twisterp2pnetworkbittorrentblockchainmicrobloggingipv6social-networkdhtdecentralizedtwister-servertwister-ipv6twister-coretwisterarmyp2p-network
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
919 lines
30 KiB
919 lines
30 KiB
11 years ago
|
/*
|
||
|
|
||
|
Copyright (c) 2003-2012, Arvid Norberg
|
||
|
All rights reserved.
|
||
|
|
||
|
Redistribution and use in source and binary forms, with or without
|
||
|
modification, are permitted provided that the following conditions
|
||
|
are met:
|
||
|
|
||
|
* Redistributions of source code must retain the above copyright
|
||
|
notice, this list of conditions and the following disclaimer.
|
||
|
* Redistributions in binary form must reproduce the above copyright
|
||
|
notice, this list of conditions and the following disclaimer in
|
||
|
the documentation and/or other materials provided with the distribution.
|
||
|
* Neither the name of the author nor the names of its
|
||
|
contributors may be used to endorse or promote products derived
|
||
|
from this software without specific prior written permission.
|
||
|
|
||
|
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
||
|
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||
|
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
||
|
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
|
||
|
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
|
||
|
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
|
||
|
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
||
|
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
|
||
|
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
||
|
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
||
|
POSSIBILITY OF SUCH DAMAGE.
|
||
|
|
||
|
*/
|
||
|
|
||
|
#include "libtorrent/pch.hpp"
|
||
|
|
||
|
#include <vector>
|
||
|
#include <boost/limits.hpp>
|
||
|
#include <boost/bind.hpp>
|
||
|
#include <stdlib.h>
|
||
|
|
||
|
#include "libtorrent/web_peer_connection.hpp"
|
||
|
#include "libtorrent/session.hpp"
|
||
|
#include "libtorrent/identify_client.hpp"
|
||
|
#include "libtorrent/entry.hpp"
|
||
|
#include "libtorrent/bencode.hpp"
|
||
|
#include "libtorrent/alert_types.hpp"
|
||
|
#include "libtorrent/invariant_check.hpp"
|
||
|
#include "libtorrent/io.hpp"
|
||
|
#include "libtorrent/version.hpp"
|
||
|
#include "libtorrent/aux_/session_impl.hpp"
|
||
|
#include "libtorrent/parse_url.hpp"
|
||
|
#include "libtorrent/peer_info.hpp"
|
||
|
|
||
|
using boost::shared_ptr;
|
||
|
using libtorrent::aux::session_impl;
|
||
|
|
||
|
namespace libtorrent
|
||
|
{
|
||
|
enum
|
||
|
{
|
||
|
request_size_overhead = 5000
|
||
|
};
|
||
|
|
||
|
web_peer_connection::web_peer_connection(
|
||
|
session_impl& ses
|
||
|
, boost::weak_ptr<torrent> t
|
||
|
, boost::shared_ptr<socket_type> s
|
||
|
, tcp::endpoint const& remote
|
||
|
, std::string const& url
|
||
|
, policy::peer* peerinfo
|
||
|
, std::string const& auth
|
||
|
, web_seed_entry::headers_t const& extra_headers)
|
||
|
: web_connection_base(ses, t, s, remote, url, peerinfo, auth, extra_headers)
|
||
|
, m_url(url)
|
||
|
, m_received_body(0)
|
||
|
, m_range_pos(0)
|
||
|
, m_block_pos(0)
|
||
|
, m_chunk_pos(0)
|
||
|
, m_partial_chunk_header(0)
|
||
|
{
|
||
|
