There are equivalent sizes, which are those whose sizing system contains the same numbers, although in different positions. For the equivalent sizes one must select the one with the highest stock, discarding their equivalences. The objective is to show those product sizes that have Stock.
We assume that all product sizes doesn't start by 0. For example 001 will be not a valid inbound data.
On the other hand we encourage you to develop your own datasets. You could place your own CSV on this folder. You could copy/paste a gherkins test case definition replacing the existing CSV filenames by your own ones. In order to run the acceptance test you could run mainAcceptanceClass or execute mvn test -P acceptance
.
.
├── README.md
├── complexityChart.png
├── logo.png
├── pom.xml
├── src
│ ├── acceptance-test
│ │ ├── java
│ │ │ ├── RunCukesTest.java
│ │ │ └── org
│ │ │ └── indietxt
│ │ │ └── acceptanceTest
│ │ │ ├── CalculateStockIT.java
│ │ │ ├── CommonsHooks.java
│ │ │ └── utils
│ │ │ ├── ClassPattern.java
│ │ │ ├── CsvFileReader.java
│ │ │ ├── Pattern.java
│ │ │ └── PatternMatching.java
│ │ └── resources
│ │ ├── basicSolution.csv
│ │ ├── features
│ │ │ └── ExerciseAcceptance.feature
│ │ ├── loadProductSizeCaseBasic.csv
│ │ ├── loadProductSizeCaseMessy.csv
│ │ ├── loadProductSizeCaseSizeSystem.csv
│ │ ├── loadProductSizeCaseSizeSystemComplex.csv
│ │ ├── loadProductSizeCaseSizeSystemReverse.csv
│ │ ├── loadStockEntryCaseBasic.csv
│ │ ├── loadStockEntryCaseMessy.csv
│ │ ├── loadStockEntryCaseSizeSystem.csv
│ │ ├── loadStockEntryCaseSizeSystemComplex.csv
│ │ ├── sizeSystemSolution.csv
│ │ ├── sizeSystemSolutionComplex.csv
│ │ └── sizeSystemSolutionMessy.csv
│ ├── main
│ │ └── java
│ │ └── org
│ │ └── indietxt
│ │ ├── model
│ │ │ ├── ProductSize.java
│ │ │ └── StockEntry.java
│ │ ├── service
│ │ │ ├── StockManager.java
│ │ │ ├── StockManagerServiceLocator.java
│ │ │ └── impl
│ │ │ └── StockManagerReferenceImpl.java
│ │ └── storage
│ │ ├── AbstractDAO.java
│ │ ├── ProductSizeDAO.java
│ │ ├── ProductSizePredicates.java
│ │ ├── StockEntryDAO.java
│ │ └── StockEntryPredicates.java
│ ├── microbenchmark
│ │ ├── java
│ │ │ └── org
│ │ │ └── indietxt
│ │ │ └── performance
│ │ │ ├── BenchmarkRunner.java
│ │ │ └── StockManagerPerformance.java
│ │ └── resources
│ │ ├── loadProductSizeEighHundredItems.csv
│ │ ├── loadProductSizeFiftyItems.csv
│ │ ├── loadProductSizeFiveHundredItems.csv
│ │ ├── loadProductSizeFourHundredItems.csv
│ │ ├── loadProductSizeHundredItems.csv
│ │ ├── loadProductSizeNineHundredItems.csv
│ │ ├── loadProductSizeSevenHundredItems.csv
│ │ ├── loadProductSizeSixHundredItems.csv
│ │ ├── loadProductSizeTenItems.csv
│ │ ├── loadProductSizeThousandItems.csv
│ │ ├── loadProductSizeThreeHundredItems.csv
│ │ ├── loadProductSizeTwoHundredItems.csv
│ │ ├── loadStockEighHundredItems.csv
│ │ ├── loadStockFiftyItems.csv
│ │ ├── loadStockFiveHundredItems.csv
│ │ ├── loadStockFourHundredItems.csv
│ │ ├── loadStockHundredItems.csv
│ │ ├── loadStockNineHundredItems.csv
│ │ ├── loadStockSevenHundredItems.csv
│ │ ├── loadStockSixHundredItems.csv
│ │ ├── loadStockTenItems.csv
│ │ ├── loadStockThousandItems.csv
│ │ ├── loadStockThreeHundredItems.csv
│ │ └── loadStockTwoHundredItems.csv
│ └── test
│ └── java
│ └── org
│ └── indietxt
│ ├── service
│ │ ├── ServiceLocatorTest.java
│ │ └── StockManagerReferenceImplTest.java
│ └── storage
│ ├── ProductSizeDaoTest.java
│ └── StockEntryDaoTest.java
└── suply-chain.iml
We ask you to solve it using the Java programming language. You are only allowed to use test libraries (JUnit, TestNG, Hamcrest,...) and nothing else.
