andmeformaate

Andmeformaate are the conventions used to encode data for storage and transfer. They specify how information is structured, typed, and serialized so software can reliably read and write it across systems and over time. The term covers both human-readable formats and compact binary representations, as well as schemas and validation rules that accompany data.

Common examples of text-based formats include JSON, XML, CSV, and YAML; newline-delimited formats such as NDJSON; and binary formats such as Protocol Buffers, Avro, Parquet, ORC, and MessagePack. Some formats are self-describing (for example JSON, XML) while others rely on separate schemas (for example Avro, Parquet) to express data types and structure.

Key characteristics include readability, schema support, immutability and forward/backward compatibility, compression, performance for parsing and serialization, and suitability for streaming or random access. The choice of format often reflects a balance between human accessibility and machine efficiency, as well as requirements for schema evolution and data integrity.

Use cases include data interchange between services, configuration files, logs, API payloads, and storage in databases and data lakes. The choice of format affects interoperability, evolution of data structures, and processing efficiency in data pipelines and software applications. When selecting a format, practitioners weigh readability, schema evolution, data volume, tooling, and ecosystem support; they often combine formats in a system—for configuration (YAML), API messages (JSON), and storage (Parquet).