JPH086759A - Data compression processor and data expansion processor, and data compressing method and data expanding method for data file - Google Patents

Abstract

PURPOSE:To compress the data of the data file efficiently with high compressibility by an easy method by providing a compressing means which has a record exclusive OR operation part and a record string-length compression part. CONSTITUTION:When an application program issues a compression command, a compressing process is started. An input data read-in part 2 reads data out of the input file specified with the compression command into an input buffer 7. The record EOR part 3 EORs a record taken out of the input buffer 7 with the last record held in a work buffer. When the taken-out record is the head record, this EOR process 18 bypassed. The record string-length compression part 4 performs the string-length compression of the record itself when the taken-out record is the head record or the EOR result when the record is the 2nd or succeeding record. As a result of the string-length compression, the compressed record is written in an output buffer 9. A compressed data write part 5 writes the compressed data of the output buffer 9 to a compression file 6.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data compression processing device, a decompression processing device, a compression processing method and a decompression processing method for a data file.

[0002]

2. Description of the Related Art Conventionally, in order to obtain high compression efficiency, a complicated compression processing method in which data control is performed by using a dictionary is generally known, and a continuous length compression processing for compressing consecutive same characters. Has the disadvantage of poor compression efficiency. FIG. 13 is a diagram showing a processing flow of a conventional continuous length compression. First, the record is read into the input area (step S1), then the record in the input area is subjected to continuous length compression (step S2) and output to the output area (step S3), and whether or not all records have been subjected to continuous length compression. Is checked (step S4), if no, the same processing is performed, and if all records are compressed in continuous length, the process ends. The input area and output area exist on the memory.

For example, "C1C1C1C1C1C2C2C2C2C2C3C3C3C3
It is assumed that a 15-byte record in hexadecimal notation "C3" is subjected to continuous length compression. In the EBCDIC code system, C1 represents the letter A, C2 represents the letter B, and C3 represents the letter C. `` C1C1C1C1C1C2
If a 15-byte record "C2C2C2C2C3C3C3C3C3" is compressed in continuous length, 9-byte compressed data "C1α04C2α04C3α04" is obtained. "C1α04" means that five characters C1 are connected, "C2α04" means that five characters C2 are connected, and "C3α04" has five characters C3. It means that. “Α” is a 1-byte control character.

FIG. 14 is a diagram showing a processing flow of a conventional continuous length expansion. First, the record is read into the input area (step S1), then the record in the input area is expanded continuously (step S2) and output to the output area (step S3), and whether all records are expanded continuously. (Step S
4), in the case of no, the same processing is performed, and if all the records are continuously expanded, the end is determined. For example, "C1α
When the 9-byte compressed data "04C2α04C3α04" is continuously expanded, a 15-byte record "C1C1C1C1C1C2C2C2C2C2C3C3C3C3C3C3" is obtained.

FIG. 15 is a diagram for explaining the conventional storage of compressed data in a block area. FIG. 15 (a) shows a first conventional example. This method is called a pattern 1 method. Pattern 1
In the method, the main limit pointer M and the sub limit pointer S are set in the buffer for outputting compressed data, and when the actual pointer P in the buffer after outputting the compressed data becomes equal to or more than the sub limit pointer S, , Output control for stopping the output processing of the compressed data to the output buffer. For this reason, when the compression source record length is large, a large empty area is generated in the block, which reduces the compression efficiency.

The above-mentioned main limit pointer M is a pointer that serves as a block delimiter in the output buffer. The sub limit pointer S is a pointer at a position obtained by subtracting the worst size from the main limit pointer M. The worst size is a worst value determined from the data size of the compression source, because some data may become long when compressed.

FIG. 15 (b) shows a second conventional example. This method is called a pattern 2 method. The pattern 2 method was devised in order to eliminate the drawbacks of the pattern 1 method. In the pattern 2 method, it is possible to output the compressed data to the work area once, compare the compressed data size with the size of the empty area in the buffer (value of M-value of P), and output to the output buffer. The compressed data in the work area is transferred to the output buffer under the condition. The pattern 2 method can improve the compression efficiency as compared with the pattern 1 method, but since compressed data must be transferred to the work area for each record, another problem that the processing time is significantly reduced occurs.

FIG. 16 is a diagram for explaining the conventional restoration of compressed data when a medium abnormality occurs. The compressed data is stored in, for example, a magnetic disk. When restoring the compressed data, the compressed data stored in the magnetic disk is read into the input buffer in block units and the compressed data is restored, but a media error occurred when reading the compressed data from the magnetic disk. Sometimes, in conventional technology,
The process after that was terminated. In the illustrated example, the number of stored records per block is 4, and it is assumed that a medium abnormality has occurred when reading block 2. Under such a state, the conventional technique does not restore the compressed data forming the blocks 2, 3, and 4. That is, 12 records are lost.

