Idea

• Provide access to JSON data via JSONPath like queries (like XMLBeam does)
• Speed up JSONPath evaluation by evaluating all queries at once during document parsing
• Avoid creation of DOM like structures, use indexes instead

Current state: version 0.0.2

• Highly experimental
• I'm still evaluating different index data structures (tries & hash based indexes)
• There are a lot of not yet implemented corner cases
• Type conversion is not implemented yet

Design goals:

JSON

• support deep data structures by using a nonrecursive parser
• small memory foodprint for large files by creating a compressed index instead of an object graph
• high performance by avoiding object allocation and buffer copies when possible
• support for compact json variants (nonquoted keys & values)

Data Projection

• read support by using a variant of JSONPath
• write support for deletes & updates
• write support for JSON structures

Read & write operations

• try to minimise output changes to get a clean diff

Java

• Java 8 and higher are supported

Dual License

• GPL V3 for open source projects
• Comercial license available later (version 1.0.0 and higher)
• Backports for pre Java 8 available under comercial licence later (version 1.0.0 and higher)