JPH08320877A - Document retrieval device - Google Patents

Abstract

PURPOSE: To handle a document file in record and field structure even when index retrieval from plural document files is performed by using a sectioning result file which contains the positions and sizes of records and fields of the document file. CONSTITUTION: A sectioning result generation part 4 is provided and then the document file in a document file storage part 2 is sectioned into records and fields by using sectioning conditions inputted at a sectioning condition input part 3. The positions and sizes of the sectioned records and fields are recorded in a sectioning result file storage part 5, and the sectioning result file is used to generate indexes for retrieval by an index generation part 6, thereby retrieving the document file having the structure.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document retrieving apparatus for retrieving a required document from a large number of document files at high speed using a computer.

[0002]

2. Description of the Related Art In recent years, with the spread of word processors and personal computers and the practical use of character recognition by computers, the number of electronic documents created by these has increased. For this reason, interest in a document database for accumulating a large amount of document information and searching for the document information as needed is increasing. Further, in a document database, a full-text search technique that can search for a required document without adding a keyword to the document has attracted attention. In this technique, the search condition given by the user is compared with all the information of the stored documents, and the document satisfying the search condition is output. In the search condition,
A character string such as a sentence may be used in addition to a word such as a conventional keyword. However, it takes time to collate the search condition with the document information when the data becomes large. Therefore, an index file for search is created and the search is performed from this index file to shorten the search time. With this method,
It was necessary to comply with the format specified by the index creation method as the target document file format.

A conventional document retrieval device will be described below. FIG. 11 shows the configuration of a conventional document search device. In FIG. 11, reference numeral 1101 denotes a document registration unit 1102.
Is a document file storage unit, 1103 is an index creation unit, 110
4 is an index file storage unit, 1105 is a search processing unit, 11
Reference numeral 06 is a search condition input unit, 1107 is an index search unit, and 110
Reference numeral 8 is a search result display section.

The operation of the document retrieval apparatus configured as described above will be described below. First, in the document registration unit 1101, the index creation unit 1103 creates a search index for the document file in the document file storage unit 1102 and stores it in the index file storage unit 1104. Next, in the search processing unit 1105, the index search unit 1107 searches the index file 1104 with the search condition input in the search condition input unit 1106, and the search result display unit 1108 displays the search result.

[0005]

However, in the above-described conventional document search apparatus, even when a search can be performed from a plurality of files, the entire file is used as a search unit. Therefore, when searching for a plurality of records, Since one record is created as one file, there is a disadvantage that the number of files increases, and even if multiple records can be handled in one file, multiple files cannot be handled and multiple files need to be one file. There is a drawback that one file becomes large. Further, conventionally, since indexes have been created for all fields, there is a drawback that the index capacity becomes large. In addition, conventionally, since all the created index files are searched, it is necessary to search even data that does not need to be searched, which results in an additional search time. Further, conventionally, as a method of designating a document file to be searched, a method of designating all of the document files to be searched has been performed. When the number of files is very large, the document file is designated. Was difficult. In addition, when the document file is updated, it is difficult to search immediately.

The present invention solves the above-mentioned problems of the prior art. It is possible to create a search index file for a document file having a structure such as a record or a field among a plurality of document files. Also, instead of creating an index for every field, you can selectively create an index from specified fields to reduce the size of the index file.
Instead of searching from all index files, only the specified document file can be selected and searched, and when specifying the document file name to be searched, it is possible to specify all files under the directory or specify the suffix It is an object of the present invention to provide a document search device capable of easily specifying a search target document file name.

[0007]

In order to achieve the above object, the present invention relates to a document storing a plurality of document files in order to correspond to a document file having a structure such as a record or a field in a plurality of document files. A file storage part, a delimiter condition input part that inputs the conditions for dividing the document file into records and fields, a delimiter result creation part that determines the position and size of the record and field in the document file according to the delimiter conditions, and a delimiter A delimiter result file storage that stores the information created by the result creation unit, and the position and size of the required records and fields are calculated from the delimiter result file, and the data is read from the document file to create an index for searching. Section, an index file storage section that stores the index file created by the index creation section, A search condition input unit for inputting search conditions, an index search unit for performing an index search from the document file and the index file, and a search result display unit for displaying a result for a document hit by the index search unit It is a thing.

