# Data directory

`entities.csv` is the canonical round-1 ecosystem address book. `ip-addresses.csv` begins the address-level map and deliberately separates registration, routing, operation, and service.

## Columns

| Column | Meaning |
|---|---|
| `id` | Stable lowercase identifier |
| `name` | Published entity or system name |
| `category` | Controlled taxonomy value |
| `scope` | `global`, `regional`, or `regional-community` |
| `region` | Service/community region, not merely headquarters |
| `homepage` | Official or canonical public homepage |
| `directory_url` | Best public contact, membership, data, or discovery entry point |
| `role` | Concise functional description |
| `source_url` | Primary evidence for the role |
| `status` | Evidence status defined in the method document |
| `verified_at` | Date checked, UTC (`YYYY-MM-DD`) |
| `notes` | Important caveat; blank when none |

CSV was chosen for round 1 because it is inspectable in Git, opens in spreadsheets, and requires no custom parser. If nested relationships become central, introduce versioned YAML or JSON without deleting CSV history.

The IP-specific columns are documented in [`docs/04-ip-address-model.md`](../docs/04-ip-address-model.md).

## Authoritative IANA IPv4 datasets

- `iana-ipv4-top-level-ranges.csv` normalizes all 256 records from the IANA IPv4 Address Space Registry. Together they cover the complete IPv4 address space at `/8` granularity.
- `iana-ipv4-special-ranges.csv` normalizes and expands all IANA IPv4 Special-Purpose Address Registry entries, including source/destination, forwarding, global-reachability, and protocol-reservation properties.

Both datasets include source URL, verification date, and the corresponding Markdown filename. They are generated by [`scripts/sync_iana_ipv4.py`](../scripts/sync_iana_ipv4.py); do not treat top-level registry authority as exact-address ownership or current route operation.