#!/usr/bin/python3

# a HTTP server for sharing files

# Copyright 2019-2020 Anton Khirnov <anton@khirnov.net>
#
# fshare is free software: you can redistribute it and/or modify it under the
# terms of the GNU General Public License as published by the Free Software
# Foundation, either version 3 of the License, or (at your option) any later
# version.
#
# fshare 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 General Public License for more details.
#
# You should have received a copy of the GNU General Public License along with
# fshare.  If not, see <http://www.gnu.org/licenses/>.

import argparse
import hmac
import json
import os
import os.path
from  http import HTTPStatus
import http.server as hs
import logging
import logging.handlers
import secrets
import shutil
import socket
import sys
import tempfile
from urllib import parse as urlparse

# TODO: detect and store mime types

class PersistentState:
    key = None

    def __init__(self, path):
        if not os.path.exists(path):
            data = { 'key' : secrets.token_hex(16) }
            with open(path, 'x') as f:
                json.dump(data, f, indent = 4)

        with open(path, 'r') as f:
            data = json.load(f)

        self.key = bytes.fromhex(data['key'])

class HTTPChunkedRequestReader:

    _stream = None
    _eof    = False

    _logger = None

    def __init__(self, stream, logger):
        self._stream = stream
        self._logger = logger

    def read(self):
        if self._eof:
            return bytes()

        l = self._stream.readline().decode('ascii', errors = 'replace')
        self._logger.debug('reading chunk: chunksize %s', l)

        try:
            chunk_size = int(l.split(';')[0], 16)
        except ValueError:
            raise IOError('Invalid chunksize line: %s' % l)
        if chunk_size < 0:
            raise IOError('Invalid negative chunksize: %d' % chunk_size)
        if chunk_size == 0:
            self._eof = True
            return bytes()

        data      = bytes()
        remainder = chunk_size
        while remainder > 0:
            read = self._stream.read(remainder)
            if len(read) == 0:
                raise IOError('Premature EOF')

            data      += read
            remainder -= len(read)

        term_line = self._stream.readline().decode('ascii', errors = 'replace')
        if term_line != '\r\n':
            raise IOError('Invalid chunk terminator: %s' % term_line)

        return data

class HTTPRequestReader:

    _stream    = None
    _remainder = 0
    _eof       = False

    def __init__(self, stream, request_size):
        self._stream    = stream
        self._remainder = request_size
        self._eof       = request_size == 0

    def read(self):
        if self._eof:
            return bytes()

        read = self._stream.read1(self._remainder)
        if len(read) == 0:
            raise IOError('Premature EOF')

        self._remainder -= len(read)
        self._eof        = self._remainder <= 0

        return read

class FShareRequestHandler(hs.BaseHTTPRequestHandler):
    # required for chunked transfer
    protocol_version = "HTTP/1.1"

    _logger = None

    def __init__(self, *args, **kwargs):
        server = args[2]
        self._logger = server._logger.getChild('requesthandler')

        super().__init__(*args, **kwargs)

    def _process_path(self, encoded_path):
        # decode percent-encoding
        path = urlparse.unquote(encoded_path, encoding = 'ascii')

        # normalize the path
        path = os.path.normpath(path)

        # make sure the path doesn't point outside of our root
        if path.startswith('..'):
            raise PermissionError('Invalid path')

        return path

    def _log_request(self):
        self._logger.info('%s: %s', str(self.client_address), self.requestline)
        self._logger.debug('headers:\n%s', self.headers)

    def do_GET(self):
        self._log_request()

        fname = self._process_path(self.path)
        path = '/'.join((self.server.data_dir, fname))
        self._logger.info('serve file: %s', fname)

        try:
            infile = open(path, 'rb')
        except OSError:
            return self.send_error(HTTPStatus.NOT_FOUND)

        try:
            stat = os.fstat(infile.fileno())

            self.send_response(HTTPStatus.OK)
            self.send_header('Content-Length', str(stat.st_size))
            self.end_headers()

            shutil.copyfileobj(infile, self.wfile)
        finally:
            infile.close()

    def do_POST(self):
        self._log_request()

        src_fname = self._process_path(self.path)
        ext   = os.path.splitext(src_fname)[1]