In order to selectively create an index from designated fields, a search target field input section is provided in front of the index retrieval section so that only designated document files can be selected and searched. Has a search target file name input section for inputting a file name to be searched, and an index position detection section for detecting the position of the index of the specified file from the delimitation result file.
It is provided with a document file selection unit that can specify a directory and suffix when designating a document file, and a document file change determination unit that determines whether the document file has been updated.

[0009]

According to the present invention, with the above configuration, a delimiter result creating unit records a position of record, size position, and size information from a plurality of document files according to a delimiter condition and creates an index using the delimiter result file. By creating an index file in the creating unit, it is possible to handle multiple files having a structure such as records and fields at the same time.

Since it is not necessary to create an index for all fields by providing a field to be searched and inputting a field to be searched and creating an index only for the field designated by the index creating section, The file size can be reduced. Further, by using the delimiter result file and the index position detection unit, the position of the document file to be actually searched in the index file can be obtained. Therefore, it is sufficient to perform the index search only for the specified document file, which is wasteful. Since there is no need to search the index, time can be reduced and search garbage can be reduced. Furthermore, since the directory and suffix can be selected in the document file selection section, the document file to be searched can be easily specified.

[0011]

【Example】

(Embodiment 1) Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing the configuration of a document search device according to the first embodiment of the present invention. In FIG. 1, 1 is a document registration unit, 2 is a document file storage unit, 3 is a delimiter condition input unit, 4 is a delimitation result creation unit, 5 is a delimitation result file storage unit, 6 is an index creation unit, and 7 is an index file storage. Reference numeral 8 is a search processing unit, 9 is a search condition input unit, 10 is an index search unit, and 11 is a search result display unit.

In the above configuration, for the document file stored in the document file storage unit 2, each document file is divided into a record and a field according to the delimiter condition input from the input unit 3 in the delimiter condition, and the record and the field are separated. The delimiter result creating unit 4 creates a delimiter result for efficiently storing the head position and size information and the document file name, and stores the delimiter result in the delimiter result file storage unit 5. Next, the position and size of each field are extracted from the delimiter result, the actual data is extracted from the document file storage unit 2, the index for document search is created by the index creation unit, and the created index file is stored in the index file. Store in part 7. In the search processing unit 8, the index file in the search file storage unit 7 is read according to the search condition input in the search condition input unit 9 and the index search unit 10 performs a search, and the result hit by the condition is displayed in the search result display unit 11 To display the search results.

By providing the delimiter result creation unit 4 and the delimiter result file storage unit 5 in this way, it is possible to create an index for data having a record or field structure for a plurality of document files and perform a search. it can.

FIG. 2 shows the structure of a delimiter result file which is an element of the first embodiment. In FIG. 2, 201 is a delimited result file header, 202 is a first file field table, 203 is a first file record table, 204 is a second file field table, and 2
Reference numeral 05 is a record table of the second file, 206 is a portion indicating these sequences, 207 is a field table of the nth file, 208 is a record table of the nth file, 209 is a file name management block, and 210 is a path name storage area. is there.

Reference numeral 201 denotes a delimiter result file header in which information on the number of files included in the delimiter result file, the maximum field size and the maximum record byte size is written, and 202 is written the field position and size information for each file. 203 is a record table in which information on the position and size of the record for each file is written, and 209 stores the number of records of each file name, the position of the field table, the position of the record table, and the file name. Is a file name management block in which the information of the tree structure is written, and 210 is a path name storage area which is an area for storing each file name.

The record table and the field table are arranged in the order from the first file to the n-th file. However, when they are created from the beginning, this is easier to create and any order may be used. It may be located after the path name storage area.

By constructing the delimiter result file in this way, even if the document file is updated, it is sufficient to correct only the record table and field table that have been modified, rather than recreating the delimiter result file from the beginning. It's easy.

