Number: Difference between revisions

From Bohemia Interactive Community
Jump to navigation Jump to search
(Added proper number match regex)
(Remove Stub)
(5 intermediate revisions by 2 users not shown)
Line 1: Line 1:
=Number Type=
{{SideTOC}}
A real number, i.e. 1, -25, 6.52, 1805.6352
A number is, depending on scope, either a [http://en.wikipedia.org/wiki/Single-precision_floating-point_format single precision floating-point number] (when talking about scripting) or a range of numerical types when talking about config context.


Numbers are stored as [http://en.wikipedia.org/wiki/Single-precision_floating-point_format single precision floating point numbers], which allow for a large range of values with limited precision (generally accurate to 6 or 7 decimal places). The number [[Data Types|data type]] is also sometimes referred to as "scalar" (for example, returned by the [[supportInfo]] script command).


The largest real positive number that can be entered via script is:  3.4028235e38
== Scripting ==


The largest real negative number that can be entered via script is: -3.4028235e38
In SQF, there are multiple accepted number formats.
However, all of them will result in the same  {{Inline code|SCALAR}} ([[typeName]]) value type.


Larger numbers: In scripts it is possible to generate a representation of an infinite positive or negative number which compares even larger or smaller than the above two floating point limits;
The '''<span style="color: #008000;">largest positive</span>''' number that can be archived is {{Inline code|3.4028235e38}} and the '''<span style="color: #800000;">largest negative</span>''' is {{Inline code|-3.4028235e38}}.<br>
It also is possible to generate positive or negative infinite values using either {{Inline code|1e39}} (string representation: {{Inline code|1.#INF}}) or {{Inline code|-1e39}} (string representation: {{Inline code|-1.#INF}}).


Positive infinity  1e39  = "1.#INF"
To check, if a number is finite, one can use the [[finite]] operator.


Negative infinity -1e39  = "-1.#INF"
{{Informative | Regex to match all numbers in [[SQF_syntax|SQF]] is {{Inline code|<nowiki>(((\$|0x)[0-9a-fA-F]+)|(\.[0-9]+))|(\b[0-9]+(\.[0-9]+|[eE][-+]?[0-9]+)?)\b</nowiki>}}}}
{{Important | Due to technical limitations, the precision of floating-point numbers is limited. Please refer to [https://en.wikipedia.org/wiki/IEEE_754 IEEE 754] for more info about the technical details.}}


Indeterminate (NaN) = "-1.#IND"
=== Decimal (Base 10) ===


A decimal number is your normal {{Inline code|0.5}} syntax stuff with one extra: You may omit the initial pack of digits.


The [[finite]] command can be used to verify if a number represents infinity or NaN.
Some Examples would be:
* {{Inline code|5.197}}
* {{Inline code| 0.47}}
* {{Inline code| 16.0}}
* {{Inline code|.8314}}
* {{Inline code|12345}}


===Technical Specification===
A regex catching these kind of numbers could look like this {{Inline code|<nowiki>((\.[0-9]+)|(\b[0-9]+(\.[0-9]+)?))\b</nowiki>}}
A number is a sequence of characters starting either with a digit, an '''$''' or a '''.'''. If it starts with a digit it may be an integer (e.g. 3), a floating point number (e.g. 3.2) or a hexadecimal number (see below). If it starts with a '''$''' it is also a hexadecimal number and if it starts with a '''.''' it is a floating point number smaller than 0 (e.g. .1 = 0.1).
If the number is not hexadecimal it may contain exactly one period and ends before the first non-digit character (except that one period of course). The only exception from that is the scientific notation.


The scientific notation in SQF has the syntax <code>([0-9]+.)?[0-9]+[eE][+-]?[0-9]+</code>
==== Scientific Notation ====
{{Informative|Regex to match all numbers is <code><nowiki>((\$[0-9a-fA-F]+)|(\.[0-9]+))|(\b[0-9]+(\.[0-9]+|[eE][-+]?[0-9]+)?\b)</nowiki></code>}}
The [https://en.wikipedia.org/wiki/Scientific_notation Scientific Notation] is a way of expressing numbers that are too big or too small to be conveniently written in decimal form.


That means it consists of a pre-factor that can be any number (integer or floating point). Then a '''e''' or '''E''' followed by an optional sign ('''+''' or '''-''') and at the end an <b>integer</b> (a whole number). This will then be interpreted in the following way
It starts of like a normal decimal and then gets expressed by an {{Inline code|E}} (Not case-sensitive, thus {{Inline code|e}} is also valid) followed by an '''optional''' {{Inline code|+}} or {{Inline code|-}} sign and ends with a range of digits.


<code>aEb = a * 10^b</code>
Some examples would be:
<b>Examples:</b>
* {{Inline code|1.23E4 ⇔ 1.23 ⋅ 10<sup>4 </sup> ⇔ 12300}}
<code>1e2 = 1E2 = 100<br>5e-2 = 5E-2 = 0.05<br>1.234e4 = 1.234E4 = 12340</code>
* {{Inline code|5e-2   ⇔    5 ⋅ 10<sup>-2</sup> ⇔  0.05}}


If omitted the exponent '''b''' will be considered to be zero and therefore
A regex catching these kind of numbers could look like this  {{Inline code|([0-9]+.)?[0-9]+[eE][+-]?[0-9]+}}
<pre>aE = a</pre>
Although the result is being calculated the engine will issue an error on that becasue the exponent is not optional!
If the pre-factor '''a''' is omitted it is no longer a number as it doesn't start with a digit or a period. If one tries to use a hexadecimal number as the pre-factor the '''e''' will simply be interpreted as part of that hexadecimal number and therefore it is no longer a scientific notation.


Essentially numbers in SQF are in the format the '''atof''' function in C(++) expects numbers to be (See http://www.cplusplus.com/reference/cstdlib/atof/) with the exception that hexadecimals can also start with a '''$''' as well.
=== Hexadecimal (Base 16) ===


==Hex numbers==
In [[SQF_syntax|SQF]], [https://en.wikipedia.org/wiki/Hexadecimal hexadecimal] (also base 16, or hex) is a positional numeral system with a base of 16.
They start either with {{Inline code|0x}} or with a single {{Inline code|$}}.


Arma supports hexadecimal numbers in both scripts and configs. Most common notation for hex numbers is '''0x''' followed by the hexadecimal value. Alternatively symbol '''$''' can be used for the same thing. It is also possible to mix and match notations:
This gets followed by one of the following characters: {{Inline code|0 1 2 3 4 5 6 7 8 9 A B C D E F}}.<br>
Note that casing does not matter, thus both {{Inline code|0xa}} and {{Inline code|0xA}} are valid.


<code>[[hint]] [[str]] 0xFF; //255
Some Examples would be:
[[hint]] [[str]] $FF; //255
* {{Inline code|0xa5}}
[[hint]] [[str]] (0xFF + $FF + 255); //765</code>
* {{Inline code|$5C}}
Hexadecimal numbers are case-insensitive:
* {{Inline code|$FFFFFF}}
<pre>0xFF = 0xff</pre>
* {{Inline code|0x123ABC}}


==Degrees==
A regex catching these kind of numbers could look like this {{Inline code|<nowiki>((\$|0x)[0-9a-fA-F]+)\b</nowiki>}}


Degrees are a poetic label used to indicate a number returned from functions like [[acos]] and [[asin]]


It's special properties are that it will always supply a value from 0 to 360
== Config ==


==Radians==
Unlike scripting, configs actually do allow for multiple number types ([[Integer]], [[Float]], etc). They are properly stored with their corresponding type in mind.
{{Informative | In a config, a [[Number]] can also mean a [[Boolean]]: in [[Description.ext]] for example, {{Inline code|disabledAI {{=}} 1;}} means [[true]].}}


