JPH0869400A - Recording method for compressed data - Google Patents

Abstract

PURPOSE: To perform random access in a file by compressing uncompressed data in an uncompressed data file and recording the uncompressed data as a compressed data file by attaching managing information for the random access of the uncompressed data at every uncompressed data unit. CONSTITUTION: The compressed data file 1 is constituted of compressed segment data CSF[o] to [n] in which the uncompressed data UFS[o] to [n] in the uncompressed data file 2 are compressed and the managing information CM[o] to [n] for the random access of the uncompressed data UFS[o] to [n] are attached at every uncompressed data UFS[o] to [n]. The uncompressed data UFS[o] to [n] are the one in which the uncompressed data file 2 is divided by fixed length UFS. Each of the compressed segment data CSF [o] to [n] is provided with the same constitution. In such a way, the uncompressed data is recorded as the compressed data file by attaching the managing information for the random access of the uncompressed data at every uncompressed data unit.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a compressed data recording method for recording compressed digital data on a recording medium such as an optical disk, and more particularly to a lossless compression method that does not cause data loss. The present invention relates to a data recording method.

[0002]

2. Description of the Related Art Due to the rise of multimedia in recent years, the capacity of an external storage device such as an optical disk is increasing at a faster rate.
The capacity required by software or data also increases, and the importance of data compression technology tends to increase.

Among the above-mentioned data compression techniques, in particular, the lossless data compression technique which does not cause data loss is a method of apparently increasing the capacity of the external storage device or more effectively using the communication line. It is attracting attention as a method.

In the method of compressing with this lossless data compression technique (hereinafter referred to as lossless compression method), when the compressed data is expanded, the data before compression is completely reproduced and information is lost at all. There is a great advantage of not being forgotten.

On the other hand, however, the compression rate at the time of data compression varies depending on the content of the data to be compressed, and it is not easy to use until all the data to be compressed is completely compressed. There is. Such a property is extremely inconvenient for a data storage that makes random access.

As a lossless compression system in which data compression is applied to data storage in consideration of these properties,
For example, a method of performing file-based data compression at the application level, a method of performing data compression at the device driver level, an operating system (hereinafter referred to as OS:
It is called Operating System. ) There is a method of performing data compression as a part of the function.

The method of compressing data on a file-by-file basis at the application level is a method of compressing an input file and compressing the data using a file-by-file compression tool that outputs the compressed file. The method using such a compression tool is used not only for the external recording device but also for reducing the communication cost. There is also an extension of this compression tool that has the function of sequentially compressing unused files. As a result, the capacity of the external storage device can be apparently increased.

The method of performing data compression at the device driver level is a method of performing compression at the device driver portion by software called disk compression.

Normally, the host side recognizes the storage capacity of the drive through the device driver. In this method,
The host side is caused to recognize the storage capacity information that the drive has a capacity larger than the physical capacity that it originally has, for example, twice the capacity. That is, the host side is made to recognize the apparent logical storage capacity of the drive, not the actual storage capacity of the drive.

Then, the host side performs all the data reading or recording on the apparent recording area of the drive. At this time, this software called disk compression is performed by the data management information BIF before compression and the data management information AI after compression as shown in FIG.
A conversion table CM (C
The apparent address of the recording area of the drive is converted to the address of the actual recording area of the drive by using the compression map.

The uncompressed data management information BIF is a CC (Compr) that divides the uncompressed data into fixed lengths.
This is information indicating the number of the session cluster block. The OS reads and records data in this CC block unit. For example, when the size of the CC block is 8 Kbytes and the physical storage capacity of the drive is 80 Mbytes, the number of CC blocks is 80,000 / 8 = without data compression.
It will be 10,000 pieces. So if you do data compression,
Since the number of CC blocks is determined by the apparent logical drive capacity, for example, the data is 1 / average.
When compressed to 2, the logical drive capacity is 160
It becomes Mbyte. Therefore, the number of CC blocks is 1
60,000 / 8 = 20,000.

The compressed data management information AIF is
This is information indicating how the data of the CC block (hereinafter referred to as CC block data) is recorded on the disc. That is, the compressed data of the CC block data is divided into fixed length CS (Compres).
section Sector) block number and its CS
This is information indicating the number of blocks.

For example, as shown in FIG. 11, the CC block number = 0, the leading CS block number = 0, and C
When the number of S blocks = 4, the CC block data [0] is CS block [0] to CS block [3].
It indicates that the data is recorded over the four CS blocks.