FIG. 3 shows the structure of a field table which is one of the structures of the delimitation result file shown in FIG. In FIG. 3, 301 is the field information of the first record,
302 is the field information of the second record, 303 is the third
The field information of the record, 304 is a portion indicating these continuations, and 305 is the field information of the nth record. 306 is the position of the first field of the first record, 30
7 is the size of the 1st field of the 1st record, 308 is the position of the 2nd field of the 1st record, 309 is the size of the 2nd field of the 1st record, 310 is a part indicating these continuations, 311 is the n1st field of the 1st record Position 312 is the size of the n1st field of the first record.

The field table starts from the field information 301 of the first record, the field information 302 of the second record, and the field information 3 of the third record in order.
03, ..., It becomes the field information 305 of the nth record. Further, first, the position 306 of the first field of the first record, the size 30 of the first field of the first record
7, second field position 308 and size 30
9 followed by the nth field position 311 and size 31
Continues to 2.

By constructing the field table in this way, the position and size of each field can be obtained, so that it is easily obtained when creating an index for each field or when displaying a specific field as a search result. be able to.

FIG. 4 shows the structure of a record table which is one of the structures of the delimitation result file shown in FIG. FIG.
, 401 is the byte size of the first record, 40
2 is the byte size of the field table of the first record, 403 is the total byte size of the first record and the second record, 404 is the total byte size of the field table of the first record and the second record, and 405 is these. A portion indicating continuity, 406 indicates a total byte size from the first record to the nth record, and 407 indicates a total byte size of the field table from the first record to the nth record.

By constructing the record table in this way, the size of each record can be easily obtained, and further, the position of the field table and the number of fields of each record can be easily obtained. You can easily ask for it when indexing or viewing the entire record.

FIG. 5 shows the structure of a file name management block which is one of the structures of the delimitation result file shown in FIG. In FIG. 5, 501 is management information of the first file,
Reference numeral 502 is management information of the second file, 503 is management information of the third file, 504 is a portion indicating the continuation of these, 50
Reference numeral 5 is management information of the nth file. 506 is the file number in the index of the first file, 507 is the length of the file name of the first file, 508 is the offset of the field table of the first file, 509 is the offset of the record table of the first file, and 510 is the first file. The number of records in the file, 511 is the update start time of the first file, 512 is the “left link” of the first file, 513 is the “right link” of the first file, 514 is the “return link” of the first file, 515 Is the “parent link” of the first file, 516 is the “skip byte length” of the first file,
517 is the “inspection mask” of the first file, 518 is the first
Offset to "skip string" of file, 519
Is the offset of the "remaining character string" in the first file.

The information of 506 to 519 shows the management information of the first file in detail, and the information of 506 to 519 similarly follows from the second file to the nth file. The tree structure from 512 to 519 is "Patri".
The method for managing a character string called "cia tree" is improved to the file name management, and the first to nth files indicate the nodes of the tree. The tree structure is a binary tree, and information 512 to 515 is used to represent the binary tree. In the "Patricia tree", the position to be compared is expressed in bit units, but the 8-bit unit is easier to process, so the comparison is performed after skipping with the "skip byte length" that indicates how many bytes to skip to start the comparison. The "inspection mask" indicating a bit is used instead. Further, the offset 518 to the "skip character string" and the offset 519 to the "remaining character string" indicate the offset to the path name storage area 210 in FIG. 2, and using these two offsets, You can create a file name.
Further, by managing the file names using this tree structure, it is possible to sort the file names in alphabetical order and check at a high speed whether or not the specified document file name is included.

In the present embodiment, the improvement of the "Patricia tree" is used as a method of storing the file name of the delimiter result file in the delimiter result file storage unit 5. However, the file name itself is used instead of the "Patricia tree". It goes without saying that it may be stored.

(Second Embodiment) Next, a second embodiment of the present invention will be described with reference to the drawings. FIG. 6 is a diagram showing the configuration of a document search device according to the second embodiment of the present invention. In FIG. 6, 601 is a document registration unit, 602 is a document file storage unit, 603 is a delimitation condition input unit, and 604.
Is a delimitation result creation unit, 605 is a delimitation result file storage unit, 606 is a search target field input unit, 607 is an index creation unit, 608 is an index file storage unit, 609 is a search processing unit, 610 is a search condition input unit, and 611 is Index search section,
Reference numeral 612 is a search result display unit. The difference from the configuration of FIG. 1 is that a search target field input unit 606 is added.

