libervia-backend/libervia/backend/plugins/plugin_misc_attach.py

#!/usr/bin/env python3

# SàT plugin for attaching files
# Copyright (C) 2009-2026 Jérôme Poisson (goffi@goffi.org)

# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.

# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU Affero General Public License for more details.

# You should have received a copy of the GNU Affero General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.

from collections import namedtuple
import mimetypes
from pathlib import Path
import shutil
import tempfile
from typing import Callable, Optional, cast

from libervia.backend.plugins.plugin_misc_upload import UploadPlugin, UploadResult

from twisted.internet import defer

from libervia.backend.core import exceptions
from libervia.backend.core.constants import Const as C
from libervia.backend.core.core_types import SatXMPPEntity
from libervia.backend.core.i18n import _
from libervia.backend.core.log import getLogger
from libervia.backend.plugins.plugin_misc_upload import UploadPlugin, UploadResult
from libervia.backend.tools import utils
from libervia.backend.tools import image

log = getLogger(__name__)


PLUGIN_INFO = {
    C.PI_NAME: "File Attach",
    C.PI_IMPORT_NAME: "ATTACH",
    C.PI_TYPE: C.PLUG_TYPE_MISC,
    C.PI_MODES: C.PLUG_MODE_BOTH,
    C.PI_DEPENDENCIES: ["UPLOAD"],
    C.PI_MAIN: "AttachPlugin",
    C.PI_HANDLER: "no",
    C.PI_DESCRIPTION: _("""Attachments handler"""),
}


AttachmentHandler = namedtuple("AttachmentHandler", ["can_handle", "attach", "priority"])


class AttachPlugin:

    def __init__(self, host):
        log.info(_("plugin Attach initialization"))
        self.host = host
        self._u = cast(UploadPlugin, host.plugins["UPLOAD"])
        host.trigger.add("sendMessage", self._send_message_trigger)
        host.trigger.add("sendMessageComponent", self._send_message_trigger)
        self._attachments_handlers = {"clear": [], "encrypted": []}
        self.register(self.default_can_handle, self.default_attach, False, -1000)

    def register(self, can_handle, attach, encrypted=False, priority=0):
        """Register an attachments handler

        @param can_handle(callable, coroutine, Deferred): a method which must return True
            if this plugin can handle the upload, otherwise next ones will be tried.
            This method will get client and mess_data as arguments, before the XML is
            generated
        @param attach(callable, coroutine, Deferred): attach the file
            this method will get client and mess_data as arguments, after XML is
            generated. Upload operation must be handled
            hint: "UPLOAD" plugin can be used
        @param encrypted(bool): True if the handler manages encrypted files
            A handler can be registered twice if it handle both encrypted and clear
            attachments
        @param priority(int): priority of this handler, handler with higher priority will
            be tried first
        """
        handler = AttachmentHandler(can_handle, attach, priority)
        handlers = (
            self._attachments_handlers["encrypted"]
            if encrypted
            else self._attachments_handlers["clear"]
        )
        if handler in handlers:
            raise exceptions.InternalError(
                "Attachment handler has been registered twice, this should never happen"
            )

        handlers.append(handler)
        handlers.sort(key=lambda h: h.priority, reverse=True)
        log.debug(f"new attachments handler: {handler}")

    async def attach_files(self, client, data):
        """Main method to attach file

        It will do generic pre-treatment, and call the suitable attachments handler
        """
        # we check attachment for pre-treatment like large image resizing
        # media_type will be added if missing (and if it can be guessed from path)
        attachments = data["extra"][C.KEY_ATTACHMENTS]
        tmp_dirs_to_clean = []
        for attachment in attachments:
            if attachment.get(C.KEY_ATTACHMENTS_RESIZE, False):
                path = Path(attachment["path"])
                try:
                    media_type = attachment[C.KEY_ATTACHMENTS_MEDIA_TYPE]
                except KeyError:
                    media_type = mimetypes.guess_type(path, strict=False)[0]
                    if media_type is None:
                        log.warning(
                            _(
                                "Can't resize attachment of unknown type: {attachment}"
                            ).format(attachment=attachment)
                        )
                        continue
                    attachment[C.KEY_ATTACHMENTS_MEDIA_TYPE] = media_type

                main_type = media_type.split("/")[0]
                if main_type == "image":
                    report = image.check(self.host, path)
                    if report["too_large"]:
                        tmp_dir = Path(tempfile.mkdtemp())
                        tmp_dirs_to_clean.append(tmp_dir)
                        new_path = tmp_dir / path.name
                        await image.resize(
                            path, report["recommended_size"], dest=new_path
                        )
                        attachment["path"] = new_path
                        log.info(
                            _(
                                "Attachment {path!r} has been resized at {new_path!r}"
                            ).format(path=str(path), new_path=str(new_path))
                        )
                else:
                    log.warning(
                        _(
                            "Can't resize attachment of type {main_type!r}: {attachment}"
                        ).format(main_type=main_type, attachment=attachment)
                    )