INVARIANT_CHECK;
|
||
|
|
||
|
if (!ses.settings().report_web_seed_downloads)
|
||
|
ignore_stats(true);
|
||
|
|
||
|
shared_ptr<torrent> tor = t.lock();
|
||
|
TORRENT_ASSERT(tor);
|
||
|
|
||
|
// we always prefer downloading 1 MiB chunks
|
||
|
// from web seeds, or whole pieces if pieces
|
||
|
// are larger than a MiB
|
||
|
prefer_whole_pieces((std::max)((1024 * 1024) / tor->torrent_file().piece_length(), 1));
|
||
|
|
||
|
// we want large blocks as well, so
|
||
|
// we can request more bytes at once
|
||
|
// this setting will merge adjacent requests
|
||
|
// into single larger ones
|
||
|
request_large_blocks(true);
|
||
|
|
||
|
#ifdef TORRENT_VERBOSE_LOGGING
|
||
|
peer_log("*** web_peer_connection %s", url.c_str());
|
||
|
#endif
|
||
|
}
|
||
|
|
||
|
void web_peer_connection::disconnect(error_code const& ec, int error)
|
||
|
{
|
||
|
boost::shared_ptr<torrent> t = associated_torrent().lock();
|
||
|
peer_connection::disconnect(ec, error);
|
||
|
if (t) t->disconnect_web_seed(this);
|
||
|
}
|
||
|
|
||
|
boost::optional<piece_block_progress>
|
||
|
web_peer_connection::downloading_piece_progress() const
|
||
|
{
|
||
|
if (m_requests.empty())
|
||
|
return boost::optional<piece_block_progress>();
|
||
|
|
||
|
boost::shared_ptr<torrent> t = associated_torrent().lock();
|
||
|
TORRENT_ASSERT(t);
|
||
|
|
||
|
piece_block_progress ret;
|
||
|
|
||
|
ret.piece_index = m_requests.front().piece;
|
||
|
ret.bytes_downloaded = m_block_pos % t->block_size();
|
||
|
// this is used to make sure that the block_index stays within
|
||
|
// bounds. If the entire piece is downloaded, the block_index
|
||
|
// would otherwise point to one past the end
|
||
|
int correction = m_block_pos ? -1 : 0;
|
||
|
ret.block_index = (m_requests.front().start + m_block_pos + correction) / t->block_size();
|
||
|
TORRENT_ASSERT(ret.block_index < int(piece_block::invalid.block_index));
|
||
|
TORRENT_ASSERT(ret.piece_index < int(piece_block::invalid.piece_index));
|
||
|
|
||
|
ret.full_block_bytes = t->block_size();
|
||
|
const int last_piece = t->torrent_file().num_pieces() - 1;
|
||
|
if (ret.piece_index == last_piece && ret.block_index
|
||
|
== t->torrent_file().piece_size(last_piece) / t->block_size())
|
||
|
ret.full_block_bytes = t->torrent_file().piece_size(last_piece) % t->block_size();
|
||
|
return ret;
|
||
|
}
|
||
|
|
||
|
void web_peer_connection::write_request(peer_request const& r)
|
||
|
{
|
||
|
INVARIANT_CHECK;
|
||
|
|
||
|
boost::shared_ptr<torrent> t = associated_torrent().lock();
|
||
|
TORRENT_ASSERT(t);
|
||
|
|
||
|
TORRENT_ASSERT(t->valid_metadata());
|
||
|
|
||
|
bool single_file_request = t->torrent_file().num_files() == 1;
|
||
|
|
||
|
if (!single_file_request)
|
||
|
{
|
||
|
// handle incorrect .torrent files which are multi-file
|
||
|
// but have web seeds not ending with a slash
|
||
|
if (m_path.empty() || m_path[m_path.size() - 1] != '/') m_path += "/";
|
||
|
if (m_url.empty() || m_url[m_url.size() - 1] != '/') m_url += "/";
|
||
|
}
|
||
|
else
|
||
|
{
|
||
|
// handle .torrent files that don't include the filename in the url
|
||
|
if (m_path.empty()) m_path += "/" + t->torrent_file().name();
|
||
|
else if (m_path[m_path.size() - 1] == '/')
|
||
|
{
|
||