[0009]

The conventional continuous length compression system as shown in FIG. 13 has a drawback that the compression efficiency is poor. Further, the conventional method of storing compressed data in the block area as shown in FIG. 15 (a) has a drawback that the data storage efficiency is poor, and the method of FIG. 15 (b) has processing efficiency. Has the drawback of being bad. Furthermore, FIG.
The conventional method of decompressing compressed data upon occurrence of a medium abnormality as shown in (1) has a drawback that a large number of records are lost.

The present invention was created in view of this point, and it is a compression processing apparatus for efficiently compressing data in a data file with a high compression rate by a simple method that does not use other resources such as a dictionary. And to provide a method. It is another object of the present invention to provide a decompression processing device and method capable of efficiently decompressing compressed data. Another object of the present invention is to provide a compression processing device and method capable of efficiently storing compressed data in an output buffer. Another object of the present invention is to provide a decompression processing device and method capable of decompressing compressed data up to a location where an abnormality occurs when a medium abnormality occurs during data input processing.

[0011]

FIG. 1A is an explanatory view of the principle of the invention of claim 1. A data compression processing apparatus according to claim 1, wherein an input buffer for storing one or a plurality of records, a compression processing means having a record exclusive OR operation section and a record run length compression section, and one or a plurality of records. The compression processing means fetches a record from the input buffer, stores the fetched record in the work area, and stores the fetched record when the fetched record is the first one. The continuous compression is performed by the long compression unit, the result of the continuous compression is stored in the output buffer, and if the retrieved record is a record following the first record, the one stored in the retrieved record and the work area is stored. The previous exclusive-OR operation is performed on the previous record by the exclusive-OR operation unit, and the result of the exclusive-OR operation is performed by the record run length compression unit. And run-length encoding Te, is characterized in that stored in the output buffer results in run-length encoding writes the retrieved records in the work area.

FIG. 1B is an explanatory view of the principle of the invention of claim 2. 3. The data decompression processing device according to claim 2, wherein an input buffer for storing one or a plurality of compressed data, a decompression processing means having a record continuous length decompression unit and a record exclusive OR operation unit, and one or a plurality of units. And an output buffer for storing the record of the record. Length-decompress, write the result of run-length decompression to the work area and output buffer, and if the extracted compressed data is subsequent to the first compressed data, then the record run-length decompressor extends that compressed data. Then, the record obtained as a result of the continuous length expansion and the record in the work area are subjected to exclusive OR operation by the record exclusive OR operation part, and the exclusive OR operation is performed. The results of calculation is characterized in that the writing in the work area and the output buffer.

According to a third aspect of the present invention, there is provided the data compression processing apparatus according to the first aspect, wherein the record consecutive length compression section performs specific control of a character string formed of a plurality of predetermined characters. Character string and a character indicating the number of the predetermined character following the specific control character, and a character string configured as a continuous sequence of a plurality of other characters,
It is characterized by being represented by the character, another specific control character following the character, and a character indicating the number of the character following the character.

A data decompression processing device according to a fourth aspect is the data decompression processing device according to the second aspect, wherein when the record consecutive length decompression unit has a specific control character appearing in the compressed data,
When a predetermined character is output by the number +1 indicated by the character following the specific control character and another specific control character appears in the compressed data, the character following the other specific control character is output. It is characterized by outputting the character existing immediately before the other specified control character by the indicated number.

FIG. 2A is a diagram for explaining the principle of the invention of claim 5. The data compression processing device according to claim 5 comprises an output buffer to which compressed data obtained by compressing records is written, a work buffer, a main limit pointer indicating a block delimiter in the output buffer, and a main limit pointer. A sub-limit pointer indicating a position obtained by subtracting a size determined by the data size of the compression source record, an actual pointer indicating a data writing position in the output buffer, and compression processing means for compressing the record and writing the record in the output buffer. If the real pointer is smaller than the sub-limit pointer, the compression processing means writes the compression result of the record in the output buffer, then updates the real pointer, and when the real pointer is equal to or larger than the sub-limit pointer. Writes the compression result of the record to the working buffer, then to the real pointer The feature is that the compression result of the work buffer is stored in the output buffer, and then the actual pointer is updated, provided that the sum of the data sizes of the compression results is less than or equal to the main limit pointer. .

FIG. 2B is an explanatory view of the principle of the invention of claim 6. A data decompression processing device according to claim 6, wherein an input buffer for storing compressed data obtained by compressing a record;
Abnormal point pointer that stores the value that shows the medium abnormal point, real pointer that shows the character reading position in the block of the input buffer, compressed data is read in block units from the data file, stored in the input buffer, and block read When a medium abnormality occurs in the process of, the data read processing means for setting the relative position in the block indicating the position of the medium abnormality to the abnormal point pointer, and the decompression processing means for decompressing the compressed data are provided. The means decompresses all the compressed data contained in the block if the medium abnormality has not occurred in the process of reading the block, and if the medium abnormality has occurred, the actual pointer becomes the abnormal point pointer. Characterized by decompressing the compressed data contained in the block until reaching Than it is.