By the conversion table CM in which the data compression information as described above is recorded, the OS on the host side reads the data into the drive which is logically apparent, or
Make a record and read the data, or record all,
It will be converted into the one on the actual drive through the software called disk compression.

That is, from the user's point of view, the above-described method of performing data compression in file units at the application level is an image in which a compressed file is recorded on a normal drive, whereas The method of performing data compression at the device driver level gives an image of using a virtual drive determined by the number of entries in the conversion table CM and the size of the CC block.

The method of performing data compression as a part of the function of the above OS is somewhat similar to the method of performing data compression at the device driver level described above, but this method uses CV.
F (Compressed Volume File)
Create a large file called a
Recognizes the CVF and performs data compression.

The CVF is one large file in which compressed files are collected, that is, a drive dedicated to compressed data. Further, the CVF has a conversion table CM created for a virtual drive as shown in FIG. 11 and can perform random access in the CVF.

The amount of data before compression that can be recorded in this CVF, that is, the capacity of the CVF is determined by the conversion table CM. However, when the size of the conversion table CM is not so large, the actual CVF is The capacity of is determined to an arbitrary size. Therefore, the apparent capacity is not insufficient, but the actual capacity is not insufficient. That is, if the compression rate is not low, the actual capacity will be large even if the apparent capacity of the CVF is the same.

In the method of performing data compression at the device driver level described above, the conversion table CM and the data compression are virtual drives over the entire drive or a part of the drive partitioned in the usual way. With this method, only the CVF is converted between the virtual drive and the real drive. A major feature of this method is that data compression is performed as part of the OS. Therefore, the OS is
It is possible to optimize the space on the disk for data compression.

[0020]

However, since the data compression function operates at the application level in the above-mentioned method of performing data compression in file units at the application level, it is impossible to randomly access the inside of the file.

For example, in the case of an uncompressed file, if the offset in the file to be accessed is known, random access can be performed using that offset.

However, when the file is compressed, the compression rate differs depending on the file, and the file is not uniformly compressed even in the same file. Further, the information indicating the correspondence between the data before compression and the data after compression like the conversion table CM described above cannot be added to the file at the application level.

Therefore, the relative relationship between the offset before the compression and the offset after the compression cannot be recognized, and when the target data is accessed, it is compressed. It was necessary to decompress the existing file from the beginning to the above target data. Thus, it was not possible to randomly access the compressed file. With the method of performing data compression at the device driver level, it is not possible to see from the OS file system how data is actually recorded on the disk. That is, since the OS file system can only see the state of data recording on the virtual drive, whether or not the data actually causes fragmentation, or the file system allocates data on the disk. Even if the optimization was performed, it was not possible to recognize whether or not the optimization was actually performed, and basically it was left to the compression tool.

Further, in this method, the capacity of the drive is determined by the size of the conversion table CM, but since there is no guarantee in the compression rate, data equivalent to the capacity of the virtual drive can be actually recorded. Not necessarily. For example, if the free capacity is actually 10 Kbytes and the host side recognizes that the free capacity is 20 Kbytes, the host side will get 2
There was a situation in which an attempt was made to record 0 Kbytes of data, and the data could not be compressed to 10 Kbytes or less and the data was lost.

Also, for example, in the conversion table CM shown in FIG. 11, the CC block size is 8 Kby.
If the size of the te, CS block is 4 Kbytes, and the CC block data of 8 Kbytes is compressed to 4 Kbytes or less, one CS is actually used.
Since blocks are used, half the space on the disk is used as compared with the case where no compression is used.

However, when the size of the compressed CC block data is 5 Kbytes, two CS blocks are used, and the space used on the disk does not change as compared with the case where compression is not used. It will be. That is, in such a case, the compression effect could not be obtained at all, although it was actually compressed considerably.

Further, in order to sufficiently obtain the effect of data compression, it is basically necessary to reduce the size of the CS block, but if the CS block is small, the number of CS blocks for one CC block data is large. Become. Therefore, a large number of bits are required for the information indicating this. That is, if the size of the CS block is reduced in order to sufficiently obtain the effect of data compression, the conversion table CM becomes large and the size of the manageable file is limited.

In the method of compressing data as a part of the function of the OS, compressed data and uncompressed data are apparently recorded in different drives. For this reason, compressed data and uncompressed data cannot be mixed in one directory.

Since the apparent capacity of the CVF is determined by the initial setting at the beginning, it cannot be changed arbitrarily after the initial setting. Further, since the conversion table CM, which is management information for data compression, is shared by all the files included in the CVF, for example, when a certain file in the CVF is compressed and transmitted through the communication line, the CM However, if it is added and sent, it takes a lot of time and labor to process it, and it is not possible to process it at high speed.