The operation of the document search apparatus configured as described above will be described below. The basic operation is the same as in the first embodiment. The difference from the first embodiment is that by providing a search target field input unit 606, the index is created only for the fields designated by the index creating unit 607, and only all fields are created. Instead, it is only necessary to create an index for a specified field, so the index capacity can be reduced.

In this embodiment, the index creating section 60
In the case of No. 7, an index is created for specified fields, but an index may be created for multiple fields, an index may be created for all fields, or an index may be created for the entire record. Good.

(Embodiment 3) Next, a third embodiment of the present invention will be described with reference to the drawings. FIG. 7 is a block diagram showing the arrangement of a document search device according to the third embodiment of the present invention. In FIG. 7, 701 is a document registration unit, 702 is a document file storage unit, 703 is a delimiter condition input unit, and 704.
Is a delimitation result creation unit, 705 is a delimitation result file storage unit, 706 is an index creation unit, 707 is an index file storage unit, 708 is a search processing unit, 709 is a search target file name input unit, 710 is an index position detection unit, 711. Is a search condition input unit, 712 is an index creation unit, and 713 is a search result display unit. The difference from the configuration of FIG. 1 is that a search target file name input unit 709 and an index position detection unit 710 are added.

The operation of the document search apparatus configured as described above will be described below. The basic operation is the same as in the first embodiment. The difference from the first embodiment is that in the first embodiment, the document file storage unit 70
In contrast to the case where the search is performed for all the files stored in No. 2, the index search is performed only for the file specified by the search target file name input unit 709 in the present embodiment. The index position detection unit 710 detects at which position in the delimiter result file of the delimiter result file storage unit 705 the file name designated by the search target file name input unit 709 is located. The position information can be transferred to the index search unit 712, and the index search unit 712 can start the search from the specified position and transfer the search result to the search result display unit 713 for display.

In this embodiment, the search target file name input unit 709 can input search target files, but a plurality of files may be specified, and the file name extension (suffix) can be specified or delimited. All files included in the result file may be specified.

(Fourth Embodiment) Next, a fourth embodiment of the present invention will be described with reference to the drawings. FIG. 8 is a diagram showing the configuration of the document search device according to the first embodiment of the present invention. In FIG. 8, 801 is a document registration unit, 802 is a document file storage unit, 803 is a delimiter condition input unit, and 804.
Is a delimiter result creation unit, 805 is a delimiter result / index file storage unit, 806 is an index creation unit, 807 is a search processing unit,
808 is a search condition input unit, 809 is an index search unit, 810
Is a search result display section. In this embodiment, the delimiter result / index file storage unit 805 is formed by combining the delimiter result file storage unit 5 and the index file storage unit 7 of FIG.

The operation of the document search device configured as described above will be described below. The basic operation is the same as in the first embodiment. The difference from the first embodiment is that the delimiter result created by the delimiter result creation unit 804 and the document search index created by the index creation unit 806 are stored in the delimiter result / index file storage unit 805. The index information is stored after the path name storage area of the delimitation result file. By doing so, the number of files is reduced, and management can be simplified.

(Fifth Embodiment) Next, a fifth embodiment of the present invention will be described with reference to the drawings. FIG. 9 is a block diagram showing the arrangement of a document search device according to the fifth embodiment of the present invention. In FIG. 9, 901 is a document registration unit, 902 is a document file storage unit, 903 is a delimiter condition input unit, and 904.
Is a document file selection section, 905 is a delimitation result creation section, 9
Reference numeral 06 is a delimiter result file storage unit, 907 is an index creation unit, 908 is an index file storage unit, 909 is a search processing unit, 910 is a search condition input unit, 911 is an index search unit, 9
Reference numeral 12 is a search result display unit. The difference from the configuration of FIG. 1 is that a document file selection unit 904 is added.

