JPH0789336B2 - Library management processing method - Google Patents

Description

DETAILED DESCRIPTION [Overview] This is an improved method of library management processing for a computer system.

In a system that manages files in a library, in addition to single-level access by library name and file name, a single-level file name is indexed from the hierarchical name column and file name of the file management system that is classified into multiple levels. Means is provided, and it is processed as a single-level access by the indexed file name.

With this management processing method, it is possible to economically configure a system having interfaces of both single-layer and multi-layer library management systems for the same shared file.

[Industrial application field]

The present invention relates to a file library management processing method in a computer system.

The number of systems that distribute processing to workstations composed of so-called personal computers tends to increase, but even in that case, files such as shared programs and data are stored and managed in a host computer.
Improve the efficiency of the entire system by configuring a network that connects workstations to host computers.

If there are multiple models of workstations that make up such a network and each has a different file management system, and if those users use the programs and data on the host, then each workstation It is desirable to be able to operate as under the same management as that used in.

[Problems to be solved by conventional technology and invention]

In many computer systems, files of programs and various data are appropriately grouped into a library, and one file can be designated by the library name that is the identification name of the library and the file name that identifies the file in the library. Such library management is used.

FIG. 4 is a diagram showing a configuration example of a library. For example, each library identified by the library name 5 is provided with a library control information block 1 and a hash table 2 which hold management information of the entire library.

The library management system indexes the hash table 2 by the hash value obtained from the file name by an appropriate predetermined hash function.

That is, for example, a bit string representing a file name is divided into a plurality of bit strings having an appropriate bit length, and an exclusive OR of the bit strings is calculated, or the bit string representing a file name is regarded as an integer, and the remainder of division by a certain integer is used. The hash value obtained by the operation of a widely known hash function, such as the value obtained by the operation of obtaining the hash value or the combination thereof is shown as the relative position in the table of the term of the hash table 2. By using it as a value, one item of the hash table 2 is determined.

The hash table 2 points to the file control information block 3 holding the management information about the file having the file name corresponding to the item by the pointer held in each item.

The file control information block 3 has an item including, for example, a file name and an address indicating the location of the file. As is widely known about hashing techniques,
In the above example, since the same hash value is generally obtained from a plurality of different file names, one file control information block 3 corresponds to a plurality of file names for which those same hash values are obtained, and A plurality of terms indicating "address" are provided, and the library management system processes the access to the target file 4 by the address of the term having the same file name.

When the host 10 of the computer network shown in FIG. 3 is composed of a computer for performing such library management and is used from a workstation 11 for performing similar library management, the user of the workstation 11 is Since access can be performed by the same procedure as when accessing a file by local processing in 11 and when using data or the like stored in the host 10, the operation can be easily performed.

However, depending on the type of workstation, there are those that perform multi-level library management.

Here, multi-level library management is to manage the files in the library by classifying the files in the library into appropriate hierarchies, such as business headquarters, business departments, departments, sections, etc. is there.

In this case, each file is identified in the form of "library name-layer name of first layer-layer name of second layer -...- File name".

Such a library management system for workstations allows a user who needs to access a file, for example, to first display a list of library names, specify a desired library in a so-called menu selection format, and then Select the desired hierarchy from the display of the hierarchy name list of the first hierarchy, and then select the desired hierarchy from the display of the hierarchy name list of the second hierarchy connected to the selected first hierarchy. Since it is possible to reach the specification of the required file, it is possible to configure a system that allows even a non-specialist computer or the like to easily operate the file.

If the workstation 12 is a model that performs such library management, when the user uses the file of the host 10, the user has to follow a completely different procedure, which makes it difficult to effectively use the shared file.

[Means for solving problems]

 FIG. 1 is a block diagram showing the configuration of the present invention.

The figure shows the structure of the library, 1 is a library control information block, 2 is a hash table, 3 is a file control information block, 4 is a file, and 21 is a file control information block 20 pointed to from the first section of the hash table 2. This is a partition information file that constitutes the multi-level library management information.

[Work]

The library management system processes a single-layer or multi-layer file specification according to, for example, the type of workstation requesting file access. In the case of a single layer, the format of "library name + file name" is used. File designation is processed in the same manner as in the conventional case.

In the case of multiple layers, the partition information file 21 is accessed by the leading library name of the input file specification, and the hash table 22 is indexed by the hash value obtained by the predetermined hash function from the input layer name,
The partition information 23 indicated by the pointer is read.

If the file is determined by the partition information, the processing in the case of the single layer is executed by the file name, and if the file is not determined, the file designation information input is supplemented.

With the above processing method, it is possible to economically construct a system that can handle both single-layer and multi-layer library management.

〔Example〕

Processing a library having the configuration shown in FIG. 1, for example, the third
It is assumed that the library management system in the host 10 in the figure processes single-layer or multi-layer format file designation according to the type of workstation requesting file access.

In the case of an access request from the workstation 11 for managing a single-level library, the file designation in the form of "library name + file name" is processed in the same manner as in the conventional case and the file 4 is processed by the configuration shown in FIG. Handle access to.