        if 'Transfer-Encoding' in self.headers:
            if self.headers['Transfer-Encoding'] != 'chunked':
                return self.send_error(HTTPStatus.NOT_IMPLEMENTED,
                                        'Unsupported Transfer-Encoding: %s' %
                                        self.headers['Transfer-Encoding'])
            infile = HTTPChunkedRequestReader(self.rfile, self._logger.getChild('chreader'))
        elif 'Content-Length' in self.headers:
            infile = HTTPRequestReader(self.rfile, int(self.headers['Content-Length']))
        else:
            return self.send_error(HTTPStatus.BAD_REQUEST)

        h = hmac.new(self.server.state.key, digestmod = 'SHA256')

        temp_fd, temp_path = tempfile.mkstemp(suffix = '.tmp', dir = self.server.data_dir)
        try:
            while True:
                data = infile.read()
                if len(data) == 0:
                    self._logger.debug('Finished reading')
                    break

                written = os.write(temp_fd, data)
                if written < len(data):
                    raise IOError('partial write: %d < %d' % (written, len(data)))

                h.update(data)

                self._logger.debug('streamed %d bytes', len(data))

            os.close(temp_fd)

            dst_fname = h.hexdigest() + ext
            self._logger.info('Received file: %s', dst_fname)

            outpath = '/'.join((self.server.data_dir, dst_fname))
            if os.path.exists(outpath):
                retcode = HTTPStatus.OK
                os.remove(temp_path)
            else:
                retcode = HTTPStatus.CREATED
                os.replace(temp_path, outpath)
        finally:
            if os.path.exists(temp_path):
                os.remove(temp_path)

        try:
            host = self.headers['host']
        except KeyError:
            host = 'host.missing'

        reply = ('https://%s/%s' % (host, dst_fname)).encode('ascii')

        self.send_response(retcode)
        self.send_header('Content-Type', 'text/plain')
        self.send_header('Content-Length', '%d' % len(reply))
        self.end_headers()

        self.wfile.write(reply)

    def do_PUT(self):
        return self.do_POST()

    def do_DELETE(self):
        self._log_request()

        fname = self._process_path(self.path)

        path = '/'.join((self.server.data_dir, fname))
        try:
            os.remove(targetpath)
        except FileNotFoundError:
            self._logger.error('DELETE request for non-existing file: %s' %
                    local_path.decode('utf-8', errors = 'backslashreplace'))
            self.send_error(HTTPStatus.NOT_FOUND)
            return

        self.send_response(HTTPStatus.NO_CONTENT)
        self.send_header('Content-Length', '0')
        self.end_headers()

class FShareServer(hs.ThreadingHTTPServer):

    data_dir = None
    state    = None

    _logger  = None

    def __init__(self, address, force_v4, force_v6, state, data_dir, logger):
        self.data_dir      = data_dir
        self.state         = state
        self._logger       = logger

        family = None
        if force_v4:
            family = socket.AF_INET
        elif force_v6:
            family = socket.AF_INET6

        if family is None and len(address[0]):
            try:
                family, _, _, _, _ = socket.getaddrinfo(*address)[0]
            except IndexError:
                pass

        if family is None:
            family = socket.AF_INET6

        self.address_family = family

        super().__init__(address, FShareRequestHandler)

# parse commandline arguments
parser = argparse.ArgumentParser('fshare server')

parser.add_argument('-a', '--address', default = 'localhost')
parser.add_argument('-p', '--port',    type = int, default = 5400)

group = parser.add_mutually_exclusive_group()
group.add_argument('-4', '--ipv4',    action = 'store_true')
group.add_argument('-6', '--ipv6',    action = 'store_true')

parser.add_argument('-l', '--loglevel', default = 'WARNING')
parser.add_argument('-d', '--debug', action = 'store_true', help = 'log to stderr')

parser.add_argument('state_file')
parser.add_argument('data_dir')

args = parser.parse_args(sys.argv[1:])

# configure logging
progname = os.path.basename(sys.argv[0])
logging.basicConfig(stream = sys.stderr, level = args.loglevel)
logger = logging.getLogger(progname)

formatter = logging.Formatter(fmt = progname + ': %(message)s')

syslog = logging.handlers.SysLogHandler('/dev/log')

handlers = [syslog]
if args.debug:
    handlers.append(logging.StreamHandler())

for h in handlers:
    h.setFormatter(formatter)
    logger.addHandler(h)

# log uncaught top-level exception
def excepthook(t, v, tb, logger = logger):
    logger.error('Uncaught top-level exception', exc_info = (t, v, tb))
sys.excepthook = excepthook

# read the state file
state = PersistentState(args.state_file)

# launch the server
server = FShareServer((args.address, args.port), args.ipv4, args.ipv6,
                      state, args.data_dir, logger)
server.serve_forever()