chupaty opened a new issue, #371:
URL: https://github.com/apache/avro-rs/issues/371
We currently have two methods to write 'raw' avro datums:
- write_avro_datum
- write_avro_datum_schemata (where the target schema depends on some other
schemata)
The second function does not scale well with the size of the schemata, as it
creates a new ResolvedSchema on every call. In my environment, where we have
~100 schemata, this creation takes ~90% of the encoding time.
We can address this with a new function (eg
fn write_avro_datum_resolved_schemata<T: Into<Value>>(
schema: &Schema,resolved_schema: &ResolvedSchema,value: T,buffer: &mut
Vec<u8>)
Here the user would be responsible for creating the ResolvedSchema from the
schemata (which may only need to happen once, or a small number of times).
Alternatively we can update the existing write_avro_datum_schemata function,
with the upside being that we have an arguably superior interface and the
downside being that we break existing users.
Is there interest for a change like this?
Thanks
Simon
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