Further, when a data-compressed disk is inserted into a drive using a file system that does not have a data compression function, necessary information may be erased by mistake.

For example, when a disk in which data is compressed by providing a special area for management on the disk is inserted into a drive using a file system that does not have a data compression function, the above-mentioned file system causes I cannot recognize the area like. Therefore, when a disk having the management information area is inserted into a drive using a file system that does not have a data compression function, the management information area of compressed data may be erased.

Therefore, the present invention has been made in view of the above-mentioned conventional circumstances, and has the following objects.

That is, an object of the present invention is to provide a compressed data recording method capable of random access within a file.

Another object of the present invention is to provide a compressed data recording method capable of performing random access within a file at high speed.

Another object of the present invention is to provide a compressed data recording method which can be applied to a disk to be inserted into a drive using a file system which does not have a data compression function.

Another object of the present invention is to provide a compressed data recording method capable of varying the size of a manageable file.

[0037]

In order to solve the above-mentioned problems, a compressed data recording method according to the present invention compresses uncompressed data of an uncompressed data file and performs random access of the uncompressed data. It is characterized in that management information is added to each uncompressed data unit and recorded as a compressed data file.

Further, in the compressed data recording method according to the present invention, the non-compressed data is obtained by dividing the non-compressed data file in fixed length units, and the non-compressed data is divided in random access units and randomly Compressed for each access unit, the compressed data in the random access unit is divided and recorded in the recording unit, and the management information is information indicating the correspondence between the data divided in the random access unit and the compressed data divided in the recording unit. Is recorded.

Further, in the compressed data recording method according to the present invention, the management information includes bitmap information indicating a usage status of each recording area of the compressed data divided into recording units in the compressed data file. And

Further, in the compressed data recording method according to the present invention, the management information is arranged at equal intervals from the head of the compressed data file in an order corresponding to the order of the uncompressed data of the uncompressed data file. Characterize.

Further, the compressed data recording method according to the present invention is characterized in that the compression rate of the compressed data file is determined after the first non-compressed data is compressed.

The compressed data recording method according to the present invention is characterized in that the compressed data divided into recording units corresponding to the data divided into one random access unit is recorded in a continuous recording area.

The compressed data recording method according to the present invention is characterized in that an empty area is provided in the recording area of the data obtained by compressing the non-compressed data.

[0044]

In the compressed data recording method according to the present invention, the uncompressed data in the uncompressed data file is compressed. The management information for random access of the uncompressed data is added to each uncompressed data unit and recorded as a compressed data file.

In the compressed data recording method according to the present invention, the non-compressed data obtained by dividing the non-compressed data file into fixed length units is divided into random access units. Then, compression is performed for each random access unit.
The compressed data of the random access unit is divided into recording units. Information indicating the correspondence between the data divided in random access units and the compressed data divided in recording units is recorded in the management information. Then, the management information is added to the compressed data divided into recording units and recorded as a compressed data file.

Further, in the compressed data recording method according to the present invention, the management information is recorded with bitmap information indicating the usage status of each recording area of the compressed data divided into recording units in the compressed data file.

Further, in the compressed data recording method according to the present invention, the management information is arranged at equal intervals from the head of the compressed data file in an order corresponding to the order of the uncompressed data of the uncompressed data file.

Further, in the compressed data recording method according to the present invention, the compression rate of the compressed data file is determined after the first non-compressed data is compressed.

Further, in the compressed data recording method according to the present invention, the compressed data divided into recording units corresponding to the data divided into one random access unit is recorded in a continuous recording area.

Further, in the compressed data recording method according to the present invention, an empty area is provided in the recording area of the data obtained by compressing the non-compressed data.

[0051]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

First, a compressed data recording method according to the first embodiment of the present invention will be described.

The compressed data file 1 compressed by applying the above compressed data recording method is, for example, as shown in FIG. 1, the uncompressed data UFS of the uncompressed data file 2.
[0] to [n] are compressed and the uncompressed data UFS is obtained.
Management information C for random access from [0] to [n]
M [0] to [n] (CM: Compression M
ap) is compressed segment data CSF [0] to [n] added to each of the uncompressed data UFS [0] to [n].
(CFS: Compressed File Segm
ent)).

The uncompressed data UFS [0] to [n] are the fixed length UFS (Un of the uncompressed data file 2).
It is divided by a compressed file segment).

The compressed segment data CSF [0]-
[N] have the same configuration. The compressed segment data CSF [0] will be described below.

