JPH0660120A - Variable length data storage method and retrieval device - Google Patents

Abstract

PURPOSE:To reduce the retrieval time considerably by generating an auxiliary file group storing retrieval information or index information. CONSTITUTION:An external storage device 1 stores various files in a database and a central processing unit 2 references each file in the external storage device 1 to retrieve required information actually. A terminal equipment 3 informs the result of retrieval to the user and uses a printer 4 to print out the result of retrieval. Variable length data stored in the external storage device 1 and data files being text files in which variable length data is processed as a format as an auxiliary file groups for the retrieval, an index information file corresponding one to one to the data file, a retrieval information file not always corresponding one to one to the data file and a data base definition file are provided.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of storing variable length data and a search device, and more particularly to a method of storing data and a search device that uses a formatted text file as a data file.

[0002]

2. Description of the Related Art FIG. 29 shows a first conventional example as a data storage method and a retrieval device for handling variable length data. In FIG. 29, 191 is an item definition file that defines the format of a database that includes the number of data items, the number of items, the item length of each item, and the item name, and 192 is a data file that stores actual data. In FIG. 29, by allocating a fixed-length data area having the maximum length of variable-length data as the item length in the data file, pseudo-variable-length data can be handled. However, in this method, the storage area of the maximum length of the variable-length data is always secured, which causes waste of the storage area, and in many cases in order to retain the item length, Is 255 bytes =
There is a limitation in the item length of 127 characters, and there is a problem that variable length data having a length longer than that cannot be stored.

FIG. 30 shows a second conventional example in which the problem of the conventional example shown in FIG. 29 is avoided (a modified example of Japanese Patent Laid-Open No. 59-90142). In FIG. 30, 201,2
Reference numeral 02 denotes an item definition file and a data file having the same format as in FIG. 29. However, when the item length is 0, the item is treated specially as variable length data. In this example, only item x is variable length data. A variable-length data storage file 203 stores variable-length data. In FIG. 30, when the item x is referred to, the item length is 0, so it is recognized as variable length data. In that case, the data file 202
Since the head position of the corresponding variable length data in the variable length data storage file 203 is stored in the item x of
Based on this, actual data is taken out from the variable length data storage file 203.

The point common to the first and second conventional examples is that variable-length data is treated as if it were fixed-length data, and its purpose is in the storage area of each item of each data. It can be said that the position of is simplified as obtained by the following equation (1). [Start address of all data] + [Size of one data] x [Data number] + [Size up to item number] (1)

On the other hand, as an example of searching variable length data by another method, FIG. 31 shows an example of using a filter program which is generally called "stream editor" and capable of operation control such as SED or AWK. In FIG. 31, 211 is a search formula description file that stores search conditions set by the user, and 212 is a search formula description file 211.
A search script to be converted by 213 is a stream editor A that converts the search expression description file 211 into a search script 212. A search expression conversion script 215 describes conversion operation from the search expression description file 211 to the search script 212. A data file that is a collection of variable-length data, 216 is a search result, and 217 is a data file 215 from the data file 215 according to the search script 214
Is a stream editor B for searching. The stream editor A 213 and the stream editor B 217 may be the same program.

In the third conventional example, as compared with the first conventional example and the second conventional example described above, a special program for newly creating, adding, updating, deleting, browsing and outputting data is provided. Since there is no need to prepare it, the data retrieval method and the burden of building the device are reduced, and at the same time, the user does not have to use the program used to edit the data or change the definition of the database. It has the advantage that it does not need to remember the operation method of the new data editing device, since it only requires a text editor.

[0007]

As described above, the conventional data storage method and search device for handling variable-length data have the following problems. That is, in the first and second conventional examples, as described above,
The problem is that it is necessary to prepare a special program for each of new creation, addition, update, deletion, browsing, and output of data, and the user needs to remember how to edit new data and operate the search device. There is. Further, in the third conventional example, since the basic operation of the above-described filter program is the operation of sequentially scanning the text file to be searched from the beginning, the original operation other than the part of the text file related to the search target item is originally performed. In order to search for a part other than other items that do not need to be searched, it takes a long time to search, and in order to copy the data itself, the storage area of an external storage device (not shown) that stores the data is compressed. In addition, there is a problem that the updated matter for the original data / the duplicated data is not reflected in the duplicated data / the original data.

The present invention has been made in view of the above circumstances, and preliminarily generates an auxiliary file group holding index information or search information from a text file containing variable-length data according to definition information of a data format. In this way, the search time should be greatly shortened, and at the same time, the poor maintainability, which is one of the harmful effects of using auxiliary files, should be avoided as much as possible, and as a result, the usability of the user should be improved. It is an object of the present invention to provide a variable length data storage method and a search device.