|
std::string tmp = t->torrent_file().files().at(0).path;
|
||
|
#ifdef TORRENT_WINDOWS
|
||
|
convert_path_to_posix(tmp);
|
||
|
#endif
|
||
|
m_path += tmp;
|
||
|
}
|
||
|
else if (!m_url.empty() && m_url[m_url.size() - 1] == '/')
|
||
|
{
|
||
|
std::string tmp = t->torrent_file().files().at(0).path;
|
||
|
#ifdef TORRENT_WINDOWS
|
||
|
convert_path_to_posix(tmp);
|
||
|
#endif
|
||
|
m_url += tmp;
|
||
|
}
|
||
|
}
|
||
|
|
||
|
torrent_info const& info = t->torrent_file();
|
||
|
|
||
|
std::string request;
|
||
|
request.reserve(400);
|
||
|
|
||
|
int size = r.length;
|
||
|
const int block_size = t->block_size();
|
||
|
const int piece_size = t->torrent_file().piece_length();
|
||
|
peer_request pr;
|
||
|
while (size > 0)
|
||
|
{
|
||
|
int request_offset = r.start + r.length - size;
|
||
|
pr.start = request_offset % piece_size;
|
||
|
pr.length = (std::min)(block_size, size);
|
||
|
pr.piece = r.piece + request_offset / piece_size;
|
||
|
m_requests.push_back(pr);
|
||
|
size -= pr.length;
|
||
|
}
|
||
|
|
||
|
proxy_settings const& ps = m_ses.proxy();
|
||
|
bool using_proxy = (ps.type == proxy_settings::http
|
||
|
|| ps.type == proxy_settings::http_pw) && !m_ssl;
|
||
|
|
||
|
if (single_file_request)
|
||
|
{
|
||
|
request += "GET ";
|
||
|
// do not encode single file paths, they are
|
||
|
// assumed to be encoded in the torrent file
|
||
|
request += using_proxy ? m_url : m_path;
|
||
|
request += " HTTP/1.1\r\n";
|
||
|
add_headers(request, ps, using_proxy);
|
||
|
request += "\r\nRange: bytes=";
|
||
|
request += to_string(size_type(r.piece) * info.piece_length() + r.start).elems;
|
||
|
request += "-";
|
||
|
request += to_string(size_type(r.piece) * info.piece_length() + r.start + r.length - 1).elems;
|
||
|
request += "\r\n\r\n";
|
||
|
m_first_request = false;
|
||
|
m_file_requests.push_back(0);
|
||
|
}
|
||
|
else
|
||
|
{
|
||
|
std::vector<file_slice> files = info.orig_files().map_block(r.piece, r.start
|
||
|
, r.length);
|
||
|
|
||
|
for (std::vector<file_slice>::iterator i = files.begin();
|
||
|
i != files.end(); ++i)
|
||
|
{
|
||
|
file_slice const& f = *i;
|
||
|
if (info.orig_files().internal_at(f.file_index).pad_file)
|
||
|
{
|
||
|
m_file_requests.push_back(f.file_index);
|
||
|
continue;
|
||
|
}
|
||
|
|
||
|
request += "GET ";
|
||
|
if (using_proxy)
|
||
|
{
|
||
|
request += m_url;
|
||
|
std::string path = info.orig_files().file_path(info.orig_files().internal_at(f.file_index));
|
||
|
#ifdef TORRENT_WINDOWS
|
||
|
convert_path_to_posix(path);
|
||
|
#endif
|
||
|
request += escape_path(path.c_str(), path.length());
|
||
|
}
|
||
|
else
|
||
|
{
|
||
|
std::string path = m_path;
|
||
|
path += info.orig_files().file_path(info.orig_files().internal_at(f.file_index));
|
||
|
#ifdef TORRENT_WINDOWS
|
||
|
convert_path_to_posix(path);
|
||
|
#endif
|
||
|
request += escape_path(path.c_str(), path.length());
|
||
|
}
|
||
|
request += " HTTP/1.1\r\n";
|
||
|
add_headers(request, ps, using_proxy);
|
||
|
request += "\r\nRange: bytes=";
|
||
|
request += to_string(f.offset).elems;
|
||
|
request += "-";
|
||
|
request += to_string(f.offset + f.size - 1).elems;
|
||
|
request += "\r\n\r\n";
|
||
|
m_first_request = false;
|
||
|