The compressed segment data CSF [0]
Is header information HD [0] and compressed data SH [0].
(SH: Sector Heap).

The header information HD [0] is a bit indicating the use status of the recording area of the management information CM [0] and the compressed data SH [0] for random access in the uncompressed data UFS [0]. Map information BM (Bit Ma
p) [0].

The compressed data SH [0] has a fixed number h
Of CS blocks [0] to [h-1].
At the end of the compressed data SH [0], an unused area (hereinafter referred to as a margin area) DMY [0] is provided.

The CS blocks [0] to [h-1] are the uncompressed data UFS [0] having a fixed length CC (Comp.
CC block data [0] to [m] divided by a block of (less cluster) are compressed in CC block units, and the compressed CC block data [0]
~ [M] is a fixed length CS smaller than the CC block
It is a block divided into (Compressed Sector). That is, the above CS blocks [0] to [h
-1] is a division of a logically continuous space managed by the file system into fixed lengths, which is recorded on the disk in CS block units.

A set of CS blocks corresponding to CC block data [0] is continuously recorded.

For example, the size of the CC block data [0] is 8 Kbytes and the size of the CS block is 512.
If it is byte, the above CC block data [0]
The compressed data is assigned to 1 to 16 CS blocks.

Since the set of CS blocks corresponding to the CC block data [0] is always expanded together, the link information and the like of the set of CS blocks is not provided and the area is always a continuous area, that is, a physical area. It is recorded in a logically contiguous area managed by the file system, not in a contiguous area on the disc.

Although not shown in the header information HD [0] in addition to the management information CM [0] and the bitmap information BM [0], information on the entire compressed data file 1, for example, Information about the compression method of the compressed data file 1, the uncompressed data UFS [0], the size of the compressed segment data CSF [0], and the like is also recorded.

The management information CM [0] is an information table showing the correspondence between the CC block data [0] to [m] and the CS blocks [0] to [h-1]. The uncompressed data UFS [0] has a form corresponding to one management information CM [0].

That is, the management information CM [0] is an information table for random access in the uncompressed data UFS [0], which is created for the uncompressed data UFS [0]. It is added to the header information HD [0] of the compressed segment data CSF [0] having the same label as the uncompressed data UFS [0].

The bitmap information BM [0] is the CS block [0] of the compressed segment data CSF [0].
~ CS block used in [h-1] and CS block not used are 1 bit or 2 bit
It is shown by. Therefore, like the management information CM [0], one piece of bitmap information BM [0] corresponds to one piece of uncompressed data UFS [0].

Next, the compressed data recording method will be described.

This compressed data recording method is a method of performing data compression processing as a part of the file system function of the operation system (hereinafter referred to as OS).

First, the uncompressed data file 2 has a fixed length U.
Divide into FS blocks. The uncompressed data UFS [0] to [n] divided into UFS units are divided into blocks of fixed length CC.

The uncompressed data UFS [0] will be described below.

CC block data [0] to [m] obtained by dividing the uncompressed data UFS [0] in CC units are compressed in CC units.

CC block data [0] to [m] compressed in CC units have a fixed length CS smaller than the CC unit.
Divide into blocks.

A fixed number h of CS blocks divided into the above CS units are collected, and a fixed number h of CS blocks [0] to [h].
-1] is recorded as compressed data SH [0].

Further, a margin area DMY [0] is provided at the end of the compressed data SH [0].

At this time, the management information CM indicating the correspondence between the CC block data [0] to [m] divided in CC units and the CS blocks [0] to [h-1] divided in CS units.
[0] and C in the CS blocks [0] to [h-1]
Bitmap information BM [0] indicating the usage status of the S block
To create.

Then, the management information CM [0] and the bitmap information BM [0] are added to the compressed data SH [0] to form compressed segment data CSF [0].

Here, when a set of CS blocks corresponding to one CC block data cannot be recorded in a continuous area, for example, a set of CS blocks corresponding to the CC block data [0] is continuously written. If the compressed segment data CSF [0] cannot be recorded in the compressed segment data CSF [0], a set of CS blocks corresponding to the CC block data [0] can be continuously recorded. ] And records it in the other searched compressed segment data CSF [x ≠ 0].

This search processing is performed not only when rewriting data, but also when the compressed data file 1 is first created.

The uncompressed data UFS [1] to [n] other than the uncompressed data UFS [0] are also compressed in the same manner as the uncompressed data UFS [0].

