Should I use an Iterator or a forloop to iterate?
I don't like the idea of having to call hasNext() and next() , and seeing how I would need a loop to call those anyways, I don't really see much point in using the iterators given to us in Lists and Maps.
Is it only if you need to pull one this from the list? Or, if you're taking one item from the list, and having to do that for an unmeasured amount of time? I couldn't find this on Google or Stack, please help me out.
I'm not talking about an enhanced for-loop specifically (for-each loop).
I also understand that foreach loops excel in performance, but this is more of a 'why does it exist' question
EDIT: Realized I was talking about collections alone and not arrays aswel. Foreach loops don't have limitations in this situation.
A foreach is equivalent to an iterator--it's syntactic sugar for the same thing. So you should always choose foreach over iterator whenever you can, simply because it's convenient and results in more concise code.
I've written about this in another answer: https://stackoverflow.com/questions/22110482/uses-and-syntax-for-for-each-loop-in-java/22110517#22110517
As stated by @RonnyShapiro, there are situations where you need to use an iterator, but in many cases a foreach should suffice. Note that a foreach is not a normal for loop. The normal for loop is needed when access to the index is required. Although you could manually create a separate index int-variable with foreach it is not ideal, from a variable-scope point of view.
Here's some more information: Which is more efficient, a for-each loop, or an iterator?
When accessing collections, a foreach is significantly faster than the basic for loop's array access. When accessing arrays, however--at least with primitive and wrapper-arrays--access via indexes is way faster. See below.
Indexes are 23- 40 percent faster than iterators when accessing int or Integer arrays. Here is the output from the below testing class, which sums the numbers in a 100-element primitive-int array (A is iterator, B is index):
[C:java_code]java TimeIteratorVsIndexIntArray 1000000
Test A: 358,597,622 nanoseconds
Test B: 269,167,681 nanoseconds
B faster by 89,429,941 nanoseconds (24.438799231635727% faster)
[C:java_code]java TimeIteratorVsIndexIntArray 1000000
Test A: 377,461,823 nanoseconds
Test B: 278,694,271 nanoseconds
B faster by 98,767,552 nanoseconds (25.666236154695838% faster)
[C:java_code]java TimeIteratorVsIndexIntArray 1000000
Test A: 288,953,495 nanoseconds
Test B: 207,050,523 nanoseconds
B faster by 81,902,972 nanoseconds (27.844689860906513% faster)
[C:java_code]java TimeIteratorVsIndexIntArray 1000000
Test A: 375,373,765 nanoseconds
Test B: 283,813,875 nanoseconds
B faster by 91,559,890 nanoseconds (23.891659337194227% faster)
[C:java_code]java TimeIteratorVsIndexIntArray 1000000
Test A: 375,790,818 nanoseconds
Test B: 220,770,915 nanoseconds
B faster by 155,019,903 nanoseconds (40.75164734599769% faster)
[C:java_code]java TimeIteratorVsIndexIntArray 1000000
Test A: 326,373,762 nanoseconds
Test B: 202,555,566 nanoseconds
B faster by 123,818,196 nanoseconds (37.437545972215744% faster)
The full testing class:
import java.text.NumberFormat;
import java.util.Locale;
/**
<P>{@code java TimeIteratorVsIndexIntArray 1000000}</P>
@see <CODE><A HREF="https://stackoverflow.com/questions/180158/how-do-i-time-a-methods-execution-in-java">https://stackoverflow.com/questions/180158/how-do-i-time-a-methods-execution-in-java</A></CODE>
**/
public class TimeIteratorVsIndexIntArray {
public static final NumberFormat nf = NumberFormat.getNumberInstance(Locale.US);
public static final void main(String[] tryCount_inParamIdx0) {
int testCount;
//Get try-count from command-line parameter
try {
testCount = Integer.parseInt(tryCount_inParamIdx0[0]);
} catch(ArrayIndexOutOfBoundsException | NumberFormatException x) {
throw new IllegalArgumentException("Missing or invalid command line parameter: The number of testCount for each test. " + x);
}
//Test proper...START
int[] intArray = new int[] {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100};
long lStart = System.nanoTime();
for(int i = 0; i < testCount; i++) {