The operation of the document search device configured as described above will be described below. The basic operation is the same as in the first embodiment. The difference from the first embodiment is that the document file selection unit 904 document file storage unit 9
When selecting the 02 document file, the files under the directory are searched for, all are regarded as the document files, and the path information of the selected file is passed to the delimiter result creation unit 905 to simplify the designation of the document file. be able to. Also recursively traverse the directory,
It is also possible to search for all the files and use them as the search target document files. In addition, if a document file and other files are mixed under the directory and the file extension (suffix) differs between the document file and the other files, search only the document file extension. It becomes possible to specify the document file, and the specification can be simplified compared to specifying all document file names.

In the present embodiment, the document file selection unit 904 selects the document file based on a certain standard. As a standard, the document file is newer or older than the date of creation or update of the document file, or within a certain range. The document file may be selected by time.

(Embodiment 6) Next, a sixth embodiment of the present invention will be described. The structure is the same as that of the fifth embodiment.
The operation is different in that the document file selection unit 904 passes the document file names to the delimiter result creation unit 905 in the alphabetical order of the file names, in the order of the file creation date, or in the file extension. By sorting in order and passing the path of the document file to the delimiter result creation unit 905, it is possible to control the order of the search results when searching.

In the present embodiment, the document files are sorted by the document file selection unit 904 according to a certain standard. However, the criteria may be sorted in the order of file size, file update date, or access date. Needless to say, the order of sorting may be either ascending or descending.

(Embodiment 7) Next, a seventh embodiment of the present invention will be described with reference to the drawings. FIG. 10 is a block diagram showing the arrangement of a document search device according to the seventh embodiment of the present invention. In FIG. 10, reference numeral 1001 denotes a document registration unit, 10
Reference numeral 02 is a document file storage unit, 1003 is a delimiter condition input unit, 1004 is a document file change determination unit, 1005 is a delimitation result creation unit, 1006 is a delimitation result file storage unit, 1007 is an index creation unit, 1008 is an index file storage unit, Reference numeral 1009 is a search processing unit, 1010 is a search condition input unit, 1011 is an index search unit, and 1012 is a search result display unit. The difference from the configuration of FIG. 1 is that a document file change determination unit 1004 is added.

The operation of the document search apparatus configured as described above will be described below. The basic operation is the same as in the first embodiment. The difference from the first embodiment is that the document file change determining unit 1004 determines whether or not the document file has been changed, and if so, automatically creates a delimiter result file and an index file. , The integrity of the document file and the index can be maintained.

[0041]

As described above, according to the present invention, when a document search for a plurality of files is performed using a delimiter result file, an index for document search is created even for a file having a plurality of records and a plurality of fields. It is possible to perform high-speed search using the index. Further, since it is sufficient to create an index for searching only for the designated field, it is possible to reduce the capacity of the index. Furthermore, instead of searching all files included in the index at the time of index search, you can search by reading the specified file position from the delimited result file, so do not search for unnecessary files. The search time can be shortened. In addition, when specifying a document file, you can specify all files under the directory, you can recursively follow the directory and specify the extension of the file, so enter all the file names to be searched. You don't have to.
Furthermore, by providing a document file selection part when inputting into the delimiter result file, alphabetical order,
It becomes possible to specify the order of file creation date, file size, and file extension. In addition, by providing the document file change judgment unit, when the document file is changed, the delimiter result file and the index file are automatically created, so that the latest index file can always be searched. It is possible to realize an excellent document search device capable of instantly searching.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of a document search device according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram showing the structure of a delimitation result file according to the first embodiment of the present invention.

FIG. 3 is a schematic diagram showing the structure of a field table in the first embodiment of the present invention.

FIG. 4 is a schematic diagram showing the structure of a record table in the first embodiment of the present invention.

FIG. 5 is a schematic diagram showing the structure of a file name management block in the first embodiment of the present invention.

FIG. 6 is a block diagram showing the configuration of a document search device according to a second embodiment of the present invention.

FIG. 7 is a block diagram showing a configuration of a document search device according to a third embodiment of the present invention.

FIG. 8 is a block diagram showing a configuration of a document search device according to a fourth embodiment of the present invention.

FIG. 9 is a block diagram showing the configuration of a document search device according to fifth and sixth embodiments of the present invention.

FIG. 10 is a block diagram showing the configuration of a document search device according to a seventh embodiment of the present invention.

