JPH01290036A - Compression system for character string of source member - Google Patents

Abstract

PURPOSE:To compress each source member with the highest efficiency by selecting automatically the character strings proper to the source members and compressing these character strings. CONSTITUTION:The title system consists of a character string counting means 1, a compressed character string table production means 2, and a character string compression/expansion means 3. The source members are previously inputted and the merging frequency of each character string in a source program is counted for production of a character string count table. Then the character strings which are effectively compressed are selected out of said count table for production of a compressed character string table. The character strings of a text record are converted into the corresponding character string compression codes. Then the character string compression codes of a compressed text record are converted into the corresponding character strings. As a result, the data storing efficiency is improved.

Description

[Detailed description of the invention] [Industrial field of application] In computer systems, creation and creation of source members
This article relates to a string compression method for source members in registration.

[Conventional technology]

Conventionally, when compressing source members, character strings that can be compressed are limited to specific character strings such as spaces and COBOL reserved words.

[Problem to be solved by the invention]

In conventional source member compression, only a very limited number of specific strings are compressed, so other strings, even if they appear repeatedly in the source member, remain as long as they are. It was stored in a library or on magnetic tape. As a result, the source member is not compressed efficiently, resulting in poor data storage efficiency.

[Means to solve the problem]

According to the present invention, source members are input in advance.

A string counting means that counts the number of occurrences of each string that appears in a source program and creates a 9-character string count table, and a string count table created by this string counting means is effectively compressed. A compressed string table creation means that selects a character string and creates a compressed string table, and matches the character string of the text record by searching the compressed string table created by this compressed string table creation means. A string compression method for a source member is characterized in that it has a string compression/expansion means for converting a string compression code of a compressed text record into a corresponding string. can get.

〔Example〕

Next, two aspects of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a diagram showing the configuration of the present invention, which consists of a nine-character string counting means 1, a compressed character string table creation means 2, and a nine-character string compression/expansion means 3.

Next, 2 character string counting means 1. The compressed character string table creation means 27 and the character string compression/expansion means 3 will be explained with reference to FIGS. 2, 3, 4, and 5, respectively.

In FIG. 2, in step A1, a code is input from the source member to be compressed. In step A2, it is determined whether or not this is the end of the input source member record. At the end, in step A3.

Set the string length as the first key and the string as the second key.

Sort the string count table. After sorting.

Finish the process.

The format of the string count table is as shown in FIG. 6, with a length of 9 strings and a string column E2. It consists of the appearance count column E3.

Next, when 9 character strings are compressed to n bytes in step A4, character strings of n+1 characters or more are extracted from the record. In step A5, it is determined whether the record contains a character string of n+1 characters or more.

If there is no character string, move to step A1. If there is a character string, the character string extracted in step A6 is.

Determine whether it has already been registered in the character string count table.

If it has not been registered yet, in step A7, the length of 9 character strings and the number of occurrences (initial value=1) are set in the 9 character string count table, and the process moves to Stella 7'A4.

If it has already been registered, in step A8, the value in the number of appearance column of the corresponding character string in the 9-character string count table is incremented by 1, and the process moves to step A4. The process of inputting the record of the source member to be compressed and counting the number of occurrences of every nine character strings is the nine character string counting means 1.

In FIG. 3, in step B1, it is determined whether compression or expansion is being performed. At the time of compression, in step B2.

One line of information created by the string counting means (
Enter the character string length (9 characters, number of occurrences).

When the two-character string count table reaches the end in step 'B3, the process ends. In step B4, the string is compressed based on the string length and number of occurrences of the nine strings.

Determine whether the effective compression standard value is exceeded.

The standard value for effective compression is the number of compressed bytes of 5 character strings ((character string length - number of compressed bytes n) x number of occurrences), and the compression additional information (character string length + 1°, details explained in step B6). Set a certain standard value by subtracting the number of bytes.

For example, assume that the number of compressed bytes n is 3 and the reference value is 5.

In the case of "CALL" in the first line of Figure 6, (4-3
)X10-(4+1)-5, which does not exceed the standard value of 5. Similarly, in the case of "'PROC'" on the second line, (4-3
)X15-(4+1)=10, which exceeds the oak standard value of 5.

If the reference value 5 is not exceeded in step B4, step B
Move on to 2.

9 character string length and character string respectively in step B5.

As shown in FIG. 7, the compressed character string length field Fl and the compressed character string field F2 of the compressed character string table are set in this order. In the compression code column F3, compression codes corresponding to nine character strings are uniquely set in ascending order (for example, values starting from 2 bytes 0000 in hexadecimal are set). After setting, proceed to step B2.

Next, in step 'Bl, at the time of expansion processing, step B6
Input the compressed string group record from the compressed source member. As shown in Figure 8, the format of the compressed string group record consists of 9 character string length fields Gl and 9 character string fields G2 for one compressed character string, and 9 character string length fields G1 and 9 character string fields G2 for each compressed character string. Created. This record is used as compression additional information when compression is applied to the source member.

It is created by sequentially extracting characters from the compressed character string length column and the compressed character string column of the compressed character string table.

What is the length of compressed additional information for one compressed character string?