        if client.encryption.is_encryption_requested(data):
            handlers = self._attachments_handlers["encrypted"]
        else:
            handlers = self._attachments_handlers["clear"]

        for handler in handlers:
            can_handle = await utils.as_deferred(handler.can_handle, client, data)
            if can_handle:
                break
        else:
            raise exceptions.NotFound(
                _("No plugin can handle attachment with {destinee}").format(
                    destinee=data["to"]
                )
            )

        await utils.as_deferred(handler.attach, client, data)

        for dir_path in tmp_dirs_to_clean:
            log.debug(f"Cleaning temporary directory at {dir_path}")
            shutil.rmtree(dir_path)

        return data

    async def upload_files(
        self, client: SatXMPPEntity, data: dict, upload_cb: Optional[Callable] = None
    ):
        """Upload file, and update attachments

        invalid attachments will be removed
        @param client:
        @param data(dict): message data
        @param upload_cb(coroutine, Deferred, None): method to use for upload
            if None, upload method from UPLOAD plugin will be used.
            Otherwise, following kwargs will be used with the cb:
                - client
                - filepath
                - filename
                - options
            the method must return a tuple similar to UPLOAD plugin's upload method,
            it must contain:
                - progress_id
                - a deferred which fire download URL
        """
        if upload_cb is None:
            upload_cb = self._u.upload

        uploads_d = []
        to_delete = []
        attachments = data["extra"]["attachments"]

        for attachment in attachments:
            if "url" in attachment and not "path" in attachment:
                log.debug(f"attachment is external, we don't upload it: {attachment}")
                continue
            try:
                # we pop path because we don't want it to be stored, as the file can be
                # only in a temporary location
                path = Path(attachment.pop("path"))
            except KeyError:
                log.warning("no path in attachment: {attachment}")
                to_delete.append(attachment)
                continue

            if "url" in attachment:
                url = attachment.pop("url")
                log.warning(
                    f"unexpected URL in attachment: {url!r}\nattachment: {attachment}"
                )

            try:
                name = attachment["name"]
            except KeyError:
                name = attachment["name"] = path.name

            attachment["size"] = path.stat().st_size

            extra = {"attachment": attachment}
            extra["hash_algo"] = "sha-256"
            progress_id = attachment.pop("progress_id", None)
            if progress_id:
                extra["progress_id"] = progress_id
            check_certificate = self.host.memory.param_get_a(
                "check_certificate", "Connection", profile_key=client.profile
            )
            if not check_certificate:
                extra["ignore_tls_errors"] = True
                log.warning(
                    _("certificate check disabled for upload, this is dangerous!")
                )

            upload_result = await upload_cb(
                client=client,
                filepath=path,
                filename=name,
                extra=extra,
            )
            uploads_d.append(upload_result.download_d)

        for attachment in to_delete:
            attachments.remove(attachment)

        upload_results = await defer.DeferredList(uploads_d)
        for idx, (success, upload_result) in enumerate(upload_results):
            attachment = attachments[idx]

            if not success:
                # upload_result is a failure here
                log.warning(f"error while uploading {attachment}: {upload_result}")
                continue

            attachment["url"] = upload_result.url
            if upload_result.hash and "hash" not in attachment:
                attachment["hash"] = upload_result.hash
                attachment["hash_algo"] = upload_result.hash_algo

        return data

    def _attach_files(self, data, client):
        return defer.ensureDeferred(self.attach_files(client, data))

    def _send_message_trigger(
        self, client, mess_data, pre_xml_treatments, post_xml_treatments
    ):
        if mess_data["extra"].get(C.KEY_ATTACHMENTS):
            post_xml_treatments.addCallback(self._attach_files, client=client)
        return True

    async def default_can_handle(self, client, data):
        return True

    async def default_attach(self, client, data):
        await self.upload_files(client, data)
        # TODO: handle xhtml-im
        body_elt = data["xml"].body
        if body_elt is None:
            body_elt = data["xml"].addElement("body")
        attachments = data["extra"][C.KEY_ATTACHMENTS]
        if attachments:
            body_links = "\n".join(a["url"] for a in attachments)
            if str(body_elt).strip():
                # if there is already a body, we add a line feed before the first link
                body_elt.addContent("\n")
            body_elt.addContent(body_links)