[0009]

In order to achieve the above object, the present invention provides (1) at least one data file for storing variable-length data, which is an aggregate of a plurality of variable-length items, and the data file An index information file having a one-to-one correspondence with the position information of each item of each data in the file, and at least one search information file holding search information indicating the location of the variable length data, One database definition file that holds the definition of the database including the definition of the item of data, and a database search and output means, wherein the variable-length item is a variable-length item body and immediately before the variable-length item body. And an item identifier that defines the attribute of the variable-length item body and an item end identifier that is located immediately after the variable-length item body and that defines the end position of the variable-length item body. Each character that is composed of children and that constitutes the variable length item is a character found in a normal text file, and (2) In addition to the configuration of the variable length item, the beginning or end of the data file Alternatively, include ancillary data of indefinite length located between the item end identifier of each item and the item identifier,
Furthermore, (3) In the above (1) or (2), the index information file includes, in addition to the position information of each item in the data file, the index number from the beginning of the file, which is the line number of each data in the file. Retaining the number of line feed codes, and further, (4) In (1), (2) or (3) above, the search information file corresponds to each data file and index information file in a one-to-one correspondence. A set file, a mother set file that holds information of all registered data, an unspecified number of intermediate set files that are set according to the main search themes, and a user set that is created and deleted according to the user's inquiry (5) In any one of the above (1) to (4), the database definition file includes a definition of a data item and a population file. Holding an intermediate set generation expression for generating the intermediate set file from the file, or (6)
At least one data file storing variable-length data, which is an aggregate of a plurality of variable-length items, has a one-to-one correspondence relationship with the data file, and holds position information in the file of each item of each data. A file group consisting of an index information file, at least one search information file that holds search information indicating the location of variable-length data, and one database definition file that holds the definition of the database that contains the definition of the data item A variable-length data search device comprising an external storage device for storing, a central processing unit for performing a search operation, and a terminal for accepting a question from a user or outputting an answer, in a format specified in the central processing unit. Auxiliary file generation that generates the corresponding index information file and search set file from the formatted text file (7) In (7) above, in addition to the auxiliary file generating means, a population set file updating means for updating the population file is provided, and (8) above (6) ) Or (7), an intermediate set update means for updating the intermediate set file from the intermediate set according to the intermediate set generation formula of the database definition file is provided, and (9) the above (6), (7) or In (8), the variable length data in the corresponding data file is updated based on the information about the arbitrary data in the search information file, and the index information file is reconstructed from the updated data file. A file updating means, and further,
(10) In any one of the above (6) to (9), a database definition updating means for updating the database definition file and reconstructing an index information file from the updated database definition file is provided. Is.

[0010]

[Operation] From the data file that stores the variable length data,
By extracting the index information of each item in advance, it is possible to directly retrieve the search target data. Further, by separately managing each data and its index information and search information, it is possible to generate, edit, and delete a mother set, an intermediate set, a data set, and a user set with the same program. Even if the database definition is changed, the database can be reused only by rebuilding the index information. Further, since the search formula for the intermediate set file is held in the database definition file, the intermediate set file can be automatically updated when data is registered. Further, since the line number of each data is held in the index information file, the target data can be quickly called when updating the data.

[0011]

Embodiments will be described below with reference to the drawings. FIG. 1 is a block diagram for explaining an embodiment of a variable-length data retrieval device according to the present invention, in which 1 is an external storage device, 2 is a central processing unit, 3 is a terminal, and 4 is a printing device. is there. The external storage device 1 stores various files of the database shown in FIG. The central processing unit 2 has the internal structure shown in FIG. 8, and refers to each file in the external storage device 1 to actually perform a search. The terminal 3 is provided with at least a text screen and a keyboard, receives a question from the user, or informs the user of the search result performed by the central processing unit 2. The printer 4 finally prints the search result.

FIG. 2 shows the structure of variable-length data stored in the external storage device 1 and an auxiliary file group for searching the variable-length data. In FIG. 2, 11 is a database definition file that stores k item names, 1 (lowercase L) macro replacement expressions and m intermediate file generation expressions, and 12 is search information for all registered data in 1. Mother set file to hold, 13l
~ 13n is a group of n data files which are a plurality of formatted text files, 14l ~ 14n are 13l ~ 13n
And n index information files corresponding to one to one, 15l ~
15n is an n number of data set files having the same correspondence as 14l to 14n, 16l to 16n are intermediate set files derived from 12 by the intermediate set generation formula in 11, 17o
Reference numerals 17x to 17x are user set files that are created and deleted according to the user's question and hold the search history. Hereinafter, the structure of the file group and an example when it is implemented as a “book database” will be described with reference to FIGS. 3 to 7.

