Attempt to fix invalid way geometries

Author: mojodna

CHANGELOG.md

@@ -24,6 +24,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Adds `riverbank`, `stream_end`, `dam`, `weir`, `waterfall`, and `pressurised`
   to the list of waterway features
 - Populates `nds` and `members` for deleted elements from the previous version
+- Attempt to fix invalid way geometries by buffering them (this will result in
+  more way geometries for certain datasets)
 
 ### Fixed
 

src/main/scala/vectorpipe/internal/package.scala

@@ -5,6 +5,7 @@ import java.sql.Timestamp
 import org.locationtech.jts.{geom => jts}
 import geotrellis.vector._
 import org.apache.log4j.Logger
+import org.apache.spark.internal.Logging
 import org.apache.spark.sql._
 import org.apache.spark.sql.catalyst.encoders.RowEncoder
 import org.apache.spark.sql.catalyst.expressions.GenericRowWithSchema
@@ -17,14 +18,12 @@ import vectorpipe.functions.asDouble
 import vectorpipe.functions.osm._
 import vectorpipe.relations.{MultiPolygons, Routes}
 
-package object internal {
+package object internal extends Logging {
   val NodeType: Byte = 1
   val WayType: Byte = 2
   val RelationType: Byte = 3
   val MultiPolygonRoles: Seq[String] = Set("", "outer", "inner").toSeq
 
-  @transient lazy val logger: Logger = Logger.getLogger(getClass)
-
   lazy val BareElementSchema = StructType(
     StructField("changeset", LongType, nullable = false) ::
       StructField("id", LongType, nullable = false) ::
@@ -332,6 +331,16 @@ package object internal {
             }
           val geometry = geom match {
             case Some(g) if g.isValid => g
+            case Some(g) if !g.isEmpty =>
+              val buffered = g.buffer(0)
+
+              if (buffered.isValid) {
+                logWarning(s"Invalid way geometry, fixed by buffering: $id@$version ($updated): $g)")
+                buffered
+              } else {
+                logWarning(s"Invalid way geometry: $id@$version ($updated): $g)")
+                null
+              }
             case _ => null
           }
           new GenericRowWithSchema(Array(changeset, id, version, updated, geometry), BareElementSchema): Row