# CLI Reference – `kdc`

The `kdc` command-line tool provides utilities for managing the Khmer Document Corpus.

## Installation

```bash
pip install -e .
```

## Commands

### `kdc ingest`

Scan a PDF directory and create JSON metadata stubs.

```
kdc ingest [OPTIONS]

Options:
  --pdf-dir   PATH   Source PDF directory      [default: pdf]
  --meta-dir  PATH   Output metadata directory [default: metadata]
  --overwrite        Overwrite existing stubs  [default: False]
```

### `kdc preview`

Render PNG preview images from PDFs.

```
kdc preview [OPTIONS]

Options:
  --pdf-dir         PATH   PDF directory        [default: pdf]
  --out-dir         PATH   Preview output dir   [default: preview]
  --dpi             INT    Render DPI           [default: 150]
  --first-page-only        Render only page 1  [default: True]
```

### `kdc validate`

Validate metadata JSON files against the schema.

```
kdc validate [OPTIONS]

Options:
  --meta-dir  PATH   Metadata directory   [default: metadata]
```

### `kdc upload`

Upload corpus assets to HuggingFace Hub.

```
kdc upload [OPTIONS]

Options:
  --repo-id     TEXT   HuggingFace repository ID  [required]
  --token       TEXT   HuggingFace API token       [env: HF_TOKEN]
  --pdf-dir     PATH   PDF directory               [default: pdf]
  --meta-dir    PATH   Metadata directory          [default: metadata]
  --preview-dir PATH   Preview directory           [default: preview]
```

### `kdc stats`

Print corpus statistics.

```
kdc stats [OPTIONS]

Options:
  --meta-dir  PATH   Metadata directory   [default: metadata]
```