The character string length is +1. Next, in step B7, it is determined whether it is the end of the compressed character string group record. Processing ends at the end of the record. In step B8, one character string length is extracted one by one from the beginning of the record. Stella 7°B9
Determine whether this is the end of the compressed string. At the end of the compressed character string, the process moves to step B6.

In step BIO, set the 9 string length and the 7th string respectively.
As shown in the figure, the compressed character string length field Fl and the compressed character string field F2 of the compressed character string table are set in this order. Regarding the compressed code column F3, compressed codes corresponding to nine character strings are uniquely set in ascending order, similar to the case of STETSUNO B5. After set.

Proceed to step B8.

When compressing, the process of selecting strings that exceed the effective compression standard value from the string count table created by this string counting means and creating a compressed string table. The process of creating a compressed character string table from compressed character string group records is the compressed character string table creation means 2.

In FIG. 4, in step C1, the first compressed character string length is extracted from the compressed character string tape created by the compressed character string table creation means. In step C2, the length of the character string extracted from the text record is compared with the length of the compressed character string. If the string length is smaller than the compressed string length, the process ends. If the character string length is larger than the compressed character string length, the next compressed character string length from the compressed character string table is retrieved in step C3. Next, in step C4, it is determined whether the compressed character string table has been searched to the end. If the search has reached the end, the process ends. If you haven't searched to the end,
Move to step C2. If the two character string lengths and the compressed character string length are equal in step C2, the character string extracted from the text record and the compressed character string are compared in step C5. If the string is larger than the compressed string, the process ends. If the character string is smaller than the compressed character string, the process moves to step C3. If the character string and compressed character string match, in step C6, a compressed code corresponding to the compressed character string is extracted from the compressed character string table.

In step C7, one character string is replaced with a compressed identifier H1 and a compressed code H2 as shown in FIG. The compression identifier H1 corresponds to each character code system (for example, JIS, A
SCII) uses an unused pit string as a 9-character code. The compression process of the two-character string compression/expansion means 3 is a process of converting the character string extracted from this text record into a corresponding compressed code by searching the compressed character string table.

In FIG. 5, in step D1, the first compressed code is extracted from the nine compressed character string tables created by the compressed character string table creation means. Step D2
Then, the compression code extracted from the compressed record using the compression identifier as a key is compared with the compression code in the compressed character string table.

The compression code extracted from the compressed record.

If the compression code in the compression string table is not equal to 1,
From the compressed string table in step D3.

The next compressed code is extracted and the process moves to step D2.

The compression code extracted from the compressed record.

When the compression code in the compression string table is equal to 5.

In step D4, the compressed character string corresponding to the compressed code is extracted from the compressed character string table. In step D5, the compression identifier and compression code are replaced with the extracted compressed character string. The process of converting the compressed code extracted from this compressed record into a corresponding compressed character string by searching the compressed character string table is the expansion process of the nine character string compression/expansion means 3.

Next, the overall operation of this embodiment will be specifically explained. In the compressed string table, as shown in Figure 7.

As a compressed string, "PROC", "5TO"
P”.

"PAR is positive TER", "' C0NVERT-8
Assume that I(ORI"... is registered.

At this time, in Fig. 10, when compressing text record II, "C0" is registered as a compressed character string.
NVERT-8HORI”,”PARAMET
ER″′ is compressed and a compressed record E2 is created. Conversely, when expanding compressed record F2, by searching for compression identifiers I21 and I23, each compression code I 22, I24 is extracted, extended processing is performed, and text record E1 is created.

〔Effect of the invention〕

As explained above, the present invention has the following effects by automatically selecting and compressing character strings unique to the source member in the compression process of the source member.

Since character strings specific to each source member can be automatically selected and compressed most effectively, each source member can be compressed most efficiently.

This also allows records to be condensed and stored on the source library and magnetic tape, making it possible to store them in a condensed manner.
It is not only efficient in terms of access, but also efficient in access.
Effective.

[Brief explanation of the drawing]

Fig. 1 is an overall configuration diagram of the present invention, Fig. 2 is a flowchart of the string counting means, Fig. 3 is a flowchart of the compressed string table creation means, and Fig. 4 is a flowchart of the compression processing of the string compression/expansion means. , Fig. 5 is a flowchart of the expansion processing of the string compression/expansion means, Fig. 6 is the format of the string count table, Fig. 7 is the format of the compressed string table, and Fig. 8 is the format of the compressed string group record. , FIG. 9 is a diagram showing a compression format, and FIG. 10 is a diagram showing an example of compression of a text record. 1... Character string counting means, 2... Compressed character string table creation means, 3... Character string compression/expansion means. Figure 6

Claims (1)

[Claims] 1) When storing a source member containing a program or data in a file or library, input the source member in advance, count the number of occurrences of each character string that appears in the source program, A string counting means for creating a string counting table; and a compressed string for selecting strings to be effectively compressed from the string counting table created by this string counting means and creating a compressed string table. By searching the table creation means and the compressed string table created by this compressed string table creation means,
A source member characterized in that it has a string compression/expansion means that converts a string of a text record into a corresponding string compression code, and converts the string compression code of the compressed text record into a corresponding string. String compression method.