Regular expressions are a powerful way to search through and manipulate text. But as we know from Spiderman, with great power comes greatly reduced readability. In my current side project, I'm using regexes as part of a dynamic comment generator for teachers to provide meaningful feedback to students. The intent of this tool is to reuse comments that a teacher has already written and apply them mad-libs style to similarly performing students. With the goal of writing extendable, readable code that adheres to the SOLID principles, let's refactor a shamelessly rigid example in which we replace all pronouns with their grammatical form so that they can later be re-substituted with male or female pronouns.
text.gsub!(/(\bhis\b|\bher\b|\bhe\b|\bshe\b|\bhim\b)/i,
"his" => ":possessive_pronoun:",
"His" => ":possessive_pronoun:",
"her" => ":possessive_pronoun:",
"Her" => ":possessive_pronoun:",
"he" => ":personal_pronoun:",
"He" => ":personal_pronoun:",
"she" => ":personal_pronoun:",
"She" => ":personal_pronoun:",
"him" => ":dative_personal_pronoun:",
"Him" => ":dative_personal_pronoun:"
)
The gsub! method takes in a regular expression as its first argument. The second argument could be a single word so that any time a match is found a single substitution could be made. But here we want to conditionally substitute. For that, the hash of matches to their substitutes is provided as the second argument. For example, when gsub! finds "his" it looks to the value stored under the "his" key, which is ":possessive_pronoun:". The first thing to do here is to give this method a good name. How about substitute_pronouns.
def substitute_pronouns(text)
text.gsub!(/(\bhis\b|\bher\b|\bhe\b|\bshe\b|\bhim\b)/i,
"his" => ":possessive_pronoun:",
"His" => ":possessive_pronoun:",
"her" => ":possessive_pronoun:",
"Her" => ":possessive_pronoun:",
"he" => ":personal_pronoun:",
"He" => ":personal_pronoun:",
"she" => ":personal_pronoun:",
"She" => ":personal_pronoun:",
"him" => ":dative_personal_pronoun:",
"Him" => ":dative_personal_pronoun:"
)
end
Next, the regex could use a name. It's looking for pronouns, so that sounds like a reasonable name.
def substitute_pronouns(text)
pronouns = /(\bhis\b|\bher\b|\bhe\b|\bshe\b|\bhim\b)/i
text.gsub!(pronouns,
"his" => ":possessive_pronoun:",
"His" => ":possessive_pronoun:",
"her" => ":possessive_pronoun:",
"Her" => ":possessive_pronoun:",
"he" => ":personal_pronoun:",
"He" => ":personal_pronoun:",
"she" => ":personal_pronoun:",
"She" => ":personal_pronoun:",
"him" => ":dative_personal_pronoun:",
"Him" => ":dative_personal_pronoun:"
)
end
Let's make this a bit easier to digest by extracting the substitution mappings into a separate method.
def substitute_pronouns(text)
pronouns = /(\bhis\b|\bher\b|\bhe\b|\bshe\b|\bhim\b)/i
text.gsub!(pronouns, substitutions)
end
def substitutions
{
"his" => ":possessive_pronoun:",
"His" => ":possessive_pronoun:",
"her" => ":possessive_pronoun:",
"Her" => ":possessive_pronoun:",
"he" => ":personal_pronoun:",
"He" => ":personal_pronoun:",
"she" => ":personal_pronoun:",
"She" => ":personal_pronoun:",
"him" => ":dative_personal_pronoun:",
"Him" => ":dative_personal_pronoun:",
}
end
This is getting easier to read! But now the duplication is starting to stand out. Each of the keys is something we're looking for in the text. We can use the keys to dynamically generate our regular expression. In addition, we can remove the i flag for case insensitivity now that we're explicitly searching for both "Her" and "her".
def keys_as_regexes
substitutions.keys.map { |pronoun| /\b#{pronoun}\b/ }
end
def substitutions
{
"his" => ":possessive_pronoun:",
"His" => ":possessive_pronoun:",
"her" => ":possessive_pronoun:",
"Her" => ":possessive_pronoun:",
"he" => ":personal_pronoun:",
"He" => ":personal_pronoun:",
"she" => ":personal_pronoun:",
"She" => ":personal_pronoun:",
"him" => ":dative_personal_pronoun:",
"Him" => ":dative_personal_pronoun:",
}
end
Now that the keys have each been converted to a regular expression, we're ready to complete this refactor by joining them together into one regex. To do this, we use Regexp.union.
def substitute_pronouns(text)
text.gsub!(Regexp.union(keys_as_regexes), substitutions)
end
def keys_as_regexes
substitutions.keys.map { |pronoun| /\b#{pronoun}\b/ }
end
def substitutions
{
"his" => ":possessive_pronoun:",
"His" => ":possessive_pronoun:",
"her" => ":possessive_pronoun:",
"Her" => ":possessive_pronoun:",
"he" => ":personal_pronoun:",
"He" => ":personal_pronoun:",
"she" => ":personal_pronoun:",
"She" => ":personal_pronoun:",
"him" => ":dative_personal_pronoun:",
"Him" => ":dative_personal_pronoun:",
}
end
Almost there! We're very close to complying with the Open Closed Principle: the substitutions are open to extension, but closed to modification. By optionally requiring substitutions to be passed in, the method that calls substitute_pronouns can extend the substitutions without needing to modify the substitutions method that contains pronoun mappings.
def substitute_pronouns(text, substitutions = {})
text.gsub!(Regexp.union(keys_as_regexes(substitutions)), substitutions)
end
def keys_as_regexes(substitutions = {})
substitutions.keys.map { |word| /\b#{word}\b/ }
end
def substitutions
{
"his" => ":possessive_pronoun:",
"His" => ":possessive_pronoun:",
"her" => ":possessive_pronoun:",
"Her" => ":possessive_pronoun:",
"he" => ":personal_pronoun:",
"He" => ":personal_pronoun:",
"she" => ":personal_pronoun:",
"She" => ":personal_pronoun:",
"him" => ":dative_personal_pronoun:",
"Him" => ":dative_personal_pronoun:",
}
end
For instance, if we want to replace "Sally" with ":first_name:" we can call substitute_pronouns(text, substitutions.merge!({"Sally" => ":first_name:"}).
Not only does this refactoring make the code easier to read and understand, but it also allows the code to be more easily extended. Do yourself and the others who have to work with your code a favor: write shameless code, then shamelessly refactor!
