Use shorter (base-36) encoding of outputPath digest

This shortens branchnames and may make them more readable and easier to
announce: "483-s-b-n-t-pil" is easier to pronounce than "82dd7da9dfe" and
contains a more data.

clients/hadoopfs/src/main/java/io/lakefs/committer/DummyOutputCommitter.java

@@ -33,6 +33,7 @@
 import org.slf4j.LoggerFactory;
 
 import java.io.IOException;
+import java.math.BigInteger;
 import java.nio.charset.Charset;
 
 // TODO(ariels): For Hadoop 3, it is enough (and better!) to extend PathOutputCommitter.
@@ -94,9 +95,9 @@ public class DummyOutputCommitter extends FileOutputCommitter {
     protected String pathToBranch(Path p) {
         String path = p.toString();
         // TODO(ariels): Use a more compact encoding (base-36?)
-        String digest = hash.hashString(path, utf8).toString();
+        String digest = new BigInteger(hash.hashString(path, utf8).asBytes()).toString(36);
         String pathPrefix = path.length() > 128 ? path.substring(0, 128) : path;
-        pathPrefix = pathPrefix.replaceAll("[^-_a-zA-Z0-9]", "-")
+        pathPrefix = pathPrefix.replaceAll("[^-_a-zA-Z0-9]", "-");
         return String.format("%s-%s-%s", branchNamePrefix, digest, pathPrefix);
     }