As described above, the compressed segment data CSF is obtained by adding the header information HD [x] such as the management information CM [x] and the bitmap information BM [x] to the head of a fixed number of CS blocks. [X] is formed.

That is, the compressed data SH [x] obtained by compressing the uncompressed data UFS [x] and the compressed data SH.
Since all the information for data compression such as the management information CM [x] attached to [x] is included in one file, for example, the above is applied to an old version file system that does not have a data compression function. When a disc compressed in this way is inserted, it is possible to prevent erroneous deletion of a part of the necessary information only by recognizing the compressed data file as an unreadable file. Therefore, a disk compressed like the compressed data file 1 can be applied to a file system that does not have a data compression function.

On the other hand, in a file system equipped with a data compression function, for example, by recording information of a compressed / uncompressed file in a file attribute recording area in a directory of the file system, Depending on that information, the compressed file,
Alternatively, it is possible to perform processing corresponding to an uncompressed file.

Since the file is compressed in units of files, compressed / non-compressed files or files compressed by different compression methods can be mixed. That is, since the data compression is a part of the file system function of the OS, a dedicated application program interface (API: Appl) that the application supports.
ication Program Interfac
By using e), compression, non-compression, or compression method can be automatically specified, and the user can access data without being aware of compression.

Also, one compressed segment data CSF
In [x], since CS blocks corresponding to one CC block data [x] are recorded in continuous areas,
The frequency of access can be reduced. That is, it is possible to access at high speed without needing to perform useless access when decompressing data.

In this case, a maximum of 15 unused CS blocks may occur in the compressed segment data CSF [0] that could not be recorded continuously, but the processing speed is prioritized.

Further, by providing the margin area DMY [x] at the end of the compressed data SH [x], it is possible to prevent the CS block originally continuous at the time of rewriting from being recorded at a distant position.

For example, the first compressed segment data CS
CS blocks [0] to [h- included in F [x]
When the CC block data [y] corresponding to [1] is rewritten, what was originally recorded using 4 CS blocks is rewritten to be recorded using 5 CS blocks. In this case, since the spare area DMY [x] is secured, it can be recorded not in the end of the compressed data file 1 but in the same compressed segment data CSF [x] as before rewriting. As a result, data in the file can be accessed at higher speed.

Further, the management information CM which is an information table for the random access in the file as described above.
Since [x] is added, random access within the file can be performed.

Here, the management information CM [x] will be specifically described.

In this compressed data recording method, the management information C
M [x] is recorded so as to be arranged at equal intervals from the beginning of the compressed data file 1. Further, the uncompressed data UFS [x] in the uncompressed data file 2 is recorded so as to be arranged in an order corresponding to the order.

The management information CM [x] created for one uncompressed data UFS [x] is compressed segment data CSF [x] whose label (= x) matches the uncompressed data UFS [x]. Header information HD [x] of the compressed segment data CSF [x] that matches the label of the management information CM [x], and other compressed segment data CSF [x + j].
It is created so that the recorded data CS therein can also be shown.

Management information CM recorded as described above
[X] is, for example, as shown in FIG. 2, the CC block data number CCNo, the compressed segment data CSF [x] number CFSNo including the leading CS block, the leading CS block number CSNo, and the CS number. It is composed of the number of blocks CSk.

The CC block data number CCNo is a label added in correspondence with the offset in the non-compressed data file 2 from the beginning of the non-compressed data file 2.

For example, the compressed segment data CSF
The size of the compressed data SH [x], which is the portion obtained by removing the header information HD [x] from [x], is determined by what compression ratio is assumed. At this time, since the compression rate is not constant, the size of the compressed data SH [x] also changes depending on the compression rate.

As described above, since a set of CS blocks corresponding to one CC block data [x] is recorded in continuous areas, one uncompressed data UFS is recorded.
A set of CS blocks created from [x] exists over a plurality of compressed data SH [x], or the compressed data SH [y] of a label different from that of the uncompressed data UFS [x]. It may be included in (y ≠ x).

More specifically, as shown in FIG.
Uncompressed data UFS [0] and compressed segment data CS
F [0] is one, and the header information HD [0] and the compressed data SH [0] in the compressed segment data CSF [0] are the compressed segment data CSF.
If the area for one [0] is not reached, in reality,
A portion in which the compressed data SH [0] and the header information HD [0] are recorded becomes a compressed data file. That is, it is not necessary to secure one area of the compressed segment data CSF [0] as an area of the compressed data file.

