characters and another character stream of forming by five characters, form the data structure of Type B record.In this example, the zone of A type record character stream " 12 " is corresponding to the zone of A type record character stream " AI ", and character stream " 12 " is identical item with " AI ".Equally, the zone of A type record character stream " 34 " is corresponding to the zone of A type record character stream " UE ", and character stream " 34 " is identical item with " UE ".
As shown in Figure 1, from the retrieving files shown in the table T1, read a plurality of character stream to be processed one by one, as record to be processed (step S101).Then, the position of each character stream, the size of each character stream and the pointer of each character stream are set.The first address of a certain record corresponding with this character stream is represented in the position of character stream.The size of character stream is represented the number of characters in the character stream.The pointer that is used for a plurality of character streams represents that these character streams are identical items.Subsequently, with the character stream structure modify among the table T1 be the structure (step S102) in the table T2 in the storer.For example, because the character stream " AI " of the Xiang Yudi three record A of the character stream " 12 " of the first record A is identical, so, be provided for the pointer p1 of character stream " 12 " and " AI ".Equally, because the item of the character stream " UE " of the Xiang Yudi three record A of the character stream " 34 " of the first record A is identical, so be provided for the pointer p2 of character stream " 34 " and " UE ".Subsequently, repeating step S101 and S102, up to all character streams that read table 1, and these character streams all with the table T2 structure be stored in the storer till (step S103).Then, table T2 is set in the clear area of storer (space area), thereby makes table T2 (step S104).
Then, the character stream of consideration plan his-and-hers watches T1 carries out more news (the renewal operation among the step S105).For example, when plan with table 1 in the character stream " 34 " of the first record A when being updated to " 56 ", a plurality of record A among the look-up table T2, the character stream of depositing in the third line among the detection table T2 " 34 ", and character stream " 34 " changed to character stream " 56 ".At this moment, because number of characters is constant, so need not to change pointer p2.
Equally, when step S105 carries out deletion action, when the plan test example writes down the character stream " UE " of A as the 3rd of table T1, a plurality of record A of look-up table T2, the character stream of depositing in detection table T2 the 7th row during pointer p2 that in following the tracks of the third line, deposits " UE ", and certain delegation of the deletion record corresponding with character stream " UE ".Also delete the pointer p2 of the third line in addition.
As mentioned above, in first kind of common information searching apparatus, deposit operation, upgrade operation and deletion action, follow the tracks of first address, size and the pointer of character stream in each row simultaneously with table 2.
Yet, when having deposited a large amount of character stream (or lot of data) in the database, just there is such defective, promptly needs to spend long time and from all data segments of database, hunt out respectively consistent one or more data segments of character stream (or a certain keyword) with a certain appointment.So, need be to search a database at a high speed.
Simultaneously, when the renewal of in first kind of common information searching apparatus, carrying out data or deletion action, need to obtain a certain extended area, make it to equal in the storer a certain zone of new record more or deletion record, so another defective is to need one to have jumbo storer.
Equally, also need after the renewal of having carried out data or deletion, rearrange ordering (sorting) operation of record order.So another defective is to need cost to upgrade for a long time or deleted data.
Simultaneously, the record of a lastest imformation is inserted in the insertion row of table T2, in this table, many data recording dense arrangement and tight, and from table T2, remove a deletion information record in the deletion row that is placed on table T2.So, in renewal or deletion action, need to make to place to insert row or delete many later data recording of a certain position of row to be shifted, thereby another defective is to upgrade for a long time or deletion action.
Have again, though this deletion information of deletion from table T2, and deleted marker that should deletion information is set to " on (opening) " situation, its another defective is that the bar number (or data number) of record is increased.
Fig. 3 is the block scheme of second kind of common information searching apparatus.
As shown in Figure 3, second kind of common information searching apparatus 11 comprises, input block 12 is used for receiving the searching character stream that file is deposited the file that the user prepared in (or renewal) operation and received a certain appointment in the search operaqtion; Handle request receiving element 13, be used for receiving the request of depositing (or renewal) file, and receive specifying the retrieval request of searching character stream from input block 12 from input block 12; One real-time processing unit 14 according to from handling the processing request that request receiving element 13 sends, is handled this document in real time, so that prepare to be used for the index file of search operaqtion; Be used for storing the index file storage unit 15 of index file; Export the output unit 16 of a segment information, the end that index file was prepared when this information was pointed out by the index file preparation end of handling request receiving element 13 notices; An and search operaqtion performance element 17, be used for carrying out search operaqtion, thereby when from processing request receiving element 13 reception search instructions, retrieval is flowed one or more snippets consistent character data with the searching character of appointment respectively in the index file of storage from index file storage unit 15.
In said structure, when the user prepares a file and this document is input to input block 12 to deposit in real time or to upgrade this document, startup file is deposited (or renewal) operation in device 11.That is, receive from (or renewal) request of depositing of the file and the file of input block 12 transmissions by handling request receiving element 13, and the processing request of this document and this document sends to real-time processing unit 14 from handling request receiving element 13.In real-time processing unit 14, handle file in real time, and prepare the index file of this document, and index file is stored in the index file storage unit 15.Simultaneously, when the preparation of index file finish after, processing unit 14 end notification that index file is prepared is handled request receiving element 13 in real time.After, the expression true-time operation ending message that index file prepares that is through with in real time is sent to output unit 16, and with this true-time operation ending message from output unit 16 outputs.So file is deposited (or renewal) EO, and second kind of common information searching apparatus 11 is set to ask waiting status.
Subsequently, when the searching character stream with a certain appointment is input to input block 12 request search operaqtions, receive retrieval request by handling request receiving element 13, search instruction sends to search operaqtion performance element 17 from handling request receiving element 13, and carries out search operaqtion in search operaqtion performance element 17.That is, from index file storage unit 15 in the index file of storage respectively retrieval flow one or more snippets consistent character data with the searching character of appointment.When the retrieval of character data finished, the advisory that search operaqtion performance element 17 finishes the character data retrieval was handled request receiving element 13, and with this character data from output unit 16 outputs, as result for retrieval.That is, search operaqtion finishes.
Yet before the preparation of end data file, can not carry out search operaqtion, and need cost to prepare index file for a long time.That is, the operator must deposit in real time or updating file and store into data file in the index file storage unit 15 before, wait for that the searching character stream with appointment is input to input block 12.So its defective is to need cost long-time, up to second kind of common information searching apparatus 11 is set at the request waiting status.
Summary of the invention
Consider the defective of above-mentioned general information search device, first purpose of the present invention provides a kind of information searching apparatus, in this device, search database for one or more consistent with the character stream (or keyword) of appointment respectively character stream, retrieve character stream with high speed, and with high speed renewal or the deletion character stream consistent with the character stream of appointment.
Second purpose of the present invention provides a kind of information searching apparatus, this device is not as long as wait until after record, increase, renewal or the deletion action end of carrying out for index file correctly retrieving character stream, even ask search operaqtion the mid-term of the record of indexed file, increase, renewal or deletion action, also can be to carrying out the seek operations of a certain index file that is equivalent to database immediately with the consistent one or more character streams of character stream (or a keyword) of an appointment respectively.
First purpose realizes that by a kind of information searching apparatus is provided this device comprises:
The number table preparation device occurs, be used for preparing one and number table occurs, list by the one or more characters that belong to these character types in the searching for text are numbered the one or more appearance numberings that obtain by order occurring for each character types in the table;
The index file preparation device, the appearance numbering that is used for occurring listing in the appearance number table of number table preparation device preparation is built into a plurality of patterns again, wherein each one group two the appearance numberings that numbering scheme is represented two adjacent characters in the searching for text occur, be used for being divided into numbering scheme occurring how numbering scheme appears in group, wherein every group of character pattern with one group of two character types of expression is corresponding, and be used for preparing an index file, in this index file, listed respectively and an appearance numbering scheme group that character pattern is corresponding, the particular type of two adjacent characters that numbering scheme represents occurred with consistent by two character types of respective symbols modal representation by each;
The index file search device, this device is given birth to a series of designated character patterns by the designated character miscarriage of being represented by a keyword, come search index file and the retrieval specific character stream consistent with described keyword, and, select described specific the occur numbering scheme series corresponding with described designated character mode sequences specificly a series of specific characters that numbering scheme represents occurring when consistent with described specific character stream by a series of; With
The index file memory storage is used to store the described numbering scheme group that occurs that is obtained by described index file preparation device, as index file.
In said structure, the resulting one or more numerals of the one or more character numbers that belong to character types of arranging by the sequencing that takes place are referred to as one or more appearance numberings in searching for text.Also one group of two character types is defined as a character pattern, two of two adjacent characters in the searching for text is occurred being numbered one group be defined as and numbering scheme occurs.Prepare one and number table occurs by the number table preparation device is listed in one or more characters of arranging in the searching for text for each character types one or more appearance numberings occurring.Subsequently, the appearance numbering that occurs listing in the number table is built into a plurality of numbering schemes that occur again, numbering scheme to occur and be divided into how numbering scheme appears in group, each numbering scheme occurs corresponding to a character pattern, and index of reference file preparation device prepares an index file, lists each group corresponding with character pattern in this index file and numbering scheme occurs.Subsequently, by index file search device search index file, and the retrieval specific character stream corresponding with a certain keyword.Promptly, produce the character pattern of a series of regulations from the regulation character stream of representing with keyword, and, select and the corresponding a series of specific numbering schemes that occur of this serial regulation character pattern when specificly a series of specific characters that numbering scheme represents occurring when consistent with specific character stream with a series of.So, the series of specific character is retrieved as the specific character stream consistent with keyword.
Therefore, in following the tracks of each row when the first address of character stream, size and pointer, because carry out in the first general information search device to deposit operation be not adopted, but all characters in the search index all use and the numbering scheme group occurs and replace, so that the size of index file reduce.Thereby be used for the capacity of storer of recording indexes file and can do lessly.
And, do not search all appearance numbering scheme group corresponding with all character patterns, with the retrieval specific character stream consistent with keyword, but, the corresponding specific numbering scheme group that occurs of regulation character pattern series of selecting and obtaining according to keyword, and with specific a series of specific numbering scheme that occurs that retrieval is selected is appearring under a series of specific characters that numbering scheme represents situation consistent with specific character stream from specific appearance numbering scheme group.So, can reduce the quantity (volume) of the appearance numbering scheme searched, and under high speed, carry out the retrieval of the specific character stream consistent with keyword.
Second purpose realizes that by a kind of information searching apparatus is provided this device comprises:
Store the searched targets memory storage of one or more searched targets;
The real-time processing data receiving trap, be used for receiving respectively with the searched targets memory storage in one or more snippets corresponding real-time processing data of a target in the searched targets of storing;
Be used for receiving the retrieving information receiving trap of one section retrieving information and a certain search request;
The real-time processing data memory storage is used for storing the real-time processing data section that is received by the real-time processing data receiving trap;
Treating apparatus is used for each segment data in the real-time processing data section of storing in the real-time processing data memory storage is handled in real time in real time, to upgrade the searched targets of storing in the searched targets memory storage;
The searched targets search device is used for searching each searched targets of storing in the searched targets memory storage, thereby retrieves one or more snippets the consistent first information of retrieving information that is received with the retrieving information receiving trap from each searched targets; And the result for retrieval of the first information is pointed out in output;
The real-time processing data search device, be used for searching each section in one or more snippets specific real-time processing data, these data segments be can't help real-time treating apparatus and are handled but be stored in the real-time processing data memory storage, thus retrieval and consistent one or more second message segments of retrieving information in each section from specific real-time processing data section by the reception of retrieving information receiving trap;
The result for retrieval modifier is used for according to second information by the retrieval of real-time processing data search device, revises from the result for retrieval of searched targets search device output, with specific real-time processing data section in the reflection result for retrieval; And
The result for retrieval output unit is used for exporting the result for retrieval of being revised by the result for retrieval modifier.
In said structure, when receiving one or more snippets real-time processing data by the real-time processing data receiving trap, the real-time processing data section is stored in the real-time processing data memory storage.Then, carry out real-time processing by real-time treating apparatus, and upgrade the searched targets of storing in the searched targets memory storage one by one each section real-time processing data.In this case, when by the retrieving information receiving trap when the interstage of handling in real time receives the request of search and one section retrieving information, can't help one or more snippets specific real-time processing data of real-time treating apparatus processing to be retained in the real-time processing data memory storage.In this case, during specific real-time processing data section is carried out seek operations, be a specific real-time processing data section not to be handled in real time.
Subsequently, by searched targets search device retrieve stored each searched targets in the searched targets memory storage,, and obtain a result for retrieval with retrieval one or more snippets first information consistent with retrieving information.And, by the real-time processing data search device search in the specific real-time processing data section each the section, thereby retrieve one or more snippets second information consistent from each section of specific real-time processing data section with retrieving information.Subsequently, revise result for retrieval by the result for retrieval modifier, and export this result by the result for retrieval output unit according to second information.So,, in result for retrieval, also can reflect specific real-time processing data section even can't help real-time treating apparatus upgrades each specific searched targets corresponding with specific real-time processing data section.
Therefore, because seek operations is receiving the request of search no matter whether the real-time processing of each real time data was carried out when finishing, so can carry out seek operations immediately, and needn't wait for the end of real-time processing (such as record, increase, renewal or deletion action).
Have again, because seek operations is to begin in the interstage of handling in real time, because each section of specific real-time processing data is to be revised according to second information of retrieving from specific real-time processing data section by the result for retrieval modifier by search of real-time processing data search device and result for retrieval, even thereby can't help one or more snippets specific real-time processing data that real-time treating apparatus handles respectively and be retained in the real-time processing data memory storage, also can correctly carry out seek operations.
Information searching apparatus preferably also comprises;
Repeat to handle in real time to avoid device, be used for when receiving the real-time processing data of the one section repetition corresponding with same specific searched targets by the real-time processing data receiving trap, will be corresponding and be stored in one section real-time processing data in the real-time processing data memory storage according to same specific searched targets with a certain specific searched targets, change over the real-time processing data that a shed repair is corrected one's mistakes; And the real-time processing data of discarded this repetition.
In said structure,, can under high speed, handle in real time effectively because avoided the repetition of re-treatment data to handle in real time.
Information searching apparatus preferably also comprises:
Exist and judge the table preparation device, be used for preparing one and have the judgement table, be stored in all the components of the real-time processing data of one section appointment of real-time treating apparatus neutralization in this table; And
The search operaqtion control device, be used for judging existing and judge that the existence that the table preparation device is prepared judges all the components that whether has retrieving information in the table, and control real-time processing data search device stops to search the real-time processing data of appointment when making it not exist all the components of key in having the judgement table.
In said structure, because prepared existence judgement table, so, when the real-time processing data of searching appointment during with retrieval one or more snippets second information consistent with retrieving information, can judge to exist and judge all the components that whether has retrieving information in the table, and when judged result is when having all the components that does not have retrieving information in the judgement table, can stop the seek operations of being undertaken by the real-time processing data search device.Thereby can carry out seek operations at a high speed.
Information searching apparatus preferably also comprises:
Device is avoided in the repeated retrieval operation, be used for from the corresponding a plurality of real-time processing data sections of the same specific searched targets of from the real-time processing data memory storage, storing in select one section up-to-date real-time processing data corresponding with specific searched targets, thereby control real-time processing data search device is retrieved second information consistent with retrieving information from up-to-date real-time processing data, and the real-time processing of control search device does not carry out the seek operations at other real-time processing datas corresponding with same specific retrieving information.