You are free to implement any mechanism for feeding input into your solution (for example, using hard coded data within a unit test).
Remember that the automated and self-checking tests are mandatory. You should provide sufficient evidence that your solution is complete by, as a minimum, indicating that it works correctly against the supplied test data.
We kindly ask you to put your code on a public project in github.com under your account. Use Maven or Gradle as automation tool and link it to a build service likeTravis CI.
The goal is to provide us with a full understanding of your coding style and skills. We’ll pay particular attention to:
- readable and well-structured code
- design and domain modelling
- quality of tests
I request to retrieve the stock of each item(with equivalent logic - full exercise), and I got the following result. We run, 2 Rounds, 20 iterations each one.
Summary
Benchmark (iterations) Mode Cnt Score Error Units
StockManagerPerformance.stockManagerTest 1 avgt 20 0.006 ± 0.001 ms/op
StockManagerPerformance.stockManagerTenItemsTest 1 avgt 20 0.016 ± 0.001 ms/op
StockManagerPerformance.stockManagerFiftyItemsTest 1 avgt 20 0.140 ± 0.006 ms/op
StockManagerPerformance.stockManagerHundredItemsTest 1 avgt 20 0.313 ± 0.014 ms/op
StockManagerPerformance.stockManagerTwoHundredItemsTest 1 avgt 20 0.905 ± 0.013 ms/op
StockManagerPerformance.stockManagerThreeHundredItemsTest 1 avgt 20 1.806 ± 0.044 ms/op
StockManagerPerformance.stockManagerFourHundredItemsTest 1 avgt 20 3.077 ± 0.139 ms/op
StockManagerPerformance.stockManagerFiveHundredItemsTest 1 avgt 20 4.550 ± 0.031 ms/op
StockManagerPerformance.stockManagerSixHundredItemsTest 1 avgt 20 6.684 ± 0.165 ms/op
StockManagerPerformance.stockManagerSevenHundredItemsTest 1 avgt 20 8.918 ± 0.173 ms/op
StockManagerPerformance.stockManagerEighHundredItemsTest 1 avgt 20 12.289 ± 0.343 ms/op
StockManagerPerformance.stockManagerNineHundredItemsTest 1 avgt 20 15.159 ± 0.759 ms/op
StockManagerPerformance.stockManagerThousandItemsTest 1 avgt 20 17.997 ± 0.246 ms/op
1 items Test (product / stock)
Result "org.indietxt.performance.StockManagerPerformance.stockManagerTest":
0.006 ±(99.9%) 0.001 ms/op [Average]
(min, avg, max) = (0.004, 0.006, 0.008), stdev = 0.001
CI (99.9%): [0.005, 0.006] (assumes normal distribution)
10 items Test (product / stock)
Result "org.indietxt.performance.StockManagerPerformance.stockManagerTenItemsTest":
0.016 ±(99.9%) 0.001 ms/op [Average]
(min, avg, max) = (0.015, 0.016, 0.017), stdev = 0.001
CI (99.9%): [0.015, 0.016] (assumes normal distribution)
50 items Test (product / stock)
Result "org.indietxt.performance.StockManagerPerformance.stockManagerFiftyItemsTest":
0.140 ±(99.9%) 0.006 ms/op [Average]
(min, avg, max) = (0.134, 0.140, 0.162), stdev = 0.007
CI (99.9%): [0.133, 0.146] (assumes normal distribution)
100 items Test (product / stock)
Result "org.indietxt.performance.StockManagerPerformance.stockManagerHundredItemsTest":
0.313 ±(99.9%) 0.014 ms/op [Average]
(min, avg, max) = (0.301, 0.313, 0.357), stdev = 0.016
CI (99.9%): [0.299, 0.326] (assumes normal distribution)
200 items Test (product / stock)
```
Result "org.indietxt.performance.StockManagerPerformance.stockManagerTwoHundredItemsTest":
0.905 ±(99.9%) 0.013 ms/op [Average]
(min, avg, max) = (0.873, 0.905, 0.929), stdev = 0.015
CI (99.9%): [0.891, 0.918] (assumes normal distribution)
```
300 items Test (product / stock)
```
Result "org.indietxt.performance.StockManagerPerformance.stockManagerThreeHundredItemsTest":
1.806 ±(99.9%) 0.044 ms/op [Average]
(min, avg, max) = (1.722, 1.806, 1.913), stdev = 0.050
CI (99.9%): [1.762, 1.850] (assumes normal distribution)
```
400 items Test (product / stock)
```
Result "org.indietxt.performance.StockManagerPerformance.stockManagerFourHundredItemsTest":
3.077 ±(99.9%) 0.139 ms/op [Average]
(min, avg, max) = (2.972, 3.077, 3.536), stdev = 0.160
CI (99.9%): [2.938, 3.216] (assumes normal distribution)
```
500 items Test (product / stock)
```
