This repository results from the need to implement a more dynamic way to assign a document/item to the user for processing. It uses a pluggable pipeline architecture with scoring steps, capacity adjustment, and constraint-based allocation.

How it works

The engine distributes items across users in a configurable pipeline of steps:

1. Load scoring — each user receives a base score: (Capacity − CurrentLoad) / Capacity. Users with more remaining capacity score higher.
2. Working-hours capacity (optional) — for part-time users, the step recalculates an effective capacity proportional to their working hours and re-scores accordingly.
3. Priority & activity scoring (optional) — adjusts the base score using the user's priority (lower value = higher priority) and recency of their last activity.
4. Fair allocation — distributes items one at a time in round-robin fashion, always picking the highest-scored eligible user. Constraints (e.g. max items per user) are enforced at each round.

The result contains:

• PrimaryUserId — the top-ranked user recommendation.
• AlternativeUserId — a deterministic second-choice fallback.
• Allocation — full map of every user -> number of items assigned.
• TotalAllocated — sum of all allocated items.
• Trace — audit trail showing each scoring step's output per user.

In case you wish to use it in your project, you can install the package from nuget.org or specify what version you want:

Install-Package ItemDistribution -Version x.x.x.x

Content

1. USING
2. CHANGELOG
3. BRANCH-GUIDE