testIterator(intArray);
}
long lADuration = outputGetNanoDuration("A", lStart);
lStart = System.nanoTime();
for(int i = 0; i < testCount; i++) {
testFor(intArray);
}
long lBDuration = outputGetNanoDuration("B", lStart);
outputGetABTestNanoDifference(lADuration, lBDuration, "A", "B");
}
private static final void testIterator(int[] int_array) {
int total = 0;
for(int i = 0; i < int_array.length; i++) {
total += int_array[i];
}
}
private static final void testFor(int[] int_array) {
int total = 0;
for(int i : int_array) {
total += i;
}
}
//Test proper...END
//Timer testing utilities...START
public static final long outputGetNanoDuration(String s_testName, long l_nanoStart) {
long lDuration = System.nanoTime() - l_nanoStart;
System.out.println("Test " + s_testName + ": " + nf.format(lDuration) + " nanoseconds");
return lDuration;
}
public static final long outputGetABTestNanoDifference(long l_aDuration, long l_bDuration, String s_aTestName, String s_bTestName) {
long lDiff = -1;
double dPct = -1.0;
String sFaster = null;
if(l_aDuration > l_bDuration) {
lDiff = l_aDuration - l_bDuration;
dPct = 100.00 - (l_bDuration * 100.0 / l_aDuration + 0.5);
sFaster = "B";
} else {
lDiff = l_bDuration - l_aDuration;
dPct = 100.00 - (l_aDuration * 100.0 / l_bDuration + 0.5);
sFaster = "A";
}
System.out.println(sFaster + " faster by " + nf.format(lDiff) + " nanoseconds (" + dPct + "% faster)");
return lDiff;
}
//Timer testing utilities...END
}
I also ran this for an Integer array, and indexes are still the clear winner, but only between 18 and 25 percent faster.
For a List of Integers , however, iterators are faster. Just change the int-array in the above code to
List<Integer> intList = Arrays.asList(new Integer[] {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100});
and make the necssary changes to the test-function ( int[] to List<Integer> , length to size() , etc)
[C:java_code]java TimeIteratorVsIndexIntegerList 1000000
Test A: 3,429,929,976 nanoseconds
Test B: 5,262,782,488 nanoseconds
A faster by 1,832,852,512 nanoseconds (34.326681820485675% faster)
[C:java_code]java TimeIteratorVsIndexIntegerList 1000000
Test A: 2,907,391,427 nanoseconds
Test B: 3,957,718,459 nanoseconds
A faster by 1,050,327,032 nanoseconds (26.038700083921256% faster)
[C:java_code]java TimeIteratorVsIndexIntegerList 1000000
Test A: 2,566,004,688 nanoseconds
Test B: 4,221,746,521 nanoseconds
A faster by 1,655,741,833 nanoseconds (38.71935684115413% faster)
[C:java_code]java TimeIteratorVsIndexIntegerList 1000000
Test A: 2,770,945,276 nanoseconds
Test B: 3,829,077,158 nanoseconds
A faster by 1,058,131,882 nanoseconds (27.134122749113843% faster)
[C:java_code]java TimeIteratorVsIndexIntegerList 1000000
Test A: 3,467,474,055 nanoseconds
Test B: 5,183,149,104 nanoseconds
A faster by 1,715,675,049 nanoseconds (32.60101667104192% faster)
[C:java_code]java TimeIteratorVsIndexIntList 1000000
Test A: 3,439,983,933 nanoseconds
Test B: 3,509,530,312 nanoseconds
A faster by 69,546,379 nanoseconds (1.4816434912159906% faster)
[C:java_code]java TimeIteratorVsIndexIntList 1000000
Test A: 3,451,101,466 nanoseconds
Test B: 5,057,979,210 nanoseconds
A faster by 1,606,877,744 nanoseconds (31.269164666060377% faster)
In one test they're almost equivalent, but still, iterator wins.
For each was added at Java 5 for easier iteration over collections. However, it does not replace iterators as only with iterators you can modify the collection while iterating over it (via the iterator interface). Attempting to addremove object from a collection within a for each will cause a ConcurrentModificationException.
If you're just reading values, foreach is probably better.
Well, Java Language Spec 8(http://docs.oracle.com/javase/specs/jls/se8/jls8.pdf),14.14.2:
The enhanced for statement is equivalent to a basic for statement of the form:
for (I #i = Expression.iterator(); #i.hasNext(); ) {
{VariableModifier} TargetType Identifier =
(TargetType) #i.next();
Statement
}
So, for compiler that's the same. Earlier versions of this standard contain the same description of 'foreach' statement
链接地址: http://www.djcxy.com/p/52988.html上一篇: 在PHP中使用foreach循环时查找数组的最后一个元素
下一篇: 我应该使用迭代器还是循环迭代器？