In said structure, because under a plurality of real-time processing datas corresponding with same specific searched targets are stored in situation in the real-time processing data memory storage, up-to-date real-time processing data is carried out seek operations, and because be other real-time processing data sections corresponding with same specific searched targets not to be carried out any seek operations, so can carry out seek operations at a high speed.
Information searching apparatus preferably also comprises:
Be used for receiving the ending request receiving trap of ending request;
Force the end operation request unit, be used for receiving under the situation of ending request at the ending request receiving trap, ask real-time treating apparatus to force and stop a processing in real time, make the real-time processing data of one or more snippets appointment that real-time treating apparatus is not handled remain in the real-time processing data memory storage; And
End situation memory storage is used for storing the real-time processing data section by the appointment that forces the end operation request unit to keep.
In said structure, even a large amount of real-time processing data of storage in the real-time processing data device, because the real-time processing at one section current real-time processing data when receiving ending request forces end, so do not exist the user to wait for that long time finishes the situation at the real-time processing of all real-time processing datas.Thereby the user can carry out the operation of another kind of type immediately with this information searching apparatus.
Information searching apparatus preferably also comprises:
Receive the startup request receiving trap that starts request; And
The start-up operation request unit is used for asking real-time treating apparatus to restart by the real-time processing that forces end operation end that request unit forces when startup request receiving trap receives the startup request.
In said structure, even the real-time processing at all real-time processing datas does not also finish, because one or more snippets in the real-time processing data of untreated appointment is stored in the end situation memory storage, so when receiving the startup request, can handle the real-time processing data section of untreated appointment.Thereby, no matter the quantity of real-time processing data how, can correctly be carried out the real-time processing at all real-time processing datas.
Information searching apparatus preferably also comprises:
Retrieval time, measurement mechanism was used for measuring by the specific searched targets of being stored in the memory storage of searched targets search device search retrieval order border or one period retrieval time that specific real-time processing data is required of being searched the specific searched targets of being stored in the real-time processing data memory storage by the real-time processing data search device;
Update time, measurement mechanism was used for measuring the update time of being upgraded the specific searched targets of storing in the searched targets memory storage by real-time treating apparatus;
Upgrade the operational notification device, be used for judging whether measurement mechanism retrieval time is later than measurement mechanism update time measured update time measured retrieval time, and when be later than retrieval time update time, prepare one section lastest imformation, this segment information is pointed out the renewal of specific searched targets operation is carried out later on having searched specific searched targets by searched targets search device or real-time processing data search device; And
The lastest imformation output unit is used for being later than under the situation of retrieval time in update time, and output is by identifier that upgrades specific searched targets that the operation information device obtains and lastest imformation.
In said structure, because judged whether be later than update time retrieval time, so, also can notify the user with one section lastest imformation even specific searched targets is to upgrade later in the seek operations that specific searched targets is carried out.
Description of drawings
After in conjunction with the accompanying drawings the present invention having been carried out following description, the reader will be well understood to purpose of the present invention, characteristic and advantage.
Fig. 1 is the process flow diagram of the process of carrying out in first kind of general information search device;
Fig. 2 is provided to show raw data table data stream, first kind of general information search device and the change tables of data that writes down in the raw data table;
Fig. 3 is the block scheme of second kind of general information search device;
Fig. 4 illustrates the appearance number table of a plurality of appearance numberings of listing the character that is used for each type;
Fig. 5 illustrates and lists the one or more index files that numbering scheme occurs that are used for each character pattern;
Fig. 6 is the information searching apparatus block scheme according to the present invention first, second and the 3rd embodiment;
Fig. 7 deposits operational flowchart according to the index file that method for searching is carried out;
Fig. 8 illustrates one or more index files that numbering scheme occurs of listing at each character pattern;
Fig. 9 is the seek operations process flow diagram of carrying out according to a kind of method for searching;
What Figure 10 described is mutual close a pair of designated character pattern and a plurality of the occur numbering schemes corresponding with each designated character pattern in the keyword;
Figure 11 illustrates second embodiment of process flow diagram do to upgrade operation according to to(for) the primary index file of depositing in the memory storage;
Figure 12 illustrates the appearance number table of preparing from the searching for text of revising according to second embodiment;
Figure 13 illustrates according to one or more corresponding with each character pattern of second embodiment and the appearance numbering scheme table of numbering scheme occurs;
Figure 14 illustrates one or more that list at each character pattern according to second embodiment and the index file of numbering scheme occurs;
Figure 15 illustrates the process flow diagram of making the part deletion action according to the index file for storage in the index file storage unit 28 of the 3rd embodiment;
Figure 16 illustrates the appearance number table of preparing from the searching for text of revising according to the 3rd embodiment;
Figure 17 illustrates the one or more numbering scheme appearance numbering scheme tables corresponding with each character pattern that occur according to the third embodiment;
Figure 18 illustrates the index file of listing one or more appearance numberings according to the 3rd embodiment at each character pattern;
Figure 19 is the block scheme according to the information searching apparatus of the present invention the 4th and the 5th embodiment;
Figure 20 is that the deal with data group and that one section real-time processing data of record is shown is handled the process flow diagram of asking in real time;
Figure 21 illustrates a routine deal with data group;
Figure 22 illustrates the deal with data group with a pair of sign;
Figure 23 illustrates a plurality of untreated deal with data group of storing in the real-time processing data storage unit and the tabulation of a plurality of deal with data groups of having handled;
Figure 24 illustrates one section video data structure;
Figure 25 illustrates an example of display data segment;
Figure 26 is the retouching operation process flow diagram that the index file of storing in the index file storage unit is shown;
Figure 27 illustrates the indexed file structure of storing in the index storage unit;
Figure 28 A illustrates and a plurality of character patterns (C1, C2) one or more character information patterns of each the character pattern correspondence in (PC1, primary index file DN);
Figure 28 B illustrates and a plurality of character patterns (C1, C2) one or more character information patterns (PC1, the index file of modification DN) of each the character pattern correspondence in;
Figure 29 is illustrated in not the process flow diagram that carries out search operaqtion under the situation that all deal with data groups are handled in real time;
Figure 30 illustrates an example of being revised the retouching operation that carries out the unit by result for retrieval shown in Figure 19;
Figure 31 be illustrate according to the fifth embodiment of the present invention to corresponding two process flow diagrams that the deal with data group is handled in real time of same specific searching for text;
Figure 32 A is the example according to the real-time update operation of the 5th embodiment;
Figure 32 B is the example according to the real-time deletion action of the 5th embodiment;
Figure 32 C is the example according to the real-time increase operation of the 5th embodiment;
Figure 33 is the block scheme according to the information searching apparatus of the sixth embodiment of the present invention;
Figure 34 deposits the process flow diagram of operation according to the 6th embodiment deal with data group;
Figure 35 is the structure that there is the judgement table in character;
Figure 36 is the example that there is the judgement table in character;
Figure 37 is the process flow diagram that carries out the character stream search operaqtion when adopting character to have the judgement table;
Figure 38 is the block scheme according to the information searching apparatus of seventh embodiment of the invention;
Figure 39 illustrates according to the 7th embodiment avoiding the operation of repeated retrieval when operation repeated retrieval to avoid the process flow diagram of the character stream search operaqtion that the unit carries out;
Figure 40 illustrates and avoids unit and result for retrieval to revise the unit according to the operation of the repeated retrieval of the 7th embodiment retrieving example with retouching operation;
Figure 41 is the block scheme according to the information searching apparatus of the eighth embodiment of the present invention;
Figure 42 is the process flow diagram that forces end operation according to the eighth embodiment of the present invention;
Figure 43 illustrates the structure of one section termination condition data;
Figure 44 illustrates the example of three sections termination condition data;
Figure 45 is the process flow diagram that illustrates according to the start-up operation of eighth embodiment of the invention;
Figure 46 is the block scheme according to the information searching apparatus of ninth embodiment of the invention;
Figure 47 is the process flow diagram that the record of retrieval time is shown;
The structure that Figure 48 illustrates a data number and organizes retrieval time;
What Figure 49 described is the example of organizing a data number and retrieval time;
Figure 50 is the process flow diagram that the record of update time is shown;
The structure that Figure 51 illustrates a data number and organizes update time;
The example that Figure 52 illustrates a data number and organizes update time;
Figure 53 is the process flow diagram that illustrates according to renewal or the output function of non-lastest imformation of the 9th kind of embodiment; And
Figure 54 is from the renewal of lastest imformation output unit output and the example of non-lastest imformation.
Embodiment
Preferred embodiment according to information searching apparatus of the present invention is described with reference to the accompanying drawings.
Method for searching according to the index file of first kind of embodiment search of the present invention one or more character streams consistent with a certain keyword (or designated character stream) is at first described.
Fig. 4 describes is one and number table occurs, listed a plurality of appearance numberings of the character of each type in the table, and Fig. 5 is an index file, has listed one or more numbering schemes that occur of each character pattern in the file.
As shown in Figure 4, a plurality of characters of arranging at searching for text are divided according to a plurality of different types, to belong to one or more characters of same character types from 1 volume on number by the order that takes place, thereby make one and number table occurs, in the table, list the appearance numbering of each character types with the order that increases progressively.For example, if the character stream of being made up of three kinds of character types " A ", " I " and " N " " AAIAAIANNAANAI " that writes down in the searching for text, the structure that number table occurs is as described below.The character of character stream " A " is by order from 1 to 8 numbering that takes place, the character of character stream " I " is by order from 1 to 3 numbering that takes place, the character of character stream " N " is by order from 1 to 3 numbering that takes place, character " I " with 1 numbering is positioned at between 2 and 3 characters " A " of numbering, character " I " with 2 numberings is positioned at between 4 and 5 characters " A " of numbering, character " N " with 1 and 2 numberings is positioned at between 5 and 6 characters " A " of numbering, character " N " with 3 numberings is positioned at between 7 and 8 characters " A " of numbering, and follows in character " A " back with 8 numberings with the characters " I " of 3 numberings.
Simultaneously, be character pattern with the mode-definition of one group of two character types, and two mode-definitions that numbering occurs of two adjacent characters that type is corresponding with a kind of character pattern are for numbering scheme occurring.For example, from part character stream " AAIA " obtain three kinds of character patterns (A, A), (A, I) and (I A), numbering scheme (1,2), (2,1) and (1,3) occur and obtain three kinds from part character stream " AAIA ".
Then, as shown in Figure 5, will gather with the corresponding respectively one or more numbering schemes that occur of a kind of character pattern by order that numbering increases progressively occurring, and the numbering scheme table appears in preparation, as the index file of each searching for text each character pattern.
Subsequently, when designated character being flowed " AIA ", when for example preparing as a keyword, just from this keyword obtain the first designated character pattern (A, I) and the second designated character pattern (I, A).The retrieval and the consistent one or more character streams of this keyword from searching for text in order to associate with a series of character patterns of searching for text by the designated character pattern that makes keyword, from appearance numbering scheme table as shown in Figure 5, pick up corresponding to the designated character pattern (A of keyword, I) and (I, A) one or more numbering schemes that occur of Yi Zhi each character pattern.Therefore, occur numbering scheme (2,1), (4,2) and (8,3) corresponding to first character pattern (A, I), and occur numbering scheme (1,3) and (2,5) corresponding to second character pattern (I, A).Like this, with the first character pattern (A, I) corresponding one specific occur second element of numbering scheme with corresponding to the second character pattern (I, A) another be specific when the first element unanimity of numbering scheme occurring, and be consistent with keyword corresponding to the character stream of the specific searching for text that numbering scheme occurs.In this example, second element of numbering scheme (2,1) occurs and first element (1) unanimity of numbering scheme (1,3) occurs, and second element (2) and first element (2) unanimity that numbering scheme (2,5) occurs of numbering scheme (4,2) occur.Therefore, there are two character streams consistent in the searching for text with keyword.Subsequently, referring to appearance number table as shown in Figure 4, from searching for text, retrieve two character streams that numbering scheme occurs corresponding to specific.
Fig. 6 is the information searching apparatus block scheme according to the present invention first, second and the 3rd embodiment.
As shown in Figure 6, information searching apparatus 21 comprises a text database 22, is used for storing a plurality of searching for texts of being prepared as database by the operator; Text record generation unit 23 is used for reading one by one the searching for text of storage in the text database 22, and produces the text record corresponding to each searching for text; Be used for the keyword reading device of the keyword that the read operation person prepares; Be used for modification text that the read operation person prepares and produce the modification text reading unit 25 of revising the text record; The appearance number table preparatory unit 26 of number table appears in preparation, and a plurality of appearance numberings that are used for the character of each type listed in the modification text record that text record that produces for text record generation unit 23 in the table or modification text reading unit 25 produce; Index file preparatory unit 27 is used for preparing the numbering scheme table to occur, lists one or more appearance numberings for each character pattern in the table, as the appearance number table index file that produces from number table preparatory unit 26 occurs; Index file storage unit 28 is used for storing and the index file (or the numbering scheme table occurring) prepared in the appearance number table that produces and the index file preparatory unit 27 occurs in the number table preparatory unit 26; The expansion area is provided with unit 29, is used for being provided with in addition the one or more expansion area Re1 behind corresponding with each character pattern in the index file one or more posting field that numbering scheme occurs; Character pattern generation unit 30 is used for producing a plurality of designated character patterns according to the keyword that keyword reading unit 24 reads; Index file is searched unit 31, consistent by a series of three-character doctrines that make index file with the designated character pattern of a pair of keyword, at consistent with keyword respectively one or more searching character streams, the index file of storage in the search index file storage unit 28; Seek operations judging unit 32, be used for judging index file search the unit whether at searched the text record corresponding to the numbering scheme that occurs of a character pattern that produces in the character pattern generation unit 30; Result for retrieval output unit 33, be used for temporary transient storage with search in the unit 31 selecteed character pattern to corresponding a series of three-character doctrines, and the output index file is searched the searching character stream that obtains in the unit 31 as result for retrieval; And numbering scheme updating block 34 appears, is used for according to revising the appearance numbering scheme that modification text that text reading unit 25 reads upgrades index file in the text retouching operation.
In said structure, the seek operations that the index file that one or more searching for texts are deposited as one or more index files is deposited an operation and an index file that flows according to consistent with a certain keyword respectively one or more searching characters of method for searching search has been described.So, revise text reading unit 25 or numbering scheme updating block 34 occurs noting be used among first embodiment, but be used among the second and the 3rd embodiment.
Fig. 7 deposits operational flowchart according to the index file that method for searching is carried out.
As shown in Figure 7, one by one a plurality of searching for texts are read the text record generation unit 23 from text database 22, and recording mechanism Ni is appended on each searching for text, produce the text record (step S201) corresponding with each searching for text.Subsequently, for example current text record of general " AAIAAIANNAANAI " is sent to and number table preparatory unit 26 occurs.In number table preparatory unit 26 occurring, with the initial zero setting of the appearance numbering N (C) of each type character.Here, the symbol C among the N (C) represents the type (C=A represents the type of character A, and C=I represents the type of character I, and C=N represents the type of character N) of character.Subsequently, character pointer P1 is initially set to the beginning character (step S202) of current text record.Then, when character made character pointer P1 displacement one by one, the order by each character types takes place will belong to one or more character numbers (step S203) of identical characters type since 1.So, each character types has been listed a plurality of appearance numbering N (C).Subsequently, judge that whether character pointer P1 is set to the end character (step S204) of current text record, if character pointer is not arranged on the end character of current text record, just continues step S203.On the contrary, if count down to the appearance numbering N (C) of current text record end character, number table appears in the part of just preparing current text record as shown in Figure 4, and this table is stored in the index file storage unit 28.Then, judge in the text record generation unit 23 whether remain with untreated text record (step S205) among step S202 and the S203.If remain with the text record of undressed back, the text record of untreated back among treatment step S202 and the S203 in an identical manner just, and number table appears in another part of preparing the text record of back, and this table is stored in the index file storage unit 28.After in step S202 and S203, having handled all text records, in number table preparatory unit 26 occurring, finish the appearance number table of all text records and prepare, and the appearance number table of all text records is stored in the index file storage unit 28.
