Architecture
High-Level Data Flow
The platform is bidirectional. One FileSpec in the registry drives both directions — it is the single source of truth for parsing, correction, validation, and file generation.
FileSpec has two optional sections: the existing inbound fields/rules, and a new outbound: OutboundConfig block. A spec can be inbound-only, outbound-only, or fully bidirectional — with zero code changes. See Events → File Pipeline for the full design.
Transformation Pipeline (Detail)
Each record flows through five deterministic stages. No stage throws — errors accumulate on the record.
Error Severity Routing
Open/Closed Extension Model
Add a parser or writer by implementing one interface + @Component. Spring discovers it automatically — zero changes to ParserRegistry or TransformationPipeline.
Key Classes
| Class | Package | Role |
|---|---|---|
FileSpec | core.spec.model | Root spec — drives everything |
ParsedRecord | core.spec.model | Universal record flowing through the pipeline |
ParserRegistry | core.spec.registry | Auto-discovers parsers and routes by format |
TransformationPipeline | core.pipeline | Orchestrates the full flow |
CorrectionEngine | core.transformers | Applies the CorrectionRule list |
ValidationEngine | core.validators | Applies the ValidationRule list |
KafkaRecordWriter | core.writers | Writes ParsedRecord → Kafka as JSON |
TransformService | api.service | Bridge: HTTP request → pipeline |
SpecService | api.service | In-memory spec store (replace with JPA for production) |
OutboundConfig | core.spec.model | Outbound section of FileSpec — field mappings, generator format, rules |
EventMapper | core.transformers | Projects ParsedRecord → FileRecord using FieldMapping list |
FileGenerator | core.generators | Interface for file format writers (NACHA, CSV, SWIFT…) |
GeneratorRegistry | core.generators | Auto-discovers @Component FileGenerator beans |
EventToFilePipeline | core.pipeline | Orchestrates map → correct → validate → generate |