Add the dependency:

"org.databrary" %% "iteratees" % "0.1"

You can see examples of usage in the test cases.

These functions provide an interface to java.util.zip.ZipOutputStream in the form of an Enumeratee. There are two main functions providing a higher-level and lower-level interface. Both follow proper reactive semantics, never blocking (except possibly on IO). However, errors may still be delivered by thrown exceptions.

Also provided are some classes that provide more convenient ways of creating ZipEntries: DirEntry, StoredEntry and DeflatedEntry.

The lower-level interface, org.databrary.ZipFile.zip, provides an Enumeratee[Either[ZipEntry, Array[Byte]], Array[Byte]] that converts an incoming sequence of Left(ZipEntry), each followed by any number of Right(data) blocks, into a zip file. This is equivalent to calling this sequence of ZipOutputStream.setNextEntry and ZipOutputStream.write on each item, so all the restrictions of those interfaces still apply.

The higher-level interface, org.databrary.ZipFile.flatZip, provides a more convenient interface for creating zip streams out of different files or data sources. It provides an Enumeratee[ZipFile.StreamEntry, Array[Byte]], where a StreamEntry is a ZipEntry amended with an Enumerator[Array[Byte]].

Two easy ways of creating StreamEntries are provided: ZipFile.DirEntry (which simply has no data content), and ZipFile.fileEntry which generates a StreamEntry from a file on disk. You can also use ZipFile.DeflatedStreamEntry or define your own.

Similar to a composition Enumeratee.take and Enumeratee.drop, org.databrary.Enumeratee.range is designed to work on Array[Byte] streams and counts individual bytes, allowing you to select an interior range of a byte stream.

Used in the construction of zip files, org.databrary.Enumeratee.filterOutputStream allows you to wrap a java.io.FilterOutputStream in an Enumartee efficiently. Unlike Enumerator.outputStream, this follows correct Enumeratee semantics, so will push back and only accept input when more is needed.