Another poetic label for a Number. Used in angular math computations with commands like [[rad]] and [[deg]]
<syntaxhighlight lang="cpp">
class MyClass
{
myInt = 1; // integer;
myFloat = 0.01; // float;
};


==Scripting vs Addons==
// description.ext example
===Integers and Floats===
disabledAI = 1; // boolean: 0 = false, anything else = true (usually 1)
Note that unlike config.cpp's (addons), a ''Number'' in scripting language is '''ANY''' numeric entity. Floats, or integers. It is NOT the same as a config's [[Integer]] or [[Float]]. Number covers both types.
</syntaxhighlight>


===Booleans===


Note also, unlike config.cpp's, [[Boolean]] is a real type in scripting language. In addons, it is a poetic licence for a zero/non zero [[Integer]]. In Arma 3 command [[parseNumber]] has been extended to accept booleans:
<code>[[hint]] [[str]] [[parseNumber]] [[true]]; //1</code>
[[Category: Data Types]]
[[Category: Data Types]]

Revision as of 19:46, 11 March 2020

Template:SideTOC A number is, depending on scope, either a single precision floating-point number (when talking about scripting) or a range of numerical types when talking about config context.


Scripting

In SQF, there are multiple accepted number formats. However, all of them will result in the same SCALAR (typeName) value type.

The largest positive number that can be archived is 3.4028235e38 and the largest negative is -3.4028235e38.
It also is possible to generate positive or negative infinite values using either 1e39 (string representation: 1.#INF) or -1e39 (string representation: -1.#INF).

To check, if a number is finite, one can use the finite operator.

Regex to match all numbers in SQF is (((\$|0x)[0-9a-fA-F]+)|(\.[0-9]+))|(\b[0-9]+(\.[0-9]+|[eE][-+]?[0-9]+)?)\b
Due to technical limitations, the precision of floating-point numbers is limited. Please refer to IEEE 754 for more info about the technical details.

Decimal (Base 10)

A decimal number is your normal 0.5 syntax stuff with one extra: You may omit the initial pack of digits.

Some Examples would be:

  • 5.197
  • 0.47
  • 16.0
  • .8314
  • 12345

A regex catching these kind of numbers could look like this ((\.[0-9]+)|(\b[0-9]+(\.[0-9]+)?))\b

Scientific Notation

The Scientific Notation is a way of expressing numbers that are too big or too small to be conveniently written in decimal form.

It starts of like a normal decimal and then gets expressed by an E (Not case-sensitive, thus e is also valid) followed by an optional + or - sign and ends with a range of digits.

Some examples would be:

  • 1.23E4 ⇔ 1.23 ⋅ 104 ⇔ 12300
  • 5e-2 ⇔ 5 ⋅ 10-2 ⇔ 0.05

A regex catching these kind of numbers could look like this ([0-9]+.)?[0-9]+[eE][+-]?[0-9]+

Hexadecimal (Base 16)

In SQF, hexadecimal (also base 16, or hex) is a positional numeral system with a base of 16. They start either with 0x or with a single $.

This gets followed by one of the following characters: 0 1 2 3 4 5 6 7 8 9 A B C D E F.
Note that casing does not matter, thus both 0xa and 0xA are valid.

Some Examples would be:

  • 0xa5
  • $5C
  • $FFFFFF
  • 0x123ABC

A regex catching these kind of numbers could look like this ((\$|0x)[0-9a-fA-F]+)\b


Config

Unlike scripting, configs actually do allow for multiple number types (Integer, Float, etc). They are properly stored with their corresponding type in mind.

In a config, a Number can also mean a Boolean: in Description.ext for example, disabledAI = 1; means true.
class MyClass
{
	myInt = 1;		// integer;
	myFloat = 0.01;	// float;
};

// description.ext example
disabledAI = 1;		// boolean: 0 = false, anything else = true (usually 1)