Result "org.indietxt.performance.StockManagerPerformance.stockManagerFiveHundredItemsTest":
4.550 ±(99.9%) 0.031 ms/op [Average]
(min, avg, max) = (4.491, 4.550, 4.656), stdev = 0.036
CI (99.9%): [4.519, 4.581] (assumes normal distribution)
```
600 items Test (product / stock)
```
Result "org.indietxt.performance.StockManagerPerformance.stockManagerSixHundredItemsTest":
6.684 ±(99.9%) 0.165 ms/op [Average]
(min, avg, max) = (6.576, 6.684, 7.394), stdev = 0.190
CI (99.9%): [6.519, 6.849] (assumes normal distribution)
```
700 items Test (product / stock)
```
Result "org.indietxt.performance.StockManagerPerformance.stockManagerSevenHundredItemsTest":
8.918 ±(99.9%) 0.173 ms/op [Average]
(min, avg, max) = (8.719, 8.918, 9.522), stdev = 0.199
CI (99.9%): [8.745, 9.091] (assumes normal distribution)
```
800 items Test (product / stock)
Result "org.indietxt.performance.StockManagerPerformance.stockManagerEighHundredItemsTest":
12.289 ±(99.9%) 0.343 ms/op [Average]
(min, avg, max) = (11.650, 12.289, 13.479), stdev = 0.395
CI (99.9%): [11.945, 12.632] (assumes normal distribution)
900 items Test (product / stock)
```
Result "org.indietxt.performance.StockManagerPerformance.stockManagerNineHundredItemsTest":
15.159 ±(99.9%) 0.759 ms/op [Average]
(min, avg, max) = (14.204, 15.159, 17.892), stdev = 0.874
CI (99.9%): [14.401, 15.918] (assumes normal distribution)
```
1000 items Test (product / stock)
Result "org.indietxt.performance.StockManagerPerformance.stockManagerThousandItemsTest":
17.997 ±(99.9%) 0.246 ms/op [Average]
(min, avg, max) = (17.761, 17.997, 18.681), stdev = 0.283
CI (99.9%): [17.752, 18.243] (assumes normal distribution)
Based on our studies we concluded that the current basic implementation StockManagerReferenceImpl has a exponential cyclomatic complexity level. On the first hand it's quite surprising if you have a look the method code, because if you consider a stream as a loop, then I will say that the complexity level should be 3 * O(n).
Code snippet:
@Override
public List<StockEntry> retrieveStock() {
List<ProductSize> pSize = productSizeDAO.filterEntities(ProductSizePredicates.alwaysTrue()).collect(Collectors.toList());
List<Integer> ids = pSize.stream().map(ProductSize::getId).collect(Collectors.toList());
return stockEntryDAO.filterEntities(StockEntryPredicates.byIds(ids.toArray(new Integer[ids.size()]))).collect(Collectors.toList());
}
But the data never lie you (this could be the title of a movie ;) ). This exercise it's just an example of Java8 features, however if you think to move on "this code" to production you should analise the behavior with a profiler in order to understand, where the bottle neck is?. Could be memory, could be CPU... etc.
Finally an other conclusion could be that nowadays with lambdas, static imports or default methods that inject code into your class make more fuzzy your code, in order to understand your "cyclomatic complexity" this is why is always a good reason to test your code with frameworks or libraries that " throw a little light on your darkness". Also the numbers are not important at all (depends on your laptop, the services or programs that your are running at this moment), however the shape of your graph it's very important.
Well Thank you if you are following this exercise, and remember that you could reuse the acceptance and performance test defined in this project with your own StockManager implementation. Just remember to add your implementation to the ServiceLocator and also update your acceptance test given step in order to add your "algorithm" name.
switch(algorithm) {
case "yourAlgorithmName"
default:
stockManagerSelected = stockManagerReference;
break;
}
Finally you need to update the acceptance test definition and replace the keyword "ReferenceImpl" by your algorithm name. By this way you will set your implementation as a default acceptance test implementation.
Well if you have any question, or you just wanna chat about technology, you can contact me at anytime.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent