1. The datatype numbers
If you want to calculate with variables (be more specific: with its content) you must use a data type which is used for numbers. One lazy way is to write a number in string and calculate with these strings.
<?php
echo "5"+"9";
?>
This will generated 14 as expected. However php didn't calculated with
numbers but with strings. This doesn't work normally (or what do you expected for
"hello" + "xyz"?), but the automatically typecasting
take place here. The strings are got casted to real numbers first.
In php there exists two types, the integer and float
numbers.
2. Integer numbers
Integer numbers are any whole numbers, they don't have any decimal places. E.g
5, 31 and 199 are integer numbers, but also
-1, -10 and -44. In php you write numbers
as they are, the php interpreter reads the sequence of digits as a number.
<?php
$var = 5;
$var2 = 31;
$var3 = -10;
?>
Its also possible to write integer numbers in hexadecimal or octal notation, if you
know how to calculate them. For hexadecimal notation you use the prefix 0x,
for octal notation the prefix 0 (normal integer numbers starts
with the highest digit or are zero anyway).
<?php
$var = 0xFF; // hexadecimal (255), often use for protocols
$var2 = 0763; // octal (499), often used for chmod
$var3 = 0; // normal zero, it doesn't matter if hex-, oct- or decimal
?>
3. Float numbers
Float numbers are values with decimal places like 5.7,
12.25 or -4.07, but also 10.0, 3.1E-10
or 4.5e3.
<?php
$var = 5.7;
$var2 = 3.1E-10;
$var = .1;
?>
Float numbers can be easily recognized by the decimal point ., so you dont
use the comma (,) for the decimal places. As the dot .
is also used for string concatenation you must be very carefully if you concat
float numbers with strings. Sometimes you must use parentheses.
<?php
echo 'Percent value:'.77.3.'%'; // won't work
echo 'Percent value:'.(77.3).'%'; // works
echo 'Percent value:'. 77.3 .'%'; // works, these spaces "helps" php to parse this line correct
?>
4. Memory usage of integers and floats
There may be a question why to use integer and float numbers as float numbers
can do more. Both types use (basically) the same memory usage but
use a different usage of it. Integer values are saved very easily. Imagine you
save an integer value into a 64 bit field. One bit is reserved for the sign
(more or less, check Two's complement).
The remaining 63 bits are for the significations. A bit with the value
1 means the integer number got this mentioned signification. The sum
of all results in the value of the integer number. These significations starts
from 2^0 (1) and goes up to 2^62 (4611686018427387904). As an
example the number 50 got the significations 32, 16 and 2. On bit level these are the
bits 1, 4 and 5. These are put into a 64 bit field.
However, integer numbers are saved in the Two's complement.
Float numbers are saved in a different way. As we can save decimal places too we cannot use the same significations as for integer numbers. Float numbers are normally saved in the IEEE 754 format. For a 32 bit float number one bit is used for the sign, 8 bits for the exponent and 23 bits for the mantissa. The mantissa builds a number between around 0 and 1. On this value a base-2-exponentation is multiplied. This exponentation is build with the mentioned 8 bits. Depending on the choosen standard the mantissa is moved left (if the factor is big, like 2^50) or moved right (if the factor is small, like 2^-50). With the sign bit you get a complete number. So you cannot save any number with decimal places but only a specific number of digits of the number. But thats normally no problem as we use it all the time like giving the weight of something in 31.6 tons instead of 31.6346800235 tons.