TORRENT_ASSERT(f.file_index >= 0);
|
||
|
m_file_requests.push_back(f.file_index);
|
||
|
}
|
||
|
}
|
||
|
|
||
|
#ifdef TORRENT_VERBOSE_LOGGING
|
||
|
peer_log("==> %s", request.c_str());
|
||
|
#endif
|
||
|
|
||
|
send_buffer(request.c_str(), request.size(), message_type_request);
|
||
|
}
|
||
|
|
||
|
// --------------------------
|
||
|
// RECEIVE DATA
|
||
|
// --------------------------
|
||
|
|
||
|
namespace
|
||
|
{
|
||
|
bool range_contains(peer_request const& range, peer_request const& req, int piece_size)
|
||
|
{
|
||
|
size_type range_start = size_type(range.piece) * piece_size + range.start;
|
||
|
size_type req_start = size_type(req.piece) * piece_size + req.start;
|
||
|
return range_start <= req_start
|
||
|
&& range_start + range.length >= req_start + req.length;
|
||
|
}
|
||
|
}
|
||
|
|
||
|
bool web_peer_connection::maybe_harvest_block()
|
||
|
{
|
||
|
peer_request const& front_request = m_requests.front();
|
||
|
|
||
|
if (int(m_piece.size()) < front_request.length) return false;
|
||
|
TORRENT_ASSERT(int(m_piece.size() == front_request.length));
|
||
|
|
||
|
// each call to incoming_piece() may result in us becoming
|
||
|
// a seed. If we become a seed, all seeds we're connected to
|
||
|
// will be disconnected, including this web seed. We need to
|
||
|
// check for the disconnect condition after the call.
|
||
|
|
||
|
boost::shared_ptr<torrent> t = associated_torrent().lock();
|
||
|
TORRENT_ASSERT(t);
|
||
|
buffer::const_interval recv_buffer = receive_buffer();
|
||
|
|
||
|
incoming_piece(front_request, &m_piece[0]);
|
||
|
m_requests.pop_front();
|
||
|
if (associated_torrent().expired()) return false;
|
||
|
TORRENT_ASSERT(m_block_pos >= front_request.length);
|
||
|
m_block_pos -= front_request.length;
|
||
|
cut_receive_buffer(m_body_start, t->block_size() + request_size_overhead);
|
||
|
m_body_start = 0;
|
||
|
recv_buffer = receive_buffer();
|
||
|
// TORRENT_ASSERT(m_received_body <= range_end - range_start);
|
||
|
m_piece.clear();
|
||
|
TORRENT_ASSERT(m_piece.empty());
|
||
|
return true;
|
||
|
}
|
||
|
|
||
|
bool web_peer_connection::received_invalid_data(int index, bool single_peer)
|
||
|
{
|
||
|
if (!single_peer) return peer_connection::received_invalid_data(index, single_peer);
|
||
|
|
||
|
// when a web seed fails a hash check, do the following:
|
||
|
// 1. if the whole piece only overlaps a single file, mark that file as not
|
||
|
// have for this peer
|
||
|
// 2. if the piece overlaps more than one file, mark the piece as not have
|
||
|
// for this peer
|
||
|
// 3. if it's a single file torrent, just ban it right away
|
||
|
// this handles the case where web seeds may have some files updated but not other
|
||
|
|
||
|
boost::shared_ptr<torrent> t = associated_torrent().lock();
|
||
|
file_storage const& fs = t->torrent_file().files();
|
||
|
|
||
|
// single file torrent
|
||
|
if (fs.num_files() == 1) return peer_connection::received_invalid_data(index, single_peer);
|
||
|
|
||
|
std::vector<file_slice> files = fs.map_block(index, 0, fs.piece_size(index));
|
||
|
|
||
|
if (files.size() == 1)
|
||
|
{
|
||
|
// assume the web seed has a different copy of this specific file
|
||
|
// than what we expect, and pretend not to have it.