3A and 3B are a book database definition file "SHOSEKI.RAT" which is a format (FIG. 3A) and an example (FIG. 3B) of the database definition file. The database definition file is a text file. In the case of the example, the first line "#RAT" is the item definition start mark, the ninth line "#MCR" is the macro definition start mark, and the eighteenth line "#TD".
B "is the start mark of the intermediate set definition, and line 20"#END"is the end mark of the definition. That is, the number of items is 7, the number of macro definitions is 8, and the number of intermediate set definitions is 1.

4A and 4B show the data file 1
3n format (Figure (a)) and example (Figure (b)) “OHX — 9205.DA
It is a dump list (excerpt) of "T". The dump list is a typical display format used when visually displaying the contents of a file containing undisplayable character codes. It is also one of the few display formats that can clearly describe the position of each character included in the file. The leftmost 8 digits of each line are hexadecimal notation for the position in the file, and the middle 16 two-digit English characters. The number is the hexadecimal notation of the character code of each character, and the cutoff of the character string following it is the previous 1
It is a format that represents a character string when a displayable character is displayed among the 6 characters. In the example, positions 00000000 (only the last 4 digits are described) to 0029 are notes, and 002C to 00
3l is item name 1 of data 1, 0032 to 004B is item 1 of data 1, 0050 to 0055 is item name 2 of data 1, 0056 to 005B is item 2 of data 1, (omitted), 00D0 to 00D5 is data 2 Item name 1, 00D6 to 00EF indicates item 1 of data 2 (hereinafter omitted).

5 (a) and 5 (b) show the format (FIG. (A)) and example (FIG. (B)) of the index information file 14n, “OHX — 9205.
IDX ”dump list (excerpt). In the example, 000
0 to 0071 is the database definition file name for database name matching, and 0081 to 001B is the corresponding data file “OHX.
_ 9205. Total number of data existing in “DAT”, 001C to 001F
Is the position information of item 1 of data 1, (Omitted), 0034 to 0037 is the position information of item 7 of data 1, 0038 to 003B is the line number of item 7 of data 1, and 003C to 003F of item 1 of data 2. It is the position information, (abbreviated below).

FIGS. 6A and 6B show the population file 1
2, data set file 15n, intermediate set file 16
m, the format (Fig. (a)) and example (Fig.
(b)) A dump list of "KAGEYAMA.TDB". In the example, 0000 to 0017 are database definition file names for database name verification, 0018 to 001B are the total number of pieces of emotion information existing in this search information file, and 001C to 0033 are search information 1
Index information file name, 0034 to 0037 are data numbers in the index information file, and 001C to 0037 can clearly indicate the location of the variable length data. 0038 ~ 00
4F is the index information file name of the search information 2, 0050 to 0053 are the same data numbers, and are omitted below. Note that search information 3 (0070)
~ 008B), it is possible that there is a search information file that spans multiple data files.

7A to 7C are based on FIGS. 4 to 6 and show how the variable length data corresponding to the first search information of the intermediate set file "KAGEYAMA.TDB" is referred to. It is shown. In FIG. 7, an arrow refers to a reference (from a search information file to an index information file by a data number, from an index information file to a data file depends on position information), a white frame refers to a data number or position information, and a black frame refers to it. Shows the location of the file.

FIG. 8 is a diagram showing the internal structure of the central processing unit. In the figure, 201 is an internal storage device used for temporary storage of various data from each circuit in the central processing unit 2, and 202 is an internal storage device. A memory manager that arbitrates a request for a storage area from each circuit to the internal storage device 201 and secures and releases the storage area. Reference numeral 203 denotes an external storage used for reading and writing a file stored in the external storage device 1. A device interface, 204 is an expression conversion circuit that performs macro expansion and inverse Polish conversion on a normal expression, and 205 uses a high-speed alignment algorithm generally known as "quick sort" to align with an external comparison circuit. A general-purpose alignment circuit that aligns arbitrary data according to the length and the number of data per target data, 206 is two I
An ID row comparison circuit used as an external comparison circuit of the general-purpose alignment circuit 205 for comparing the size of D rows, and 210 is an auxiliary file generation for generating a corresponding index information file and a data set file from a text file of a proper format. A circuit 211 is a set alignment circuit that performs alignment on the data set on the internal storage device 201 using the first item, which is its ID item, as a key.

