JPH04104323A - Bit length unit conversion system - Google Patents

Abstract

PURPOSE:To fetch n-bit unit data at high speed from data composed of an m-bit unit by integrating a main program described by means of a high-level language with a processing routine as a sub-routine. CONSTITUTION:When a system enters a step converting a nine-bit length unit into an eight-bit length unit, th main program fetches data to be converted in the block unit from a data file 11. When data is fetched by one block, the sub-routine 13 is called data in a nine-bit unit into data in the eight-bit unit and is outputted to the output area P4 of the program 12. Then, a method removing the one head bit of nine bits and converting it into data of eight bits is shown. In the case, one bit is shifted to left from the state of a different diagram (d) and 24 bits are shifted to right. This state is shown in diagrams (g) and (h). Then, a method removing the three head bits of nine bits and converting them to data of six bits is shown in diagrams (i) and (j).

Description

[Detailed Description of the Invention] [Summary] Regarding a bit length unit conversion method that converts data handled in different bit length units into data in a desired bit length unit, the bit length unit conversion is summarized in a subroutine format using a high-level language. In the bit length unit conversion method, which converts data composed of m bits into data composed of n bits, the purpose is to perform conversion in desired bit length units at high speed using a program that is
The number of bits from a predetermined reference position to the bit immediately before the first bit of arbitrary m-bit unit data is divided by n to obtain an integer quotient W1 and an integer remainder W2, and from the reference position W 1
* Take out bits consisting of integral multiples of n with n + 1 bit as the leading bit, and shift the bits in this register to the left by W2 bits,
After further shifting to the left by 2 (0≦ffi<m) bits, km
+ shift to the right by n bits, - to the left of this register
The configuration is such that n-bit unit data is generated by entering a predetermined value into a bit string up to m+42 bits.

[Industrial application field]

The present invention relates to a bit length unit conversion method for converting data handled in different bit length units into data in a desired bit length unit.

The bit length units currently handled by computers are often 8 bits or an integer multiple of 8 bits. However, computers that use data in 9-bit length units or 6-bit length units are also used. Data with different bit lengths like this are divided into 8
When used in a computer that handles data in bit length units based on focus, the bit length units are converted into 8-bit units.

[Conventional technology]

An example of such data conversion will be explained using FIG. 6.

Figure 6 shows the ASCII code expressed in 9-bit units.
EBCDIC code (BCD code extended to 8 bits) expressed in bits.
In Japanese, katakana is written in katakana, which makes it difficult to create even with an assembler, and it is also difficult to maintain when making modifications. Furthermore, when writing in the high-level language C0BOL, there are problems with byte boundary adjustment and slow processing performance of the C0BOI- bit processing library.

The present invention has been made in view of the above problems, and
The purpose of the present invention is to provide a bit length unit conversion method that compiles the bit operation part for converting bit length units in a subroutine format and enables easy and high-speed conversion to a desired bit length unit using a high-level language such as C0BOL. .

[Means to solve the problem]

FIG. 1 is a flow diagram showing the principle of the present invention, and FIG. 2 is a diagram explaining the contents of FIG. 1. Step 1 in Figure 1
, Pl is the address P as shown in FIG. 2(a).
3 represents the number of bits from the predetermined reference position to the previous bit in units of m bits to be converted, and Wl is P
This shows a specific example of converting l into an integer quotient divided by n bits, which is the bit unit to be converted, W2 is an integer remainder (including 0), and W3 is expressed in n bit units (ζ can be processed without a foot code). (a) shows an example of conversion of characters representing alphabets, numbers, and kana symbols. In this case, the MS of 9-bit data
Since the D (most significant digit) bit is 0, this 0 is removed and converted to 8-bit data.

(b) represents a kanji in 18 bits, and the MSD bit and the 9th bit are 0, so 8 bits excluding these two x 2
An example of converting to 2-byte data consisting of: (C)
The MSD bit represents the code S, and the remaining 8 bits represent fixed-point data that represents a number.
An example is shown in which the code is converted to Hyde and the code S is written in each bit of the second Hyde.

Conventionally, there are two types of data converters (data conversion programs) that perform such conversion: one for general files written in an assembler language, and one for lists processed in a high-level language (for example, COBOL).

[Problem to be solved by the invention]

By the way, in the case of a program written in an assembler language, in order to handle a bit string with a byte machine, it is special, so it indicates the address closest to the unit of m bits to be converted. Address W
From step 3, import the data for bits, which is the length of the register of the computer to be used, into this register (step 2).
Then, the bits in this register are shifted to the left by W2 bits, and the first bit of the m-bit unit to be converted is moved to the leftmost bit position of this register (step 3).

This situation is shown in FIG. 2(b). More I! , (0≦l
<m) Shift bits to the left. This l is an m-bit MS
From the D (most significant digit) bit! It represents the process of discarding up to bits, and represents the process explained in FIGS. 6(a) and (b). This situation is shown in FIG. 2(C). The case of ρ-〇 represents the case where all m bits are converted into n bit units. Next -m+jl! When the bits are shifted to the left, the LSD (least significant digit) bits of m bits to be converted match the right end of the register (step 4).

This situation is shown in FIG. 2(d). In (d), -m+j! from the left end! The bits up to are binary zeros, but signs may be written as shown in FIG. 6(C). The contents of this register are output in units of n bits (step 5).

