Repository navigation
Support ChunkedDecoder to read big xml data chunked #132
Description
The latest release of kotlin serialization 1.5.0 added a new interace called ChunkedDecoder which support decoding large data with the help of okio in smaller chunks. This prevents OutOfMemory Exceptions and some performance improvements.
Kotlin/kotlinx.serialization#2012
At the moment it is only supported by the JsonDecoder. Are there any plans to support this also in your XML Serialization Library?
Activity
pdvrieze commentedon Apr 6, 2023
There is "partial" support in the dev branch. This is partial in that it is supported by serialization but not supported in the parsers (yet). This means that the parser will provide whatever string/text size it does, and the serializer provides it in chunks. There is an implication in the parsing that strings are given contiguous, but this is not a hard requirement. If they chunk then chunked parsing will support that. I haven't put such functionality in the platform independent parser yet (and for DOM it makes no sense at all).
slavonnet commentedon Sep 13, 2023
kotlin have buffered streams . its sane but better )
pdvrieze commentedon May 15, 2025
At this point changing the parser to support a push model on text reading (needed for chunked text at that level) is not a priority. Only the existing partial solution is kept.