According to a seventh aspect of the data compression processing method, a record is fetched from an input buffer storing one or a plurality of records, and if the fetched record is the first one, the fetched record is placed in a work area. In the output buffer and the result of the continuous compression is stored in the output buffer. If the retrieved record is a record following the first record, the retrieved record and the one stored in the work area are stored. An exclusive-OR operation is performed on the preceding record, the result of the exclusive-OR operation is run-length compressed, the fetched record is written to the work area, and the run-length compressed result is stored in the output buffer. To do.

According to the data decompression processing method of claim 8, the compressed data is taken out from the input buffer storing one or a plurality of compressed data, and if the taken out compressed data is the first one, the compressed data is extracted. Perform continuous length expansion, write the result of continuous length expansion to the work area and output buffer, and if the extracted compressed data follows the first compressed data, expand the compressed data by continuous length and expand the continuous length. The record obtained as a result of the above and the record in the work area are subjected to an exclusive OR operation, and the result of the exclusive OR operation is written to the work area and the output buffer.

According to a ninth aspect of the data compression processing method of the seventh aspect, in the data compression processing method of the seventh aspect, a character string formed of a plurality of consecutive predetermined characters is used as a specific control character and the specific control character. A character string that is represented by the following character that indicates the number of the predetermined characters and that is composed of a plurality of consecutive characters other than the above character, and that character, another specific control character that follows the character string, and the following character It is characterized by being expressed by a character indicating the number of the character.

The data decompression processing method according to a tenth aspect is the data decompression processing method according to the eighth aspect, wherein when a specific control character appears in the compressed data, it is indicated by a character following the specific control character. The number of specified characters is output by the number +1 that is displayed, and when another specific control character appears in the compressed data, the number of characters specified by the character following the other specific control character It is characterized in that the character existing immediately before the character is output.

According to the data compression processing method of claim 11, a main limit indicating a block delimiter in the output buffer.
Set a sub-limit pointer that indicates the position where the size determined by the data size of the compression source record is subtracted from the pointer and main limit pointer, and a real pointer that indicates the data write position in the output buffer, and read the record from the input buffer. If the actual pointer is less than the sub-limit pointer, write the compressed result of the read record to the output buffer, then update the actual pointer. If the actual pointer is greater than or equal to the sub-limit pointer, read record Write the result of compression to the work buffer, then store the result of compression in the work buffer in the output buffer, provided that the sum of the data size of the result of compression and the actual pointer is less than or equal to the main limit pointer, and then The feature is that the real pointer is updated.

According to a twelfth aspect of the present invention, in the data decompression processing method, one block of compressed data is read from the data file on the external storage device into the input buffer, and if a medium error occurs during the reading process, where is stored in the input buffer. The value indicating whether the compressed data has been read normally is written to the abnormal point pointer, and the fact that a medium error has occurred is stored.If there is no medium error during the block reading process, it is included in the block. If all the compressed data is decompressed and a media error occurs, the compressed data contained in the block is expanded until the actual pointer reaches the abnormal location pointer. Is.

[0023]

The operation of the data compression processor according to claim 1 will be described. In the data compression processing device according to claim 1,
For the first record, run length compression is performed and the result is stored in the output buffer. For the second record, the EOR of the second record and the first record is taken, and EO
The result of R is length-compressed and the result is stored in the output buffer. For the third record, the EOR of the third record and the second record is taken, the result of the EOR is subjected to continuous length compression, and the result is stored in the output buffer. Hereinafter, similar processing is repeated. The operation of the data compression processing method according to claim 7 is almost the same.

The operation of the data decompression processing device according to claim 2 will be described. The data decompression processing device of claim 2 decompresses the compressed data compressed by the data compression processing device of claim 1. The first compressed data is subjected to continuous length expansion, the continuous length expansion result data is stored in the work area, and then stored in the output buffer. For the second compressed data, the continuous length expansion is performed, the EOR of this continuous length expansion result and the data in the work area is taken, the EOR result is stored in the work area, and then stored in the output buffer. For the third and subsequent compressed data, the same processing as the processing for the second compressed data is performed. The operation of the data decompression processing method according to claim 8 is almost the same.

The operation of the data compression processor of claim 3 will be described. The data compression processing apparatus according to claim 3, wherein a character string configured as a plurality of continuous sequences of a predetermined character is arranged in a row and compressed with a specific control character (for example, "β") and a character indicating a predetermined number of characters. To do. For example, a character string consisting of five predetermined characters is "β0
Compressed as 4 ". The predetermined character is, for example, "00" in hexadecimal notation. When the EOR of the i-th record and the i + 1-th record is taken, the appearance frequency of "00" is high. The operation of the data compression processing method according to claim 9 is almost the same.

