Operators and String Formatting in Python
Operators and String Formatting
Terms in This Chapter
 Format directives
 Hexdump
 Key
 Keyword
 Literal
 Modulus
 Operator precedence
 Boolean value
 Class
 Concatenation
 Conversion
 Dictionary
 Directive
 Field
 Flag
 Operator (%, Arithmetic, Bitwise, Comparison, Conditional, Logical, Sequence, Shift)
 String
 Tuple
 Variable
In this chapter, we'll cover operators and string formatting. Python string formatting controls the creation of strings. Done correctly, it makes the production of these strings simple and straightforward.
I've said it before, and I'll say it again: If you're a beginning programmer, remember that the only way to learn programming is by programming, so try to follow along with the interactive sessions throughout the chapter. The interactive interpreter mode will give you a handson understanding of Python operators and string formatting. If you have trouble with an Advanced Topic section, just skim over it; don't let it slow you down.
As in Chapter 2, most of the concepts in this chapter act as building blocks for more complex ideas. Don't worry if something seems unclear to you at this point; you might understand it later, in a different context. For example, logical and comparison operators may not be easily grasped here, but wait until Chapter 4, where we deal with the if statement, which makes frequent use of these operators and so should clear things up.
If you've programmed before, most of this chapter will be familiar. For example, operators and string formatting in Python and C are very similar. If you have indepth programming experience, you can probably just skim this material, especially if you're comfortable with C, Java, and/or Visual Basic. Do, however, pay attention to the following sections:
"Arithmetic with Strings, Lists, and Tuples"
"% Tuple String Formatting"
"Advanced Topic: Using the %d, %i, %f, and %e Directives for Formatting Numbers"
Also read the "For Programmers" sidebar (see pages 50–51).
Operators
Recall from Chapter 2 our definition of expressions as simple statements that return a value. In Python, many expressions use operators, such as +, –, *, and =. The following subsections describe each operator type, and each section contains a table of the type's operators along with sample interactive sessions illustrating their use. If you feel as if you've been this way before, you have—we've been using operators since Chapter 1.
Arithmetic Operators
Arithmetic operators work with the numeric types Float, Int, and Long. Table 3–1 describes them, including three we have yet to encounter: modulus (%), which gives the remainder; exponential (**), which raises one number to the power of another number; and abs, which gives a number's absolute value.
One example of modulus is 3/2, which gives the remainder of 1 (^{ 3 }/_{ 2 }= 1^{ 1 }/_{ 2 }). Another is 10/7, which gives a remainder of 3 (^{ 10 }/_{ 7 }= 1^{ 3 }/_{ 7 }). In Python, we express the previous sentence as
>>> 10 % 7 3 >>> 3 % 2 1 >>>
Once you understand modulus, the divmod() function, which we'll discuss in a later chapter, should come easily to you.
Table 3–1 Arithmetic Operators
Operator 
Description 
Interactive Session 
+ 
Addition 
>>> x = 1 + 2 


>>> print (x) 


3 
– 
Subtraction 
>>> x = 2 – 1 


>>> print (x) 


1 
* 
Multiplication 
>>> x = 2 * 2 


>>> print (x) 


4 
/ 
Integer division returns an Integer type; float division returns a float type 
Integer division: 


>>> x = 10 / 3 


>>> print (x) 


3 


Float division: 


>>> x = 10.0 / 3.3333 


>>> print (x) 


3.000030000300003 
% 
Modulus—gives the remainder; typically used for integers 
>>> x = 10 % 3 


>>> print (x) 


1 
** 
Exponential 
>>> x = 10**2 


>>> print(x) 


100 
divmod 
Does both of the division operators at once and returns a tuple; the second item in the tuple contains the remainder. divmod(x,y) is equivalent to x/y,x%y 
This: 


>>> divmod (10,3) 


(3, 1) 


Is the same as this: 


>>> 10/3,10%3 


(3, 1) 


This: 


>>> divmod (5,2) 


(2, 1) 


Is the same as this: 


>>> 5/2, 5%2 


(2, 1) 
abs 
Finds the absolute value of a number 
>>> abs(100) 


100 


>>> abs(100) 