[For production]

By configuring as above, it is possible to convert data composed of m bits into data composed of n bits, and this processing routine can be incorporated as a subroutine into the main program written in a high-level language. This makes it possible to extract data in units of n bits at high speed from data configured in units of m bits.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings. Third
The figure shows a subroutine according to an embodiment of the present invention and a main program that uses the subroutine to perform data conversion and the like. In the figure, 11 is a data file storing data in 9-pinto units, 12 is a main program that performs data conversion, etc., and 13 is a subroutine of this embodiment that converts 9-pinto unit data into 8-pinto units. . 14 is a parameter for receiving and passing information.

FIG. 4 shows a diagram of the bit length conversion carried out in this embodiment. As shown in FIG. 5A (b), Pl is the number of focuses from the beginning of the block to the first bit of the nine bits to be converted, and Wl
The number of hides from the beginning of the flock closest to the first focus of the cigarette, W2 is the number of bits representing the difference between this Pl and W1*8, and W3 is the address closest to the beginning of the 9 bits to be converted expressed in bytes. show. Next, 4 bytes of data from the position indicated by address W3 are taken into the computer's general-purpose register (step 42). The reason why it is set to 4 bytes is because this general-purpose register has a size of 4 bytes. This state is shown in FIG. 5A (C). In this case, it is the W2-1 bit. Next, the bits in the register are shifted to the left by W2 bits (step 43). This situation is shown in FIG. 5B (d). This shift brings the first focus of the 9 bits to be converted to the left end of the register. Next step 44
to determine the processing mode. In this case, there are three processing modes as described above. The first processing mode is step 4
4 shows a case where 9-pin data is converted to 8-bit data. In this case, it is a 23-pin to the right diagram, and FIG. 5 is an explanatory diagram of the conversion process of the general-purpose register in the computer at the time of conversion.

In this embodiment, data 1010 in units of 9 pips
When converting 000 (11) directly to 8-bit unit data, if you remove MSD bit 1 of this data and convert it to 8-pin unit data, remove 3 bits including the MSD bit of this data and convert it to 6-bit data. Three cases of converting this data into 8-bit unit data will be described.

In FIG. 3, when the main program enters the step of converting a 9-bit length into an 8-bit length, it first extracts the data to be converted from the data file 11 in blocks. In this case, l block consists of 3 records. Each record has an R CW (
Record Control Word) is provided. When one block of this data is taken in, the subroutine 13 of this embodiment is called, and the data in units of 9 bits is converted into data in units of 8 bits, and the data is output to the output area P4 of the main program 12. Next, in FIG. 4, step 41
The meaning of each symbol is to move 9-bit data to the right end of the shift register by shifting (step 45),
The third byte data 00000001 and the fourth byte data 01000001 are obtained (step 45). This situation is shown in FIG. 5B (e). Extract this data to the output area 14 as 2-byte data (step 48)
. This situation is shown in Figure 8).

Next, step 46 shows a method of removing the first bit of the 9 bits and converting them into 8-bit data.

In this case, from the state shown in FIG. 5B(d), a 1-bit shift is made to the left and then a 24-bit shift is made to the right (step 46).

This state is shown in Fig. 5C (g5.ω). Step 47 shows a method of removing the first 3 bits of 9-bit focus and converting it to 6-bit data, and this process is shown in Figure 5C (
Shown in i) and (j).

As is clear from the above explanation, it is possible to directly convert 9-bit unit data into 2 bytes of 8-bit data, or convert the first 1
The bits can be discarded and converted to 8-bit data to make 8-bit unit data of 1 byte, or the first 3 bits can be discarded to make 6-bit data and 8-bit unit data can be expressed as 1 byte of 8-bit data. Although the present embodiment has described converting 9-bit long grass position data to 8-bit long grass position data, m-bit length unit data can be similarly easily converted to n-bit length unit data. This allows bit string operations that are difficult to handle on byte-based machines to be performed all at once in a subroutine format, making it possible to easily reference data in units of arbitrary bit length using a high-level language such as COBOL. As computer systems become more open in the future, it is expected that there will be an increase in the number of passing processes needed to exchange data between different types of devices.In this case, high-speed bit string references can be realized using C0BOL, which is easy to maintain, so data conversion can be performed efficiently. etc. can be developed.

〔Effect of the invention〕

According to the present invention, data in units of m-bit length can be converted to data in units of n-bit length, so by converting this method into a subroutine and calling it from the main program, C0BOL
It can be easily converted into data in units of desired bit length at high speed using a high-level language such as .

[Brief explanation of drawings]

FIG. 1 is a flow diagram showing the principle of the present invention, FIG. 2 is a diagram explaining the contents of FIG. 1, and FIG. 3 is a diagram showing the relationship with the main program when this embodiment is a subroutine program. FIG. 4 is a flowchart showing the operation of the embodiment; FIGS. 5A, 5B, and 5C are diagrams explaining register shift operations during implementation of this embodiment; FIG. 6 is a 9-bit machine data FIG. 3 is a diagram showing a specific example of converting the data into 8-bit machine data. (a> Figure 1 showing the principle of the present invention Figure 2 explaining the contents of Figure 1

Claims (1)

[Claims] 1. In the bit length unit conversion method that converts data composed of m-bit units to data composed of n-bit units, the bits from a predetermined reference position to the bit immediately before the first bit of arbitrary m-bit unit data Divide the number by n to obtain an integer quotient W1 and an integer remainder W2, and from the reference position W
Take out k bits consisting of an integer multiple of n with the 1*n+1 bit as the first bit into a register consisting of k bits, shift the bits in this register to the left by W2 bits, and further shift l (0≦l<m) bits. After shifting to the left, km+l
Shift bits to the right, k−m+ from the left side of this register
A bit length unit conversion method characterized in that n-bit unit data is generated by entering a predetermined value into a bit string up to l bits.