The operation of the data decompression processing device according to claim 4 will be described. In the data decompression processing device according to the fourth aspect, for example, when a character string “β04” is present, it is decompressed into a character string composed of five predetermined characters. The operation of the data decompression processing method according to claim 10 is almost the same.

The operation of the data compression processing device of claim 5 will be described. The data compression processing device according to claim 5, wherein the actual pointer is smaller than the sub-limit pointer,
If the compressed data is written directly to the output buffer and the actual pointer is greater than or equal to the secondary limit pointer, the compressed data is written to the working buffer once, and the size of this compressed data plus the current actual pointer is the main limit. -The compressed data in the work buffer is written to the output buffer on condition that the pointer is not exceeded. The operation of the data compression processing method according to claim 11 is almost the same.

The operation of the data decompression processing device according to claim 6 will be described. In the data decompression processing device according to claim 6, when a medium abnormality occurs in the block reading process, the abnormal point pointer E is set to the relative position within the block indicating the position of the medium abnormality, and the actual pointer P is set to the abnormal point pointer. The decompression process of the block is performed until E is reached. The operation of the data expansion processing method according to claim 12 is as follows:
It is almost the same.

[0029]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 3 is a functional block diagram showing an outline of a data compression processing device of the present invention. In the figure, 1 is an input file, 2 is an input data reading unit, and 3 is a record EOR.
Reference numeral 4 indicates a record continuous length compression unit, 5 indicates a compressed data writing unit, 6 indicates a compressed file, 7 indicates an input buffer, 8 indicates a work buffer, and 9 indicates an output buffer.

The input file 1 exists on a magnetic disk, for example. The same applies to the compressed file 6. The input data reading unit 2, the record EOR unit 3, the record continuous length compression unit 4, and the compressed data writing unit 5 exist as functions in the computer. The input buffer 7, the work buffer 8, and the output buffer 9 exist on the memory of the computer.

When a compression command is issued from the application program, the compression process is started. Input data reading unit 2
Reads data from the input file specified by the compression command into the input buffer 7. The record EOR section 3 is
EOR of the record fetched from the input buffer 7 and the previous record held in the work buffer is calculated. If the retrieved record is the first record, this EOR process is bypassed. In the case of the first record, the record run length compression unit 4 changes the record itself to the second record.
In the case of the record and thereafter, the EOR result is compressed for a continuous length. The result of the continuous length compression is written in the output buffer 9. The compressed data writing unit 5 writes the compressed data in the output buffer 9 into the compressed file 6.

FIG. 4 is a functional block diagram showing an outline of the decompression processing device of the present invention. In the figure, 11 is a compressed file, 12 is a compressed data reading unit, 13 is a record continuous length expansion unit, 14 is a record EOR unit, 15 is an output data writing unit, 16 is an output file, 17 is an input buffer, 1
Reference numeral 8 is a work buffer, and 19 is an output buffer.

The compressed file 11 exists on a magnetic disk, for example. The same applies to the output file 16. The compressed data reading unit 12, the record continuous length expansion unit 13, the record EOR unit 4, and the output data writing unit 15 exist as functions in the computer. The input buffer 17, the work buffer 18, and the output buffer 19 exist in the memory of the computer.

When a decompression command is issued from the application program, decompression processing is started. Compressed data reading unit 1
2 reads the compressed data from the file designated by the decompression command into the input buffer 17. The record run length expansion unit 13 runs length expansion of the compressed data extracted from the input buffer 17. The record EOR unit 14 takes the EOR of the result of the continuous length expansion and the previous record held in the work buffer 18. If the extracted compressed data is the head data, this EOR process is bypassed. The decompression result is written in the work buffer 18 and the output buffer 19. The output data writing unit 15 writes the record of the output buffer 9 to the file 16.

FIG. 5 is a diagram showing a processing flow of an example of the compression processing of the present invention. In step S1, the record is read into the input area and then copied into the work area. Step S2
Then, the input record is output after it is compressed. In step S3, it is checked whether or not the process has ended. If no, the process proceeds to step S4, and if yes, the process ends.

In step S4, the next record is read into the input area. In step S5, the record stored in the input area and the previous record stored in the work area E
Perform an OR operation. In step S6, the result of the EOR operation is run length compressed and output to the output area. In step S7, the input record is copied to the work area. Step S8
Then, it is checked whether or not it is finished. If no, step S4
Return to and end if yes.