Then, prepare to occur the numbering scheme table by index file preparatory unit 27.Know clearly it, numbering scheme (step S207) is appearred in the top of current text record N1 that the appearance number table of numbering scheme pointer P2 initial setting storage to the index file storage unit 28 occurs, when making to pattern pattern pointer P2 displacement one by one, list one or more numbering schemes (step S208) that occur of pointing out by pattern pointer P2 one by one at the index file that is used for each character pattern.Then, whether judgment model pointer P2 is arranged to the end that the current text of number table record N1 occurs and numbering scheme (step S209) occurs.Numbering scheme do not occur if pattern pointer P2 is arranged to the end of current text record N1, then continue step S208.On the contrary, numbering scheme appears if listed the end of current text record N1, so just be each character pattern of current text record N1, two expansion area Re1 (step S210) be set in addition in one or more posting field back that numbering scheme occurs.The increase of expansion area Re1 is provided with unit 29 by the expansion area to be carried out, and the end corresponding with each character pattern numbering scheme occurs and be recorded in once more in the first expansion area Re1, and in second expansion area of back, first expansion area record not whatever.The second expansion area Re1 is arranged the additional appearance numbering that obtains in the operation of recording indexes file modification.So the appearance numbering scheme table that is used for current text record N1 is prepared by index file preparatory unit 27 as shown in Figure 8, and the index file N1 of conduct deposits in the indexed file storage unit 28.Then, whether untreated text record is retained in the text record generation unit 23 (step S211) among determining step S207 and the S208.If remain with the undressed text record of back, just handle the text record of back in the same way to S210 at step S207, and the numbering scheme table appears in another of text record of preparing the back, and it is stored in the index file storage unit 28, as another index file.After in step S207 and S208, having handled all text records, with regard to the preparation (step S212) of one or more index files of finishing all text record in the indexed file preparatory unit 27, and this index file is stored in the index file storage unit 28.That is, finish the index file that all texts are write down and deposit operation.
Like this, (Ai, first and second elements Bi) (or occur numbering) Ai and Bi just the series arrangement of numbering occurred and have got up by increasing the appearance numbering scheme corresponding with each character pattern.That is, satisfy concern A1<A2<...<Am and B1<B2 ...<Bm.
Method for searching according to the index file of consistent with a certain keyword respectively one or more character streams of the search of method for searching is then described.
Fig. 9 is the process flow diagram of the seek operations of carrying out according to this method for searching.
As shown in Figure 9, read a certain keyword of preparing by the operator (step S221), and produce the character pattern { C1 (p), C2 (p) } of a series of appointments, { C1 (p+1), C2 (p+1) } (step S222) from this keyword by generation device 30.Here, symbol C1 (p) represents first character in p the designated character pattern of this keyword, and symbol C2 (p) represents second character in p the designated character pattern of this keyword.For example, from keyword " AIA ", obtain first designated character pattern " AI " and second designated character pattern " IA ".Subsequently, by searching unit 31,, and represent the character pattern (step S224) of p appointment of this keyword with a keyword pointer P3 to recording mechanism N1 of current searching for text recording setting.First character (p=1) of representing this keyword with pointer P3 at the beginning.Subsequently, select p the designated character pattern { C1 (p), C2 (p) } of this keyword and the character pattern { C1 (p+1), C2 (p+1) } of p+1 appointment, as the character pattern (step S225) of two appointments relevant with search operaqtion.For example, at first be provided with two uppermost character patterns { C1 (1), C2 (1) } of this keyword=(A, I) and { C1 (2), C2 (2) }=(I, A).By keyword reading unit 24 execution in step S221 to S225.
Then, as shown in figure 10, in the index file from be stored in index file storage unit 28 will with p designated character pattern { C1 (p), C2 (p) } corresponding appearance numbering scheme { N1 (p, j), N2 (p, j) } read index file and search unit 31 (step S226).Here, symbol N1 (p, j) character pattern { C1 (p) of expression and p appointment, C2 (p) } corresponding a plurality of j appearance numbering N1 that first element of numbering scheme occurs in the numbering scheme that occur, and symbol N2 (p, j) the individual appearance numbering N2 that second element of numbering scheme occur of expression j.From index file, read at the beginning and numbering scheme { N1 (p, 1), N2 (p, 1) }=(2,1) occurs.Equally, as shown in figure 10, will number with the appearance of (p+1) individual appointment C1 (p+1), and the appearance numbering scheme that C2 (p+1) is corresponding { N3 (p+1, k), N4 (p+1, k) } read in the index file search unit 31 (step S227).Here, symbol N3 (p+1, k) expression and p+1 designated character pattern { C1 (p+1), C2 (p+1) } corresponding a plurality of k appearance numbering N3 that first element of numbering scheme occurs in the numbering scheme that occur, and symbol N4 (p, k) the individual appearance numbering N4 that second element of numbering scheme occur of expression k.From index file, read at the beginning and numbering scheme { N3 (p, 1), N4 (p, 1) }=(1,3) occurs.
Then, by index file search unit 31 judge occur numbering scheme N1 (p, j), N2 (p, j) } second element appearance numbering N2 whether with occur numbering scheme N3 (p+1, k), N4 (p+1, k) } the appearance numbering N3 unanimity (step S228) of first element.If numbering N2 is consistent with appearance numbering N3, then with numbering scheme { N1 (p appears, j), N2 (p, and { N3 (p+1 j) }, k), N4 (p+1, the series of being made up of three characters of k} correspondence is just lined up in current searching for text record N1, and temporarily count N1 with the record of current searching for text and be stored in (step S229) in the result for retrieval output unit 33, and program proceeds to step S230, retrieval and designated character pattern { C1 (p) from current searching for text N1, C2 (p) } and another corresponding three-character doctrine series of { C1 (p+1), C2 (p+1) }.On the contrary, if it is different with appearance numbering N3 to occur numbering N2 in step S228, then program proceeds to step S230, does not export any three-character doctrine series.
In step S230, in searching unit 31, judge occur numbering scheme N1 (p, j), N2 (p, j) } the appearance numbering N2 of second element whether be higher than occur numbering scheme N3 (p+1, k), N4 (p+1, k) } the appearance numbering N3 of first element.If this appearance numbering N3 is higher than numbering N2 appears, the value of j is increased progressively (j ← j+1), with check the next one occur numbering scheme N1 (p, j), N2 (p, j) } (step S231), and judge occur numbering scheme N1 (p, j), N2 (p, j) } (whether p j) is higher than appearance numbering N2 (p, j-1) (step S232) to the appearance of first element numbering N2.That is, judge occur numbering scheme N1 (p, j), N2 (p, j) } whether be recorded in the first expansion area Re1.If numbering N2 (p, j) be higher than occur numbering N2 (p, j-1), then because occur numbering scheme N1 (p, j), N2 (p, j) } be not to be recorded among the Re1 of expansion area, so program is got back to step S228.On the contrary, if numbering N2 (p, j) be equal to or less than occur numbering N2 (p, j-1), then because occur numbering scheme N1 (p, j), N2 (p, j) } be recorded among the first expansion area Re1, so program proceeds to step S237.Equally, be lower than appearance numbering N2 if numbering N3 in step S230, occurs, the value of k is increased progressively (k ← k+1) (step S233), and judge and numbering scheme { N3 (p+1 occurs, k), N4 (p+1, k) } the appearance numbering scheme N3 (p+1 of first element, k) whether be higher than appearance numbering N3 (p+1, k-1) (step S234).That is, judge occur numbering scheme N3 (p+1, k), N4 (p+1, k) } whether be recorded among the first expansion area Re1.(p+1, (p+1, k-1), then program is got back to step S228 k) to be higher than appearance numbering N3 if number N3.On the contrary, (p+1, (p+1, k-1), then program proceeds to step S237 k) to be equal to or less than appearance numbering N3 if number N3.Equally, equal to occur numbering N2, then make the value of j and k increase progressively (k ← k+1 if numbering N3 in step S230, occurs, j ← j+1) (step S235), and judge occur numbering scheme N1 (p, j), N2 (p, j) } (whether p j) is higher than and numbering N2 (p occurs for the appearance numbering N2 of second element, j-1) and judge and numbering scheme { N3 (p+1 to occur, k), N4 (p+1, k) } the appearance numbering N3 (p+1 of first element, k) whether be higher than appearance numbering N3 (p+1, k-1) (step S236).That is, judge occur numbering scheme N1 (p, j), N2 (p, j) } whether be recorded among the first expansion area Re1, and judge occur numbering scheme N3 (p+1, k), N4 (p+1, k) } whether be recorded among the Re1 of expansion area.(p j) is higher than and numbering N2 occurs (p, j-1), and (p+1 k) is higher than and numbering N3 occurs (P+1, k-1), then program is got back to step S228 numbering N3 to occur if numbering N2.On the contrary, if numbering N2 (p, j) be equal to or less than occur numbering N2 (p, j-1) or numbering N3 occurs (p+1 k) is equal to or less than and numbering N3 occurs (p+1, k-1), program proceeds to step S237 so.
For example, if keyword reading unit 24 reads keyword " AIA ", the appearance numbering scheme (2 that writes down among the first expansion area Re1 then, 1), (4,2) and (8,3) and numbering scheme (8 occurs, 3) corresponding to first (p=1) designated character pattern (A, I), and the appearance numbering scheme that in the expansion area, writes down (1,3) and (2,5) and occur numbering scheme (2,5) corresponding to second (p=2) designated character pattern (I, A).So the appearance numbering of first element of numbering scheme (1,3) to occur consistent with first (k=1) to judge appearance numbering that second element of numbering scheme (2,1) appear in first (j=1) in step S228.Because numbering scheme (2 appears in first (j=1), 1) numbering scheme (1 appears in the appearance of second element numbering " 1 " and first (k=1), 3) appearance of first element numbering " 1 " unanimity, so in step S229, number (2 with appearance, 1) and (1,3) corresponding three characters " AIA " series temporarily be stored in the output unit 33.Subsequently, the value of j and k is increased progressively, and in step S228, judge the appearance that first element of numbering scheme (2,5) appears in appearance numbering " 2 " and second (k=2) that second element of numbering scheme (4,2) appear in second (j=2) and number " 2 " unanimity.So, in step S229, will consistent three-character doctrine " AIA " series temporarily be stored in the output unit 33 with numbering scheme (4,2) and (2,5) occurring.Then, the value of j and k is increased progressively, and judge the appearance numbering that first element of numbering scheme (2,5) appears in second (k=20) at step S236 and equal the appearance numbering " 2 " that first element of numbering scheme (2,5) appears in the 3rd (k=3).So program proceeds to step S237.
In step S237, because searched with corresponding to p designated character pattern of keyword the numbering scheme that occurs and corresponding to (p+1) individual designated character pattern the numbering scheme that occurs all characters of relevant current searching for text, whether be the last designated character pattern in this keyword so judge (p+1) individual designated character pattern.In other words, judge whether to have checked the appearance numbering scheme corresponding by searching unit 31 with all designated character patterns that obtain from this keyword.If (p+1) individual designated character pattern is not the last designated character pattern in this keyword, then the value of p is increased progressively (p ← p+1), and program is got back to step S225, retrieval and another one or more pairs of characters to the consistent current searching for text of designated character pattern.On the contrary, if (p+1) individual designated character pattern is last the designated character pattern in this keyword, so because checked the appearance numbering scheme corresponding, so judge in step S229, whether each storage to the three-character doctrine series of designated character pattern to be existed (step S239) with all designated character patterns that obtain from this keyword.If in step S229, stored each three-character doctrine series to the designated character pattern, then produce one or more searching character streams by connecting each three-character doctrine series to the designated character pattern, and the segment information that output expression searching character stream exists and the recording mechanism N1 of current searching for text are as the result for retrieval (step S240) from result for retrieval output unit 33.On the contrary, if in step S229, do not store each three-character doctrine series to the designated character pattern, then output expression searching character flows the recording mechanism N1 of a non-existent segment information and current searching for text, as the result for retrieval (step S241) from result for retrieval output unit 33.
As mentioned above, in first embodiment, each searching for text record Ni is prepared to occur number table in number table preparatory unit 26 occurring, and be stored in the index file storage unit 28, when unit 29 being set two expansion area Re1 being set by the expansion area in addition for each designated character pattern, appearance number table according to storage in the index file storage unit 28 in the index file preparatory unit 27 for preparing one, each searching for text record Ni the numbering scheme table occurred, and this table is stored in the index file storage unit 28 index file as each searching for text record, indexed file is searched and to be found out in the unit 31 in the keyword that is read by keyword reading unit 24 and the mutual close corresponding a pair of numbering scheme that occurs of a pair of designated character pattern, indexed file is searched and to be judged in the unit 31 near each whether the designated character pattern being existed mutually in the keyword numbering scheme to occur right, and a stream existence of output expression searching character or a non-existent segment information, as result for retrieval from result for retrieval output unit 33.So, each searching for text of storage in the searching text database 22 substantially, and can retrieve with the consistent respectively one or more searching characters of keyword and flow.
In addition, be to get up because the appearance numbering scheme of each character pattern in the numbering scheme table occurs by the series arrangement that numbering increases occurring, so, can sequential search that numbering increases occur each numbering scheme occurs by making.Therefore, can search the occur numbering scheme corresponding at a high speed with a pair of designated character pattern of keyword.
In first embodiment, as shown in Figure 5, the appearance numbering scheme that each character pattern in the numbering scheme table occurs is the tactic of numbering increase to occur.Yet also can adopt the order that the appearance numbering is reduced to arrange and numbering scheme occur.
In addition, also the second expansion area Re1 is set in the indexed file for each character pattern.Yet, a plurality of second expansion areas are set for each character pattern in can indexed file.
In addition, retrieve the one or more character streams consistent by in information searching apparatus 21, searching searching for text with keyword.But first embodiment is not the retrieval that only is confined to character stream.That is, can retrieve the one or more data stream with certain pattern consistent with a certain keyword.
Renewal operation according to the index file of storage in the index file storage unit 28 of second embodiment of the invention is described below.
As shown in Figure 6, when reading the modification text of preparing by the operator and revising the text record that produces modification in the text reading unit 25, the modified number table that occurs is to produce according to the modification text record that occurs in the number table preparatory unit 26, and is stored in the index file storage unit 28.Subsequently, prepare to be used to revise the modified numbering scheme table that occurs of text for each character types, as in the index file preparatory unit 27 through revising index file N1R, and the primary index file N 1 of storage in the index of reference file storage unit 28 replaces the index file N1R that revises.Like this, when the modified quantity that numbering scheme occurs of modified index file N1R specific character pattern is higher than the quantity of appearance numbering scheme of primary index file N1, just add modified expansion area Re2, thereby write down final modified the occur numbering scheme corresponding once more with each specific character pattern of modified index file N1R.
Describe the renewal operation in detail below with reference to Figure 11.
Figure 11 describes is renewal operational flowchart according to the index file of storage in the index file storage unit 28 of second embodiment.
As shown in figure 11, the modified searching for text " AAIAIIANAAINAI " that obtains read by revising text reading unit 25 by revising the original searching for text " AAIAAIANNAANAI " that files as primary index file N1, and recording mechanism N1R is attached on the modified searching for text, produces a modified text record (step S301).Subsequently, in number table preparatory unit 26 occurring,, and character pointer P1 is initially set to the initial character (step S302) that modified text writes down N1R with the initial zero setting of the appearance numbering N (C) of each type character.Subsequently, with character pointer P1 displacement, one or more characters that will belong to the identical characters type are numbered (step S303) by the order that each character types occurs from 1 by character by character.So, listed a plurality of appearance numbering N (C) at each character types.Then, judge whether character pointer P1 is placed the last character (step S304) of modified text record N1R.If character pointer P1 is not placed in the last character of modified text record N1R, just continue step S303.On the contrary, if counted the appearance numbering N (C) of modified text record N1R last character, so just prepare the modified number table that occurs of modified text record N1R as shown in figure 12, and it is stored in the index file storage unit 28.
Subsequently, find out the primary index file corresponding by index file preparatory unit 35 from index file storage unit 28 with text recording mechanism N1, initially place the top of primary index file N1 top character pattern numbering scheme to occur with numbering scheme pointer P2 occurring, and numbering scheme pointer P2 displacement will appear, to point out the appearance numbering scheme (step S305) of primary index file N1 one by one.Also more new model pointer P4 initially places and points out the modified modified numbering scheme that occurs that the number table top occurs, and more new model pointer P4 displacement is to point out the modified modified numbering scheme (step S306) that occurs that number table occurs one by one.Then, by numbering scheme updating block 34 occurring, replace among the primary index file N1 one by one by the appearance numbering scheme (step S307) that numbering pointer P2 represents occurring by the modified numbering scheme that occurs that new model pointer P4 more points out with modified occur in the number table one by one.
Subsequently, judge the quantity (step S308) whether the modified quantity that numbering scheme occurs of each character pattern becomes and be higher than the appearance numbering scheme among the primary index file N1.If the quantity (referring to Figure 13) that the modified quantity that numbering scheme occurs of specific character pattern becomes and is higher than the appearance numbering scheme of specific character pattern among the primary index file N1, then by the expansion area unit 29 being set increases one or more recording areas, to write down all modified numbering schemes (step S309) that occur of specific character pattern among the modified index file N1R.Subsequently, whether judgment model pointer P4 is arranged to modified last modified numbering scheme (step S310) that occurs that number table occurs.If pattern pointer P2 is not arranged to modified last modified numbering scheme that occurs that number table occurs, then continue step S307.On the contrary, if modifiedly occur that last of number table is modified the index file N1R that numbering scheme has been listed in modification to be occurred and suffered, then each character pattern among the index file N1R that revises is provided with in addition and follows at one or more modified modification expansion area Re2 (step S311) that the numbering scheme back occurs.Interpolation through revising expansion area Re2 is provided with unit 29 by the expansion area to be carried out, and the last modified numbering scheme that occurs is recorded in the modified expansion area Re2 once more.Like this, may further be each character pattern setting follows at one or more modified a plurality of expansion area Re2 that the numbering scheme back occurs.So as shown in figure 14, the modified numbering scheme that occurs of modified text record N1R is revised unit 34 preparations by numbering scheme occurring, and is stored in the index file storage unit 28.
Among Figure 14, each character pattern is listed last appearance numbering scheme for twice, and (A I) is provided with modified expansion area Re2 to character pattern.And, be lower than the quantity of the appearance numbering scheme of specific character pattern among the primary index file N1 if the modified quantity that numbering scheme occurs of a certain specific character pattern becomes, then the one or more recording areas among the primary index file N1 changed over one or more expansion area Re2.For example, (A is A) with (A, N) each pattern in recording areas is altered to a modified expansion area Re2 with character pattern.And, although if in primary index file N1, have a specific character pattern one or morely numbering scheme occurs but do not have the modified numbering scheme that occurs of specific character pattern, then the one or more recording areas with specific character pattern among the primary index file N1 are altered to one or more expansion area Re2.For example, character pattern (N, I) and (N, N) in a recording areas of each pattern be altered to a modified expansion area Re2.
