Better relation embedder cli (#2788)

* start creating CLIMain

* make a local CLIMain work

* enable CLI to read from live database

* make pipeline data an env var

* make pipeline date an argument

* Apply auto-formatting rules

* remove debugging code

* de-akkafy reading from stdin

* better handling of dodgy input

* try removing internal model from batcher

* improve StdInNDJSON interface

* improve commentary

* report parse failure

---------

Co-authored-by: Github on behalf of Wellcome Collection <wellcomedigitalplatform@wellcome.ac.uk>

pipeline/relation_embedder/relation_embedder/src/main/scala/weco/pipeline/relation_embedder/lib/StdInBatches.scala

@@ -1,15 +1,41 @@
 package weco.pipeline.relation_embedder.lib
+import grizzled.slf4j.Logging
 import weco.json.JsonUtil._
 import weco.pipeline.relation_embedder.models.Batch
 
 import scala.io.Source.stdin
+import scala.util.{Failure, Success, Try}
 
-trait StdInBatches {
+/** Trait to deal with Newline Delimited JSON being provided on STDIN.
+  *
+  * Each JSON object in the input is transformed to an instance of T, according
+  * to jsonToInstance (provided by the extending class) and used to populate the
+  * instances Iterator.
+  */
+
+trait StdInNDJSON[T] extends Logging {
+  protected def jsonToInstance(str: String): Try[T]
   private val stdInStrings: Iterator[String] = stdin.getLines()
 
-  private def toBatch(jsonString: String) =
-    fromJson[Batch](jsonString).get
+  private def toInstance(jsonString: String): Option[T] =
+    jsonToInstance(jsonString) match {
+      case Failure(exception) =>
+        error(exception.getMessage)
+        None
+      case Success(value) => Some(value)
+    }
+
+  protected val instances: Iterator[T] =
+    stdInStrings
+      .flatMap(
+        toInstance
+      )
+
+}
+
+trait StdInBatches extends StdInNDJSON[Batch] {
+  def jsonToInstance(jsonString: String): Try[Batch] =
+    fromJson[Batch](jsonString)
 
-  protected val batches: Iterator[Batch] =
-    stdInStrings.map(toBatch)
+  protected val batches: Iterator[Batch] = instances
 }