Therefore, in this case, the size of the uncompressed data file is equal to or less than the fixed length UFS, and the size of the compressed data file after compression is also the compressed segment data CSF.
Since the area is equal to or smaller than one area of [0], the header information HD
Both the management information CM [0] of [0] and the bitmap information BM [0] have a sufficient entry. The management information CM [0] has the management information CM [0].
Segment data CSF [0] that matches the label of
Information indicating the inner CS block will be recorded.

By the way, when a file having the same size as the uncompressed file shown in FIG. 3 is compressed at a low compression rate, the compressed data file after compression has two compressed segment data as shown in FIG. It records over CSF [0] and compressed segment data CSF [1]. At this time, the number of entries of the management information CM [0] of the header information HD [0] and the management information CM [1] of the header information HD [1] is determined. In this case, since the size of the uncompressed data file before compression is equal to or less than the fixed length UFS, all the information of the compressed data file after compression is recorded in the management information CM [0] of the header information HD [0] at the beginning. . And 2
Only the area is reserved for the second header information HD [1], and the header information HD [0] and the header information HD [1] are arranged at equal intervals.

Therefore, the management information CM [0] includes not only the information indicating the CS block in the compressed segment data CSF [0] that matches the label of the management information CM [0] but also the compressed segment data CSF [1]. The information indicating the CS block in [] will be recorded.

As shown in FIG. 5, the uncompressed data file is divided into two uncompressed data UFS [0] and uncompressed data UFS [1], and the respective compression rates are different, that is, The size of the compressed data UFS [0] is equal to or smaller than the area of one compressed segment data CSF [0], and the compressed data of the uncompressed data UFS [1] is the compressed segment data CSF.
Even when recording is performed over the empty area of [0] and the compressed segment data CSF [1], the header information HD [0] and the header information HD [1] are arranged at equal intervals. Then, the compression information regarding the uncompressed data UFS [0] is recorded in the management information CM [0] of the first header information HD [0], and the management information C of the second header information HD [1] is recorded.
In M [1], compression information regarding the uncompressed data UFS [1] is recorded.

Therefore, in the management information CM [0], information indicating the record data CS in the compressed segment data CSF [0] that matches the label of the management information CM [0] is recorded, and the management information CM [[ 1] includes management information CM [1]
Segment data CSF [1] that matches the label of
Not only the information indicating the CS block in the inside, but also the information indicating the CS block in the compressed segment data CSF [2] is recorded. Alternatively, as shown in FIG. 6, data obtained by compressing the uncompressed data UFS [0] is compressed segment data CSF [0] and compressed segment data CSF.
[1] is recorded over and compressed data of uncompressed data UFS [1] is compressed segment data CSF.
Even when recording is performed over the empty area of [1] and the compressed segment data CSF [2], the header information HD [0], the header information HD [1], and the header information HD [2] are evenly spaced. Deploy.

Then, the compression information relating to the uncompressed data UFS [0] is recorded in the management information CM [0] of the header information HD [0] at the beginning, and the second header information H is recorded.
The uncompressed data U is stored in the management information CM [1] of D [1].
The compression information on FS [1] is recorded, and only the area is reserved for the management information CM [2] of the third header information HD [2].

Therefore, in the management information CM [0], not only the information indicating the CS block in the compressed segment data CSF [0] that matches the label of the management information CM [0] but also the compressed segment data CSF [1]. Information indicating the recording data CS in [] is also recorded, and the management information CM [1] is compressed in which the information indicating the CS block in the compressed segment data CSF [0] matches the label of the management information CM [1]. Information indicating the CS block in the segment data CSF [1] will be recorded.

The format states shown in FIGS. 3 to 6 do not directly correspond to physical positions on the disk, but correspond to logically continuous spaces managed by the file system. FIG.

As described above, C in the compressed segment data CSF [x] that matches the label of the management information CM [x].
Not only S block but other compressed segment data CS
Since the CS block in F [x + j] can also be shown, the position of the management information CM [x] in the compressed data file 1 can be obtained by a simple calculation. Furthermore, a specific CS of certain management information CM [x]
The position of the block in the compressed data file 1 can be obtained by a simple calculation. That is, random access within a file can be performed at high speed.

For example, a case of reading the data from the first 1,500,000 bytes in the non-compressed data file 2 will be described with reference to the flowchart shown in FIG. Here, the size of the CC block data is set to 8 Kbyte.
When the size of the e and CS blocks is 512 Kbytes and the number of entries of the management information CM [x] is 128, the size of the uncompressed data UFS [x] is 8192 × 128 = 1,048,576 bytes.

First, the number of CC block data from the beginning of the non-compressed data file 2 in which the target data is included is determined (step S1).