FIG. 11 is a block diagram showing the configuration of a conventional document search device.

[Explanation of symbols]

 1 Document Registration Section 2 Document File Storage Section 3 Separation Condition Input Section 4 Separation Result Creation Section 5 Separation Result File Storage Section 6 Index Creation Section 7 Index File Storage Section 8 Search Processing Section 9 Search Condition Input Section 10 Index Search Section 11 Search Results Display unit 601 Document registration unit 602 Document file storage unit 603 Separation condition input unit 604 Separation result creation unit 605 Separation result file storage unit 606 Search target field input unit 607 Index creation unit 608 Index file storage unit 609 Search processing unit 610 Search condition input unit 610 Part 611 Index search part 612 Search result display part 701 Document registration part 702 Document file storage part 703 Separation condition input part 704 Separation result creation part 705 Separation result file storage part 706 Index creation part 707 Index file storage part 708 Search processing part 709 Search versus File input unit 710 Index position detection unit 711 Search condition input unit 712 Index search unit 713 Search result display unit 801 Document registration unit 802 Document file storage unit 803 Separation condition input unit 804 Separation result creation unit 805 Separation result / Index file storage unit 806 Index creation unit 807 Search processing unit 808 Search condition input unit 809 Index search unit 810 Search result display unit 901 Document registration unit 902 Document file storage unit 903 Separation condition input unit 904 Document file selection unit 905 Separation result creation unit 906 Separation result file storage Part 907 Index creation part 908 Index file storage part 909 Search processing part 910 Search condition input part 911 Index search part 912 Search result display part 1001 Document registration part 1002 Document file storage part 1003 Separation condition input part 1004 Document file selection unit 1005 Separation result creation unit 1006 Separation result file storage unit 1007 Index creation unit 1008 Index file storage unit 1009 Search processing unit 1010 Search condition input unit 1011 Index search unit 1012 Search result display unit 1101 Document registration unit 1102 Document file storage Part 1103 Index creation part 1104 Index file storage part 1105 Search processing part 1106 Search condition input part 1107 Index search part 1108 Index result display part

Claims (7)

[Claims] 1. A document file storage unit for storing a plurality of document files, a delimiter condition input unit for inputting a condition for dividing the document file into records and fields, and a delimiter condition for the document file. A delimiter result creation unit for obtaining the positions and sizes of records and fields, a delimiter result file storage unit for storing the information created by the delimiter result creation unit, and a required record and field position and size from the delimiter result file. An index creation unit that obtains data from the document file and creates an index for search, an index file storage unit that stores the index file created by the index creation unit, and a search condition for inputting search conditions An input unit, an index search unit for performing an index search from the document file and the index file, and the index A document search device having a search result display unit for displaying a result for a document hit by the search unit. 2. The index capacity is reduced by providing a search field input section for inputting a search field in front of the index creation section and creating an index only for the field designated by the index creation section. The document search device according to claim 1, wherein 3. A search target file input unit for inputting a search target file and an index position detection unit are provided in front of the index search unit, and at which position of the index file the search target file is located in the index position detection unit. 2. The document search device according to claim 1, wherein the search time is shortened by detecting and searching the index. 4. The delimiter result file and the index file are 1
The document retrieval device according to claim 1, wherein a delimiter result / index file storage unit for storing in one file is provided to facilitate management of the index file. 5. For a document file in a document file storage unit, all files under a directory are designated, or all files are designated by recursively tracing a directory under the directory, or files are designated. A document file selection section that selects whether to specify only the extension (suffix) of the document file or to specify the document file based on the creation or update date and time of the document file is provided so that the document file to be searched can be easily specified. The document search device according to claim 1, 6. The document file selection section stores the delimiter result information of the document files in the delimitation designation file in the alphabetical order of file names, in the order of file creation date, in the order of file size, or in the order of file size. 6. The document search device according to claim 5, wherein the extension order can be designated. 7. A document file change determination unit for determining whether a document file has been changed after the time when the delimitation result file and the index file were created, and when there is a change, the delimitation result file and the index are automatically added. The delimiter result file and the index file can be always kept up to date by creating a file.
Document retrieval device described.