Subsequently, with first embodiment (Fig. 9) in same mode, carry out the one or more character streams consistent with a certain keyword searched for modified index file N1R seek operations.
Therefore, even the quantity of the modified appearance of the specific character pattern numbering quantity of the appearance numbering scheme that is higher than specific character pattern in the primary index file that becomes also can be listed the modified numbering scheme that occurs immediately among the modified index file N1R in one or more recording areas and an expansion area Re1.So, can carry out renewal operation at a high speed to index file.
In addition, even the quantity of the recording areas of a certain specific character pattern and expansion area Re1 is lower than the modified quantity that numbering scheme occurs of specific character pattern, but because modifiedly also be provided with a modified expansion area Re2 in addition when numbering scheme occurring in revising index file N1R, having listed one at every turn, so can in modified index file N1R, list all modified numbering schemes that occur of specific character pattern easily.
The part deletion action of the text record of storing as the primary index file according to a deletion part in the indexed file storage unit 28 of third embodiment of the invention is then described below.
As shown in Figure 6, when reading when deleting the modified text that a certain original text N1 obtains and produce modified text record in modified text reading unit 25 by part, the modified number table that occurs occurs according to modified that modified text record produces and is stored in the index file storage unit 28 in the number table preparatory unit 26.Subsequently, for each character pattern is prepared the modified numbering scheme table that occurs of modified text, as the modified index file N1D in the index file preparatory unit 27, and the primary index file N1 of storage in the index of reference file storage unit 28 replaces the index file N1D of modification.Like this, because the quantity that numbering scheme quantity must be equal to or less than the appearance numbering scheme of primary index file N1 appears in the modified of a certain specific character pattern of modified index file N1D, any expansion area is not set in addition.
Describe the part deletion action in detail below with reference to Figure 15.
Figure 15 illustrates to carry out the process flow diagram of part deletion action according to the 3rd embodiment for being stored in index file in the index file storage unit 28.
As shown in figure 15, read modified searching for text " AAIAAIAN " by modified text reading unit 25, it is by obtaining as the original searching for text " AAIAAIANNAANAI " that index file files by the part deletion in the indexed file storage unit 28, and recording mechanism N1D is attached on the modified searching for text, produce modified text record N1D (step S401).Subsequently, as shown in figure 16, with the step S302 of second embodiment to the identical mode of S304, prepare the modified number table that occurs of modified text record N1D, and this table be stored in (step S402 is to S404) in the index file storage unit 28.
Subsequently, with with the step S305 of second embodiment and S306 in identical mode, when making pattern pointer P2 and P4 displacement (step S405 and S406), by numbering scheme updating block 34 occurring, numbering scheme occurs and replace one by one and occur the represented appearance numbering scheme (step S407) of numbering scheme pointer P2 among the primary index file N1 with modified pointed modified of new model pointer P4 more that occur in the number table.Like this, obtain, replace any numbering scheme that occurs that writes down among the Re1 of expansion area so numbering scheme need not modifiedly occur because modified searching for text is deleted original searching for text by part.
Then, whether judgment model pointer P4 is set to modified last modified numbering scheme (step S408) that occurs that number table occurs.If pattern pointer P2 is not set to modified last modified numbering scheme that occurs that number table occurs, then continue step S407.On the contrary, if in modified index file N1D, listed modified last the modified numbering scheme that occurs that number table occurs, then obtained the modified numbering scheme table that occurs as shown in figure 17.Then, the last appearance numbering scheme through a modification that writes down in the recording areas that once more will be corresponding with each character pattern is recorded among an at the heel recording areas or the expansion area Re1 (step S409).Then, by index file search unit 31 judge modified occur whether existing in the numbering scheme table do not appear at character pattern in the modified searching for text, obsolete (step S410).If exist and do not appear in revising searching for text, obsolete character pattern, then corresponding with obsolete character pattern top recording areas just becomes an expansion area Re3, with the appearance numbering scheme that in the recording areas of top, writes down of deletion, and one section invalid mode represented information stores (step S411) in the Re3 of expansion area.That is, one or more recording areas and the expansion area Re1 with Re3 back, expansion area treats as the dead space.For example, as shown in figure 18 because do not exist in the modified searching for text character pattern (N, A), so (N A) treats as obsolete character pattern, and the top recording areas becomes an expansion area Re3 character pattern.And (N N) treat as an obsolete character pattern, and the top recording areas corresponding with obsolete character pattern becomes an expansion area Re3 with character pattern.
Then, with first embodiment (Fig. 9) in identical mode, the seek operations that the one or more character streams consistent with a certain keyword are searched modified index file N1D.Like this, even record identical last occur maintaining index file in the recording areas of a pair of recording areas back of numbering scheme or the expansion area and produce the specific numbering scheme of listing in the operation (Fig. 7) that occurs, because seek operations is so that order that numbering increases occurs and arrange under the situation of the modified appearance numbering scheme that each character pattern in the numbering scheme table occurs and carry out, so do not carry out the seek operations that numbering scheme occurs to specific.For example, although with character pattern (A, A) correspondence, in the second and the 3rd recording areas, write down and occurred after the numbering scheme (3,4), in the Re1 of expansion area, write down and numbering scheme (6 occurs, 7), but do not carry out the appearance numbering scheme (3,4) that writes down among the 3rd recording areas and the expansion area Re1 and the seek operations of (6,7).Same reason is not carried out seek operations to any numbering scheme that occurs that writes down in the dead space.
Therefore, because in the part deletion action, do not delete all unnecessary appearance numbering schemes, can carry out the part deletion action at a high speed.
The information searching apparatus of realizing second purpose according to the 4th embodiment is then described.
Figure 19 is the block scheme according to the information searching apparatus of the present invention the 4th and the 5th embodiment.
As shown in figure 19, information searching apparatus 41 comprises:
Request and Data Receiving unit 42, be used for receiving the deal with data group of one section real-time processing data and operate, increase in real time operation or the real-time processing request of deletion action in real time according to real-time processing data request real-time update, and the retrieval request of the retrieval group of reception searching character stream and request search operaqtion
Real-time processing data deposit unit 43 is used for depositing the deal with data group of real-time processing data and according to the real-time processing request of real-time processing request,
Real-time processing data storage unit 44 is used for depositing a plurality of deal with data groups that position (BtmNew) storage is deposited one by one by deposit unit 43 in data;
Real-time processing data condition administrative unit 45, be used for managing in the real-time processing data storage unit 44 service condition of the deal with data group of storage, and control real-time processing data storage unit 44 is deposited each deal with data group of locating to be deposited by deposit unit 43 in position (BtmNew) to be stored in data;
Video data generation unit 46 is used for asking according to corresponding the processing in real time, according to each real-time processing data of being deposited by real-time processing data deposit unit 43, produces one section video data;
The video data storage unit 47 of storage video data;
The display 48 that shows video data;
Index file storage unit 49 is used for storing index file, the in store one or more original searching for text of forming by the character that is serial arrangement respectively substantially in the index file, and each original searching for text discerned by data number,
Real-time processing unit 50, be used under the control of real-time processing data condition administrative unit 45, carry out real-time processing one by one in real time to each data set in the deal with data group of storage in the real-time processing data storage unit 44, and in the index file storage unit 49 storage index file in the deal with data group file, to revise index file
Index file fine unit 51; Be used for the index file of storage in the search index file memory cell 49; Thereby when request receives the retrieval group of retrieval request and searching character stream with data reception unit 42; No matter in real-time processing unit 50 to real-time processing data memory cell 44 in the real-time processing of all deal with data groups of storage whether finish; Retrieval receives the consistent one or more character streams of searching character stream that unit 42 receives with request and data; And output is flowed the result for retrieval that consistent character flows with searching character
Character stream retrieval unit 52, be used for searching all deal with data groups, when receiving the retrieval group of retrieval request and searching character stream by request and Data Receiving unit 42, all deal with data groups that search is stored in the real-time processing data storage unit 44 and real-time processing unit 50 is not handled flow consistent one or more character streams thereby retrieve respectively with the searching character that request and Data Receiving unit 42 receive from the deal with data group;
Result for retrieval is revised unit 53, be used for according to the deal with data search result that obtains in the character stream retrieval unit 52, be modified in index file and search the result for retrieval that obtains in the unit 51, thereby whether the real-time processing that real-time 50 pairs of deal with data groups of processing unit are carried out finishes all deal with data groups that all reflection request and Data Receiving unit 42 are received in result for retrieval; And
Result for retrieval output unit 54 is used for exporting result for retrieval and revises the modified result for retrieval that unit 53 is obtained.
In said structure, described the deal with data group of one section real-time processing data and handled depositing of request in real time with reference to Figure 20.As shown in figure 20, when a certain user asks one section real-time processing data do processing in real time by importing the deal with data group of being made up of real-time processing data and real-time processing request, the deal with data group of being made up of real-time processing data group and processing request is in real time received by request and Data Receiving unit 42, and deposit (step S501), and information searching apparatus 41 is set to treatment conditions (step S502) by real-time processing data condition administrative unit 45 by real-time processing data deposit unit 43.For example, as shown in figure 21, by real-time processing data and the deal with data group formed of the request of handling in real time point out with data number to be that the renewal that the one piece of data of D4 is updated to character stream " EOK " is operated.Then, check that by real-time processing data deposit unit 43 data in the real-time processing data storage unit 44 that real-time processing data condition administrative units 45 are provided with deposit position (BtmNew) (step S503), and by real-time processing data and the deal with data group formed of the request of handling in real time be stored in data in the real-time processing data storage unit 44 and deposit position (BtmNew) and locate (step S504).Like this, as shown in figure 22, a pair of sign F 1 and F2 are added on the deal with data group of being made up of real-time processing data and processing request in real time by real-time processing data condition administrative unit 45.Sign F1 is a retrieval sign of pointing out search condition, and another sign F2 is a real-time processing sign of pointing out to upgrade, increase or delete situation.So, as shown in figure 23, be stored in the real-time processing data storage unit 44 with sign with real-time processing data and the real-time a plurality of deal with data groups of handling request that Data Receiving unit 42 receives by request one by one.Then, data are deposited displacement to lower position (BtmNew ← BtmNew+1) (step S505).
Then, when real-time processing request was pointed out to upgrade or increased operation, the real-time processing data of being deposited by the real-time processing data deposit unit 43 in the video data generation unit 46 produced one section video data.Subsequently,, then replace one section old video data of storage in the video data storage unit 47,, then in video data storage unit 47, increase video data (step S506) if increase operation with video data if upgrade operation.And,, then in video data storage unit 47, delete the one piece of data of representing data number by real-time processing data if deletion action is pointed out in the request of handling in real time.For example, the one section video data that produces as shown in figure 24 with video data generation unit 46, and as shown in figure 25 is that the data segment of D4 changes over " EOK " with data number.Subsequently, with the expression real-time processing data and the one piece of data that has finished of depositing of handling in real time the processing of request data set deposit ending message and be sent to display 48 (step S507) from real-time processing data deposit unit 43, finish thereby notify the user to deposit operation.Subsequently, with the condition (step S508) of pending requests such as information searching apparatus 41 is set to.
Therefore, when the user asked another to handle (step S509) in real time, repeating step S501 was to step S508.
Below, with reference to Figure 26 describe to be stored in the real-time processing data storage unit 44 by real-time processing data and the deal with data group formed of the request of handling in real time handle and the modification of the index file carried out.
Figure 26 illustrates being stored in the process flow diagram that index file is made amendment and operated in the index file storage unit 49.
As shown in figure 26, by real-time processing unit 50 by 45 pairs of real-time processing data condition administrative units by one section real-time storage data and the deal with data groups formed of the request of handling in real time whether be stored in and judge (step S511) in the real-time processing data storage unit 44.If the deal with data group is not stored in the real-time processing data storage unit 44, repeating step S511 after the stand-by period then.On the contrary, if one or more deal with data groups have been stored in the real-time processing data storage unit 44, then the deal with data group is placed in the real-time processing data storage unit 44 from the top and handles on position (TopNew) handles processing position the position (BtmNew) to the bottom, the top is handled the position and is represented with handling pointer P5, and under the control of real-time processing data condition administrative unit 45, be sent to real-time processing unit 50 (step S512) being stored in the top deal with data group of locating processing position, top (TopNew), and the real-time processing sign F2 of top deal with data group is arranged to " on (unlatching) " condition (step S513).For example, as shown in figure 23, the top deal with data group of one a section real-time processing data " AIU " and a real-time update request is sent to real-time processing unit 50.Subsequently,, top deal with data group is handled in real time, to revise the index file (step S514) of storage in the index file storage unit 49 according to top deal with data group.For example, replace the one piece of data of storing with data number D5 in the index file with real-time processing data " AIU ".
The structure of index file as shown in figure 27.As shown in figure 27, (C1, C2), each character pattern is represented the mutual close first character C1 and the second character C2 in the searching for text to list a plurality of character patterns in the indexed file.To each character pattern in the index file, (PC1, DN), they represent the position PC1 of the first character C1 in the searching for text and the data number DN of identification searching for text respectively also to list one or more character information patterns.For example, shown in Figure 28 A, if original searching for text with character stream " AIAAIUAU " expression data number DN=1, then by the character number of original searching for text being determined the character position of each character from 1 by the order that takes place, and in the primary index file, list corresponding to character pattern (A, A) character information pattern (3,1), corresponding to character pattern (A, I) two character information patterns (1,1) and (4,1), corresponding to character pattern (A, U) character information pattern (7,1), corresponding to character pattern (I, character information pattern (2,1) A), corresponding to character pattern (I, U) character information pattern (5,1) with corresponding to character pattern (U, character information pattern (6,1) A).Subsequently, if the character stream " AIU " with data number DN=5 in step S514 is added on the original searching for text, so because character stream " AIU " is divided into two character pattern (A, I) and (I, U), corresponding to character pattern (A, I) character information pattern (1,5) and corresponding to character pattern (I, character information pattern (2,5) U) just adds to the primary index file.Thereby, shown in Figure 28 B, revised the primary index file.
And,, so just deleted and a certain deleted data number all relevant character information patterns if the primary index file is carried out deletion action.And, if the primary index file is upgraded operation, so just deleted all character information patterns relevant, and one or more renewal character information patterns just added to the primary index file with a certain renewal data number.
Then, after the real-time processing to top deal with data group finishes, the real-time processing sign F2 of top deal with data group is configured to " off (shutoff) " condition (step S515), and (TopNew ← TopNew+1) is to handle (step S516) in real time to next data set just to be displaced to a certain lower position by the position of handling the real-time processing data storage unit 44 that pointer P5 points out.So, point out from the next deal with data group of second the position placement in processing position at top with handling pointer P5.Equally, point out to have placed the processing position, previous top of top deal with data group treated in step S514 with deletion pointer P6, and be set to the real-time processing end position (BtmOld) (step S517) of bottom by processing position, real-time processing data condition administrative unit 45 previous tops.Thereby treated a plurality of deal with data groups (or a plurality of treated data set) just are placed on to be handled end position (TopOld) in real time and handles position the end position (BtmOld) to the bottom in real time from the top.
Then, judge that by real-time processing data condition administrative unit 45 being placed on the top handles the retrieval sign F1 that end position (TopOld) locates in real time and handle sign F2 in real time and whether be arranged to " off " condition (step S518) together.If the retrieval sign F1 of treated data set and processing in real time indicate that F2 is arranged to " off " condition together, so because the real-time processing of treated data set has finished and treated data set is not arranged to search the search condition of deal with data group with character stream retrieval unit 52, from real-time processing data storage unit 44, delete (step S519) so the treated data set that end position (TopOld) locates to place is handled at the top in real time, and the position of the real-time processing data storage unit of being pointed out by deletion pointer P6 44 just is displaced to lower position (TopOld ← TopOld+1) (step S520).On the contrary, if the retrieval sign F1 of the treated data set that end position (TopOld) locates to place is handled at the top in real time or handle sign F2 in real time and be arranged to " on " condition, because still need treated data set, processing procedure jumps to step S520 so.Subsequently, repeating step S518 is to S520, and up to having deleted all treated data sets, and the top is handled end position (TopOld) in real time and arrived deal with data group top and handle position (TopNew) (step S521).Subsequently, repeating step S511 is to S522, up to the real-time processing that all deal with data groups are carried out all finish and deal with data group tip position (TopNew) arrive data to deposit position (BtmNew) till (step S522).