|
||
|
int fi = files[0].file_index;
|
||
|
int first_piece = fs.file_offset(fi) / fs.piece_length();
|
||
|
// one past last piece
|
||
|
int end_piece = (fs.file_offset(fi) + fs.file_size(fi) + 1) / fs.piece_length();
|
||
|
for (int i = first_piece; i < end_piece; ++i)
|
||
|
incoming_dont_have(i);
|
||
|
}
|
||
|
else
|
||
|
{
|
||
|
incoming_dont_have(index);
|
||
|
}
|
||
|
|
||
|
peer_connection::received_invalid_data(index, single_peer);
|
||
|
|
||
|
// if we don't think we have any of the files, allow banning the web seed
|
||
|
if (num_have_pieces() == 0) return true;
|
||
|
|
||
|
// don't disconnect, we won't request anything from this file again
|
||
|
return false;
|
||
|
}
|
||
|
|
||
|
void web_peer_connection::on_receive(error_code const& error
|
||
|
, std::size_t bytes_transferred)
|
||
|
{
|
||
|
INVARIANT_CHECK;
|
||
|
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded() + bytes_transferred < size_t(INT_MAX));
|
||
|
int dl_target = m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded() + bytes_transferred;
|
||
|
#endif
|
||
|
|
||
|
if (error)
|
||
|
{
|
||
|
m_statistics.received_bytes(0, bytes_transferred);
|
||
|
#ifdef TORRENT_VERBOSE_LOGGING
|
||
|
peer_log("*** web_peer_connection error: %s", error.message().c_str());
|
||
|
#endif
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
|
||
|
boost::shared_ptr<torrent> t = associated_torrent().lock();
|
||
|
TORRENT_ASSERT(t);
|
||
|
|
||
|
for (;;)
|
||
|
{
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded() + int(bytes_transferred)
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
|
||
|
buffer::const_interval recv_buffer = receive_buffer();
|
||
|
|
||
|
int payload;
|
||
|
int protocol;
|
||
|
bool header_finished = m_parser.header_finished();
|
||
|
if (!header_finished)
|
||
|
{
|
||
|
bool failed = false;
|
||
|
boost::tie(payload, protocol) = m_parser.incoming(recv_buffer, failed);
|
||
|
m_statistics.received_bytes(0, protocol);
|
||
|
TORRENT_ASSERT(int(bytes_transferred) >= protocol);
|
||
|
bytes_transferred -= protocol;
|
||
|
|
||
|
if (failed)
|
||
|
{
|
||
|
m_statistics.received_bytes(0, bytes_transferred);
|
||
|
#ifdef TORRENT_VERBOSE_LOGGING
|
||
|
peer_log("*** %s", std::string(recv_buffer.begin, recv_buffer.end).c_str());
|
||
|
#endif
|
||
|
disconnect(errors::http_parse_error, 2);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
|
||
|
TORRENT_ASSERT(recv_buffer.left() == 0 || *recv_buffer.begin == 'H');
|
||
|
|
||
|
TORRENT_ASSERT(recv_buffer.left() <= packet_size());
|
||
|
|
||
|
// this means the entire status line hasn't been received yet
|
||
|
if (m_parser.status_code() == -1)
|
||
|
{
|
||
|
TORRENT_ASSERT(payload == 0);
|
||
|
TORRENT_ASSERT(bytes_transferred == 0);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded() + int(bytes_transferred)
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
break;
|
||
|
}
|
||
|
|
||
|
if (!m_parser.header_finished())
|
||
|
{
|
||
|
TORRENT_ASSERT(payload == 0);
|
||
|
TORRENT_ASSERT(bytes_transferred == 0);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded() + int(bytes_transferred)
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
break;
|
||
|
}
|
||
|
|
||
|
m_body_start = m_parser.body_start();
|
||
|
m_received_body = 0;
|
||
|
}
|
||
|
|
||
|
// we just completed reading the header
|
||
|
if (!header_finished)
|
||
|
{
|
||
|
#ifdef TORRENT_VERBOSE_LOGGING
|
||
|
peer_log("*** STATUS: %d %s", m_parser.status_code(), m_parser.message().c_str());
|
||
|
std::multimap<std::string, std::string> const& headers = m_parser.headers();
