Skip to content

Concepts

These pages explain how HIPO works — the on-disk layout, the in-memory model, and the class layering that ties them together. Read them once and the recipes and API pages will make a lot more sense.

Pages

  • Architecture — the layered class hierarchy (reader / writer → record → event → bank → schema) and a step-by-step trace of what happens during a read and a write.
  • Data model — records, events, banks, schemas, and the columnar Structure-of-Arrays layout. Covers schema identification by (group, item), buffer reuse, self-describing files, and event tags.
  • File format — byte-level reference for the file header, record header, event header, and structure entries. Use this when you need to debug a corrupt file or interoperate with a different reader.

Where to go next