Skip to content

RION Ops for Java is a toolkit for reading and writing RION encoded data. RION is a compact, fast binary data format.

Notifications You must be signed in to change notification settings

nanosai/rion-ops-java

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RION Ops for Java

Introduction

RION Ops for Java is a toolkit for reading and writing the compact, fast, binary data format RION.

RION Ops Tutorial

This README file only contains an introduction to what RION and RION Ops are. To learn how to use RION Ops, see the RION Ops for Java Tutorial .

RION - Raw Internet Object Notation

RION is short for Raw Internet Object Notation. By "Raw" we mean "in its raw form - as encoded in bytes". RION itself will not be described here. Here is the RION Documentation

What We Mean by "Object" in RION.

The term "Object" in "Raw Internet Object Notation" has a slightly different meaning from what it means in JSON. By "object" we mean "any individual piece of data that it makes sense to represent / read / write / exchange individually". According to that interpretation, a single integer field is an object. A Key + value pair can be interpreted as two objects even though they logically represent a composite object (key + value). A set of key + value pairs can be considered separate, small objects even though they logically represent one big object with properties with each property consisting of a key + value pair.

As you can see in the RION field list below, RION has a special type of field called an "Object" which can contain nested RION fields. The "Object" field type is a commonly used RION field type, but it is not this field type we refer to in the name RION. It is totally fine for a RION file or stream to contain multiple, separate non-"Object" RION fields. This is different from JSON, where a JSON file can only contain either a full JSON object or array. This is not so for RION. A RION file could contain a single integer RION field, or multiple fields, including an "Object" field if that makes sense for the given use case.

RION Field Types

RION contains the following field types:

  • Bytes
  • Boolean
  • Integer, positive
  • Integer, negative
  • Float
  • UTF-8
  • UTF-8 Short
  • UTC date time
  • Reference (new)
  • Array (*)
  • Table
  • Object
  • Key
  • Key Short
  • Extended

The Array, Table and Object field types are composite field types which can contain other RION fields inside them. Actually, this is also possible with the Bytes field type - although it is intended to contain raw bytes (e.g. a file, audio, video etc.). The rest of the field types are simple field types which contain a single data value.

Since you can model an Array as a single column Table we are currently considering whether the Array field is really necessary, or if it superfluous.

TION - Textual Internet Object Notation

We are planning to create a textual version of RION, which we will call TION - Textual Internet Object Notation. RION is thus the raw form of Nanosai ION (RION), whereas TION is the textual form. The purpose of TION is to make it easier to read and write RION in a text editor. Simply convert RION to TION and open it in a text editor. You can then edit the TION data, save it, and convert it back to RION.

RION vs. ION

RION was first released under the name ION, but since its release Amazon has released its own binary data format named ION. To avoid confusiong we are currently using the term "Raw Internet Object Notation" (RION) to clearly distinguish it from Amazon ION. Actually, Nanosai ION and Amazon ION are similar in structure, but we believe Nanosai ION has a few advantages over Amazon ION, especially when working with the data in its raw form (binary form).

Note: RION is our current "working title", but we may change the name in the future. However, the most important is not the name of the format, but it's capabilities!

Maven Dependency

If you want to use RION Ops with Maven, the Maven dependency for RION Ops looks like this:

<dependency>
    <groupId>com.nanosai</groupId>
    <artifactId>rion-ops</artifactId>
    <version>0.11.0</version>
</dependency>

Remember to substitute the version with the version of RION Ops you want to use. See the RION Ops version history in the next section.

Version History

Version Java Version Change
0.11.0 Java 8+ Second round of cyclic graph support - now inside Table fields too. A few bug fixes and minor enhancements too.
0.10.0 Java 8+ First implementation of the RION Reference field - used in object graph serialization / deserialization (issue #11). A few bug fixes of object read / write functionality.
0.9.1 Java 8+ More null support, related bug fixes, improved tests and code cleanup (issue #14).
0.9.0 Java 8+ Improved support for writing null fields in RionWriter. Bug fix of RionWriter.writeUtf8() which serialized empty string as UTF-8-Short null - not UTF-8 with length 0 (issue #14).
0.8.0 Java 8+ RionObjectWriterBuilder and RionObjectReaderBuilder classes added.
0.7.0 Java 8+ RionObjectWriter and RionObjectReader + support classes added.
0.6.0 Java 8+ RionToHexConverter added. Minor enhancements to RionWriter for writing Array and Table fields.
0.5.3 Java 8+ Bug fix of reading UTF-8 fields into Java Strings (issue #3)
0.5.2 Java 8+ First release

About

RION Ops for Java is a toolkit for reading and writing RION encoded data. RION is a compact, fast binary data format.

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages