Specs2: compare files in two directories, filtering out some line based on regex (for ex when they contains datetime "now", random id, etc)

specs2_regex_filtered_file_content_match.scala

package test

/**
 * Specs2 matcher are wonderful. You're almost sur to find one for your need, whatever the need. 
 * Mine was to compare the output of an algo writting a directory tree with template filled
 * with a reference one. And great! Specs2 has a:
 *    
 *     actualDir must haveSameFilesAs(expectedDir)
 * 
 * See: https://etorreborre.github.io/specs2/guide/SPECS2-3.6.6/org.specs2.guide.Matchers.html#optional
 * in section "File Content"
 * 
 * But I needed to filter out some line comparison in my file, because they were filled with 
 * non predictible data (like generated uuid or "System.currentTimeMillis")
 * 
 * So I created a pre-filter for the content matcher which filter out lines read from the 
 * files in directories based on regex, in both the actual and expected directories. 
 * 
 * And it was trivial to do it. 
 */

// need an extra import: org.specs2" %% "specs2-matcher-extra" % "3.6.6" % "test" 
import org.specs2.matcher.ContentMatchers
import org.specs2.io.FileLinesContent
import org.specs2.mutable.Specification
import org.specs2.text.LinesContent
//other imports..

class CompareDirectoryContentSpec extends Specification with ContentMatchers {

  // the regex filter allowing to filter out line of files
  case class RegexFileContent(regex: List[String]) extends LinesContent[File] {
    val patterns = regex.map(_.r.pattern)

    override def lines(f: File): Seq[String] = {
      FileLinesContent.lines(f).filter { line => !patterns.exists { _.matcher(line).matches() } }
    }

    override def name(f: File) = FileLinesContent.name(f)
  }

  // The matcher which uses the content filter to read files
  def ignoreSomeLinesMatcher(regex: List[String]): LinesPairComparisonMatcher[File, File] = {
    LinesPairComparisonMatcher[File, File]()(RegexFileContent(regex), RegexFileContent(regex))
  }

// some code, initialise things, do write your output

 "My generated files in generatedDir" should {
   "have exactly the same content that the ones in expectedDir (modulo some lines)" in {
      generatedDir must haveSameFilesAs(expectedDir)
        //this filter already exists in specs2, if you want to completly ignore files
        .withFilter { f => f.getName != "rudder_promises_generated" }
        //and here, our content matcher which ignores some lines
        .withMatcher(ignoreSomeLinesMatcher(
             """.*rudder_node_config_id" string => .*"""
          :: """.*string => "/.*/configuration-repository.*"""
          :: Nil
        ))
   }
 }
}