Further, reference numeral 212 is a set file update circuit for updating the mother set file and the intermediate set file from the designated data set file, and 213 is the data set on the internal storage device 201 and the conversion processing is completed by the expression conversion circuit 204. A keyword search circuit that performs a keyword search from the keyword search formula and leaves the result in the second data set on the internal storage device 201. Reference numerals 214 to 216 denote the internal storage device 201.
An inter-set logic (sum / product / negation) circuit 217 is input by the user to obtain the logical sum / logical product / logical negation of the first and second data sets and leave the result in the third data set. An inter-set operation circuit for performing an inter-set operation according to a recording formula, 21
Reference numeral 8 is a data file update circuit for updating a data file including specified search information including related index information files and search aggregate file groups, and 219 is a database definition update circuit for updating a file group related to the database definition file. Reference numeral 297 is a data bus which connects 201 to 299 to each other, and 299 is a control circuit which controls the central processing unit 2 as a whole.

Internal storage device 201 to ID row comparison circuit 20
No. 6 is not a main part of the present invention, so it will not be described further, but the following auxiliary file generation circuit 210
-In the description of the database definition update circuit 219, each storage area in the internal storage device 201 has a property of being secured by the memory manager 202 at an appropriate timing and being discarded after use, and when reading and writing a file. , The external storage device interface 203 is used at an appropriate timing, and the search expression has already undergone macro expansion, error checking, and reverse Polish conversion by the expression conversion circuit 204.

FIG. 9 is a diagram showing the internal structure of the auxiliary file generating circuit, and FIG. 10 is a flow chart showing the operating procedure thereof. In FIG. 9, in the internal storage device 201, a text file name M1 which is a character string representing the name of a text file having an appropriate format to be processed and data for reading the data from the text file line by line are used. A row memory M2, an index information memory M3 that holds index information for one data item, and a header memory M4 that holds search information for one data item or a header portion added to the beginning of the index information / search information file. , An item name memory M5 that is a character string type array that holds item names that act as item identifiers, and a search information memory M that is used when reading the entire search information file that has been generated once for sorting
6 is secured. Further, inside the auxiliary file generation circuit 210, a position register that holds the position information of the text file up to the present, a line counter that holds the physical line number,
There is a register group consisting of an item counter that holds the number of the item currently searched for and a registered number counter that holds the number of data found up to now.

FIG. 10 is a flow chart showing the operation procedure of the auxiliary file generation circuit. First, the corresponding index information file name and data set file name are generated from M1, and a file is created with that name. Next, a dummy header including a database definition file name (not shown) and a data number 0 is generated in M4 and written in the index information file and the data set file (S1). Next, the register group inside the auxiliary file generation circuit 210 is initialized to 0 (S2). Next, the following work is performed until the end of the text file is reached (S3). First, the current position of the file is assigned to the position register, one line is read into M2, and the line counter is incremented by 1 (S4). Next, the head part of M2 is compared with the item name at the item counter number (S5), and if they match, the contents of the position register are stored in the index information memory at the item counter number and the item counter is incremented by 1 (S6). ), If the item counter matches the maximum number of items not shown
(S7), the row counter is stored in the row information storage area of M6, the index information file name and the registration number counter are stored in M4, M6 is added to the index information file, M4 is added to the data set file, and the registration number counter is added. The value is incremented by 1, and the item counter is initialized to 0 (S8). When the end of the text file has been reached (S3), a real header consisting of the database definition file name and the registration number counter is generated in M4 and written in the head portion of the index information file and data set file (S9). Next, the entire data set file is read into M3 (S10), and the set sorting circuit 2 shown in FIG.
The first item, which is an ID item, is sorted in ascending order by 11 and written back to the data set file (S1).
1, S12). With the above, the operation of the auxiliary file generation circuit 210 ends.

FIG. 11 is a diagram showing the internal structure of the set alignment circuit, and FIG. 12 is a flow chart showing the operating procedure thereof. In FIG. 11, in the internal storage device 201, a search information memory M1 that stores a search information array to be sorted
Then, an ID row memory M2 for storing the first item of each search information of M1 and an array of ID rows consisting of serial numbers is secured. Further, inside the set sorting circuit 211, there is a number register that holds the number of pieces of search information to be sorted.

FIG. 12 is a flow chart showing the operation procedure of the set alignment circuit. First, the number of records recorded in the header part of M1 is assigned to the number of registers (S1). Then M
The first item of each search information of No. 1 is read to M2 (S2), M
A serial number is assigned to each element of 2 (S3). Next, the ID row comparison circuit 206 is designated as the comparison circuit, the content of the number register is designated as the size of the number of IDs, the size of the ID row is designated, and the general-purpose sorting circuit 205 is operated on M2 to perform the sorting. (S4). Finally, the search information in M1 is arranged in the same order as M2 by the serial number given to M2 (S5). With this, the operation of the set sorting circuit 211 ends.

