JPS63124147A - Directory managing system for file system - Google Patents

Abstract

PURPOSE:To minimize the number of necessary stages of directory in a state where files are sorted with plural categories by using a directory structure expressing method which allows a single life to belong to plural directories. CONSTITUTION:In a file system consisting of a directory structure expressing method 2, a file producing means 3, a file retrieving means 4 and a file deleting means 5, the method 2 expresses such a directory structure that allows a single file to belong to plural directories. The means 3 produces a single file under plural directories. The means 4 retrieves a files which belongs to plural directories and has plural file names with designation of a single directory name/file name via a descriptor. Then the means 5 deletes a designated file out of all directories including the relevant file. Thus it is possible to minimize the number of desired stages of a directory in a state where files are sorted with plural categories.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a directory management method in a file system of a computer system.

[Prior Art] In directory management in a file system of a computer system, the task of storing files in a large number of directories can be summarized as follows.

This is file classification work. When considering file classification, it is generally possible to classify files into multiple categories. For example, when classifying source files, you can classify them by the category of work content, and at the same time, by the language in which they are written. in this way,
Typically, files are classified into multiple categories and each group is stored under a corresponding directory.

Conventional directory management systems use a tree structure with nodes called directories to classify files as described above.

[Problem to be solved by the invention] However, in the tree-structured pointer) IJ management method.

The directory hierarchy is always deeper depending on the type of category. For this reason, when specifying a file, it is necessary to specify several levels of directory salts, which is cumbersome. Also, there are problems such as remembering the file name of a certain file, but forgetting which directory it is under.

One option is to temporarily give the file a different name, but you still have to specify a long file name when giving the new name. Also, position the directory,
I can think of a way to specify a directory/file name below that directory, but if you want to specify a file that is branched into a different directory in a directory above the first directory, it would be too long. A file name must be specified.

Furthermore, either method is powerless against the problem of forgetting which directory a file is under.

Therefore, an object of the present invention is to provide a directory management system that prevents the directory hierarchy from becoming deeper than necessary.

[Means for solving the problem] Since the work of storing files under a directory structure is the work of classifying files, it is natural to want to classify them into multiple categories. However, when specifying a file, it is not always necessary to specify items in all categories.

If that information is necessary and sufficient, it is preferable to specify a smaller amount of information.

Therefore, in the present invention, one file is allowed to belong to multiple directories.

Directories corresponding to items in each category are created at the same depth, and each file is created so that it belongs to the directory of the corresponding item in each category at the same time. In this way, a file can be accessed by simply specifying the directory salt that corresponds to one category, the number of directories will not be unnecessarily deep, and files can be classified according to multiple categories. It becomes possible.

In other words, the directory management method of the file system of the present invention requires a method of representing a directory structure that allows one file to belong to multiple directories, a file creation means that creates multiple pointers for one file under IJ, and , multiple pointers) Files that belong to IJ and have multiple file names are searched for in one directory, file descriptor is searched by specifying the file name, and the specified file is searched for in all directories containing that file. and file deletion means for deleting files from the computer.

[Example] Next, the present invention will be explained in detail with reference to the drawings.

Figure 1 is an explanatory diagram of a file system according to an embodiment of the present invention. File creation means 3 to create under the directory.

a file search means 4 for searching for a file descriptor of a file belonging to multiple directories and having multiple file names by specifying one directory salt and file name;
and file deletion means 5 for deleting a specified file from all directories containing the file.

Next, the operation of the present invention will be explained.

FIG. 2 is a diagram showing the directory structure representation method 2 of FIG. 1. Directory structure representation method 2 is a directory descriptor 21 that has a pointer to the directory salt/file descriptor pointer table of each directory, and a file descriptor pointer table 22 that has a pointer to the file name/file descriptor of each file. , a file descriptor table 23 containing a pointer to the file space of each file and a pointer to the directory descriptor pointer table, a directory descriptor pointer table 24 containing pointers to each directory, and a file space 25. The file descriptor pointer table 22 pointed to by the pointer to the file descriptor pointer table in the installed directory descriptor 21 contains the file names and file descriptions of files belonging to that directory that are visible when viewed from that directory. A pointer to the child is filled in.

In the directory descriptor pointer table 24 pointed to by the pointer to the directory descriptor pointer table in the file descriptor table 23, a pointer to the directory descriptor of the directory containing the file is entered.

FIG. 3 is a diagram showing the flow of the file creation means 3 of FIG. 1. In the file creation means 3, a file descriptor initialization step 31 searches for an empty entry from the file descriptor table and initializes it, a directory descriptor pointer table securing step 32 secures a directory descriptor pointer table, and a directory descriptor table is secured. search stella f33
Search for the directory descriptor of the directory specified in step f3 and create a directory descriptor pointer table.
Step 4 adds a pointer to the directory descriptor to the directory descriptor pointer table, and step 35 adds the file description obtained in the file descriptor initialization step to the file descriptor pointer table that the directory has. Adds a pointer to the child and sets it to the specified file name.

The directory descriptor search step 33, the directory descriptor pointer table creation step 34, and the file descriptor pointer registration step 35 are repeated for all specified directory names and file names.

FIG. 4 is a diagram showing the flow of the file search means 4 of FIG. 1. The file search means 4 searches for a directory descriptor from the directory name specified in the IJ descriptor search step 41 (IJ descriptor search step 41), and searches the file descriptor pointer table that the directory descriptor has in the file descriptor search step 42. Finds a pointer to the file descriptor corresponding to the file name.

FIG. 5 is a diagram showing the flow of the file deletion means 5 of FIG. 1. The file deletion means 5 obtains a file descriptor from the directory name/file name specified in the file search step 51, and extracts the directory descriptor from the directory descriptor pointer table held by the file descriptor in the IJ descriptor search step 52. Get the file descriptor? In the interface deletion step f53, the pointer to the file name/file descriptor to be deleted is deleted from the file descriptor pointer table held by the directory descriptor, and in the directory descriptor pointer table release step 54, the directory descriptor pointer table is released. ,
The file descriptor is released using the file descriptor release Stella 7'55.

Directory descriptor search Stella 7'52 and file descriptor pointer deletion step f53 are repeated for all directory descriptors registered in the directory descriptor pointer table.

FIG. 6 shows an example of actual use in a file system using the directory management method of the present invention. For example, to specify file A.

FORTRAN/A or WORK3/A may be used.

[Effects of the invention] As can be seen from the usage example above, by allowing one file to belong to multiple directories and creating each file so that it belongs to the directory of the corresponding item in each category at the same time.

It is now possible to classify files into multiple categories without making the number of directories deeper than necessary.

[Brief explanation of the drawing]

FIG. 1 is an explanatory diagram of a file system according to an embodiment of the present invention, FIG. 2 is a diagram showing directory structure representation method 2, and FIG.
Figure 4 shows the flow of the file creation means 3, Figure 4 shows the flow of the file search means 4, Figure 5 shows the flow of the file deletion means 5, and Figure 6 shows an example of use. be. In FIG. 1, 2 is a directory structure representation method, 3 is a file creation means, 4 is a file search means, and 5 is a file deletion means. Figure 1 File space Figure 3 Figure 4 Accessible Figure 5

Claims (1)

[Claims] 1. In a computer file system, there is a method of expressing the directory structure that allows one file to belong to multiple directories, a file creation method that allows one file to be created under multiple directories, and a method that allows one file to belong to multiple directories. It consists of a file search means that searches for a file descriptor of a file with multiple file names by specifying a single directory name or file name, and a file deletion means that deletes the specified file from all directories containing that file. A file system directory management method characterized by: