Support stripping by blob Id on the command line

Helps a little with issue #12. People can get a list of blob-ids using
"git rev-list --all --objects", then grep to list all files in
directories they want to nuke, and pass that to the BFG, as noted by
@Fryguy:

#12 (comment)

src/main/scala/com/madgag/git/bfg/GitUtil.scala

@@ -35,10 +35,6 @@ import com.madgag.git.bfg.cleaner._
 import org.eclipse.jgit.api.Git
 import org.eclipse.jgit.treewalk.filter.{AndTreeFilter, TreeFilter}
 
-object ObjectId {
-  def apply(str: String) = org.eclipse.jgit.lib.ObjectId.fromString(str)
-}
-
 trait CleaningMapper[V] extends Cleaner[V] {
   def isDirty(v: V) = apply(v) != v
 
@@ -82,6 +78,10 @@ object GitUtil {
     }
   }
 
+  implicit class RichString(str: String) {
+    def asObjectId = org.eclipse.jgit.lib.ObjectId.fromString(str)
+  }
+
   implicit class RichRevTree(revTree: RevTree) {
     def walk(postOrderTraversal: Boolean = false)(implicit reader: ObjectReader) = {
       val treeWalk = new TreeWalk(reader)

src/main/scala/com/madgag/git/bfg/cli/CLIConfig.scala

@@ -21,7 +21,6 @@
 package com.madgag.git.bfg.cli
 
 import java.io.File
-import com.madgag.git.bfg.GitUtil._
 import com.madgag.git.bfg.cleaner._
 import com.madgag.git.bfg.cleaner.TreeBlobsCleaner.Kit
 import com.madgag.git.bfg.textmatching.RegexReplacer._
@@ -34,10 +33,11 @@ import org.eclipse.jgit.storage.file.FileRepository
 import protection.ObjectProtection
 import scopt.immutable.OptionParser
 import scala.Some
-import com.madgag.git.bfg.GitUtil.SizedObject
+import com.madgag.git.bfg.GitUtil._
 import com.madgag.git.bfg.model.{FileName, TreeBlobEntry}
 import com.madgag.git.bfg.textmatching.{Glob, Literal, TextMatcher}
 import com.madgag.inclusion._
+import org.eclipse.jgit.lib.ObjectId
 
 
 object CLIConfig {
@@ -49,6 +49,9 @@ object CLIConfig {
       intOpt("B", "strip-biggest-blobs", "NUM", "strip the top NUM biggest blobs") {
         (v: Int, c: CLIConfig) => c.copy(stripBiggestBlobs = Some(v))
       },
+      opt("bi", "strip-blobs-with-ids", "<blob-ids-file>", "strip blobs with the specified Git object ids") {
+        (v: String, c: CLIConfig) => c.copy(stripBlobsWithIds = Some(Source.fromFile(v).getLines().map(_.trim).filterNot(_.isEmpty).map(_.asObjectId).toSet))
+      },
       opt("D", "delete-files", "<glob>", "delete files with the specified names (eg '*.class', '*.{txt,log}' - matches on file name, not path within repo)") {
         (v: String, c: CLIConfig) => c.copy(deleteFiles = Some(FileMatcher(v)))
       },
@@ -99,6 +102,7 @@ case class CLIConfig(stripBiggestBlobs: Option[Int] = None,
                      filenameFilters: Seq[Filter[String]] = Nil,
                      filterSizeThreshold: Int = BlobTextModifier.DefaultSizeThreshold,
                      replaceBannedStrings: Traversable[String] = List.empty,
+                     stripBlobsWithIds: Option[Set[ObjectId]] = None,
                      blobCharsetDetector: BlobCharsetDetector = QuickBlobCharsetDetector,
                      strictObjectChecking: Boolean = false,
                      sensitiveData: Option[Boolean] = None,
@@ -138,6 +142,8 @@ case class CLIConfig(stripBiggestBlobs: Option[Int] = None,
       }
   }
 
+  lazy val blobsByIdRemover = stripBlobsWithIds.map(new BlobRemover(_))
+
   lazy val blobRemover = {
     implicit val progressMonitor = new TextProgressMonitor()
 
@@ -171,7 +177,7 @@ case class CLIConfig(stripBiggestBlobs: Option[Int] = None,
 
   lazy val commitNodeCleaners = Seq(new CommitMessageObjectIdsUpdater(objectIdSubstitutor)) ++ formerCommitFooter
 
-  lazy val treeBlobCleaners = Seq(blobRemover, fileDeletion, blobTextModifier).flatten
+  lazy val treeBlobCleaners = Seq(blobsByIdRemover, blobRemover, fileDeletion, blobTextModifier).flatten
 
   lazy val definesNoWork = treeBlobCleaners.isEmpty
 

src/test/scala/com/madgag/git/bfg/cli/MainSpec.scala

@@ -24,9 +24,10 @@ import com.madgag.git.bfg._
 import com.madgag.git.bfg.GitUtil._
 import scala.collection.convert.wrapAsScala._
 import org.specs2.mutable._
-import org.eclipse.jgit.lib.Repository
+import org.eclipse.jgit.lib.{ObjectId, Repository}
 import org.eclipse.jgit.revwalk.RevCommit
 import org.specs2.matcher.Matcher
+import scalax.file.Path
 
 class MainSpec extends Specification {
   "CLI" should {
@@ -59,7 +60,29 @@ class MainSpec extends Specification {
 
       commitHist must (not(haveFolder("secret-files"))).forall
     }
+
+    "strip blobs by id" in {
+      implicit val repo = unpackRepo("/sample-repos/example.git.zip")
+      implicit val (revWalk, reader) = repo.singleThreadedReaderTuple
+
+      def haveCommitWhereObjectIds(boom: Matcher[Traversable[ObjectId]]): Matcher[RevCommit] = boom ^^ {
+          (c: RevCommit) => c.getTree.walk().map(_.getObjectId(0)).toList
+        }
+
+      val badBlobs = Set(abbrId("db59"),abbrId("86f9"))
+
+      val blobIdsFile = Path.createTempFile()
+      blobIdsFile.writeStrings(badBlobs.map(_.name()),"\n")
+
+      commitHist must haveCommitWhereObjectIds(containAllOf(badBlobs.toSeq)).atLeastOnce
+
+      run(s"--strip-blobs-with-ids ${blobIdsFile.path}")
+
+      commitHist must (not(haveCommitWhereObjectIds(containAnyOf(badBlobs.toSeq)))).forall
+    }
   }
 
+
+
   def commitHist(implicit repo: Repository) = repo.git.log.all.call.toSeq.reverse
 }

src/test/scala/com/madgag/git/bfg/package.scala

@@ -52,5 +52,5 @@ package object bfg {
   }
 
   def commitThatWasFormerly(id: ObjectId): RevCommit => Boolean =
-    _.getFooterLines.exists(f => f.getKey == FormerCommitFooter.Key && ObjectId(f.getValue) == id)
+    _.getFooterLines.exists(f => f.getKey == FormerCommitFooter.Key && f.getValue.asObjectId == id)
 }