100 
, 
Sign 
>>> 1, 1, +1, +1 


(1, 1, 1, 1) 
Numeric Conversion Operators
Many times we need to convert from one numeric type to another. The three operators that perform this conversion are Int(x), Long(x), and Float(x), where x is any numeric value. To illustrate, in the example that follows we create three numeric types: 1 (Long), f (Float), and i (Integer).
>>> l,f,i=1L, 1.0, 1
The output is
>>> l,f,i (1L, 1.0, 1)
The next three examples in turn convert i to Float, f and i to Long, and l and f to Integer.
>>> float (i) 1.0 >>> float(l) 1.0 >>> long(f), long(i) (1L, 1L) >>> int(l), int(f) (1, 1) >>>
Logical Operators, Comparison Operators, and Boolean Values
Logical operators are a way to express choices, such as "This one and that one or that one but not this one." Comparison operators are a way to express questions, such as "Is this one greater than that one?" Both work with Boolean values, which express the answer as either true or false. Unlike Java, Python has no true Boolean type. Instead, as in C, its Booleans can be numeric values, where any nonzero value must be true and any zero value must be false. Thus, Python interprets as false the following values:
 None
 Empty strings
 Empty tuples
 Empty lists
 Empty dictionaries
 Zero
and as true all other values, including
 Nonempty strings
 Nonempty tuples
 Nonempty lists
 Nonempty dictionaries
 Not zero
Table 3–2 describes the logical operators. They return 1 for a true expression and 0 for a false expression. Table 3–3 describes the comparison operators. They return some form of true for a true expression and some form of false for a false expression.
Logical and comparison operators often work together to define application logic (in English, application logic simply means decision making).When they do, they're often used with if and while statements. Don't worry about if and while just yet; we'll get into them in detail in Chapter 4. For now, a simple way to visualize them is to imagine that you like vanilla and chocolate ice cream but hate nuts, and you want to express your preference in a way that Python will understand, like this:
if (flavor == chocolate or flavor == vanilla and \ not nuts and mycash > 5): print("yummy ice cream give me some") while(no_vanilla_left and no_chocolate_left ): print ("no more ice cream for me")
Table 3–2 Logical Operators
Operator 
Description 
Interactive Session 
and 
And two values or comparisons together 
x,y = 1,0 


x and y 


0 


x,y = 1,1 


x and y 


1 
or 
Or two values together 
x,y = 1,0 


x or y 


1 


x,y = 0,0 


x or y 


0 
not 
Inverse a value 
x,y = 0,0 


not x 


1 


>>> not y 1 


x = 1 


not x 


0 
Table 3–3 Comparison Operators
Operator 
Description 
Interactive Session 
== 
Equal to 
>>> x,y,z=1,1,2 


>>> x==y 


1 


>>> x==z 


0 
>= 
Greater than or equal to 
>>> z>=x 


1 


>>> x>=z 


0 
<= 
Less than or equal to 
>>> z<=x 


0 


>>> x<=z 


1 
> 
Greater than 
>>> x>z 


0 


>>> z>x 


1 
< 
Less than 
>>> x<z 


1 


>>> z<x 


0 
!= 
Not equal to 
>>> x!=y 


0 


>>> x!=z 


1 
<> 
Not equal to 
>>> x<>y 


0 


>>> x<>z 


1 
is 
Object identity 
>>> str = str2 = "hello" 


>>> str is str2 





1 


>>> str = "hi" 


>>> str is str2 


0 
is not 
Negated object identity 
>>> s1 = s2 = "hello" 


>>> s1 is not s2 


0 


s1 = s2 + " Bob" 


s1 is not s2 


1 
in 
Checks to see if an item 
>>> list = [1,2,3] 

is in a sequence 


>>> 1 in list 


1 





>>> 4 in list 


0 

not in 
Checks to see if an item is 
>>> list = [1,2,3] 

