The Natural Order of Refactoring Under the Microscope Part 3: Extract Method
- Mariusz Sieraczkiewicz
- Software development , Code refactoring
- June 11, 2011
Table of Contents
Analyzing Class and Method Responsibilities
In the next step, we examine the responsibilities of individual classes and methods, checking if they align with the intended responsibility of the class. It is best to analyze all methods and group them based on performing similar operations. We look for a place for these in other classes or a new class. Remember: if there is a significant private method in a class (longer than 3-4 lines), it should be moved to another class.
Example: Extracting Methods for Clarity
In our example, the methods mainly revolve around discrete operations aimed at obfuscating text and ensuring operations are performed with a certain probability. We can extract classes corresponding to these operations: TextObfuscatorMethods and Probability.
Refactored Code Example
The code after this change might look like this:
public class TextObfuscator
{
// ...
public string Obfuscate(string text)
{
List<TextPart> parts = methods.ParseTextForWordsAndNonWords(text);
methods.ChangeWordsOrderRandomly(parts);
methods.AddMeaninglessWordsRandomly(parts);
methods.RemoveSeparators(parts);
methods.AddSeparatorsRandomly(parts);
methods.RemoveSpaces(parts);
// ...
}
}
public class TextObfuscatorMethods
{
// ...
public virtual void RemoveSpaces(List<TextPart> parts)
{
for (int i = 0; i < parts.Count; i++)
{
RemoveSpacesFromTextPart(parts, i);
RemoveTextPartIfEmpty(parts, i);
}
}
private static void RemoveSpacesFromTextPart(List<TextPart> parts, int i)
{
parts[i] = parts[i].ReplaceContents(" ", "");
}
public virtual void RemoveSeparators(List<TextPart> parts)
{
for (int i = 0; i < parts.Count; i++)
{
RemoveSeparatorsFromTextPart(parts, i);
RemoveTextPartIfEmpty(parts, i);
}
}
public virtual void AddSeparatorsRandomly(List<TextPart> parts)
{
for (int i = 0; i < parts.Count; i++)
{
if (ShouldAddSeparator())
{
parts.Insert(i, TextPart.NewSeparator(DrawSeparator())); i++;
}
}
}
private bool ShouldAddSeparator()
{
return probability.ShouldBeDone(ADDING_SEPARATOR_PROBABILITY);
}
// ...
}
public class Probability
{
private Random random = new Random();
public bool ShouldBeDone(double probability)
{
if (probability < 0.0 || probability > 1.0)
{
throw new ProbabilityException("Probability should be in range [0.0, 0.1]");
}
return random.NextDouble() < probability;
}
}
This refactoring approach helps to maintain a clean codebase by ensuring that each class has a single responsibility and methods are organized according to their functionalities.
(Text translated and moved from original old blog automatically by AI. May contain inaccuracies.)