Therefore, the a plurality of deal with data groups that received by request and Data Receiving unit 42 are stored in data one by one as a plurality of deal with data groups and deposit position (BtmNew) and locate, make data deposit position (BtmNew) displacement simultaneously, the deal with data group is placed on the position of depositing position (BtmNew) to data from processing position, top (TopNew), when making the top handle position (TopNew) displacement, handle in real time being placed on a deal with data group of locating processing position, top (TopNew), the a plurality of treated data set that finishes that the real-time processing of deal with data group is obtained is placed on handles end position (TopOld) from the top in real time and handle position the end position (BtmOld) to the bottom in real time, and when all deal with data of processing, delete all treated data one by one.So, can real time modifying index file storage unit 49 in the index file of storage.
The search operaqtion of carrying out under the situation that all deal with data groups of storing in not to real-time processing data storage unit 44 according to first embodiment handle in real time is described below.
Figure 29 is the process flow diagram that carries out search operaqtion under situation about all deal with data groups not being handled in real time.
As shown in figure 29, when no matter in real time in the processing unit 50 to real-time processing data storage unit 44 in the real-time processing carried out of all deal with data groups of storage whether finish and when receiving the retrieval group of the searching character stream prepared by the user and retrieval request (step S531) by request and Data Receiving unit 42, because receive retrieval request, just be sent to index file search unit 51 and character stream retrieval unit 52 (step S532) so be equivalent to the searching character stream of a certain keyword.In searching unit 51, the index file of storage in the search index file storage unit 49, with retrieval and the consistent one or more character streams (step S533) of this searching character stream, and will represent that the result for retrieval of the character stream retrieval consistent with this searching character stream outputs to result for retrieval modification unit 53 (step S534).
In detail, searching character stream is divided into the character pattern { C1 (i) of a series of appointments, C2 (i) } (i=1,2, p, p+1,), from index file, select the one or more character information patterns corresponding with the character pattern of each appointment, and when another specific character information pattern that character position PC1 compares and the second designated character pattern is corresponding of each specific character information pattern consistent with each other and corresponding with the first designated character pattern to the data number DN of specific character information pattern hangs down 1, from the character information pattern, select with mutual close each to the first designated character pattern { C1 (p), C2 (p) } and the corresponding one or more pairs of specific character information pattern of the second designated character pattern { C1 (p+1), C2 (p+1) }.That is,, so just select each to specific character information pattern if two character C1 (p) of the first and second designated character patterns and C1 (p+1) adjoin each other in proper order by that in same searching for text.So, if for there are one or more pairs of specific character information patterns in each to pattern from all designated character pattern centerings of all designated character mode producing, so just judge whether exist in the index file and one or more character streams that searching character stream is consistent.For example, if in step S531, receive searching character stream " AIU ", and the index file shown in Figure 28 B is stored in the index file storage unit 49, then searching character stream " AIU " is divided, in searching unit 51, produce the character pattern (A of two appointments, I) and (I, U), from index file, select character pattern (A with appointment, I) the corresponding characters information pattern (1,1), (4,1) and (1,5) and with the character pattern (I of appointment, U) the corresponding characters information pattern (5,1) and (2,5), because the data number DN that the specific character information pattern is right is consistent with each other and a specific character information pattern (4,1) character position PC1=4 selects a pair of specific character information pattern (4,1) and (5 than the PC1=5 of another specific character information pattern (5,1) low 1,1), because the data number DN that the specific character information pattern is right is consistent with each other and the character position PC1=1 of a specific character information pattern (1,5) selects another to specific character information pattern (1 than the PC1=2 of another specific character information pattern (2,5) low 1,5) and (2,5).So, exist two to flow " AIU " consistent character stream in the indexed file with searching character.
Then, under the control of real-time processing data condition administrative unit 45, read one or more deal with data groups of depositing the position placement the position (BtmNew) in the real-time processing data storage unit 44 from processing position, top (TopNew) to data, and this data set is sent to character stream retrieval unit 52 (step S535), and by real-time processing data condition administrative unit 45 the retrieval sign F1 of deal with data group is arranged to " on " condition (step S536) one by one.Each of " on " condition is handled data set and be can't help real-time processing unit 50 and handle.Subsequently, search each real-time processing data in the deal with data group, flow consistent one or more character streams (step S537) with searching character respectively, and the deal with data search result is outputed to result for retrieval modification unit 53 with retrieval.Then, revise index file according to the deal with data search result and search the result for retrieval that obtains in the unit 51, with all deal with data groups (step S538) that received by request and Data Receiving unit 42 in the reflection result for retrieval, the retrieval sign F1 that each of storing in the real-time processing data storage unit 44 is handled data set by real-time processing data condition administrative unit 45 is arranged to " off " condition (step S539), and from result for retrieval output unit 54 output modified result for retrieval (step S540).When the user sends the modified result for retrieval of instruction demonstration to information searching apparatus 41, also show modified result for retrieval by display 48 by video data storage unit 47.In modified result for retrieval, list one or more data number of the one or more searching for texts that have the one or more character streams consistent, and list the quantity of the character stream consistent for each data number with the character stream of appointment with the character stream of appointment.
Describe the retouching operation that carries out in the modification unit 53 in detail below with reference to Figure 30.
As shown in figure 30, by the index file of searching storage in the unit 51 search index file storage units 49, consistent with searching character stream " AIU " and be listed in a plurality of character streams in a plurality of searching for texts that data number is D1, D5, D10 and D15 and four deal with data groups as shown in Figure 30 of existence in the real-time processing data storage unit 44 with retrieval.In this case, because the first deal with data group and the character stream " AIU " of searching for text D5 is added in the result for retrieval, because the character stream " UEO " different with searching character stream is added among the searching for text D12 and the discarded second deal with data group according to the second deal with data group, use the character stream " AIU " consistent to replace the character stream of searching for text D11 because being subjected to the instruction of the 3rd deal with data group and the character stream " AIU " of searching for text D11 is increased to result for retrieval, and because the manage data set everywhere and from result for retrieval, delete searching for text D5 with searching character stream.
Therefore, although the centre of the real-time processing of carrying out at real-time processing unit 50 is input to information searching apparatus 41 with retrieval request, but because each data set of can't help in one or more deal with data groups that real-time processing unit 50 handles is searched by character stream retrieval unit 52, and revise by index file and search the result for retrieval that unit 51 obtains, so in modified result for retrieval, can reflect all deal with data groups that are input to information searching apparatus 41.Promptly, although indexed file deposit, increase, upgrade or delete the intermediate request search operaqtion, but can for consistent with searching character stream respectively one or more character streams immediately the execution index file fine operate, and needn't be by the time to the depositing of index file, increase, the end of renewal or deletion action.
In the 4th embodiment, the absolute position of character is by 45 controls of real-time processing data condition administrative unit in each searching for text.Yet the relative position of character also can be managed by real-time processing data condition administrative unit 45 in each searching for text.
In the 4th embodiment, also use two sign F1 and F2.But, will not indicate that F1 and F2 are arranged to the situation of " on " condition because do not exist, so can replace sign F1 and F2 with a sign.
The 5th embodiment is described below, in this embodiment, in request and Data Receiving unit 42, receive under the situation of a data set in the deal with data group when in real-time processing data storage unit 44, storing another deal with data group, two deal with data groups corresponding to same specific searching for text are handled in real time.
As shown in figure 19, information searching apparatus 41 also comprises:
Repeat to handle in real time to avoid unit 55, when being used for receiving with the corresponding re-treatment data set of same specific searching for text, will and being stored in original deal with data group in the real-time processing data storage unit 44 corresponding to specific searching for text and being altered to modified deal with data group corresponding to same specific searching for text in request and Data Receiving unit 42; And give up the re-treatment data set.
Be described in the real-time processing that is used in the information searching apparatus 41 of said structure corresponding to two deal with data groups of same specific searching for text below with reference to Figure 31.
Figure 31 is the process flow diagram of handling in real time corresponding to two deal with data groups of identical specific searching for text that illustrates according to fifth embodiment of the invention.
As shown in figure 31, after as shown in figure 20 step S501 receives and deposits particular procedure data set corresponding to a certain specific searching for text in the S503, because the original deal with data group corresponding to same specific searching for text is stored in the real-time processing data storage unit 44, so, by repeating to handle in real time to avoid unit 55 to judge whether these particular procedure data are equivalent to the deal with data group (step S601) of a certain repetition.If in real-time processing data storage unit 44, store and the corresponding original deal with data group of identical specific searching for text, so just judge that this particular procedure data set is equivalent to a certain re-treatment data set, and by original deal with data group being altered to the corresponding modified deal with data group of identical specific searching for text and giving up the re-treatment data set, and avoid unit 55 to carry out re-treatments avoiding operation (step S602) by repeating to handle in real time.
For example, shown in Figure 32 A, if the original deal with data group that storage increase corresponding with specific searching for text D1 and request increase character stream " AIU " is operated in real-time processing data storage unit 44, and thereby reception and depositing request are upgraded operation replaces the one piece of data of specific searching for text D1 with a certain character stream " IUE " re-treatment data set, so because add to the character stream " AIU " of specific searching for text with character stream " IUE " replacement, so original deal with data group just changes over request character stream " IUE " is added to the modified deal with data group of the increase operation of this specific searching for text D1, and gives up the re-treatment data set.
And, shown in Figure 32 B, if storage replaces with the one piece of data of specific searching for text D2 corresponding to specific searching for text D1 and request the original deal with data group of the renewal operation of character stream " UEO " in real-time processing data storage unit 44, and reception and depositing request are deleted the re-treatment data set of the deletion action of specific searching for text D2, so because this specific searching for text D2 is finally deleted, so original deal with data group just changes over the modified deal with data group of the deletion action of this specific searching for text D2 of request deletion, and the re-treatment data set is rejected.
Simultaneously, shown in Figure 32 C, if storage is corresponding to the original deal with data group of the deletion action of specific searching for text D3 and the specific searching for text D3 of request deletion in real-time processing data storage unit 44, and receive and depositing request adds to character stream " IUE " the re-treatment data set of the increase operation of specific searching for text D3, so because deleted the data of specific searching for text D3 and added character stream " IUE ", so original deal with data group changes over the renewal operation of specific searching for text D3 is replaced in request with character stream " IUE " modified deal with data group.
Then, by repeating to handle in real time to avoid unit 55 to judge whether and modified deal with data group need be stored in (step S603) in the real-time processing data storage unit 44.Store modified deal with data group if desired, so just avoid unit 55 that modified deal with data group is stored in the real-time processing data storage unit 44, and data are deposited displacement to lower position (BtmNew ← BtmNew+1) (step S605) by repeating processing in real time.Subsequently, execution step S506 as shown in Figure 20 is to S509.
Therefore, because can avoid the repetition that repeats the deal with data group is handled in real time, so can under high speed, handle in real time effectively.
The 6th embodiment is described below.In this embodiment, deposit in the operation in the deal with data group and to prepare a character and have the judgement table, this judgement table is used for judging in real-time processing data storage unit 44 existence of employed character in the real-time processing data of all deal with data groups of storage, exists the judgement table to carry out search operaqtion in character stream retrieval unit 52 with this character.
Figure 33 is the information searching apparatus block scheme according to sixth embodiment of the invention.
As shown in figure 33, information searching apparatus 61 comprises:
Request and Data Receiving unit 42; Real-time processing data deposit unit 43; Real-time processing data storage unit 44; Real-time processing data condition administrative unit 45; Index file storage unit 49; Real-time processing unit 50; Index file is searched unit 51; Character stream retrieval unit 52;
Table preparatory unit 62 exist to be judged in character, is used for preparing a character and has the judgement table, points out the existence of the character that uses in one section real-time processing data of the particular procedure data set stored in the real-time processing data storage unit 44 in the table,
Table storage unit 63 exist to be judged in character, be used for the character that store character exists judgement table preparatory unit 62 to be prepared to have the judgement table,
Judge table reset unit 64, when the real-time processing to the particular procedure data set finishes and deletes the particular procedure data set of storage in the real-time processing data storage unit 44, make the character of storage in the real-time processing data storage unit 44 exist the judgement table to reset;
Search operaqtion control module 65, when request and Data Receiving unit 42 receive the retrieval group of retrieval request and searching character stream, control character stream retrieval unit 52, all be stored in the real-time processing data storage unit 44 and can't help all deal with data groups that real-time processing unit 50 handles to search each, and when the reference character exists the character existence of judging storage in the table storage unit 63 to judge table and output deal with data search result, control character stream retrieval unit 52, with retrieval with by request and Data Receiving unit 42 from the not consistent one or more character streams of the searching character flow point of deal with data group of received;
Result for retrieval is revised unit 53 and result for retrieval output unit 54.
Describe in the said structure deal with data group below with reference to Figure 34 and deposit operation according to the 6th embodiment.
Figure 34 is a process flow diagram of depositing operation according to the deal with data group of the 6th embodiment.
As shown in Figure 34, after having carried out step S501 to S503 in identical mode as shown in figure 20, make character have the initialization of judgement table as shown in figure 35, judge that there is judgement symbol (step S610) in each in all types character corresponding to a character in the table.That is, all characters are existed in the judgement symbol each be arranged to represent not exist " off " condition of respective symbols type.Subsequently, exist judgement table preparatory unit 62 to check the type (step S611) of one section real-time processing data top character of the deal with data group that real-time processing data deposit unit 43 is deposited by character, exist in the judgement table and will exist judgement symbol to be arranged to " on " condition (step S612) with the corresponding character of character types on inspection at character.For example, deposit and be shown in deal with data group set among Figure 21, will be arranged to " on " condition with the type corresponding characters sign F (E) of character " E ".Subsequently, set character exists the judgement symbol corresponding characters whether to be arranged in the last position of real-time processing data (step S613) among judgement and the step S612.If having the judgement symbol corresponding characters with character set among the step S612 is not to be placed on the rearmost position, so just check the type (step S614) of character late in the real-time processing datas by preparatory unit 63, and repeating step S612 and S613.On the contrary, if having the judgement symbol corresponding characters with the character that is provided with among the step S612 is to be placed on the rearmost position, so because with real-time processing data in the type corresponding characters of all characters of existing exist judgement symbol to be configured to represent to exist " on " condition of respective symbols, so existing the judgement table to be stored in character character exists in the judgement table storage unit 63, the deal with data group is stored in (step S504) in the real-time processing data storage unit 44, and execution in step S505 to S509.In this example, because real-time processing data " EOK " is deposited by real-time processing data deposit unit 43 as shown in figure 36,, the character that will exist judgement symbol to be arranged to " on " condition corresponding to the character of character " E ", " O " and " K " judges that table exists judgement table preparatory unit 62 to prepare by character so existing.
Subsequently, when finishing the real-time processing of deal with data group by real-time processing unit 50 and during the deal with data group of being stored in by real-time processing data condition administrative unit 45 deletion real-time processing data storage unit 44, exist judgement symbol to reset to " off " condition by judgement table reset unit 64 with all character types corresponding characters in the real-time processing data of deal with data group.
Below with reference to the character stream search operaqtion of Figure 37 description according to the 6th embodiment, character stream retrieval unit 52 uses character to have the judgement table and carry out this operation under the control of search operaqtion control module 65.
The character stream search operaqtion process flow diagram that Figure 37 carries out when being to use character to have the judgement table.
After having carried out step S531 to S536, a certain particular procedure data set of representing the deal with data group of storage in the real-time processing data storage unit 44 is carried out the character stream search operaqtion by character stream retrieval unit 52.Promptly, as shown in figure 37, the searching character that extraction is received by request and Data Receiving unit 42 flows the character at top, as checking character (step S621), and judge and character corresponding characters on inspection exists judgement symbol whether to exist at specific character to be arranged to " on " condition in the judgement table that this table has pointed out to be used for exist (the step S622) of particular procedure data set character.
If exist judgement symbol to be configured to " on " condition, so just judge whether character on inspection is set to searching character stream rearmost position (step S623) with character corresponding characters on inspection.If character on inspection is not the rearmost position that is set to searching character stream, another character that so just extracts character back on inspection from searching character stream is as inspection character (step S624), and repeating step S622.
On the contrary, if having judgement symbol with on inspection character corresponding characters is not to be set to " on " condition, to exist judgement symbol to exist in the judgement table at specific character be not to be set to " on " condition because flow the corresponding one or more characters of all characters with searching character, so make the judgement that does not have the character stream consistent with searching character stream.So, finish character stream search operaqtion to the particular procedure data set.
And, if character in step S623 on inspection is not arranged on the rearmost position of searching character stream, exist judgement symbol to be configured to specific character to have " on " condition in the judgement table because flow the corresponding one or more characters of all characters, so make the judgement of consistent with the searching character stream respectively one or more character streams of existence with searching character.So retrieval is flowed consistent one or more character streams (step S537) with searching character respectively from one section real-time processing data of particular procedure data set, execution in step S538 is to S540 in the same way.
Therefore, because prepared to point out that there is the judgement table in the character that uses character to exist in the particular procedure data set, so when search particular procedure data set flows consistent one or more character stream with retrieval with searching character, can carry out at a high speed whether there being the judgement of the one or more character streams consistent in the particular procedure data set with searching character stream, and, can stop character stream search operaqtion by 52 execution of character stream retrieval unit when being judged as when not having the character stream consistent with searching character stream.So, can under high speed, carry out the character stream search operaqtion.
The 7th embodiment is described below, in this embodiment, under a plurality of deal with data groups corresponding with same specific searching for text are stored in situation in the real-time processing data storage unit 44, the nearest deal with data group corresponding to a certain specific searching for text is carried out the character stream search operaqtion.
Figure 38 is the block scheme according to the information searching apparatus of seventh embodiment of the invention.
As shown in figure 38, information searching apparatus 71 comprises:
Request and Data Receiving unit 42; Real-time processing data deposit unit 43; Real-time processing data storage unit 44; Real-time processing data condition administrative unit 45; Video data generation unit 46; Video data storage unit 47; Display 48; Index file storage unit 49; Real-time processing unit 50; Index file is searched unit 51; Character stream retrieval unit 52;
Unit 72 is avoided in repeated retrieval operation, be used for from real-time processing data storage unit 44 select nearest deal with data group in the corresponding a plurality of deal with data groups of the identical specific searching for text of storage corresponding to a certain specific searching for text; Control character stream retrieval unit 52, flow corresponding one or more character streams with the searching character that request and Data Receiving unit 42 receives respectively with retrieval from nearest deal with data group, and control character stream retrieval unit 52, other deal with data groups corresponding to identical specific searching for text are not carried out the character stream search operaqtion;
Result for retrieval is revised unit 53; With result for retrieval output unit 54.
Be described in below with reference to Figure 39 in the information searching apparatus 71 of said structure, when avoiding the repeated retrieval operation, operate the character stream search operaqtion of carrying out by character stream retrieval unit 52 under the control of avoiding unit 72 in repeated retrieval.
The step S531 to S536 as shown in figure 29 that carries out with identical mode in as information searching apparatus 41 is stored in a plurality of deal with data groups in the real-time processing data storage unit 44 with acquisition after, as shown in figure 39, because in a plurality of deal with data groups of in real-time processing data storage unit 44, storing, it is up-to-date that the data that are arranged in real-time processing data storage unit 44 are deposited the designated treatment data set of locating position (BtmNew), so take out the deal with data group of this appointment, as a retrieve data group (step S631), and the retrieve data group is carried out the character stream search operaqtion, from the retrieve data group, to retrieve and the consistent one or more character streams (step S632) of searching character stream.Subsequently, according to the deal with data search result of character stream retrieval unit 52, revise unit 53 by result for retrieval and revise the result for retrieval (step S538) that index file is searched acquisition in the unit 51.Then, judge whether this retrieve data group is in the processing position (TopNew) (step S633) at top.
If the retrieve data group is arranged in the processing position (TopNew) at top, so because of the character stream search operaqtion of in step S632, having carried out each data set in all deal with data groups of storage in the real-time processing data storage unit 44, so finish to avoid the operation of unit 52.On the contrary, if this retrieve data group is not arranged in the processing position (TopNew) at top, so just take out the deal with data group that is placed on retrieve data group upper position place, as corresponding to retrieve data group (step S634) by the specific searching for text of data number sign, and judge whether previous in step S632 to carrying out character stream search operaqtion (step S635) with the corresponding particular procedure data set of identical specific searching for text.
If never to carrying out the character stream search operaqtion with the corresponding particular procedure data set of identical specific searching for text, repeating step S632, S538, S633 and S634 so.That is, carry out search operaqtion of carrying out by character stream retrieval unit 52 and the retouching operation of carrying out by modification unit 53 by the reverse sequence of listing.On the contrary, if previous, just the retrieval process data set is not carried out any character stream search operaqtion, and process turns back to step S633 so to carrying out the character stream search operaqtion with the corresponding particular procedure data set of identical specific searching for text.That is, avoided the repeated retrieval operation.
Be described in detail in retrieval and the retouching operation that in character stream retrieval unit 52 and modification unit 53, carries out under the control of avoiding unit 72 below with reference to Figure 40.
As shown in figure 40, by the index file of searching storage in the unit 51 search index file storage units 49, consistent with searching character stream " AIU " and be listed in existence four deal with data groups as shown in figure 40 in a plurality of character streams in a plurality of searching for texts that data number is D1, D5, D10 and D15 and the real-time processing data storage unit 44 with retrieval.Like this, because be arranged in that the data of real-time processing data storage unit 44 deposit that the first deal with data group of locating position (BtmNew) is represented the deletion action of searching for text D5 from result for retrieval deletion searching for text D5, because being arranged in the second deal with data group at the first deal with data group upper position place represents to use the character stream " AIU " consistent with searching character stream to replace the character stream of searching for text D11 and the character stream " AIU " of searching for text D11 is added to result for retrieval, be added to searching for text D12 and give up the 3rd deal with data group that is arranged in the second deal with data group upper position place because will flow different character stream " UEO " with searching character, and manage the searching for text D5 that data set represents everywhere and give up and be arranged in the of the 3rd deal with data group upper position place and manage data set everywhere because under the control of avoiding unit 72, handled the by character stream retrieval unit 52 according to the 3rd deal with data group.
Therefore, because if be stored in the real-time processing data storage unit 44 with the corresponding a plurality of deal with data groups of identical specific searching for text and to carry out the character stream search operaqtion corresponding to the nearest deal with data group of specific searching for text, and because other deal with data groups corresponding to identical specific searching for text are not carried out the character stream search operaqtion, so can carry out the character stream search operaqtion at a high speed.
Describe below according to the 8th embodiment during the real-time processing of being undertaken by real-time processing unit 50, carry out force end operation and at the start-up operation that forces end operation to carry out later on.
Figure 41 is the block scheme according to the information searching apparatus of eighth embodiment of the invention.
As shown in figure 41, information searching apparatus 81 comprises:
Request and Data Receiving unit 42; Real-time processing data deposit unit 43; Real-time processing data storage unit 44; Real-time processing data condition administrative unit 45; Video data generation unit 46; Video data storage unit 47; Display 48; Index file storage unit 49; Real-time processing unit 50;
Force end operation request unit 82, be used for when request and Data Receiving unit 42 receive ending request, ask real-time processing unit 59 to force and finish to handle in real time;
Termination condition storage unit 83 is used for storing one or more deal with data groups of not handled by real-time processing procedure owing to force end operation request unit 82 the forcing end operation of request;
Start-up operation request unit 84 is used for when request and Data Receiving unit 42 receive the startup request, asks real-time processing unit 50 to start the real-time processing procedure of forcing end by ending request,
Index file is searched unit 51; Character stream retrieval unit 52; Result for retrieval is revised unit 53 and result for retrieval output unit 54.
Be described in the end operation that forces that in information searching apparatus 81, carries out in the said structure below with reference to Figure 42.
Figure 42 is the process flow diagram that forces end operation that illustrates according to eighth embodiment of the invention.
As shown in figure 42, when the centre of handling in real time at real-time processing unit 50 receives an ending request (step S701) by request and Data Receiving unit 42, force the real-time processing unit 50 of end operation request unit 82 requests so after having handled current deal with data group, force and finish to handle in real time (step S702).For example, if the deal with data group to as shown in figure 23 four deal with data group tops has been carried out under the situation of real-time processing when receiving ending request, force later at the EO that searching for text D5 is updated to character stream " AIU " and to finish to handle in real time.
Then, by request unit 82 one or more deal with data groups that each is stored in the real-time processing data storage unit 44 and real-time processing unit 50 is not handled are sent to termination condition storage unit 83, as one section termination condition data (step S703).As shown in figure 43, each in the termination condition data comprises such as upgrading operation, increasing data number and one piece of data that operation or deletion action etc. are handled type, the corresponding searching for text of sign.Because be through with to the renewal of searching for text D5 operation, thus as shown in figure 44, with three sections termination condition data storage in termination condition storage unit 83.That subsequently, tells that the user handles in real time forces end (step S704).
The start-up operation that forces end operation to carry out is later on described below with reference to Figure 45.
As shown in figure 45, when request and Data Receiving unit 42 receive the startup request (step S711), judge whether in storage unit 83, to have stored one or more snippets termination condition data (step S712) by start-up operation request unit 84.If one or more snippets termination condition data storage is in storage unit 83, just the termination condition data are sent to real-time processing unit 50, as one or more deal with data groups, in the deal with data group each is handled (step S713) in real time, operate as emergency start.After having carried out the emergency start operation, carry out normal start-up operation (step S714).
Therefore, even in real-time processing data storage unit 44, stored a large amount of deal with data groups, but because when receiving ending request, force the real-time processing that is through with to the deal with data group, so do not exist the user need wait for situation for a long time to the end of the real-time processing of all deal with data groups.So the user can carry out the operation of another kind of type immediately with information searching apparatus 81.
And, even the real-time processing of all deal with data groups is not finished, but because untreated one or more data sets are stored in the termination condition storage unit 83, so, untreated deal with data group can be used as emergency start and operate and handle.So, can correctly handle in real time regardless of the size of deal with data group to all deal with data groups.
Result for retrieval output function according to the 9th embodiment is described below, in this operation, if specific searching for text upgrades later on obtaining modified result for retrieval, then the modified result for retrieval of specific searching for text is shown that with a segment table lastest imformation that this specific searching for text upgrades exports.
Figure 46 is the block scheme of the information searching apparatus of ninth embodiment of the invention.
As shown in figure 46, information searching apparatus 91 comprises:
Request and Data Receiving unit 42; Real-time processing data deposit unit 43; Real-time processing data storage unit 44; Real-time processing data condition administrative unit 45; Video data generation unit 46; Video data storage unit 47; Display 48; Index file storage unit 49; Real-time processing unit 50; Index file is searched unit 51; Character stream retrieval unit 52; Result for retrieval is revised unit 53; Result for retrieval output unit 54;
Measure and record cell 92 retrieval time, be used for measuring and the record retrieval time, in this time, to in the searching for text each, the deal with data group of searching unit 51 or searching searching for texts or searching for text by index file by character stream retrieval unit 52, and by the character stream retrieval unit retrieves from searching for text and searching character stream is consistent one or more character streams;
Measure update time and record cell 93, be used for measuring and record update time, in this time,, upgrade the searching for text of storing in the index file storage unit 49 by real-time processing unit 50 in the searching for text each,
Result for retrieval output function control module 94 when request and Data Receiving unit 42 receive result for retrieval output request, is used for controlling the carrying out of result for retrieval output function;
Upgrade operational notification unit 95, be used for judging measure retrieval time and 92 pairs of searching for texts of record cell in each searching for text retrieval time of measuring whether than update time measurement and record cell 93 to record a update time slow, if and update time is slower than the retrieval time, then prepare one section lastest imformation, its expression is to carry out the search of the deal with data group of searching for text or searching for text and carry out later on being searched in unit 51 or 52 pairs of searching for texts of character stream retrieval unit each by index file to the renewal of a certain searching for text operation, and
Lastest imformation output unit 96 is used for exporting the data number that each searching for text is obtained by renewal operational notification unit 95, data and the renewal or the non-lastest imformation of searching for text.
The result for retrieval output function of being undertaken by information searching apparatus 91 in said structure is described below.
Figure 47 is the process flow diagram that the record retrieval time is shown.
When request and Data Receiving unit 41 receive the retrieval group of the searching character stream of request search operaqtion and retrieval request, with same as shown in figure 29 mode execution in step S531 to S538.Subsequently, carry out the record of retrieval time as shown in figure 47.Promptly, as shown in figure 48, to in the searching for text each, with one group of data number and by index file search retrieval time that the deal with data group of unit 51 or 52 pairs of searching for texts of character stream retrieval unit or searching for text searches and from searching for text by 52 retrievals of character stream retrieval unit with the consistent one or more character streams of searching character stream, be recorded in that measure retrieval time and record cell 92 interior (step S721).Subsequently, with Figure 29 in identical mode execution in step S539 and S540.For example, from a certain searching for text D1, retrieving at 10 o'clock and searching character stream " AIU " consistent character stream if search unit 51 by index file, so as shown in figure 49, just obtained one group of data number D1 and retrieval time 10:00.
Subsequently, carry out the record of update time.In detail, when request and Data Receiving unit 42 receive the deal with data group that the real-time processing request specific searching for text handled in real time by one section real-time processing data and request forms, to S509, be used for the deal with data group of specific searching for text with mode execution in step S501 same as shown in Figure 20 with storage and processing.Subsequently, by real-time processing unit 50 with mode execution in step S511 identical shown in Figure 26 to S522, to upgrade the specific searching for text of storage in the index file storage unit 49.And, as shown in figure 50, the update time (step S731) that measurement is upgraded the specific searching for text of storage in the index file storage unit 49 by real-time processing unit 50, and as shown in Figure 51, update time measure and record cell 93 in write down specific searching for text one group of data number and update time (step S732).For example, if in 10 minutes, a certain searching for text D1 is updated to a character stream " IUE " 10 o'clock mistakes, so shown in Figure 52, just obtain one group of data number and one retrieval time 10:10.
Subsequently, export one section lastest imformation, this information representation indexed file is searched and after this searching for text is searched in unit 51 searching for text has been carried out upgrading operation.In detail, shown in Figure 53, when request and Data Receiving unit 42 receive the result for retrieval output request of the modified result for retrieval of asking the one or more searching for texts of output (step S741), under the control of result for retrieval output function control module 94, measurement and record cell 92 were sent to and upgraded operational notification unit 95 (step S742) retrieval time of each searching for text from retrieval time, under the control of result for retrieval output function control module 94, measurement and record cell 93 were sent to and upgraded operational notification unit 95 (step S743) update time of each searching for text from update time.Then, each searching for text is carried out the judgement (step S744) that whether is later than update time retrieval time.If equal retrieval time or be later than update time, then prepare one section lastest imformation by upgrading operational notification unit 95, this information representation is not carried out any renewal operation (step S745) to searching for text after being searched unit 51 or character stream retrieval unit 52 search searching for texts by index file or being used for the deal with data group of this searching for text.On the contrary, if be later than retrieval time update time, then prepare one section lastest imformation by upgrading operational notification unit 95, this information representation is upgraded operation (step S746) to searching for text after having been searched searching for text or be used for the deal with data group of this searching for text by index file search unit 51 or character stream retrieval unit 52.Subsequently, to each searching for text, by table (step S747) of listing data and the renewal or the non-lastest imformation of data number table, searching for text of lastest imformation output unit 96 outputs.
For example, if search unit 51 or each reception and the searching character stream consistent one or more character streams " IUE " of character stream retrieval unit 52 from searching for text D1, D2 by index file, then be sent to and upgrade operational notification unit 95 in the retrieval time that step S742 will be used for searching for text D1, and step S743 will be used for searching for text D2 update time 9:00 and 9:40 update time that is used for searching for text D5 be sent to and upgrade operational notification unit 95.Subsequently, because be used for being later than the update time (Figure 52) of searching for text D1 10:00 retrieval time (Figure 49) that is used for searching for text D1, so shown in Figure 54, prepare to be used for one section lastest imformation (step S746) of searching for text D1 by upgrading operational notification unit 95.
Therefore, because made judgement update time, so, also can notify the user one section lastest imformation even upgrade searching for text later in that searching for text is carried out seek operations to whether being later than retrieval time.
In conjunction with preferred embodiment principle of the present invention is being carried out after the foregoing description, those skilled in the art should understand, and can make amendment to the arrangement and the details of the foregoing description under the situation that does not depart from above-mentioned principle.These modifications all will drop in the included scope of claim.

Claims (8)

1. information searching apparatus is characterized in that it comprises:
The searched targets memory storage is used for storing one or more searched targets;
The real-time processing data receiving trap is used for receiving one or more snippets real-time processing data, and described data are corresponding with the searched targets of storing in the searched targets memory storage respectively;
The retrieving information receiving trap is used for receiving one section retrieving information and and searches request;
The real-time processing data memory storage is used for storing the real-time processing data section that is received by the real-time processing data receiving trap;
Treating apparatus is used for each section in the real-time processing data section of storing in the real-time processing data memory storage is handled in real time, to upgrade the searched targets of storing in the searched targets memory storage in real time;
The searched targets search device, be used for searching each searched targets of storing in the searched targets memory storage, thereby retrieve one or more snippets the consistent first information of retrieving information that from each searched targets, receives with the retrieving information receiving trap, and the result for retrieval of the output expression first information;
The real-time processing data search device, be used for searching real-time treating apparatus and do not handle and be stored in each section in one or more snippets specific real-time processing data in the real-time processing data memory storage, retrieve respectively one or more snippets consistent second information of retrieving information that receives with the retrieving information receiving trap in each section with from implementation-specific the time the process segments of data;
The result for retrieval modifier is used for second information according to real-time processing data search device retrieval, revises from the result for retrieval of searched targets search device output process segments of data during with the implementation-specific in the reflection result for retrieval; And
The result for retrieval output unit is used for exporting the result for retrieval of being revised by the result for retrieval modifier.
2. information searching apparatus as claimed in claim 1, it is characterized in that, each searched targets of storing in the described searched targets memory storage is a character stream, and described character stream is with also representing with the character information pattern that the data number that identifies this character stream is pointed out with the position of first character in the character stream by one or more character patterns of the first close mutually character and second character representation and each are all corresponding with character pattern respectively.
3. information searching apparatus as claimed in claim 1 is characterized in that it also comprises:
Repeat to handle in real time to avoid device, be used for receiving by the real-time processing data receiving trap when repeating real-time processing data with identical specific searched targets corresponding a section, one section real-time processing data corresponding with specific searched targets and that be stored in the real-time processing data memory storage is changed over the one section modified real-time processing data corresponding with same specific searched targets, and give up the repetition real-time processing data.
4. information searching apparatus as claimed in claim 1 is characterized in that it also comprises:
Exist and judge the table preparation device, be used for preparing one and have the judgement table, have one section all constituents of specifying real-time processing data storing in the real-time processing data memory storage in the table; And
The search operaqtion control device, be used for judging and exist the existence of judging the preparation of table preparation device to judge all the components that whether has retrieving information in the table, and when in having the judgement table, not having all the components of retrieving information, control real-time processing data search device is to stop specifying the search of real-time processing data.
5. information searching apparatus as claimed in claim 1 is characterized in that it also comprises:
Device is avoided in the repeated retrieval operation, be used for from be stored in the real-time processing data memory storage select one section nearest real-time processing data corresponding in the corresponding multistage real-time processing data of the identical specific searched targets of storing with specific searched targets, control real-time processing data search device is to retrieve second consistent with the described retrieving information information from described nearest real-time processing data, and control real-time processing data search device, to not carrying out seek operations with corresponding other real-time processing data sections of identical specific retrieving information.
6. information searching apparatus as claimed in claim 1 is characterized in that it also comprises:
The ending request receiving trap is used for receiving ending request;
Force the end operation request unit, be used for when receiving ending request by the ending request receiving trap, ask real-time treating apparatus to force and finish to handle in real time, with one or more snippets appointment real-time processing data that in the real-time processing data memory storage, keeps real-time treating apparatus not handle; And
The termination condition memory storage is used for storing the real-time processing data section by the appointment that forces the end operation request unit to keep.
7. information searching apparatus as claimed in claim 6 is characterized in that it also comprises:
Start the request receiving trap, be used for receiving starting and ask; And
The start-up operation request unit is used for asking real-time treating apparatus to restart the real-time processing that is forced end by the end operation request unit when receiving the startup request by startup request receiving trap.
8. information searching apparatus as claimed in claim 1 is characterized in that it also comprises:
Retrieval time, measurement mechanism was used for measuring in the real-time processing data memory storage that the specific searched targets of storing in the searched targets memory storage of searched targets search device search or real-time processing data search device search the retrieval time for one section specific real-time processing data of specific searched targets storage;
Update time, measurement mechanism was used for measuring the update time of being upgraded the specific searched targets of storing in the searched targets memory storage by real-time treating apparatus;
Upgrade the operational notification device, be used for judging the update time that whether is later than measurement mechanism measurement update time retrieval time that measurement mechanism is measured by retrieval time, and when be later than retrieval time update time, prepare one section lastest imformation, be illustrated in search specific searched targets by searched targets search device or real-time processing data search device after, specific searched targets has been carried out upgrading operation; And
The lastest imformation output unit is used for being later than under the situation of retrieval time in update time, exports the identification number of specific searched targets and the lastest imformation that is obtained by renewal operational notification device.
CN 02132175 1996-03-27 1997-02-27 Information searching device for searching text to retrieve a certain character stream matching a certain keyword Expired - Fee Related CN1271549C (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
JP8072710A JP3024544B2 (en) 1996-03-27 1996-03-27 Information retrieval device
JP072710/96 1996-03-27
JP072710/1996 1996-03-27

