Home

Zipper overview

Zipper is a .NET 10 command-line tool that generates eDiscovery load files and production sets. It creates compressed ZIP archives containing placeholder native files (PDF, TIFF, EML, DOCX, XLSX, JPG) alongside corresponding load files in DAT, OPT, CSV, and EDRM-XML formats. Legal teams and eDiscovery vendors use it to produce test data that matches the structure and size of real production datasets.

The tool operates in three generation modes: Standard (default, produces a ZIP archive with load files), Loadfile-Only (generates load files directly to disk with optional chaos anomalies), and Production Set (creates a full directory tree with NATIVES/IMAGES/DATA/TEXT subdirectories and per-volume load files).

Key characteristics:

• Supports up to 100 million files per run with O(1) per-file distribution algorithms
• Channel-based producer-consumer pipeline for parallel file generation and archive writing
• Deterministic output via seeded random number generation
• Custom Roslyn analyzer (FgrFlatAccessAnalyzer) enforces architectural boundaries
• 80% test coverage gate enforced in CI

Quick links

• Architecture - system diagrams and component map
• Getting started - build, test, and run instructions
• Glossary - eDiscovery domain terms used throughout the codebase
• Three generation modes - Standard, Loadfile-Only, Production Set
• Load file pipeline - the composer/serializer/emitter seam

---

Generated by Factory