The first record is "C1C1C1C1C1C2C2C2C2
"C2C3C3C3C3C4" and the second record is "C1C1C1C1"
C1C2C2C2C2C2C3C3C3C3C3 ". 1
The hexadecimal notation "C4" represents the letter D in the EBCDIC code system. If the processing of FIG. 5 is performed under such an assumption, first, the first record “C1C1C1C1C1C2C2C2
"C2C2C3C3C3C3C4" is continuously compressed, 10-byte compressed data "C1α04C2α04C3α03C4" is generated, and this compressed data is written to the output area. Next, the second record "C1C1C1C1C1C2C2C2C2C2C3C3C3C3C3" is EORed with the first record "C1C1C1C1C1C1C2C2C2C2C3C3C3C3C4C" to generate "000000000000000000000000000007". When "000000000000000000000000000007" is consecutively compressed, 4-byte compressed data "00α0D07" is generated,
This compressed data is written to the output area.

FIG. 6 is a diagram showing a processing flow of an example of the decompression processing of the present invention. The decompression processing of FIG. 6 is to decompress the compressed data compressed by the compression processing of FIG. In step S1, the record is read into the input area. Step S
In 2, the read record is expanded by the continuous length method. In step S3, it is checked whether or not it is in the initial state (first extension). If no, go to step S4, yes
In case of, it progresses to step S5.

In step S4, an EOR operation is performed on the expanded record and the record in the work area. In step S5, the decompression result is copied to the work area. In step S6, the record of the work area is output to the output area. Step S
In step 7, it is checked whether or not the process has ended. If yes, the process ends, and if no, the process returns to step S1.

Assume that the second record is "00α0D07". When this record is expanded using the continuous length method, "00
0000000000000000000000000007 "is obtained. The record stored in the work area is "C1C1C1C1C1C2C2C2C2C2C3C3
Assuming that it is "C3C3C4", "00000000000000000000"
"0000000007" and "C1C1C1C1C1C2C2C2C2C2C3C3C3C3C4" are EORed and "C1C1C1C1C1C1C2C2C2C2C2C3C3C3C3C3C
3 ”is obtained.

FIG. 7 is a diagram showing a processing flow of an example of record length compression of the present invention. The continuous length compression of FIG.
Taking advantage of the characteristic that "00" is often found in the compressed character string when the data file is compressed by the compression process of, the two characters of the compression encoded character string "00α" are replaced with one character of "β", The compression rate is further increased. “Β” is a 1-byte control character. The process flow of FIG. 7 is a diagram illustrating in detail the processes of steps S2 and S6 in the process flow of FIG.

In step S1, the continuous character counter N is set to 0 and one character is input. In step S2, it is checked whether or not the process is completed. If yes, the process proceeds to step S13,
In the case of, it progresses to step S3. In step S3, one character is input. In step S4, it is checked whether or not the process is finished.
In the case of yes, it progresses to step S13, and in the case of no, it progresses to step S5. In step S5, it is checked whether or not the input character is continuous with the previous character. If no, the process returns to step S3, and if yes, the process proceeds to step S6.

In step S6, the continuous character counter N is incremented by 1 and one character is input. In step S7, it is checked whether or not the input character is continuous with the previous character. If yes, the process proceeds to step S8, and if no, step S9.
Proceed to. In step S8, it is checked whether or not it is finished, and yes.
If yes, go to step S9; if no, go to step S9
Return to 6.

In step S9, discontinuous characters are output. The discontinuous character output is a process of outputting a discontinuous character string that has been input until a continuous character is detected in step S5, from the work buffer (input buffer in the case of the first record) to the output buffer. . In the output, the character string for the number of discontinuous characters is output from the stored output start pointer of the discontinuous characters. In step S10, the control character "β" representing that "00" is continuous is output to the output area on condition that the continuous character string is "00".

At step S11, the continuous character string is "00".
Under the condition that is not continuous, the continuous character and the control character "α" indicating that it is continuous are output to the output area. In step S12, the value of the continuous character counter N is output to the output area and the continuous character counter N is set to 0. Then, the process returns to step S2. In step S13, the discontinuous character is output and is set to the end.

When the record "000000000000000000000000000007" is compressed by the method of FIG. 5, "00α0D07" is generated. When the record "000000000000000000000000000007" is compressed by the method of FIG. 7, "β0D07" is generated, and further One byte shorter.

FIG. 8 is a diagram showing a processing flow of an example of the record decompression processing of the present invention. The decompressing process of the record of FIG. 8 is to decompress the compressed record compressed by the compressing process of the record of FIG. 7. The processing flow of FIG. 8 is as shown in FIG.
It is a figure explaining in detail the process of step S2 in the process flow of FIG. In step S1, one character is read from the input record in the input area. In step S2, it is checked whether the read character is a control character. If yes, step S2
If No, go to step S6. In step S3, the character string up to immediately before the control character is output to the output area. In step S4, on condition that the control character = “β”, the number of times indicated by one character next to “β” +1
"00" of is output to the output area.

In step S5, one character immediately before "α" is changed to "α" on condition that the control character is "α".
Output to the output area the number of times indicated by the next character. In step S6, it is checked whether the end of the record has been reached. If the end of the record has not been reached, the process returns to step S1. If the end has been reached, the end is determined.