FIG. 13 is a diagram showing the internal structure of the aggregate file update circuit, and FIG. 14 is a flow chart showing the operation procedure thereof. 13, in the internal storage device 201, an update set memory M1 for reading the entire search information file to be updated, an update result memory M2 for storing the updated search information, and a new addition to the mother set / intermediate set. The data set memory M3 that holds the retrieved search information, the intermediate set definition M4 that is the array of the definition information of the intermediate set, the intermediate set name M5 that is the array of the intermediate set file name, and the definition of M4 are derived by M3. A subset memory M6 for storing the addition of each of the intermediate set files to be created is secured. or,
Inside the aggregate file update circuit 212, there is an intermediate aggregate counter that holds the number of the intermediate aggregate file that is currently being updated.

FIG. 14 is a flow chart showing the operation procedure of the aggregate file update circuit. First, the mother set file is read into M1 (S1), the logical sum of M1 and M3 is obtained in M2 by the inter-set logical sum circuit shown in FIG. 17 (S2), M2
To the mother set file and update the mother set file (S3). Next, initialize the intermediate set counter with 0
(S4), while the intermediate set counter is smaller than the defined number of intermediate sets, the next operation is repeated (S5, S6). Figure 15 first
Using the keyword search circuit shown in, the search information matching the intermediate set counter th M4 in M3 is extracted to M6. Next, all the search information of M5 of the intermediate set counter th is read into M1 (S7), and the logical sum of M1 and M6 is obtained in M2 by the inter-set OR circuit shown in FIG. 17 (S8), M
2 is written back to the intermediate set counter th M5 (S9, S
10). Thus, the operation of the aggregate file update circuit 212 is completed.

FIG. 15 is a diagram showing the internal structure of the keyword search circuit 213, and FIG. 16 is a flow chart showing the operating procedure thereof. In FIG. 15, the internal storage device 20
1, a search source memory M1 that holds an array of search information of a search source, a post-search memory M2 that holds an array of search information after a search, and an index information memory that stores index information of data under search. Of M3 and each item of the data being searched,
A data storage memory M4 for storing necessary items, an intermediate result stack M5 for holding the presence / absence of individual keyword data in the data, and a keyword list M6 for holding a reverse Polish notation expression for a keyword to be searched are secured. . or,
Inside the keyword search circuit 213, a number register 1 for holding the number of search sources, a number register 2 for storing the number of successful searches, a number counter for storing the number of the search information currently being searched, and a current There is a register group consisting of search item registers that store the keyword or operator being searched.

FIG. 16 is a flowchart showing the operation procedure of the keyword search circuit. First, the number of cases M1 is taken out to the case number register 1 (S1), M2 is erased, and the case number register 2 is initialized to 0 (S2). Next, while changing the number counter from 0 to the number register l-1 (lowercase letter L) (S3), the following processing is performed for each search information. First, the index information corresponding to the current search information is set to M3.
Read in (S4). Next, M4 and M5 are erased, and the search item register is reset to the head of M6 (S5). next,
Search item register M6 by search item or operator
Are scanned, and the next processing is performed until the end of M6 is reached (S6, S7). If the current contents of the search item register is an operator (S8), two intermediate results are fetched from M5, and a prescribed operation (logical sum / logical product / logical negation) is performed on the two intermediate results. , The result is loaded in M5 (S9, S
10).

If the current contents of the search item register is a search item consisting of a keyword and an item (S8), that item is read into M4 as necessary, and the result depending on the presence or absence of a keyword in the item is M5. (S11, S12).
When the end of M6 is reached, if the search is successful according to the result of M5 (S13), the current search information is added to M2 (S14), and the number register is incremented by 1 (S15). M1
After the processing for all the search information included in is completed, the case number register 2 is written in the case number holding unit in M2 (S16), and the operation of the keyword search circuit 213 ends.

FIG. 17 is a diagram showing the internal structure of the inter-set OR circuit, and FIG. 18 is a flow chart showing its operating procedure. 17, in the internal storage device 201, a search information memory A / B M1 / 2 that holds two search information arrays of operation sources, an ID row memory M3 that stores all ID rows of M2, and Row memory M4 for reading a certain row
Then, the search information memory C M5 for storing the result is secured. Further, inside the inter-set logical sum circuit 214, M1 / M2 /
There is a register group consisting of a number register A / B / C for holding the number of M5 and a number counter for holding the number of the search information currently being processed.

FIG. 18 is a flowchart showing the operation procedure of the inter-set OR circuit. First, the number of cases of M1 and M2 is taken out to the case number register A / B (S1), and when the case number register A <the case number register B (S2), M1 and M2
And the number register A and the same number B are replaced (S3). Next, copy M2 to M5 and copy the number register B to the same C (S
4), read the first item of all search information of M2 into M3 (S
5). Next, the number counter is changed from 0 to the number register A-1.
The following processing is performed for each piece of search information of M1 while changing to (S6). First, the first item of the search information being processed is read into M4 (S7). Next, it is checked whether or not there is the same item as M4 in M3 (S8). If there is not, the search information being processed is added to M5 (S9), and the number counter C is incremented by 1 (S10). After the processing of all search information in M1 is completed, the number register C is written in the number holding unit in M5 (S11), and M5 is sorted using the set sorting circuit of FIG. 11 (S12). With the above, the operation of the inter-set logical sum circuit 214 ends.