In this case, 1,500,000 / 8192 = 183.10 ... And this data is included in the 148th CC block data. Next, it is determined what number of compressed segment data CSF [x] from the head of the compressed data file 2 the 148th CC block data corresponds to (step S2).

That is, in this case, 184/128 = 1.43 ... 184-128 = 56, and the information recorded in the 56th entry of the second management information CM [1] is the target data. It becomes the information of the corresponding compressed data.

At this time, as described above, the management information CM
[X] are arranged at equal intervals from the beginning of the compressed data file, that is, the compressed segment data CSF.
Since the size of [x] is constant, the position of the second management information CM [1] can be easily obtained.

Then, the second management information C obtained above is obtained.
The data of the CS block indicated by the information of M [1] is expanded (step S3), and necessary data in the expanded data of the CS block is read (step S4).

As described above, since the data required for random access in the file can be obtained by the four arithmetic operations,
Random access within a file can be performed accurately and at high speed.

Further, in the compressed data file 1 formatted as described above, the management information CM [x] can be increased as needed. That is, the management information CM
Since it is not necessary to limit the size of the compressed data file depending on the size of [x], the size of the compressed data file can be changed.

The addresses in the above-described embodiment of random access within a file are addresses within a logically continuous space in a file managed by the file system. Therefore, when obtaining the address of the required CC block data, that is, the CS block corresponding to the information obtained by the management information CM [x] on the actual disk, the function as the original file system is used. Ask for.

Next, a compressed data recording method according to the second embodiment of the present invention will be described.

The compressed data recording method according to the second embodiment of the present invention is the compressed data file 1 shown in FIG.
Of the compressed segment data CSF [x] of the first uncompressed data UFS [x] in the uncompressed data file 2
(X = 0) is determined after compression, and in the compressed data file 1, the compressed segment data CSF [x]
Is recorded in the compressed data file 1 with a fixed size. At this time, the bitmap information BM
The size of [x] is the determined compressed segment data CS
It is constant regardless of the size of F [x].

That is, in this compressed data recording method, as shown in FIG.
As shown in, the size of the compressed data SH [x] is determined at the time when the compression of the predetermined fixed portion, for example, D bytes, is completed from the beginning of the uncompressed data file 2.

The determining method is, for example, as shown in FIG. 9, when the size nD of the data obtained by compressing Dbyte is nD =
In the case of D, the size of the compressed data SH [x] is set to UFS which is the size of the uncompressed data UFS [x]. Also, D
The size nD of the compressed data of byte is 0.75D ≦
When nD <D, the size of the compressed data SH [x] is set to 0.
If the size nD of the data compressed to 75 UFS and Dbyte is 0.5D ≦ nD <0.75D, the size of the compressed data SH [x] is set to 0.5UFS. Also, Db
The size nD of compressed yte data is nD <0.5D
, The size of the compressed data SH [x] is 0.25 UF
S.

As described above, the first uncompressed data U
After compression of FS [x], compressed data SH [x] is obtained in four stages.
, That is, the size of the compressed segment data CSF [x] is determined and made constant in the compressed data file 1.

Here, the compression rates in the same file are similar to each other to some extent. Therefore, for example, as shown in FIG. 10, an uncompressed data file spanning two uncompressed data UFS [0] and uncompressed data UFS [1] is compressed at a high compression rate, and one compressed segment after compression. Data CS
It is possible to reduce the useless portion B that occurs when the compressed data is stored in F [0].

Further, if the size of the compressed data SH [x] is changed, the number of entries of the bitmap information BM [x] will also change, but in this compressed data recording method, the bitmap information BM [x]. ], The maximum possible size is secured in advance. Then, the information indicating the number of entries of the bitmap information BM [x] actually used after the size of the compressed segment data CSF [x] is determined is recorded in the header information HD [x]. Since this bitmap information BM [x] has 1 bit for one entry, it does not affect the above-described effect of reducing unnecessary portions.

In the first and second embodiments described above, as shown in FIG. 1, the uncompressed data file 2 obtained by dividing the uncompressed data file 2 into blocks of fixed length UFS.
Although FS [0] to [n] are further divided into fixed-length CC blocks and compressed in CC units, the uncompressed data file 2 is not divided into fixed-length UFS blocks.
The uncompressed data file 2 is compressed in CC units from the beginning, and the data having a fixed length UFS is used as uncompressed data UFS [0].
The file system may recognize that the compression process has been started for [1] and create new management information CM [1] for the uncompressed data UFS [1].