|
||
|
for (std::multimap<std::string, std::string>::const_iterator i = headers.begin()
|
||
|
, end(headers.end()); i != end; ++i)
|
||
|
peer_log(" %s: %s", i->first.c_str(), i->second.c_str());
|
||
|
#endif
|
||
|
// if the status code is not one of the accepted ones, abort
|
||
|
if (!is_ok_status(m_parser.status_code()))
|
||
|
{
|
||
|
int retry_time = atoi(m_parser.header("retry-after").c_str());
|
||
|
if (retry_time <= 0) retry_time = 5 * 60;
|
||
|
// temporarily unavailable, retry later
|
||
|
t->retry_web_seed(this, retry_time);
|
||
|
std::string error_msg = to_string(m_parser.status_code()).elems
|
||
|
+ (" " + m_parser.message());
|
||
|
if (m_ses.m_alerts.should_post<url_seed_alert>())
|
||
|
{
|
||
|
m_ses.m_alerts.post_alert(url_seed_alert(t->get_handle(), m_url
|
||
|
, error_msg));
|
||
|
}
|
||
|
m_statistics.received_bytes(0, bytes_transferred);
|
||
|
disconnect(error_code(m_parser.status_code(), get_http_category()), 1);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
if (is_redirect(m_parser.status_code()))
|
||
|
{
|
||
|
// this means we got a redirection request
|
||
|
// look for the location header
|
||
|
std::string location = m_parser.header("location");
|
||
|
m_statistics.received_bytes(0, bytes_transferred);
|
||
|
|
||
|
if (location.empty())
|
||
|
{
|
||
|
// we should not try this server again.
|
||
|
t->remove_web_seed(this);
|
||
|
disconnect(errors::missing_location, 2);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
|
||
|
bool single_file_request = false;
|
||
|
if (!m_path.empty() && m_path[m_path.size() - 1] != '/')
|
||
|
single_file_request = true;
|
||
|
|
||
|
// add the redirected url and remove the current one
|
||
|
if (!single_file_request)
|
||
|
{
|
||
|
TORRENT_ASSERT(!m_file_requests.empty());
|
||
|
int file_index = m_file_requests.front();
|
||
|
|
||
|
torrent_info const& info = t->torrent_file();
|
||
|
std::string path = info.orig_files().file_path(info.orig_files().at(file_index));
|
||
|
#ifdef TORRENT_WINDOWS
|
||
|
convert_path_to_posix(path);
|
||
|
#endif
|
||
|
path = escape_path(path.c_str(), path.length());
|
||
|
size_t i = location.rfind(path);
|
||
|
if (i == std::string::npos)
|
||
|
{
|
||
|
t->remove_web_seed(this);
|
||
|
disconnect(errors::invalid_redirection, 2);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
location.resize(i);
|
||
|
}
|
||
|
t->add_web_seed(location, web_seed_entry::url_seed, m_external_auth, m_extra_headers);
|
||
|
t->remove_web_seed(this);
|
||
|
disconnect(errors::redirecting, 2);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
|
||
|
std::string const& server_version = m_parser.header("server");
|
||
|
if (!server_version.empty())
|
||
|
{
|
||
|
m_server_string = "URL seed @ ";
|
||
|
m_server_string += m_host;
|
||
|
m_server_string += " (";
|
||
|
m_server_string += server_version;
|
||
|
m_server_string += ")";
|
||
|
}
|
||
|
|
||
|
m_body_start = m_parser.body_start();
|
||
|
m_received_body = 0;
|
||
|
m_range_pos = 0;
|
||
|
}
|
||
|
|
||
|
recv_buffer.begin += m_body_start;
|
||
|
|
||
|
// we only received the header, no data
|
||
|
if (recv_buffer.left() == 0)
|
||
|
{
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
break;
|
||
|
}
|
||
|
|
||
|
size_type range_start;
|
||
|
size_type range_end;
|
||
|
if (m_parser.status_code() == 206)
|
||
|
{
|
||
|
boost::tie(range_start, range_end) = m_parser.content_range();
|
||
|
if (range_start < 0 || range_end < range_start)
|
||
|
{
|
||
|
m_statistics.received_bytes(0, bytes_transferred);
|
||
|
// we should not try this server again.
|
||
|
t->remove_web_seed(this);
|
||
|
disconnect(errors::invalid_range);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
// the http range is inclusive
|
||
|
range_end++;
|
||
|
}
|
||
|
else
|
||
|
{
|
||
|
range_start = 0;
|
||
|
range_end = m_parser.content_length();
|
||
|
if (range_end == -1)
|
||
|
{
|
||
|
m_statistics.received_bytes(0, bytes_transferred);
|
||
|
// we should not try this server again.
|
||
|
t->remove_web_seed(this);
|
||
|
disconnect(errors::no_content_length, 2);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
}
|
||
|
|
||
|
if (m_requests.empty() || m_file_requests.empty())
|
||
|
{
|
||
|
m_statistics.received_bytes(0, bytes_transferred);
|
||
|
disconnect(errors::http_error, 2);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
return;
|
||
|
}
|
||
|
|
||
|
// =========================
|
||
|
// === CHUNKED ENCODING ===
|
||
|
// =========================
|
||
|
while (m_parser.chunked_encoding()
|
||
|
&& m_chunk_pos >= 0
|
||
|
&& m_chunk_pos < recv_buffer.left())
|
||
|
{
|
||
|
int header_size = 0;
|
||
|
size_type chunk_size = 0;
|
||
|
buffer::const_interval chunk_start = recv_buffer;
|
||
|
chunk_start.begin += m_chunk_pos;
|
||
|
TORRENT_ASSERT(chunk_start.begin[0] == '\r' || is_hex(chunk_start.begin, 1));
|
||
|
bool ret = m_parser.parse_chunk_header(chunk_start, &chunk_size, &header_size);
|
||
|
if (!ret)
|
||
|
{
|
||
|
TORRENT_ASSERT(int(bytes_transferred) >= chunk_start.left() - m_partial_chunk_header);
|
||
|
bytes_transferred -= chunk_start.left() - m_partial_chunk_header;
|
||
|
m_statistics.received_bytes(0, chunk_start.left() - m_partial_chunk_header);
|
||
|
m_partial_chunk_header = chunk_start.left();
|
||
|
if (bytes_transferred == 0) return;
|
||
|
break;
|
||
|
}
|
||
|
else
|
||
|
{
|
||
|
#ifdef TORRENT_VERBOSE_LOGGING
|
||
|
peer_log("*** parsed chunk: %d header_size: %d", chunk_size, header_size);
|
||
|
#endif
|
||
|
TORRENT_ASSERT(int(bytes_transferred) >= header_size - m_partial_chunk_header);
|
||
|
bytes_transferred -= header_size - m_partial_chunk_header;
|
||
|
m_statistics.received_bytes(0, header_size - m_partial_chunk_header);
|
||
|
m_partial_chunk_header = 0;
|
||
|
TORRENT_ASSERT(chunk_size != 0 || chunk_start.left() <= header_size || chunk_start.begin[header_size] == 'H');
|
||
|
// cut out the chunk header from the receive buffer
|
||
|
TORRENT_ASSERT(m_body_start + m_chunk_pos < INT_MAX);
|
||
|
cut_receive_buffer(header_size, t->block_size() + request_size_overhead, int(m_body_start + m_chunk_pos));
|
||
|
recv_buffer = receive_buffer();
|
||
|
recv_buffer.begin += m_body_start;
|
||
|
m_chunk_pos += chunk_size;
|
||
|
if (chunk_size == 0)
|
||
|
{
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
chunk_start = recv_buffer;
|
||
|
chunk_start.begin += m_chunk_pos;
|
||
|
TORRENT_ASSERT(chunk_start.left() == 0 || chunk_start.begin[0] == 'H');
|
||
|
#endif
|
||
|
m_chunk_pos = -1;