FIG. 19 is a diagram showing the internal structure of the inter-set AND circuit, and FIG. 20 is a flow chart showing the operational taste. In FIG. 19, in the internal storage device 201, search information memories A / B M1 / 2 that hold two search information arrays of operation sources, ID row memories M3 that store all ID rows of M2, and M1 of M1 are stored. Row memory M4 for reading a certain row
Then, the search information memory C M5 for storing the result is secured. Further, inside the inter-set logical product circuit 215, M1 / M
A number register A / B / C that holds the number of 2 / M5,
There is a register group consisting of a number counter that holds the number of the search information currently being processed.

FIG. 20 is a flow chart showing the operation procedure of the inter-set AND circuit. First, the number of cases of M1 and M2 is taken out to the case number register A / B (S1, S2), and if the case number register A <the case number register B (S3), M1
And M2 and number register A and same B are replaced (S
4). Next, M5 is deleted, the number register C is initialized to 0 (S5), and the first item of all search information of M2 is read into M3 (S6). If the number-of-cases register B is 0, the operation of the inter-set AND circuit 215 ends. next,
While changing the number of items counter from 0 to the number of items register A-1 (S7), the following process is performed for each search information of M1. First, the first item of the search information being processed is read into M4 (S8). Next, it is checked whether or not there is the same item as M4 in M3 (S9). If there is, the search information being processed is added to M5 (S10) and the number counter C is incremented by 1.
(S11). After the processing of all search information in M1 is completed, the number register C is written in the number holding unit in M5 (S12). With this, the operation of the inter-set AND circuit 215 ends.

FIG. 21 is a diagram showing the internal structure of the inter-set logical NOT circuit, and FIG. 22 is a flow chart showing its operating procedure. 21, in the internal storage device 201, a search information memory A / B M1 / 2 that holds two search information arrays of operation sources, an ID row memory M3 that stores all ID rows of M2, and Row memory M4 for reading a certain row
Then, a search information memory C M5 for storing the result is secured.
In addition, inside the inter-set recording negation circuit 216, M1 / M2
There is a register group including a number register A / B / C that holds the number of / M5 and a number counter that holds the number of the search information currently being processed.

FIG. 22 is a flow chart showing the operation procedure of the collective logic NOT circuit. First, the number of cases of M1 and M2 is taken out to the case number register A / B (S1, S2). When the case number register B = 0 (S3), M1 is copied to M5.
(S4), the operation ends. If the count register B> 0, (S3) M5 is erased, the count register C is initialized to 0, and the first item of all search information of M2 is read into M3.
(S5). Next, the number counter is changed from 0 to the number register A-
While changing to 1 (S6), the following process is performed for each search information of M1. First, the first item of the search information being processed is read into M4 (S7). Next, it is checked whether or not there is the same item as M4 in M3 (S8). If there is not, the search information being processed is added to M5 (S9), and the number counter C is incremented by 1 (S10). After the processing of all search information in M1 is completed, the number register C is written in the number holding unit in M5 (S11). With the above, the operation of the inter-set logical NOT circuit 216 ends.

FIG. 23 is a diagram showing the internal structure of the inter-set arithmetic circuit, and FIG. 24 is a flow chart showing its operating procedure. In FIG. 23, in the internal storage device 201,
A search information stack M1 for storing a plurality of pieces of search information in an array first-in-first-out (LIFO) format and a search expression M2 for holding an arithmetic expression described in reverse Polish notation are secured. or,
Inside the inter-set operation circuit 217, there is a search item register that holds the search item or operator currently being processed.

FIG. 24 is a flow chart showing the operation procedure of the inter-set arithmetic circuit. First, erase M1 (S
1), each item or operator of M2 is sequentially scanned using the search item register, and the following processing is performed (S2).
First, one item or operator is retrieved from M2 into the retrieval item register (S3), and if it is an operator (S4), two retrieval sets are retrieved from M1 and the set corresponding to the operator is retrieved. Inter-logic (sum / product / negation) circuit (Figure 17 / Figure 19
/ Figure 21) is used to perform a logical operation and the result is stacked in M1.
(S5) If the search item register is not an operator, the search information having that as the search information file name is read and loaded in M1 (S6). When the scanning of M2 ends, the operation of the inter-set arithmetic circuit 217 ends.