FIG. 9 is a diagram for explaining the storage of compressed data in the block area according to the present invention. As shown in FIG. 15A, in the conventional technique, the main limit pointer M and the sub limit pointer S are provided in the compressed data output buffer.
And the actual pointer P in the buffer after the output of the compressed data is equal to or more than the sub limit pointer S, the output control (called the pattern 1 method) for stopping the output processing of the compressed data to the output buffer is performed. For this reason, when the compression source record length is large, a large empty area is generated in the block, which reduces the compression efficiency.

As a remedy for this problem, as shown in FIG. 15 (b), the compressed data is temporarily output to the work buffer, and the data size after compression and the free space size in buffer (MP) are calculated. In comparison, there is an output control method (referred to as a pattern 2 method) that transfers compressed data in the work area to the output buffer on the condition that the data can be output to the output buffer and improves compression efficiency. However, in the pattern 2 method, since the data transfer process is performed for each record, another problem that the processing time is significantly reduced occurs.

Therefore, in the present invention, when the real pointer P in the output buffer is smaller than the sub limit pointer S, compression is performed by the pattern 1 method, and the real pointer P is sub limit limit.
When the pointer becomes equal to or higher than the pointer S, control is performed to switch to the pattern 2 system to minimize the empty area in the output buffer,
In addition, it is possible to minimize the decrease in processing time. In the illustrated example, six pieces of compressed data are stored in one block area, and the transfer to the work area is completed only twice. In addition, management information indicating how many compressed data are present in the block is added to the head of the block.

FIG. 10 is a diagram for explaining the storage processing flow of the compressed data in the block area according to the present invention. Step S1
Then, the main limit pointer M and the sub limit pointer S in the output area are set. In step S2, the record is read into the input area. In step S3, it is checked whether the actual pointer P in the output area is smaller than the sub limit pointer S. If yes, go to step S4, no
In the case of, it progresses to step S5.

In step S4, the compression process is performed with the output destination of the compression result as the output area. Next, it progresses to step S8. In step S5, the output destination of the compression result is set to the work area and compression processing is performed. In step S6, the actual pointer P
It is checked whether the sum of the data size C and the compressed data size C is larger than the main limit pointer M. ye
In the case of s, the processing is ended, and in the case of no, the processing proceeds to step S7. In step S7, the compression result of the work area is output to the output area.

In step S8, the compressed data size C is added to the real pointer P to update the real pointer P. In step S9, it is checked whether or not the record still exists. If the record still exists, the process returns to step S2, and if it does not exist, the process ends.

In the illustrated example, since the actual pointer P becomes larger than the sub limit pointer S after the writing of the data 4, the data 5 and the data 6 are written in the output area by the pattern 2 method. Since the data 7 becomes larger than the main limit pointer M when the data 7 is written, the data 7 is not written in this output area. The storage process of the compressed data in the block area described with reference to FIGS. 9 and 10 is generally applicable to a method of managing the compressed data in block units of a fixed size, regardless of a specific compression algorithm.

FIG. 11 illustrates the restoration of compressed data when a medium abnormality occurs in the present invention. As shown in FIG. 16, in the conventional technique, the processing is terminated immediately when the medium abnormality occurs, so that there is a problem that most of the data is lost. According to the present invention, all possible portions in the block in which the medium abnormality has occurred are expanded, and the processing of subsequent blocks is continued, so that partial data loss is sufficient. In the example of FIG. 11, only the fourth compressed data of block 2 is invalidated, and the decompression processing is performed for other compressed data.

FIG. 12 is a diagram showing a flow of processing for restoring compressed data when a medium abnormality occurs according to the present invention. In step S1, the record is read into the input area. In step S2,
Whether or not a medium abnormality has occurred is checked. If yes, the process proceeds to step S6, and if no, the process proceeds to step S3. If a medium error occurs during the process of reading one block of data from the data file, the flag information indicating that a medium error has occurred while reading this block and what is normal in the input buffer The abnormal point pointer E indicating whether or not it has been read is set in the control table on the memory.

In step S3, normal expansion processing is performed and the result is output to the output area. In step S4, it is checked whether or not the record still exists, and if it exists, the process returns to step S1. If not, the process proceeds to step S5. In step S5, the end is set.

In step S6, the actual pointer P of the input area
Is smaller than the abnormal point pointer E in the input area. If no, the process proceeds to step S4, and if yes, the process proceeds to step S7. In step S7, one character is read from the position designated by the real pointer P in the input area,
The actual pointer P in the input area is incremented by 1. In step S8,
It is checked whether the read character is a control character. If yes, the process proceeds to step S11, and if no, the process proceeds to step S9.

In step S9, the read character is output to the output area and the decompressed data size C is incremented by one. In step S10, it is checked whether the decompressed data size C is larger than the uncompressed data size L, and yes
If yes, go to step S4; if no, go to step S4
Return to 6.

In step S11, it is checked whether or not the actual pointer P in the input area is smaller than the abnormal point pointer E in the input area. If no, the process proceeds to step S4, and if yes, the process proceeds to step S12. In step S12, one character (in this case, the consecutive number N) is read from the location designated by the real pointer P in the input area, and the real pointer P in the input area is incremented by one. In step S13, if the control character is α, it is checked whether or not the total of the read consecutive character number N and the decompressed data size C is larger than the uncompressed data size L. If yes, the step If it is no, the process proceeds to step S14. If the control character is β,
It is checked whether or not the total of the read continuous number N + 1 and the decompressed data size C is larger than the data size L before compression. If yes, the process proceeds to step S4, and if no, the process proceeds to step S14. .

In step S14, when the control character is α, consecutive characters are output to the output area N times. When the control character is β, consecutive characters are output to the output area N + 1 times. In step S15, when the control character is α, N is added to the decompressed data size C to decompress the decompressed data.
Update size C. Then, the process returns to step S6. When the control character is β, N + 1 is added to the decompressed data size C.
Is added to update the decompressed data size C. Then, the process returns to step S6.

The compressed data decompression process described with reference to FIGS. 11 and 12 is generally applicable to a system in which compressed data is managed in block units of a fixed size, regardless of a specific compression / decompression algorithm.

[0064]

As is apparent from the above description, according to the present invention, (i) the compression processing by the simple control can provide the effective compression efficiency and the high-speed compression processing. . (ii) By managing compressed data in block units of a fixed size, decompression processing in block units becomes possible.
User resources that are lost even when an abnormality occurs can be minimized compared to the conventional method. Such a remarkable effect can be achieved.

[Brief description of drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is an explanatory diagram (continuation) of the principle of the present invention.

FIG. 3 is a functional block diagram of a data compression processing device of the present invention.

FIG. 4 is a functional block diagram of a data decompression processing device of the present invention.

FIG. 5 is a diagram showing a processing flow of an example of compression processing of the present invention.

FIG. 6 is a diagram showing a processing flow of an example of decompression processing of the present invention.

FIG. 7 shows a processing flow of an example of continuous length compression of records according to the present invention.

FIG. 8 is a diagram showing a processing flow of an example of continuous length expansion of records according to the present invention.

FIG. 9 is a diagram illustrating storage of compressed data in a block area according to the present invention.

FIG. 10 is a diagram showing a flow of processing for storing compressed data in a block area according to the present invention.

FIG. 11 is a diagram illustrating the restoration of compressed data when a medium abnormality occurs in the present invention.

FIG. 12 is a diagram showing a flow of processing for restoring compressed data when a medium abnormality occurs according to the present invention.

FIG. 13 is a diagram showing a processing flow of conventional continuous length compression.

FIG. 14 is a diagram showing a processing flow of conventional continuous length extension.

FIG. 15 is a diagram illustrating storage of compressed data in a conventional block area.

FIG. 16 is a diagram for explaining the conventional decompression of compressed data when a medium abnormality occurs.

[Explanation of symbols]

 1 Input File 2 Input Data Reading Section 3 Record EOR Section 4 Record Continuous Length Compressing Section 5 Compressed Data Writing Section 6 Compressed File 7 Input Buffer 8 Work Buffer 9 Output Buffer 11 Compressed File 12 Compressed Data Reading Section 13 Record Continuous Length Expansion Section 14 Record EOR section 15 Output data writing section 16 Output file 17 Input buffer 18 Work buffer 19 Output buffer

Claims (12)