In the case of an access request from the workstation 12 of the multi-layer library management, the partition information file 21 of the library determined by the leading library name of the input file specification is read.

The partition information file 21 is accessed via the hash table 2 and the file control information block 20 by a normal single-layer library access using, for example, a fixed file name determined by the system.

That is, when there is a file designation in the format of "library name-first hierarchy name-second hierarchy name -...- file name" from the workstation of the multi-tiered library, first the first hierarchy name and below `` Library name-
The file control information block 20 is obtained by searching the hash table 2 in the same manner as described in the above-mentioned conventional case as if "a fixed file name determined by the system" was specified, and the file name and file The address of the partition information file 21 is obtained by selecting an item having a matching file name (a fixed file name defined by the system) from the item indicating the address, as described above.

The partition information file 21 is composed of a hash table 22 and partition information 23. The partition information file 21 has a file designation of “first name of first library name—first hierarchy name—second hierarchy name ... Hierarchical name ”of the following (specified in a format such as“ x1.x2.e1 ”as described below) or from the file specification information consisting of a hierarchical name string and a file name, Similar to the case of searching the hash table 2 by the file name in the description, the hash value is obtained by the predetermined hash function and the hash table 22 is indexed.

Since each item of the hash table 22 holds a pointer that points to the partition information 23, the partition information 23 pointed to by the pointer is read.

As described above, in general, there are a plurality of items corresponding to the same hash value, that is, the partition information 23 in this case. For example, as shown in the figure, the partition information 23 indicated by the pointer of the hash table 22 is at the top. Are arranged. Each partition information 23 is, for example, as shown in FIG. 2, a file designation information column 30, a type column 31, and a control information column 32 from the left. The type column 31 is the content of the file designation information column 30 (in the example of the figure, x1.x2 "," x1.x2.e1 ") indicates whether the hierarchy is specified (displayed as" Level "in the example in the figure) or up to the file name (displayed as" File "in the example in the figure).

When the type column is "File", the control information column on the right side holds the file name of the file in the single-layer library. When "Type", appropriate control information, such as the layer to be displayed next, is stored. The location information of menu information and other information are held.

The library management system refers to the type column 31 for the item in which the file designation information column 30 of the partition information 23 pointed from the hash table 22 matches the designation information, and when it is "file", the control information on the right side Assuming that the file name of the single-level library in the format of "library name-file name" is designated by the file name held in the column, the request access is processed again starting from the index of the hash table 2 by the file name.

When the type column 31 is "hierarchy", the contents of the control information column 32 are used to supplement the file designation information input by appropriate control such as displaying the layer name menu, and the above-mentioned by the partition information file 21. Perform similar processing.

〔The invention's effect〕

As is clear from the above description, according to the present invention, there is a remarkable industrial effect that it is possible to economically construct a computer system that can handle both single-layer and multi-layer library management.

[Brief description of drawings]

 FIG. 1 is a block diagram of the principle configuration of the present invention, FIG. 2 is an explanatory diagram of partition information, FIG. 3 is a diagram showing a configuration example of a computer network, and FIG. 4 is a block diagram of a conventional configuration example. In the figure, 1 is a library control information block, 2 and 22 are hash tables, 3 and 20 are file control information blocks, 4 is a file, 5 is a library name, 10 is a host, 11 and 12 are workstations, and 21 is partition information. File, 23 indicates partition information.

 ─────────────────────────────────────────────────── --- Continuation of the front page (56) References JP-A-56-90361 (JP, A) Masami Yamatani, Yoshihiro Akiyama "Virtual Machine" (Sho 53-12-25) Kyoritsu Publishing PP. 58-61 Jun Fujii, Nobuo Suzuki "Operating System" (Sho 45-7-31) Sangyo Tosho PP. 112-119 Junro Emura, "Structural Approach to Operating Systems (First Volume)," 10th Edition (Sho 47-7-1) Japan Computer Association P.P. 239-253 Nobukuni Takahashi, Norihisa Doi, Takashi Masuda, "Iwanami Course Information Science 16 Functions and Structure of Operating System" (1983-9-9) Iwanami Shoten PP. 284-295

Claims (1)

[Claims] 1. A computer system comprising a library composed of a plurality of files and having an access means for accessing the file by the identifier of the library and the identifier of the file. The identification name of the file is indexed from the hierarchy name column and the file name for sequentially specifying the hierarchy to which the file belongs from the upper hierarchy connected to the hierarchy to the lower hierarchy. When access to the file is requested under a specific condition, the file specification of the request is made by the identification name of the library and the column of the hierarchical name and the file name. Received by the index means (21) to index the identification name of the file, and the identification name of the library,
Using the identifier of the file obtained as a result of the index, access by the access means is executed, and when requested under a condition other than the specific condition, the designation of the requested file is performed to identify the library. A library management processing method, characterized in that it is configured to be received as a designation by a name and an identification name of a file, and the designation is used as it is to execute an access by the access means.