RU2615336C1 - Method for encoding and computing date using simplified format in digital devices - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: method comprises the stages, at which the K bit long integer is allocated for storing the date data, including year, month and day; a fixed date is set as the first of January of the certain year Y0; the values of the current year Y, month M, day D are introduced; the integer N is calculated for storing the date data by the formula N=D+M*31+(Y-Y0)*372-32; N number is stored in the binary format; the date is calculated, if required, by the formula D=((N mod 372) mod 31)+1, M=((N mod 372) div 31)+1, Y=Y0+(N div 372), where div - the integer division operation (division with discarding the fractional part); mod - the remainder taking operation of the integer division.

EFFECT: increasing the possible range of the date values.

Description

FIELD OF THE INVENTION

The present invention relates to computer technology and, in particular, to methods for encoding and converting dates for storage and use in digital devices: computers, terminals, etc.

State of the art

In modern digital devices, various internal data formats are used to store and use the date and time. To convert from an internal format to an external one and vice versa, a number of calculations and encoding / decoding operations are required.

Thus, a known method of encoding and converting dates in software (software) SQL Server 2005 and SQL Server 2012 from Microsoft [1, 2]. To store the date and time, two 4-byte integers are allocated (a total of 8 bytes are used). The first 4 bytes are used to store the actual date and contain the number of days before or after the reference date, which was selected on January 1, 1900 (for the datetime data type; the second 4 bytes contain the current time value, presented in the form of three hundredths of a second after midnight ) Storage of date and time is also used with less accuracy. In this option, dates are stored as a 2-byte integer representing the number of days elapsed from January 1, 1900 (for the data type smalldatetime; the next two bytes store the number of minutes elapsed after midnight).

Thus, the date can be stored as an integer and occupy at least 2 bytes (16 bits).

To directly convert the date according to the Gregorian calendar to the contents of a memory cell, the following algorithm can be used [3]:

a = (M-14) div 12,

y = Y + 4800-a,

m = M + 12 × a-3,

J = D + (153 × m + 2) div 5 + 365 × y + div div 4 div 100 + div 400-32075,

where a, y, m are intermediate variables of integer type;

M - month of the year, from 1 (January) to 12 (December);

D - day of the month, from 1 to 31;

Y is the year of the Gregorian calendar.

After that, to obtain N from the obtained value J (Julian day number for the starting date), the constant J0 is subtracted, which depends on the starting date and is the Julian day number for the starting date. However, for the inverse transformation of the integer N to the date, an even greater number of conversions and calculations are used, first get the Julian day number J, adding the constant J0, then calculate

j = J + 32044,

g = j div 146097,

dg = j mod 146097,

c = (dg div 36524 + 1) × 3 div 4,

dc = dg-s × 36524,

b = dc div1461,

db = dc mod 1461,

a = (db div 365 + 1) × 3 div 4,

da = db - a × 365,

y = g × 400 + s × 100 + b × 4 + a,

m = (da × 5 + 308) div 153-2,

d = da- (m + 4) × 153 div 5 + 122,

Y = y-4800 + (m + 2) div 12,

M = (m + 2) mod 12 + 1,

D = d + 1,

where b, c, d, j, g, da, db, dc, dg are intermediate variables of integer type.

The disadvantages of this method are the need to carry out a number of complex calculations and a large occupied volume of stored data.

In a number of digital devices, the need arises to use only the date, including the year, month and day, which allows to reduce the amount of stored data and simplify calculations.

For such devices, a known date encoding method [4] can be used, namely, that

- allocate for storing data about the date, including the year, month and day, an integer of length K bits;

- set a fixed date for the countdown as the first of January of a certain year Y0;

- enter the values of the current year Y, month M, day D;

- calculate an integer N for storing date data according to the formula

N = D + (M-1) * 32 + (Y-Y0) * 384;

- store the number N in binary format;

- calculate, if necessary, the date by the formulas

D = N mod 32,

M = (N div 32) div 12 + 1,

Y = Y0 + (N div 384),

where div is the operation of integer division (division with discarding the fractional part);

mod - operation to take the remainder of integer division.

The described method is adopted as a prototype.

The disadvantage of this method is the limitation of the range of stored dates. Despite the fact that the method wins in the simplicity of calculation and compactness of the data, some of the possible values are not used.

For comparison, consider a possible range of stored dates. As the initial date Y0, we take 01.01.1900 (January 1, 1900). The date size is 2 bytes (16 bits).

If you use the method of Microsoft, when the date is calculated in days from the original date, we get the end date 06/06/2079. If we consider only the completely incoming years, then 12/31/2078. Thus, this method provides the maximum possible range of 179 years.

If you use the method adopted for the prototype, then we get the maximum date 08/30/2070. If we consider only the completely incoming years, then we get 12/31/2069. As a result, a range of 170 years is ensured. This method contains fewer calculations, but some of the values have to be sacrificed, so the range becomes shorter by 9 years.