NOT in a sequence 


>>> 1 not in list 


0 





>>> 4 not in list 


1 

Advanced Topic: Logical Operators and Boolean Returns
Comparison operators always return either 1 or 0 of type Integer.
>>> 1 > 2 0 >>> 1 < 2 1
Logical operators can return more than the Integer types 1 or 0, as we see in the following expression, which determines if 0 or (1,2,3) is true.
>>> 0 or (1,2,3) (1, 2, 3)
Python equates 0 to false and a nonempty tuple (1, 2, 3) to true, so the logical operator returns the true statement, that is, the (1,2,3) tuple literal.
The following expression determines if 1 < 2 or the integer 5 is true:
>>> 1 < 2 or 5 1
Because 1 < 2 is true, the expression returns 1, which equates to true, but it equates 5 to true as well; however, only the first true statement in an or statement (1 above) is returned.
The next expression also determines if 1 < 2 or the integer 5 is true, but this time we swap the operands.
>>> 5 or 1 < 2 5
Once again, only the first true statement is returned, which is now 5.
Like or, and returns the first true operand. However, and is unlike or in that only the last operand can make it true.
>>> (1,1) and [2,2] [2, 2] >>> [2,2] and (1,1) (1, 1) >>> [2,2] and (3,3) and {"four":4} {'four': 4}
Conversely, the first occurrences of a false are returned by and.
>>> [1,1] and {} and () {} >>> (1,1) and [] and {} []
For Programmers: Conditional Operators in Other Languages
C, C++, and Java have a conditional operator that works conveniently as shorthand for and and or. In Java, for example, the following two if statements are equivalent:
val = boolean_test ? true_return : false_return; if (boolean_test) val = true_return; else val = false_return;
Python has no conditional operator, but you can simulate one with the form
val = (boolean_test and true_return) \ or false_return
This works because or always returns the first true statement and and always returns the last one, so these two statements are equivalent:
>>> if ( 3 > 5): ... num = 1 ... else: ... num = 2 ... >>>num 2 >>> num = (3>5 and 1) or 2 >>> num 2
The following two expressions are also equivalent:
>>> if ( 5 > 3): ... num = 1 ... else: ... num = 2 ... >>>num 1 >>>num = (5>3 and 1) or 2 >>> num 1
Be warned: This simulation works only if the expressions true_return and false_ return are not equivalent to false. If we need them to be false, we can do something like this:
>>> true_return = 0 >>> false_return = 2 >>> num = (5 > 3 and [true_return]) \ or [false_return] >>> num [0]
Now the num variable is equivalent to a list containing one element, but this isn't exactly what we want. However, since this code is returning a list, we can put the entire expression to the left of the assignment operator in parentheses, which will achieve our desired result.
>>> num = ((5 > 3 and [true_return]) or [false_return])[0] >>> num 0
To sum up, Python's bulletproof equivalent of the conditional operator is
val = ((boolean_test and [true_return]) \ or [false_return])[0]
which is no more verbose than another Python expression:
if (boolean_test): val = true_return else: val = false_return
Python's version of the conditional operator is hard to understand and use, so go easy with it. Perhaps one day Python will have a conditional operator of its own (and, I might add, its own += operator).
Advanced Topic: Bitwise and Shift Operators
If you lack experience with any programming language or with Boolean algebra, you should ignore bitwise operators. Another reason to ignore them is that they're usually associated with lowlevel programming, and you're learning Python, which is much higher level than C, C++, or even Java. If for some reason you're curious about bitwise operators, any introductory C text will tell you all you need to know. The same goes for shift operators.
Just for the sake of completeness, Table 3–4 describes both operator types. To understand it, you need to know something of hexadecimal and Boolean algebra. (See Chapter 10 for an example of a hexdump file viewer, which uses the shift operators.)
Operator Precedence
Operator precedence determines the order in which Python evaluates the parts of a statement. It generally follows the operator precedence you learned in high school algebra and is nearly identical to that used in any other common programming language. Here's an example in which y/z is processed before 2 + y, rendering x equal to 4 and not 6.
>>> x,y,z=1,4,2 >>> x = 2 + y / z >>> x 4
When in doubt as to which operator will be evaluated first, use parentheses. They prevent many a mistake if you use them to force precedence, and they enhance code readability.
You may occasionally want to force a precedence other than the algebraic default to make it more explicit. The following example shows how to do this:
>>> x,y,z = 1,4,2 >>> x = 2 + (y/z) >>> x 4
>>> x = (2+y) /z >>> x 3
The choice of precedence here depends on which expression—2 + y or y/z—is to be evaluated first. Note, though, that the value of x changes according to the grouping and ends up as either 4 or 3. The first expression, 2 + y/z, is unnecessary except to adorn the code with parentheses for clarity, which is important for code maintainability.
Visit www.python.org for a detailed description of operator precedence. For now, the following list shows all operators in their precedence order:
[], (), {}, ' 
Parentheses, string conversion 
seq[index] 
Indexing sequences or dictionaries 
integer 
.MAX_INT Attribute reference 
~I 
Bit inversion 
i, +i 
Unary minus, unary plus 
*, /, % 
Multiplication, division, modulus 
+,  
Addition, subtraction 
Table 3–4 Bitwise and Shift Operators
Operator 
Description 
Interactive Session 
<< 
Shift left 
>>> # binary 1111 1111 