[0123]

According to the compressed data recording method of the present invention,
Compress uncompressed data in uncompressed data files. The management information for random access of the uncompressed data is added to each uncompressed data unit and recorded as a compressed data file. This allows random access within the file. In addition, there is no problem even if a disk adopting the above-mentioned compressed data recording method is inserted into a drive using a file system not equipped with a data compression function.

In the compressed data recording method according to the present invention, the uncompressed data obtained by dividing the uncompressed data file in fixed length units is divided in random access units. Then, compression is performed for each random access unit.
The compressed data of the random access unit is divided into recording units. Information indicating the correspondence between the data divided in random access units and the compressed data divided in recording units is recorded in the management information. Then, the management information is added to the compressed data divided into recording units and recorded as a compressed data file. As a result, the size of the manageable file can be changed. Moreover, random access in the file can be performed at high speed.

Further, in the compressed data recording method according to the present invention, the management information is recorded with bitmap information indicating the usage status of each recording area of the compressed data divided into recording units in the compressed data file. As a result, random access within the file can be performed at high speed.

Further, in the compressed data recording method according to the present invention, the management information is arranged at equal intervals from the head of the compressed data file in an order corresponding to the order of the uncompressed data of the uncompressed data file. As a result, random access within the file can be performed at high speed.

Further, in the compressed data recording method according to the present invention, the compression rate of the compressed data file is determined after the compression of the first non-compressed data. As a result, the size of the manageable file can be changed.

Further, in the compressed data recording method according to the present invention, the compressed data divided into recording units corresponding to the data divided into one random access unit is recorded in the continuous recording area. As a result, random access within the file can be performed at high speed.

Further, in the compressed data recording method according to the present invention, an empty area is provided in the recording area of the data obtained by compressing the non-compressed data. As a result, random access within the file can be performed at high speed.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of a compressed data file in which data is compressed and recorded by applying a compressed data recording method according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a structure of management information of the compressed data file.

[FIG. 3] Both uncompressed data and compressed segment data are 1
It is a figure showing the format of the compressed data file in one case.

FIG. 4 is a diagram showing a format in which uncompressed data having the same size as the uncompressed data is subjected to data compression recording at a low compression rate.

FIG. 5 is a diagram showing a format when an uncompressed data file composed of two uncompressed data is compressed and recorded over two compressed segment data.

FIG. 6 is a diagram showing a format when an uncompressed data file composed of two uncompressed data is compressed and recorded over three compressed segment data.

FIG. 7 is a flowchart for explaining an access process within a compressed file.

FIG. 8 is a diagram for explaining a process of determining the size of compressed segment data after the end of compression of the first non-compressed data in the compressed data recording method according to the second embodiment of the present invention.

FIG. 9 is a diagram for explaining a determination method for determining the size of the compressed segment data from four stages.

FIG. 10 is a diagram showing a format in the case where an uncompressed data file spanning two uncompressed data is compressed at a high compression rate and, after compression, the compressed data is stored in one compressed segment data.

FIG. 11 is a diagram showing a configuration of a conversion table used in a method of performing data compression at a device driver level.

[Explanation of symbols]

 1 uncompressed data file 2 compressed data file UFS uncompressed data CC CC block data CFS compressed segment data HD header information SH compressed data CM management information BM bitmap information CS CS block DMY leeway area

Claims (7)

[Claims] 1. An uncompressed data file is compressed, and management information for random access of the uncompressed data is added to each uncompressed data unit and recorded as a compressed data file. Compressed data recording method. 2. The uncompressed data is obtained by dividing the uncompressed data file in fixed length units, and the uncompressed data is divided in random access units and compressed in random access units. The divided compressed data is recorded in a recording unit, and the management information is recorded with information indicating a correspondence between the data divided in the random access unit and the compressed data divided in the recording unit. 1. The compressed data recording method described in 1. 3. The compressed data recording according to claim 2, wherein the management information includes bitmap information indicating a usage status of each recording area of the compressed data divided into recording units in the compressed data file. Method. 4. The compressed data according to claim 2, wherein the management information is arranged at equal intervals from the beginning of the compressed data file in an order corresponding to the order of the uncompressed data of the uncompressed data file. Recording method. 5. The compressed data recording method according to claim 2, wherein the compression rate of the compressed data file is determined after the first uncompressed data is compressed. 6. The compressed data recording method according to claim 5, wherein compressed data divided into recording units corresponding to data divided into one random access unit is recorded in a continuous recording area. 7. The compressed data recording method according to claim 5, wherein an empty area is provided in a recording area of the data obtained by compressing the uncompressed data.