I would like to begin by saying that this internship has definitely been one of the top highlights of my Ph.D. journey. I spent most of my working hours implementing the idea of server-side virtual data processing. This means that data on the THREDDS Data Servers (TDS) could be virtually processed without actually modifying the data. As such, the data integrity would remain intact, but it could be optimized for ML/AI.
THREDDS development is now switching to version 5.0 which will require Java 8. Version 5 is a major upgrade and some of the APIs will change. Deprecated classes will be moved to a legacy jar and will not be supported. If you are a developer, you will need to test the new version against your code. We expect to have an alpha release out by July for that purpose.
A SAX (Simple API for XML) parser is a particular mechanism for parsing XML documents. Using a SAX parser has the advantage over the DOM-based parser in that it is not necessary to build the explicit DOM tree. On the other hand, it can be difficult to build a SAX parser because it requires management of complex state.
Combining SAX parsing with a GNU Bison generated parser is appealing because it allows the Bison parser to manage all of the state. Additionally, the .y file encapsulates the equivalent of a DTD but in a much more readable form. The combination makes using SAX parsing a lot simpler.