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
CN 97102954 Division CN1114877C (en) 1996-03-27 1997-02-27 Information searching apparatus for searching text to retrieve character streams agreeing with key word

Publications (2)

Publication Number Publication Date
CN1492360A CN1492360A (en) 2004-04-28
CN1271549C true CN1271549C (en) 2006-08-23

Family

ID=13497191

Family Applications (2)

Application Number Title Priority Date Filing Date
CN 97102954 Expired - Fee Related CN1114877C (en) 1996-03-27 1997-02-27 Information searching apparatus for searching text to retrieve character streams agreeing with key word
CN 02132175 Expired - Fee Related CN1271549C (en) 1996-03-27 1997-02-27 Information searching device for searching text to retrieve a certain character stream matching a certain keyword

Family Applications Before (1)

Application Number Title Priority Date Filing Date
CN 97102954 Expired - Fee Related CN1114877C (en) 1996-03-27 1997-02-27 Information searching apparatus for searching text to retrieve character streams agreeing with key word

Country Status (3)

Country Link
JP (1) JP3024544B2 (en)
KR (1) KR100234932B1 (en)
CN (2) CN1114877C (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100440906B1 (en) * 2001-02-15 2004-07-19 전석진 Method and system for indexing document
US7702666B2 (en) 2002-06-06 2010-04-20 Ricoh Company, Ltd. Full-text search device performing merge processing by using full-text index-for-registration/deletion storage part with performing registration/deletion processing by using other full-text index-for-registration/deletion storage part
JP5437557B2 (en) * 2006-10-19 2014-03-12 富士通株式会社 Search processing method and search system
JP4893805B2 (en) 2009-11-06 2012-03-07 富士通株式会社 Information processing program, information retrieval program, and information processing apparatus

Also Published As

Publication number Publication date
CN1114877C (en) 2003-07-16
JP3024544B2 (en) 2000-03-21
CN1166005A (en) 1997-11-26
KR970066950A (en) 1997-10-13
CN1492360A (en) 2004-04-28
KR100234932B1 (en) 1999-12-15
JPH09265420A (en) 1997-10-07

Similar Documents

Publication Publication Date Title
CN1253790C (en) Display device and driving method thereof
CN1194319C (en) Method for retrieving, listing and sorting table-formatted data, and recording medium recorded retrieving, listing or sorting program
CN1498367A (en) Information processing device, momery management device, memory management method and information processing method
CN1109994C (en) Document processor and recording medium
CN1290031C (en) Character information transformation processing system
CN1752947A (en) Memory storage, storage controlling method, and computer product
CN1922605A (en) Dictionary creation device and dictionary creation method
CN1552032A (en) Database
CN1674003A (en) Guidance information retrieval apparatus and guidance information retrieval system using this guidance information retrieval apparatus
CN1881903A (en) File edition management device and method and program
CN1190477A (en) Method and apparatus for modifying existing relational database schemas to reflect changes made in corresponding object model
CN1517869A (en) Processor, arithmetic processing method and priority decision method
CN1331449A (en) Method and relative system for dividing or separating text or decument into sectional word by process of adherence
CN1288584C (en) Information display control apparatus, server, recording medium which records program and program
CN1722108A (en) Disk drive, control method thereof and disk-falsification detection method
CN1310173C (en) Table format data presenting method, inserting method, deleting method, and updating method
CN1991798A (en) Semiconductor storage apparatus
CN1288304A (en) Receiving and displaying device and method, readable recording medium of recording the program computer
CN1203395A (en) File management device and method and medium for recording file management program
CN1552075A (en) Information processing device and method, recording medium, and program
CN1722138A (en) Structured-document management apparatus, search apparatus, storage method, search method and program
CN1098501C (en) simulator and method for SQL relational database
CN1737802A (en) Information processing apparatus and method, recording medium, and program
CN1532709A (en) Data sotrage cache memory and data storage cache system
CN1151558A (en) Information searching method and system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
C17 Cessation of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20060823