srowen commented on a change in pull request #24362: [SPARK-27454][ML][SQL]
Spark image datasource fail when encounter some illegal images
URL: https://github.com/apache/spark/pull/24362#discussion_r275366266
##########
File path: mllib/src/main/scala/org/apache/spark/ml/image/ImageSchema.scala
##########
@@ -133,7 +133,13 @@ object ImageSchema {
*/
private[spark] def decode(origin: String, bytes: Array[Byte]): Option[Row] =
{
- val img = ImageIO.read(new ByteArrayInputStream(bytes))
+ val img = try {
+ ImageIO.read(new ByteArrayInputStream(bytes))
+ } catch {
+ // Catch runtime exception because `ImageIO` may throw unexcepted
`RuntimeException`.
+ // But do not catch the declared `IOException` (regarded as FileSystem
failure)
+ case _: RuntimeException => null
Review comment:
Yeah, at best I would log 1 or N failures and then stop logging. While we
don't have to do it here, indeed it might be puzzling to read 1000 images and
always keep getting 998 or something with no warning?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