[Claims] 1. An input buffer for storing one or a plurality of records, a compression processing means having a record exclusive OR operation section and a record run length compression section, and one or a plurality of compressed data. And an output buffer, wherein the compression processing means fetches a record from the input buffer, and if the fetched record is the first one,
Stores the retrieved record in the work area, compresses it continuously with the record concatenation compression unit, stores the result of the concatenation length compression in the output buffer, and when the retrieved record is the record following the first record, , The record fetched and the previous record stored in the work area are subjected to exclusive OR operation by the record exclusive OR operation part, and the result of the exclusive OR operation is continuous length compressed by the record continuous length compression part. The data compression processing apparatus is characterized in that the fetched record is written in the work area and the result of the continuous length compression is stored in the output buffer. 2. An input buffer for storing one or more compressed data, a decompression processing means having a record run length decompression unit and a record exclusive OR operation unit, and one or more records. The decompression processing unit fetches the compressed data from the input buffer, and if the fetched compressed data is the first one, the decompressed data is continuously decompressed by the record concatenated decompression unit and concatenated. The result of the long decompression is written to the work area and the output buffer, and if the extracted compressed data follows the first compressed data, the record continuous length decompressor is used to continuously expand the compressed data. The record obtained as a result of the above and the record in the work area are subjected to exclusive OR operation by the record exclusive OR operation part, and the result of the exclusive OR operation is made to the work area. Data expanding device and writes to the spare output buffers. 3. The record consecutive length compression unit includes a character string configured as a plurality of consecutive sequences of a predetermined character as a specific control character and a character indicating the number of the predetermined character following the specific control character. , A character string composed of a continuous sequence of other characters is represented by the character, another specific control character that follows it, and a character that indicates the number of that character that follows it. The data compression processing device according to claim 1. 4. When the specific control character appears in the compressed data, the record continuous length decompressing unit is equal to the number +1 indicated by the character following the specific control character,
Output the specified character, and if other specific control characters appear in the compressed data,
The data decompression processing device according to claim 2, wherein the character existing immediately before the other specific control character is output by the number indicated by the character following the other specific control character. 5. An output buffer to which compressed data obtained by compressing a record is written, a work buffer, and a main limit indicating a block delimiter in the output buffer.
Pointer, main limit pointer, sub-limit pointer that indicates the position where the size determined by the data size of the compression source record is subtracted, real pointer that indicates the data write position in the output buffer, and output buffer that compresses the record When the real pointer is smaller than the sub-limit pointer, the compression processing means writes the compression result of the record to the output buffer, then updates the real pointer, and the real pointer is If it is greater than or equal to the secondary limit pointer, the compression result of the record is written to the work buffer, and then the actual pointer plus the data size of the compression result is less than or equal to the main limit pointer. The compression result is stored in the output buffer, and then the actual pointer is updated. Data compression apparatus for treating. 6. An input buffer for storing compressed data obtained by compressing a record, an abnormal point pointer for storing a value indicating a medium abnormal point, and an actual pointer for indicating a character reading position in a block of the input buffer. Read the compressed data from the data file in block units, store it in the input buffer, and when the medium error occurs in the process of reading the block, set the relative position within the block that indicates the position of the medium error in the abnormal point pointer. It comprises a processing means and a decompression processing means for decompressing the compressed data, and the decompression processing means, when a medium error has not occurred in the process of reading the block, compresses the compressed data contained in the block. When all media are expanded and a media error occurs, the actual pointer reaches the error location pointer. The data decompression processing device is characterized in that it decompresses the compressed data included in the block until the end. 7. A record is fetched from an input buffer storing one or a plurality of records, and if the fetched record is the first one, the fetched record is stored in a work area and compressed for continuous length,
Stores the result of run-length compression in the output buffer, and when the fetched record is the record following the first record, the fetched record and the previous record stored in the work area are subjected to exclusive logic. Sum operation,
A data compression processing method characterized in that a result of an exclusive OR operation is subjected to continuous length compression, a fetched record is written to a work area, and a result of the continuous length compression is stored in an output buffer. 8. The compressed data is extracted from an input buffer storing one or a plurality of compressed data, and if the extracted compressed data is the first one, the compressed data is expanded by continuous length and expanded by continuous length. The result of is written to the work area and the output buffer, and if the extracted compressed data is subsequent to the first compressed data, the compressed data is expanded continuously, and the record and work obtained as a result of continuous expansion are A data decompression processing method characterized by performing an exclusive OR operation with a record in an area and writing the result of the exclusive OR operation to a work area and an output buffer. 9. A character string composed of a plurality of consecutive predetermined characters is represented by a specific control character and a character indicating the number of the predetermined characters following the specific control character, and other characters 7. A character string configured as a plurality of continuous sequences of the above is represented by the character, another specific control character following the character, and a character indicating the number of the characters following the character. Data compression processing method. 10. When a specific control character appears in the compressed data, a predetermined character is output by the number +1 indicated by the character following the specific control character, and another specific character is output in the compressed data. If a control character appears,
9. The data decompression processing method according to claim 8, wherein the character existing immediately before the other specific control character is output by the number indicated by the character following the other specific control character. 11. A main limit pointer indicating a block delimiter in the output buffer, a sub limit pointer indicating a position obtained by subtracting a size determined by the data size of the compression source record from the main limit pointer, and data writing in the output buffer. Set the real pointer indicating the position, read the record from the input buffer, and if the real pointer is less than the secondary limit pointer, write the compressed result of the read record to the output buffer, then update the real pointer, If the actual pointer is greater than or equal to the sub-limit pointer, write the compression result of the read record into the work buffer, and then add the actual pointer plus the data size of the compression result and make sure it is less than or equal to the main limit pointer. Output the work buffer compression result as a condition. Stored in the file, then the data compression processing method characterized by updating the actual pointer. 12. An input buffer is loaded with one block of compressed data from a data file on an external storage device,
If a media error occurs during the reading process, a value indicating how much compressed data has been normally read in the input buffer is written to the error location pointer, and the media error is stored, and the block read If there is no media error in the process, decompress all the compressed data contained in the block, and if there is a media error, move to the block until the actual pointer reaches the abnormal location pointer. A method for decompressing data, comprising decompressing included compressed data.