FIG. 25 is a diagram showing the internal structure of the data file updating circuit, and FIGS. 26 (a) and 26 (b) are flowcharts showing the operating procedure. 25, in the internal storage device 201, a search information memory M1 which is an array of search information including search information corresponding to the data file to be updated, and index information for reading the index information file corresponding to the data files before and after the update. A memory M2, an ID row memory A M3 that stores an array of the first item of each data of the data file before updating, and an ID row memory B M4 that stores an array of the first item of each data of the data file after updating Is secured. Also, inside the data file update circuit 218, a data number register that holds the number of the search information corresponding to the data to be updated in M1, and a line that holds the line number information in the index information that corresponds to the data to be updated. There is a register group consisting of number registers.

FIGS. 26A and 26B are flow charts showing the operation procedure of the data file updating circuit. First, the index information file corresponding to the data to be updated in M2 is read (S1), and the line number of the data is extracted from the index information corresponding to the data to be updated in the line number register (S1).
2). Next, the first item of all the data of the data file to be updated is read into M3 based on the index information of the first item stored in M2 (S3). Next, the data file is copied to a temporary file with an appropriate name, and the temporary file is edited with a text editor using the contents of the line number register as the edit start line number (S4). Next, the corresponding index information file and data set file are generated from the edited temporary file using the auxiliary file generation circuit (FIG. 9) (S5).

Next, the index information file generated from the temporary file is read into M2 (S6), and the first item of all data of the temporary file is read into M4 according to M2 (S7).
Next, M3 and M4 are compared (S8), and only when they are exactly the same, that is, when data addition, deletion, order change or first item update has not been performed before and after updating the data file. By deleting the data file and the index information file before the update (S9) and changing the name of the index information file corresponding to the temporary file to the name of the index information file corresponding to the data file deleted earlier,
The update of the data is reflected (S10). Finally, delete the data set file generated from the temporary file (S1
1), the operation of the data file update circuit 218 ends.

FIG. 27 is a diagram showing the internal structure of the database definition updating circuit, and FIG. 28 is a flow chart showing the operating procedure thereof. In FIG. 27, the internal storage device 20
In 1 is an item definition memory M which is an array of item name character strings.
1, a macro definition memory M2 that is an array of macro definition character strings, an intermediate set definition memory A M3 that holds the definition expression of the intermediate set before updating, and an intermediate set definition that holds the definition expression of the intermediate set after updating Memory A M4 is reserved. Further, inside the database definition updating circuit 219, there is an update flag for holding the necessity of updating the intermediate set file group.

FIG. 28 is a flow chart showing the operation procedure of the database definition updating circuit. First, the database definition file is copied to a temporary file (S1) and edited with a text editor (S2). Next, the item definition part / macro definition part / intermediate set definition part of the temporary file is read into M1 / M2 / M4 respectively (S3). next,
M3 and M4 are compared (S4), and if the definition contents are the same, the update flag is set to "no", and if not, "correct" is set. Next, the index information file and the data set file are generated for all the data files existing in the external storage device 1 by using the auxiliary file generation circuit (FIG. 9) (S5, S6). Next, if the update flag is "yes" (S7), the intermediate aggregate file defined in M3 is deleted (S8), M4 is copied to M3, and the aggregate file update circuit (Fig. 13) To generate a mother set file and an intermediate set file described in M4 (S9, S1
0). Thus, the operation of the database definition update circuit 219 ends.

[0043]

Since the present invention is constructed as described above, it has the following effects. That is, in the variable-length data retrieval device, the variable-length data is stored in a data file which is a formatted text file, and in addition, there is not necessarily a one-to-one correspondence with the index information file corresponding to the data file. By using two types of auxiliary files of the search information file, it is possible to greatly reduce the data search speed, and the search device is equipped with means for automating the maintenance of the auxiliary file group. No need for complicated maintenance.

[Brief description of drawings]

FIG. 1 is a configuration diagram for explaining an embodiment of a variable length data search device according to the present invention.

FIG. 2 is a diagram showing a group of files constituting a database stored in the external storage device of FIG.

FIG. 3 is a diagram showing a format and an embodiment of a database definition file in FIG.

FIG. 4 is a diagram showing a format and an embodiment of a data file in FIG.

5 is a diagram showing a format and an embodiment of an index information file in FIG.

FIG. 6 is a diagram showing a format and an example of a search information file in FIG.

FIG. 7 is a diagram showing an example of a reference state of a search information file, an index information file and a data file in FIG.

8 is a block diagram of the central processing unit of FIG. 1. FIG.

9 is a configuration diagram of an auxiliary file generation circuit in FIG.

10 is a flowchart showing an operation procedure of the auxiliary file generation circuit in FIG.

11 is a configuration diagram of a set alignment circuit in FIG.

