app/utils/gmail_utils.py

from __future__ import annotations

import base64
import binascii
import mimetypes
import os
from email import message_from_bytes
from email.mime.application import MIMEApplication
from email.mime.audio import MIMEAudio
from email.mime.base import MIMEBase
from email.mime.image import MIMEImage
from email.mime.multipart import MIMEMultipart
from email.mime.text import MIMEText
from typing import Any, Dict, List, Optional

from app.config import get_settings
from app.models.schemas import (
    GmailComposeMessageRequest,
    GmailMessageAttachmentMeta,
    GmailParsedMessage,
)

_settings = get_settings()


class GmailMessageValidationError(ValueError):
    """Raised when a composed message fails size/content validation."""


class GmailParseError(ValueError):
    """Raised when a Gmail API message payload cannot be parsed."""


_EXTENSION_TO_MIME: Dict[str, str] = {
    ".pdf": "application/pdf",
    ".png": "image/png",
    ".jpg": "image/jpeg",
    ".jpeg": "image/jpeg",
    ".gif": "image/gif",
    ".bmp": "image/bmp",
    ".webp": "image/webp",
    ".txt": "text/plain",
    ".csv": "text/csv",
    ".json": "application/json",
    ".xml": "application/xml",
    ".zip": "application/zip",
    ".doc": "application/msword",
    ".docx": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
    ".xls": "application/vnd.ms-excel",
    ".xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
    ".ppt": "application/vnd.ms-powerpoint",
    ".pptx": "application/vnd.openxmlformats-officedocument.presentationml.presentation",
    ".mp3": "audio/mpeg",
    ".wav": "audio/wav",
    ".mp4": "video/mp4",
    ".mov": "video/quicktime",
}

_IMMUTABLE_HEADERS = frozenset({
    "from", "to", "subject", "cc", "bcc", "reply-to",
    "in-reply-to", "references", "content-type", "mime-version",
    "content-transfer-encoding",
})


def guess_mime_type(filename: str) -> str:
    """Best-effort MIME type from a filename; falls back to octet-stream."""
    mime_type, _ = mimetypes.guess_type(filename)
    if mime_type is None:
        ext = os.path.splitext(filename)[1].lower()
        mime_type = _EXTENSION_TO_MIME.get(ext, "application/octet-stream")
    return mime_type


def _create_attachment_part(filename: str, content_bytes: bytes, mime_type: str) -> MIMEBase:
    main_type, _, sub_type = mime_type.partition("/")
    sub_type = sub_type or "octet-stream"

    if main_type == "image":
        part: MIMEBase = MIMEImage(content_bytes, _subtype=sub_type)
    elif main_type == "audio":
        part = MIMEAudio(content_bytes, _subtype=sub_type)
    elif main_type == "application":
        part = MIMEApplication(content_bytes, _subtype=sub_type, Name=filename)
    else:
        part = MIMEBase(main_type or "application", sub_type)
        part.set_payload(content_bytes)

    part.add_header("Content-Disposition", f'attachment; filename="{filename}"')
    part.add_header("Content-ID", f"<{filename}>")
    return part


def build_mime_message(
    request: GmailComposeMessageRequest,
) -> str:
    """Build an RFC 2822 message from a structured compose request.

    Returns the base64url-encoded message string accepted by the Gmail API.
    The ``From`` header is intentionally omitted so Gmail fills it with the
    authenticated mailbox address (client-supplied, stateless design).
    """
    message = MIMEMultipart("mixed")

    to = request.to
    if not to:
        raise GmailMessageValidationError("At least one 'to' recipient is required.")

    message["To"] = ", ".join(to)
    message["Subject"] = request.subject

    if request.cc:
        message["Cc"] = ", ".join(request.cc)
    if request.bcc:
        message["Bcc"] = ", ".join(request.bcc)
    if request.reply_to:
        message["Reply-To"] = ", ".join(request.reply_to)
    if request.in_reply_to:
        message["In-Reply-To"] = request.in_reply_to
    if request.references:
        message["References"] = " ".join(request.references)

    if request.headers:
        for key, value in request.headers.items():
            if key.lower() not in _IMMUTABLE_HEADERS:
                message[key] = value

    alternative = MIMEMultipart("alternative")
    if request.body:
        alternative.attach(MIMEText(request.body, "plain", "utf-8"))
    if request.html_body:
        alternative.attach(MIMEText(request.html_body, "html", "utf-8"))
    elif request.body:
        alternative.attach(MIMEText(request.body, "html", "utf-8"))

    if alternative.get_payload():
        message.attach(alternative)

    if request.attachments:
        total_bytes = sum(
            len(_b64_decode(s.content_base64)) for s in request.attachments
        )
        if total_bytes > _settings.gmail_max_payload_bytes:
            raise GmailMessageValidationError(
                f"Total attachment payload of {total_bytes} bytes exceeds the "
                f"limit of {_settings.gmail_max_payload_bytes} bytes."
            )

        for spec in request.attachments:
            content = _b64_decode(spec.content_base64)
            if len(content) > _settings.gmail_max_attachment_bytes:
                raise GmailMessageValidationError(
                    f"Attachment '{spec.filename}' is {len(content)} bytes, exceeding "
                    f"the {_settings.gmail_max_attachment_bytes} byte per-file limit."
                )
            mime_type = spec.mime_type or guess_mime_type(spec.filename)
            message.attach(_create_attachment_part(spec.filename, content, mime_type))

    return encode_mime_to_base64(message)


def encode_mime_to_base64(mime_message: Any) -> str:
    """Base64url-encode a MIME message for the Gmail API."""
    return base64.urlsafe_b64encode(mime_message.as_bytes()).decode("ascii")


def _b64_decode(data: str) -> bytes:
    """Decode base64url (or standard base64) into bytes."""
    try:
        return base64.urlsafe_b64decode(data.encode("ascii"))
    except (binascii.Error, ValueError) as exc:
        raise GmailMessageValidationError(
            "Attachment content must be valid base64."
        ) from exc


def _base64_decode_text(data: str) -> str:
    if not data:
        return ""
    try:
        return _b64_decode(data).decode("utf-8", errors="replace")
    except GmailMessageValidationError:
        return ""


def parse_message(message: Dict[str, Any]) -> GmailParsedMessage:
    """Parse a Gmail ``format=full`` or ``format=metadata`` message resource.

    Handles the ``raw`` format transparently when present. Recursively extracts
    plain/HTML bodies and top-level attachment metadata.
    """
    try:
        message_id = message.get("id", "")
        thread_id = message.get("threadId", "")
        label_ids = message.get("labelIds") or []
        snippet = message.get("snippet", "")
        history_id = message.get("historyId")

        headers: Dict[str, str] = {}
        plain_body: Optional[str] = None
        html_body: Optional[str] = None
        attachments: List[GmailMessageAttachmentMeta] = []

        payload = message.get("payload") or {}
        if message.get("raw"):
            decoded = _b64_decode(message["raw"])
            parsed = message_from_bytes(decoded)
            headers = {
                h: v for h, v in parsed.items()
            }
            if parsed.get_content_type() == "multipart/mixed":
                for part in parsed.walk():
                    ctype = part.get_content_type()
                    if ctype == "text/plain":
                        plain_body = part.get_payload(decode=True).decode("utf-8", errors="replace")
                    elif ctype == "text/html":
                        html_body = part.get_payload(decode=True).decode("utf-8", errors="replace")
                    else:
                        filename = part.get_filename()
                        if filename:
                            attachments.append(GmailMessageAttachmentMeta(
                                attachment_id="",
                                filename=filename,
                                mime_type=part.get_content_type(),
                                size_bytes=len(part.get_payload(decode=True) or b""),
                            ))
            elif parsed.get_content_type() == "text/plain":
                plain_body = parsed.get_payload(decode=True).decode("utf-8", errors="replace")
        else:
            for header in payload.get("headers") or []:
                name = header.get("name", "")
                if name:
                    headers[name] = header.get("value", "")
            plain_body, html_body, attachments = _walk_payload(payload)

        subject = headers.get("Subject", headers.get("subject", ""))
        if subject and headers.get("Subject") is None and "subject" in headers:
            subject = headers["subject"]

        return GmailParsedMessage(
            id=message_id,
            thread_id=thread_id,
            label_ids=label_ids,
            snippet=snippet,
            history_id=history_id,
            headers=headers,
            sender=headers.get("From", ""),
            to=headers.get("To", ""),
            cc=headers.get("Cc", ""),
            bcc=headers.get("Bcc", ""),
            subject=subject,
            date=headers.get("Date"),
            plain_body=plain_body,
            html_body=html_body,
            attachments=attachments,
        )
    except GmailMessageValidationError as exc:
        raise GmailParseError(str(exc)) from exc


def _walk_payload(
    payload: Dict[str, Any],
) -> tuple[Optional[str], Optional[str], List[GmailMessageAttachmentMeta]]:
    """Recursively extract text bodies and attachment metadata from a payload tree."""
    plain_body: Optional[str] = None
    html_body: Optional[str] = None
    attachments: List[GmailMessageAttachmentMeta] = []

    def visit(node: Dict[str, Any]) -> None:
        nonlocal plain_body, html_body
        mime_type = node.get("mimeType", "")
        filename = node.get("filename", "")
        body = node.get("body") or {}
        parts = node.get("parts") or []

        if filename:
            attachments.append(GmailMessageAttachmentMeta(
                attachment_id=body.get("attachmentId", ""),
                filename=filename,
                mime_type=mime_type,
                size_bytes=int(body.get("size", 0) or 0),
            ))

        if mime_type == "text/plain" and plain_body is None:
            plain_body = _base64_decode_text(body.get("data", ""))
        elif mime_type == "text/html" and html_body is None:
            html_body = _base64_decode_text(body.get("data", ""))

        for child in parts:
            visit(child)

    visit(payload)
    return plain_body, html_body, attachments