rdsr commented on a change in pull request #220: Support parsing of special
characters in TypeToSchema visitor
URL: https://github.com/apache/incubator-iceberg/pull/220#discussion_r294068511
##########
File path: core/src/main/java/org/apache/iceberg/avro/TypeToSchema.java
##########
@@ -214,4 +216,58 @@ public Schema primitive(Type.PrimitiveType primitive) {
return primitiveSchema;
}
+
+ private static String sanitize(String name) {
+ if (validAvroName(name)) {
+ return name;
+ }
+ int length = name.length();
+ StringBuilder sb = new StringBuilder(name.length());
+ char first = name.charAt(0);
+ if (!(Character.isLetter(first) || first == '_')) {
+ sb.append(sanitizedValue(first));
+ } else {
+ sb.append(first);
+ }
+ for (int i = 1; i < length; i++) {
+ char character = name.charAt(i);
+ if (!(Character.isLetterOrDigit(character) || character == '_')) {
+ sb.append(sanitizedValue(character));
+ } else {
+ sb.append(character);
+ }
+ }
+ return sb.toString();
+ }
+
+ private static final Map<Character, String> SANITIZE_MAP = new ImmutableMap
+ .Builder<Character, String>()
+ .put('.', "__DOT__").put('#', "__HASH__")
Review comment:
Because of this change
`icebergSchema` will not always be equal to
`toIcebergSchema(toAvro(icebergSchema))` when special chars exists in Iceberg
schema name.
Do you think we need to reverse the mapping when converting Avro to
IcebergSchema e.g `__HASH__` converted to `#` etc. so the original names in the
Iceberg schema are preserved?
----------------------------------------------------------------
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]