12 is a flowchart showing an operation procedure of the set alignment circuit in FIG.

13 is a configuration diagram of a set file update circuit in FIG.

14 is a flowchart showing an operation procedure of the collective file update circuit in FIG.

15 is a configuration diagram of a keyword search circuit in FIG.

16 is a flowchart showing an operation procedure of the keyword search circuit in FIG.

FIG. 17 is a configuration diagram of an inter-set OR circuit in FIG.

18 is a flowchart showing an operation procedure of the inter-set logical sum circuit in FIG.

19 is a configuration diagram of a configuration and an operation procedure of the inter-set OR circuit in FIG.

20 is a flowchart showing a configuration and an operation procedure of the inter-set OR circuit in FIG.

21 is a configuration diagram of the configuration and operation procedure of the inter-set logical NOT circuit in FIG.

22 is a flowchart showing a configuration and an operation procedure of the inter-set logical NOT circuit in FIG.

23 is a configuration diagram of the configuration and operation procedure of the inter-set logical operation circuit in FIG.

24 is a flowchart showing a configuration and an operation procedure of the inter-set logical operation circuit in FIG.

25 is a configuration diagram of a configuration and an operation procedure of the data file update circuit in FIG.

26 is a flowchart showing the configuration and operation procedure of the data file update circuit in FIG.

27 is a configuration diagram of a configuration and an operation procedure of the database definition update circuit in FIG.

28 is a flowchart showing the configuration and operation procedure of the database definition update circuit in FIG.

FIG. 29 is a diagram for explaining Conventional Example 1.

FIG. 30 is a diagram for explaining a second conventional example.

FIG. 31 is a diagram for explaining Conventional Example 3.

[Explanation of symbols]

1 ... External storage device, 2 ... Central processing unit, 3 ... Terminal, 4 ...
Printing device.

Claims (10)

[Claims] 1. At least one data file for storing variable-length data, which is an aggregate of a plurality of variable-length items,
An index information file having a one-to-one correspondence with the data file and holding position information in the file of each item of each data, and at least one search information holding search information indicating the location of the variable length data A file, a database definition file that holds a definition of a database including definitions of data items, and a database search / output unit, wherein the variable-length item is a variable-length item body and the variable-length item body. And an item identifier that is located immediately before the variable length item body and that defines the attribute of the variable length item body, and an item end identifier that is located immediately after the variable length item body and that defines the end position of the variable length item body. A method for storing variable-length data, in which the individual characters that make up the variable-length item are characters found in ordinary text files. 2. In addition to the structure of the variable length item, it includes ancillary data of indefinite length located at the beginning or the end of the data file or between the item end identifier of each item and the item identifier Claim 1 characterized by
How to store the described variable length data. 3. The index information file holds, in addition to the position information of each item in the data file, the number of line feed codes from the beginning of the file, which is the line number in the file of each data. The method for storing variable length data according to claim 1 or 2. 4. The search information file is a data set file corresponding to each data file and index information file in a one-to-one correspondence, a mother set file holding information of all registered data, and a main search theme. 4. The variable length data storage method according to claim 1, comprising an unspecified number of intermediate set files that are set according to the above, and a user set file that is created and deleted in response to a user inquiry. 5. The database definition file holds an intermediate set generation formula for generating the intermediate set file from a mother set file, in addition to the definition of data items. Storage method of variable length data described in. 6. At least one data file storing variable-length data, which is an aggregate of a plurality of variable-length items,
An index information file having a one-to-one correspondence with the data file and holding position information in the file of each item of each data, and at least one search information holding search information indicating the location of the variable length data An external storage device that stores a file group including a file and a database definition file that holds a definition of a database that includes definitions of data items, a central processing unit that performs a search operation, and accepts a question from a user or In a variable-length data retrieval device composed of a terminal for outputting an answer, auxiliary file generating means for generating a corresponding index information file and a retrieval set file from a text file shaped in a prescribed format in the central processing unit is provided. An apparatus for retrieving variable-length data, comprising: 7. The variable-length data retrieving apparatus according to claim 6, further comprising, in addition to the auxiliary file generating means, a mother set file updating means for updating a mother set file. 8. The variable-length data search device according to claim 6, further comprising an intermediate set updating unit that updates the intermediate set file from the intermediate set according to the intermediate set generation formula of the database definition file. 9. Data for updating variable-length data in a corresponding data file based on information about arbitrary data in the search information file, and reconstructing an index information file from the updated data file. 9. A file updating means is provided, wherein the file updating means is provided.
Retrieval device for described variable length data. 10. The variable length according to claim 6, further comprising database definition updating means for updating the database definition file and reconstructing an index information file from the updated database definition file. Data retrieval device.