wallabag2kindle/send_to_kindle.py

#!/usr/bin/env python3
import argparse
import mimetypes
import os
import smtplib
import re
import zipfile
import xml.etree.ElementTree as ET
from email.message import EmailMessage
from pathlib import Path

DEFAULT_CONFIG = Path("./mail.conf")
DEFAULT_OUT = Path("./out")


def load_config(path: Path) -> dict:
    """Load simple KEY=VALUE config file. Lines starting with # are ignored."""
    cfg = {}
    if not path.is_file():
        return cfg
    for line in path.read_text().splitlines():
        line = line.strip()
        if not line or line.startswith("#") or "=" not in line:
            continue
        key, value = line.split("=", 1)
        cfg[key.strip()] = value.strip().strip('"').strip("'")
    return cfg


def get_value(args, cfg, attr, key, env=None, default=None):
    val = getattr(args, attr, None)
    if val not in (None, ""):
        return val
    if key in cfg and cfg[key] != "":
        return cfg[key]
    if env and os.getenv(env):
        return os.getenv(env)
    return default


def chunks(items, size):
    for i in range(0, len(items), size):
        yield items[i:i + size]


def epub_title(path: Path) -> str | None:
    """Read dc:title from EPUB metadata, if available."""
    try:
        with zipfile.ZipFile(path) as z:
            container = ET.fromstring(z.read("META-INF/container.xml"))
            ns_container = {"c": "urn:oasis:names:tc:opendocument:xmlns:container"}
            rootfile = container.find(".//c:rootfile", ns_container)
            if rootfile is None:
                return None
            opf_path = rootfile.attrib["full-path"]
            opf = ET.fromstring(z.read(opf_path))
            ns = {"dc": "http://purl.org/dc/elements/1.1/"}
            title = opf.find(".//dc:title", ns)
            if title is not None and title.text:
                return " ".join(title.text.replace("_", " ").split())
    except Exception:
        return None
    return None


def attachment_name(path: Path) -> str:
    """Use EPUB metadata title as emailed attachment filename.

    Local filenames stay unchanged. Amazon Send-to-Kindle often derives the
    displayed document title from the email attachment filename, so use a nice
    title-based attachment filename while keeping the .epub extension.
    """
    title = epub_title(path)
    if not title:
        return path.name
    name = re.sub(r'[\\/:*?"<>|]+', ' ', title)
    name = re.sub(r"\s+", " ", name).strip().rstrip('.')
    return (name[:120] or path.stem) + path.suffix.lower()


def send_to_kindle(smtp_host, smtp_port, smtp_user, smtp_pass, sender, kindle_email, file_paths):
    file_paths = [Path(p) for p in file_paths]
    for file_path in file_paths:
        if not file_path.is_file():
            raise FileNotFoundError(file_path)

    msg = EmailMessage()
    msg["From"] = sender
    msg["To"] = kindle_email
    msg["Subject"] = "Send to Kindle"
    display_names = [attachment_name(p) for p in file_paths]
    msg.set_content("Attached ebook(s):\n\n" + "\n".join(display_names))

    for file_path in file_paths:
        ctype, _ = mimetypes.guess_type(file_path)
        if ctype is None:
            ctype = "application/octet-stream"
        maintype, subtype = ctype.split("/", 1)

        with file_path.open("rb") as f:
            msg.add_attachment(
                f.read(),
                maintype=maintype,
                subtype=subtype,
                filename=attachment_name(file_path),
            )

    with smtplib.SMTP_SSL(smtp_host, int(smtp_port)) as smtp:
        smtp.login(smtp_user, smtp_pass)
        smtp.send_message(msg)


def find_epubs(out_dir: Path):
    return sorted(out_dir.glob("*.epub"))


def main():
    p = argparse.ArgumentParser(description="Send ebook(s) to Kindle via email")
    p.add_argument("file", nargs="?", help="ebook file, e.g. .epub/.pdf/.mobi. If omitted, sends all .epub files in ./out")
    p.add_argument("--config", default=str(DEFAULT_CONFIG), help=f"config file, default: {DEFAULT_CONFIG}")
    p.add_argument("--kindle", help="your Kindle email, e.g. name@kindle.com")
    p.add_argument("--smtp-host")
    p.add_argument("--smtp-port", type=int)
    p.add_argument("--smtp-user")
    p.add_argument("--smtp-pass")
    p.add_argument("--sender")
    p.add_argument("--max-attachments", type=int, default=16, help="maximum attachments per email, default: 16")
    args = p.parse_args()

    cfg = load_config(Path(args.config).expanduser())

    settings = {
        "smtp_host": get_value(args, cfg, "smtp_host", "SMTP_HOST", "SMTP_HOST"),
        "smtp_port": get_value(args, cfg, "smtp_port", "SMTP_PORT", "SMTP_PORT", "465"),
        "smtp_user": get_value(args, cfg, "smtp_user", "SMTP_USER", "SMTP_USER"),
        "smtp_pass": get_value(args, cfg, "smtp_pass", "SMTP_PASS", "SMTP_PASS"),
        "sender": get_value(args, cfg, "sender", "SMTP_SENDER", "SMTP_SENDER"),
        "kindle": get_value(args, cfg, "kindle", "KINDLE_EMAIL", "KINDLE_EMAIL"),
    }

    missing = [k for k, v in settings.items() if not v]
    if missing:
        raise SystemExit(
            "Missing: " + ", ".join(missing) +
            f"\nAdd them to {args.config} or pass them as command-line options."
        )

    files = [Path(args.file).expanduser()] if args.file else find_epubs(DEFAULT_OUT)
    if not files:
        raise SystemExit(f"No EPUB files found in {DEFAULT_OUT}")

    max_attachments = max(1, args.max_attachments)
    batches = list(chunks(files, max_attachments))
    for idx, batch in enumerate(batches, 1):
        send_to_kindle(
            settings["smtp_host"],
            settings["smtp_port"],
            settings["smtp_user"],
            settings["smtp_pass"],
            settings["sender"],
            settings["kindle"],
            batch,
        )
        suffix = f" ({idx}/{len(batches)})" if len(batches) > 1 else ""
        print(f"Sent email{suffix}: {len(batch)} attachment(s)")
        for file_path in batch:
            file_path.unlink()
            print(f"  - sent and deleted: {file_path}")


if __name__ == "__main__":
    main()