>>> x = 0xff 





>>> # z = 0011 1111 1100 


>>> z = x << 2 


>>> print (x) 


255 





>>> print (z) 


1020 
>> 
Shift right 
>>> # z = 1020 


>>> #z = 1111 1111 





>>> z = z >> 2 


>>> z 


255 
& 
Bitwise and 
>>> # y = binary 0000 1010 


>>> y = 0x0A 


>>> print (y) 


10 





>>> print (x) 


255 





>>> z = y & z 


>>> print (z) 


10 
 
Bitwise or 
>>> #continued example 


>>> z = y  x 


>>> print (z) 


255 
^ 
Bitwise XOR 
>>> z = y ^ x 


>>> z 


245 
~ 
Bitwise not 
>>> y = 0xffffffff 


>>> y 


1 





>>> z = ~y 


>>> z 


0 

<<,>> 
Bit shifting, left and right 

[bar], &, ^ 
Bitwise operator or, and, XOR 

<, >, >=, <=, ==, is not, 
Comparison operators 

is, !=, <>, in, not in 


not 
Logical not 

and 
Logical and 

or 
Logical or 
Arithmetic with Strings, Lists, and Tuples
As in Java, the addition operator (+) in Python works with string types to concatenate, that is, link strings. (Recall that we used + on strings in Chapter 1.) Unlike in Java, the multiplication operator (*) in Python is used to repeat string values. Consider the following interactive session:
>>> love = "I love my family" + (" very " * 4) + "much!" >>> love 'I love my family very very very very much!' >>> hugs_and_kisses = "XO" * 20 >>> hugs_and_kisses 'XOXOXOXOXOXOXOXOXOXOXOXOXOXOXOXOXOXOXOXO'
In the first line, we concatenate the string "I love my family" with the string "very" multiplied by 4, which gives "very very very very". If we want to sign our letter with hugs and kisses (Xs and Os), we can multiply to save time in a similar way.
Addition and multiplication also work for lists.
>>> friends = ["Joey", "Monica", "Ross", "Chandler"] >>> old_friends = ["Fonz", "Ritchie", "Potsy"] >>> strange_cast = friends + old_friends >>> strange_cast ['Joey', 'Monica', 'Ross', 'Chandler', 'Fonz', 'Ritchie', 'Potsy']
They work for tuples as well.
>>> tuple1 = (1,2,3,4) >>> tuple2 = ("5", 6L, 7.0, 0x8) >>> tuple3 = tuple1 + tuple2 >>> tuple3 (1, 2, 3, 4, '5', 6L, 7.0, 8) >>> tuple4 = tuple1 * 2 >>> tuple4 (1, 2, 3, 4, 1, 2, 3, 4)
Sequence Operators
We've been using sequence operators since Chapter 1, so you've seen most of them. However, some of the operators in Table 3–5 will be new to you.
Table 3–5 Sequence Operators
Operator 
Description 
Interactive Session 
[index] 
Get the indexed item 
>>> nums = (0,1,2,3,4,5,6,7,8,9) 


>>> nums[0] 


0 





>>> nums[1] 


1 
[:] 
Slice notation 
>>> nums = (0,1,2,3,4,5,6,7,8,9) 





>>> nums [:] 


(0, 1, 2, 3, 4, 5, 6, 7, 8, 9) 





>>> nums [5:] 


(5, 6, 7, 8, 9) 





>>> nums [:5] 


(0, 1, 2, 3, 4) 



len() 
Length 
>>> nums = (0,1,2,3,4,5,6,7,8,9) 





>>> len (nums) 


10 
max() 
Get the largest item in the sequence 
>>> nums = (0,1,2,3,4,5,6,7,8,9) 





>>> max(nums) 


9 





>>> letters = "abcdefg" 


>>> max (letters) 


'g' 



min() 
Get the smallest item in the sequence 
>>> nums = (0,1,2,3,4,5,6,7,8,9) 





>>> min(nums) 


0 


>>> letters = "abcdefg" 





>>> min (letters) 


'a' 