gists/config/visidata/README.md

# VisiData Config + Plugins

This folder contains a VisiData `config.py` (symlinked from `visidatarc`) plus a small set of local plugins under `plugins/`.

## Install

The installer links (or copies) the config and plugins into VisiData’s per-user directory.

```bash
./install.sh --link        # default, symlinks into place
./install.sh --copy        # copies into place
./install.sh --deps        # installs optional Python deps into $VD_DIR/plugins-deps
```

On VisiData 3.3, `$VD_DIR` defaults to:
- macOS: `~/Library/Preferences/visidata`
- Linux: `${XDG_CONFIG_HOME:-~/.config}/visidata`

## Plugins

Plugins are installed into `$VD_DIR/plugins/` and imported via the top-level `plugins` package.

## Showcase Demo

This repo includes a self-contained sample dataset + command log to demonstrate the local IOC/IP features:

- `showcase_ioc.tsv` (sample IOC rows)
- `showcase_ioc.vdj` (replay file)

Run it interactively from this repo root:

```bash
vd --visidata-dir "$PWD" --config "$PWD/visidatarc" --play showcase_ioc.vdj
```

What it showcases:
- custom types: `IP`, `Domain`, `URL`, `Hash`
- IP membership expressions: `src_ip * network`
- URL parsing fields: `url.host`, `url.parts.path`, `url.domain`
- hash classification: `file_hash.kind`
- IP lookups: `src_ip.ipinfo.*`, `src_ip.asn.*`, `src_ip.geo.*`, `src_ip.country()`
- provider visibility: `src_ip.geo.source`, `src_ip.asn.source`, `domain.dns.source`
- domain/network intel: `domain.dns.*`, `domain.rdap.*`
- hash intel: `file_hash.mb.*` (MalwareBazaar)
- VirusTotal lookups: `src_ip.vt.*`, `file_hash.vt.*`, `domain.vt.*`, `url.vt.*`
- local plugin command: `tke-hidecol`

Lookup notes:
- VT columns require `options.tke_vt_api_key` (or `VT_API_KEY` / `VIRUSTOTAL_API_KEY` / `~/.virustotal_api_key`).
- IPInfo/ASN/Geo columns use free providers and may be rate-limited; `options.tke_ipinfo_token` improves reliability.
- To keep replays practical with strict throttling, some heavy lookup columns are intentionally limited to a subset of rows.

### `plugins/hidecol.py`

Adds a command to hide columns that are empty or constant across all rows.

- Command: `tke-hidecol`
- Menu: `Column -> Hide -> empty/superfluous columns`

### `plugins/iptype.py`

Adds a custom IP datatype that supports:
- IPv4 + IPv6 addresses
- CIDR networks (e.g. `192.168.7.0/24`)
- Correct sorting (numeric, by version)
- Membership test operator: `ip * net` (and `net * ip`)
- Normalized lookup/enrichment properties, accessible as attributes in expressions

#### Type + Command

- Type converter: `ip(...)`
- Type name: `IP`
- Command: `type-ip` (sets `cursorCol.type=ip`)

#### Operations

Membership test:
- `ipcol * "192.168.7.0/24"` -> `True`/`False`
- `"192.168.7.0/24" * ipcol` -> `True`/`False`

#### Attributes (on `IP` typed cells)

Lookup objects expose both normalized fields and raw response data:

- `ipcol.ipinfo.country`
- `ipcol.ipinfo.data.<any_json_field>`
- `ipcol.asn.asn`, `ipcol.asn.name`, `ipcol.asn.country`
- `ipcol.asn.data.<any_json_field>`
- `ipcol.vt.verdict` (e.g. `"3/94"`), `ipcol.vt.malicious`, `ipcol.vt.total`, `ipcol.vt.category` (alias: `ipcol.vt.type`)
- `ipcol.vt.data.<any_json_field>`
- `ipcol.geo.*` (best-available geo: prefers MaxMind mmdb, else free HTTP providers)
- `ipcol.maxmind.*` (offline-only MaxMind lookup; empty if no mmdb)

#### Caching

All lookup providers cache results in a local sqlite+pickle DB (default `~/.visidata_cache.db`).

#### Lookup Providers + Keys

Options (set in `config.py` / `visidatarc`):
- `options.tke_cache_db_path="~/.visidata_cache.db"`
- `options.tke_lookup_cache_ttl=86400`
- `options.tke_lookup_error_ttl=300`
- `options.tke_lookup_timeout=10`
- `options.tke_ipinfo_token="..."` (optional; ipinfo can work without it)
- `options.tke_ipapi_key="..."` (optional)
- `options.tke_vt_api_key="..."` (required for VT lookups unless using `~/.virustotal_api_key`)
- `options.tke_maxmind_mmdb_path="/path/to/GeoLite2-City.mmdb"` (optional)

Env var equivalents:
- `IPINFO_TOKEN`, `IPAPI_KEY`
- `VT_API_KEY` or `VIRUSTOTAL_API_KEY` (also supports `~/.virustotal_api_key`)
- `MAXMIND_MMDB_PATH` or `GEOIP_MMDB_PATH`

MaxMind (offline “free” GeoLite2) support:
- Place a `GeoLite2-City.mmdb` / `GeoLite2-Country.mmdb` file in `$VD_DIR/`, or set `options.tke_maxmind_mmdb_path`.

### `plugins/iplib.py`

Pure-Python library used by `iptype.py` for:
- Normalized info classes (`IPInfo`, `ASNInfo`, `VTInfo`, `GeoInfo`)
- `JSONNode` wrapper (`.data.<field>`) for safe attribute-style access into raw dict/list JSON
- Parsing/normalization helpers for each provider’s response shape

This file intentionally does **not** import VisiData so it can be validated outside the VisiData runtime.

## Config: `visidatarc`

This repo’s `visidatarc` is intended to be installed as VisiData’s `config.py`:
- `$VD_DIR/config.py` (VisiData 3.3 default)
- and also `~/.visidatarc` as a legacy fallback

It currently contains:
- display/date format options
- a sqlite+pickle caching decorator and a set of general-purpose helpers (aggregators, timestamp parsing, “dirty” JSON parsing, etc)

### Are the `visidatarc` functions superseded?

Partially:
- The **IP-centric lookups and normalized attribute access** are now primarily handled by `plugins/iptype.py` on typed values (e.g. `ipcol.geo.country_code`).
- Many other helpers in `visidatarc` (aggregators like `avgdiff`, parsing/time conversion helpers, etc.) are still independent and useful.

### Keeping old + new side-by-side (without duplicating code)

Yes. The cleanest pattern in VisiData is:
1. Put shared logic into a module under `plugins/` (so it’s on `sys.path` via `$VD_DIR`).
2. In `visidatarc`, import and expose thin wrappers (or just import the module and use `module.func(...)` in expressions).

Concretely:
- `plugins/iplib.py` already holds parsing/normalization shared by the IP type.
- If you have legacy functions in `visidatarc` that overlap with the new IP lookups, refactor those functions into a shared module (e.g. `plugins/lookups.py`) and have both `visidatarc` and `plugins/iptype.py` call into it.

This keeps backward-compatible names available while ensuring caching/auth/provider behavior is implemented in one place.