Disclosure of invention

The technical result is to increase the range of possible date values.

For this, a method is proposed, which consists in the fact that

- allocate for storing data about the date, including the year, month and day, an integer of length K bits;

- set a fixed date for the countdown as the first of January of a certain year Y0;

- enter the values of the current year Y, month M, day D;

- calculate an integer N for storing date data according to the formula

N = D + M * 31 + (Y-Y0) * 372-32;

- store the number N in binary format;

- calculate, if necessary, the date by the formulas

D = ((N mod 372) mod 31) +1,

M = ((N mod 372) div 31) +1,

Y = Y0 + (N div 372),

where div is the operation of integer division (division with discarding the fractional part);

mod - operation to take the remainder of integer division.

In the proposed method, an integer N is calculated for storing date data according to the formula

N = D-1 + (M-1) * 31 + (Y-Y0) * 372.

N is the number of days from the original date.

When calculating N using the above formula, the maximum possible value of 31 is taken as the number of days in any month. Thus, each date within the year will have its own constant serial number from January 1 (January 1) to 31 * 12 = 372 (December 31), regardless whether the year was a leap year or not. The offset between the same day of the current and next month will be equal to 31, and between the date of the current year and the same date of the next year will be equal to 31 * 12 = 372. Extra days on February 30, February 31, or February 29 for a non-leap year are not used.

Reducing the entry for N, we finally get

N = D + M * 31 + (Y-Y0) * 372-32.

The number of bytes for storing the number N remains the same as in the prototype (2 bytes = 16 bits). Accordingly, the maximum N contains 2 ¹⁶ = 65536 values in the range from 0 to 65535.

If Y0 = 01/01/1900, then we obtain

- when N = 0, the date is 01.01.1900,

- at N = 1 - 02/01/1900,

- at N = 31 - 02/01/1900,

- at N = 372 - 01/01/1901, etc.

Using this method for the initial date Y0 = 01/01/1900, we obtain the maximum date 03/02/2076. If you take only the completely incoming years - 12/31/2075. As a result, a range of 176 years is provided, which is 6 years more than the known method adopted as a prototype, with a comparable computational speed.

For the considered option, when Y0 = 01/01/1900 is taken for the initial date, the maximum date by the method from Microsoft is 06.06.2079 (the range of 65536 days is covered, denoted by i), by the prototype method it is 30.08.2070 (62334 days, denoted by p), according to the proposed method - 03/02/2076 (64345 days, denote by v). Thus, the prototype method provides p / i = 62334 / 65536≈95.11%, and the proposed method v / i = 64345 / 65536≈98.18% of the maximum possible. Moreover, if the prototype method loses about 5% of the values of the maximum possible, then the proposed method is only about 2%.

Thus, the proposed method allows to increase the range of date values with comparable simplicity and speed of calculations, as well as compact data storage.

The implementation of the invention

Implementation of the proposed method can be implemented in a computer system capable of performing integer division operations, i.e. in almost any, including systems running, for example, running the MS Windows operating system.

To implement the method, it is sufficient to implement the implementation of the proposed operations in any programming language.

For example, in C #, an implementation might look like this:

Implement the actions of the proposed method as part of a program or function can be a specialist in programming (programmer). The application of the proposed method can have the greatest effect in systems that process date data and have strong restrictions on the amount of memory and hardware resources, as well as transmitting information about dates under conditions of restrictions on the amount of transmitted information.

Examples of such systems are systems that store and / or transmit various information, including dates (for example, the date of birth of a person, the date of departure of the goods, the date of departure of the train, etc.) in the form of a bar code. In this case, each bit saved is significant.

It is also effective to use the method in systems and databases in which there is a large amount of data with dates, in which case significant savings in storage volume can be achieved.

Information sources

1. Date and time (Transact-SQL), article at

https://msdn.microsoft.com/en-us/library/ms 187819 (% D 1% 83 = sq1.90) .aspx

2. Datetime (Transact-SQL), article at

https://msdn.microsoft.com/en-us/library/ms 187819 (% D 1% 83 = sq1. 110) .aspx

3. Julian day, article at

http://en.wikipedia.org/wiki/Julian_day

4. RF patent No. 2543961, priority dated January 16, 2013.

Claims (13)

The method of encoding and calculating the date using a simplified format in digital devices, namely, that allocate for storing data about the date, including the year, month and day, an integer of length K bits; set a fixed date for the countdown as the first of January of a certain year Y0; enter the values of the current year Y, month M, day D; calculate an integer N for storing date data according to the formula N = D + M * 31 + (Y-Y0) * 372-32; store the number N in binary format; if necessary, calculate the date by the formulas D = ((Nmod372) mod 31) +1, M = ((Nmod372) div 31) +1, Y = Y0 + (Ndiv372), where div is the operation of integer division (division with discarding the fractional part); mod - operation to take the remainder of integer division.