Thanks for the clarification, Doug. Could adding length prefix be considered feasible enough to file a feature request?
2014-10-15 21:15 GMT+04:00 Doug Cutting <[email protected]>: > There is no method to read individual records as binary since they're > not delimited nor length-prefixed. Instead you can get a block of > records as binary and the count of records in the block and pass that > to a deserializer that parses individual records. > > Doug > > On Wed, Oct 15, 2014 at 4:47 AM, Ivan Balashov <[email protected]> wrote: >> Hi, >> >> Imagine a scenario where one reads avro file and distributes >> individual records (aka datums). Suppose the schema is known, and we >> just need to pass records as bytes to another system for further >> processing. >> >> Given the above, is there a way to read individual records as bytes >> without deserialising them first? >> >> As far as the documentation goes it can be easily done with blocks, >> but it doesn't mention records. >> >> Just to be more clear, there is a method which allows to write encoded >> datum to stream >> org.apache.avro.file.DataFileWriter#appendEncoded >> >> Is there a similar method for reading records as bytes? >> >> Thanks,
