# Kramerius PDF Downloader

Download page images from Digitalniknihovna / Kramerius and assemble them into a PDF.

The script:

- discovers all page UUIDs for a document
- downloads each page image
- embeds the downloaded JPEGs into a PDF
- can resume interrupted downloads by skipping existing images
- supports authenticated downloads with a Kramerius bearer token
- can resolve other `digitalniknihovna.cz/<library>/...` links automatically

## Requirements

Python 3.10 or newer is recommended. The script uses only the Python standard library, so no `pip install` step is needed.

## Basic Usage

```bash
python3 download_pdf.py "https://www.digitalniknihovna.cz/mzk/view/uuid:0abf59c0-794e-11e4-8ce5-005056827e52?page=uuid:e57e2820-8547-11e4-a321-5ef3fc9bb22f" \
  --output-dir pages \
  --pdf book.pdf
```

To test only the first few pages:

```bash
python3 download_pdf.py "URL_HERE" \
  --limit 20 \
  --output-dir test_pages \
  --pdf test.pdf
```

If a run is interrupted, run the same command again. Existing images are skipped unless you pass `--overwrite`.

## Authenticated Downloads

Some pages require a logged-in account. The Digitalniknihovna web app sends protected image requests with:

```text
Authorization: Bearer <token>
Client: www.digitalniknihovna.cz
```

Use either `--auth-token`:

```bash
python3 download_pdf.py "URL_HERE" \
  --auth-token "PASTE_TOKEN_HERE" \
  --output-dir pages \
  --pdf book.pdf
```

Or put the token in `auth-token.txt` and use `--auth-file`:

```bash
python3 download_pdf.py "URL_HERE" \
  --auth-file auth-token.txt \
  --output-dir pages \
  --pdf book.pdf
```

`auth-token.txt` may contain either the raw token or a copied header like:

```text
Authorization: Bearer eyJ...
```

## How To Get The Auth Token

1. Log in on `https://www.digitalniknihovna.cz`.
2. Open the protected book page in the browser.
3. Open browser developer tools:
   - Chrome / Edge: `F12` or `Ctrl+Shift+I`
   - Firefox: `F12`
4. Go to the **Application** tab.
5. Open **Local storage** in the left sidebar.
6. Select:

```text
https://www.digitalniknihovna.cz
```

7. Find the key for your library. For MZK it is:

```text
auth.token.mzk
```

8. Copy the value and save it in `auth-token.txt`, or pass it directly with `--auth-token`.

For another library, the key usually follows the same pattern:

```text
auth.token.<library-code>
```

For example, a URL containing `/mzk/view/...` uses `auth.token.mzk`.

## Using Other Digitalniknihovna URLs

Paste the document URL directly:

```bash
python3 download_pdf.py "https://www.digitalniknihovna.cz/LIBRARY_CODE/view/uuid:..." \
  --auth-file auth-token.txt \
  --output-dir pages \
  --pdf book.pdf
```

The script reads the library code from the URL and tries to resolve the correct Kramerius API backend automatically. If that fails, pass the API backend manually:

```bash
python3 download_pdf.py "URL_HERE" \
  --library mzk \
  --base-url "https://api.kramerius.mzk.cz" \
  --auth-file auth-token.txt \
  --output-dir pages \
  --pdf book.pdf
```

## Useful Options

```text
--limit N          Download only the first N pages
--overwrite        Redownload images that already exist
--delay SECONDS    Wait between image downloads
--cookies FILE     Send cookies from a cookies.txt export
--auth-token TOKEN Send Authorization: Bearer TOKEN
--auth-file FILE   Read token or headers from a file
--client VALUE     Override the Client header
```

## Security Notes

Treat `auth-token.txt`, `cookie-dl.txt`, and copied request headers like passwords. Do not share them and do not commit them to version control.

This repository's `.gitignore` already ignores common token, cookie, header, PDF, and downloaded-page files.