lordcheng10 commented on code in PR #8095:
URL: https://github.com/apache/inlong/pull/8095#discussion_r1208768400
##########
inlong-sort/sort-flink/sort-flink-v1.13/sort-connectors/iceberg/src/main/java/org/apache/inlong/sort/iceberg/sink/multiple/SchemaChangeUtils.java:
##########
@@ -51,49 +53,81 @@ public class SchemaChangeUtils {
static List<TableChange> diffSchema(Schema oldSchema, Schema newSchema) {
List<String> oldFields =
oldSchema.columns().stream().map(NestedField::name).collect(Collectors.toList());
List<String> newFields =
newSchema.columns().stream().map(NestedField::name).collect(Collectors.toList());
- int oi = 0;
- int ni = 0;
+ Set<String> oldFieldSet = new HashSet<>(oldFields);
+ Set<String> newFieldSet = new HashSet<>(newFields);
+
+ Set<String> intersectColSet = Sets.intersection(oldFieldSet,
newFieldSet);
+ Set<String> colsToDelete = Sets.difference(oldFieldSet, newFieldSet);
+ Set<String> colsToAdd = Sets.difference(newFieldSet, oldFieldSet);
+
List<TableChange> tableChanges = new ArrayList<>();
- while (ni < newFields.size()) {
- if (oi < oldFields.size() &&
oldFields.get(oi).equals(newFields.get(ni))) {
- oi++;
- ni++;
- } else {
- NestedField newField = newSchema.findField(newFields.get(ni));
+
+ // step0: Unknown change
+ if (!colsToDelete.isEmpty() && !colsToAdd.isEmpty()) {
Review Comment:
> example first [a, b, c] -> delete c [a, b] -> add d [a, b, d]
If two messages are received one after another: (a1,b1,c1)->(a1,b1,d1),
through the diff schema, you can get an add(d) and delete(c), so after
executing add and delete , becomes [a, b, d] ?
##########
inlong-sort/sort-flink/sort-flink-v1.13/sort-connectors/iceberg/src/main/java/org/apache/inlong/sort/iceberg/sink/multiple/SchemaChangeUtils.java:
##########
@@ -51,49 +53,81 @@ public class SchemaChangeUtils {
static List<TableChange> diffSchema(Schema oldSchema, Schema newSchema) {
List<String> oldFields =
oldSchema.columns().stream().map(NestedField::name).collect(Collectors.toList());
List<String> newFields =
newSchema.columns().stream().map(NestedField::name).collect(Collectors.toList());
- int oi = 0;
- int ni = 0;
+ Set<String> oldFieldSet = new HashSet<>(oldFields);
+ Set<String> newFieldSet = new HashSet<>(newFields);
+
+ Set<String> intersectColSet = Sets.intersection(oldFieldSet,
newFieldSet);
+ Set<String> colsToDelete = Sets.difference(oldFieldSet, newFieldSet);
+ Set<String> colsToAdd = Sets.difference(newFieldSet, oldFieldSet);
+
List<TableChange> tableChanges = new ArrayList<>();
- while (ni < newFields.size()) {
- if (oi < oldFields.size() &&
oldFields.get(oi).equals(newFields.get(ni))) {
- oi++;
- ni++;
- } else {
- NestedField newField = newSchema.findField(newFields.get(ni));
+
+ // step0: Unknown change
+ if (!colsToDelete.isEmpty() && !colsToAdd.isEmpty()) {
Review Comment:
> example first [a, b, c] -> delete c [a, b] -> add d [a, b, d]
@hejiay
If two messages are received one after another: (a1,b1,c1)->(a1,b1,d1),
through the diff schema, you can get an add(d) and delete(c), so after
executing add and delete , becomes [a, b, d] ?
--
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.
To unsubscribe, e-mail: commits-unsubscr...@inlong.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
