shenodaguirguis commented on a change in pull request #911: Support for types
in Samza SQL UDF
URL: https://github.com/apache/samza/pull/911#discussion_r254915133
##########
File path:
samza-sql/src/main/java/org/apache/samza/sql/data/SamzaSqlExecutionContext.java
##########
@@ -52,16 +54,19 @@ private SamzaSqlExecutionContext(SamzaSqlExecutionContext
other) {
public SamzaSqlExecutionContext(SamzaSqlApplicationConfig config) {
this.sqlConfig = config;
- udfMetadata =
-
this.sqlConfig.getUdfMetadata().stream().collect(Collectors.toMap(UdfMetadata::getName,
Function.identity()));
+ udfMetadata = new HashMap<>();
+ for(UdfMetadata udf : this.sqlConfig.getUdfMetadata()) {
+ udfMetadata.putIfAbsent(udf.getName(), new ArrayList<>());
+ udfMetadata.get(udf.getName()).add(udf);
+ }
}
public ScalarUdf getOrCreateUdf(String clazz, String udfName) {
return udfInstances.computeIfAbsent(udfName, s -> createInstance(clazz,
udfName));
}
public ScalarUdf createInstance(String clazz, String udfName) {
- Config udfConfig = udfMetadata.get(udfName).getUdfConfig();
+ Config udfConfig = udfMetadata.get(udfName).get(0).getUdfConfig();
Review comment:
I see. Thanks for the clarification & the comment. Don't we still need to
guard code against erroneously empty list?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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