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
|
||
|
size_type left_in_response = range_end - range_start - m_range_pos;
|
||
|
int payload_transferred = int((std::min)(left_in_response, size_type(bytes_transferred)));
|
||
|
|
||
|
torrent_info const& info = t->torrent_file();
|
||
|
|
||
|
peer_request front_request = m_requests.front();
|
||
|
|
||
|
TORRENT_ASSERT(m_block_pos >= 0);
|
||
|
|
||
|
#ifdef TORRENT_VERBOSE_LOGGING
|
||
|
peer_log("*** payload_transferred: %d [ %d:%d = %d ]"
|
||
|
, payload_transferred, front_request.piece
|
||
|
, front_request.start, front_request.length);
|
||
|
#endif
|
||
|
m_statistics.received_bytes(payload_transferred, 0);
|
||
|
TORRENT_ASSERT(int(bytes_transferred) >= payload_transferred);
|
||
|
bytes_transferred -= payload_transferred;
|
||
|
m_range_pos += payload_transferred;
|
||
|
m_block_pos += payload_transferred;
|
||
|
if (m_range_pos > range_end - range_start) m_range_pos = range_end - range_start;
|
||
|
|
||
|
#if 0
|
||
|
std::cerr << "REQUESTS: m_requests: " << m_requests.size()
|
||
|
<< " file_requests: " << m_file_requests.size() << std::endl;
|
||
|
#endif
|
||
|
|
||
|
int file_index = m_file_requests.front();
|
||
|
peer_request in_range = info.orig_files().map_file(file_index, range_start
|
||
|
, int(range_end - range_start));
|
||
|
|
||
|
// request start
|
||
|
size_type rs = size_type(in_range.piece) * info.piece_length() + in_range.start;
|
||
|
// request end
|
||
|
size_type re = rs + in_range.length;
|
||
|
// file start
|
||
|
size_type fs = size_type(front_request.piece) * info.piece_length() + front_request.start;
|
||
|
#if 0
|
||
|
size_type fe = fs + front_request.length;
|
||
|
|
||
|
std::cerr << "RANGE: r = (" << rs << ", " << re << " ) "
|
||
|
"f = (" << fs << ", " << fe << ") "
|
||
|
"file_index = " << file_index << " received_body = " << m_received_body << std::endl;
|
||
|
#endif
|
||
|
|
||
|
// the http response body consists of 3 parts
|
||
|
// 1. the middle of a block or the ending of a block
|
||
|
// 2. a number of whole blocks
|
||
|
// 3. the start of a block
|
||
|
// in that order, these parts are parsed.
|
||
|
|
||
|
bool range_overlaps_request = re > fs + int(m_piece.size());
|
||
|
|
||
|
if (!range_overlaps_request)
|
||
|
{
|
||
|
incoming_piece_fragment((std::min)(payload_transferred
|
||
|
, front_request.length - m_block_pos));
|
||
|
m_statistics.received_bytes(0, bytes_transferred);
|
||
|
// this means the end of the incoming request ends _before_ the
|
||
|
// first expected byte (fs + m_piece.size())
|
||
|
disconnect(errors::invalid_range, 2);
|
||
|
return;
|
||
|
}
|
||
|
|
||
|
// if the request is contained in the range (i.e. the entire request
|
||
|
// fits in the range) we should not start a partial piece, since we soon
|
||
|
// will receive enough to call incoming_piece() and pass the read buffer
|
||
|
// directly (in the next loop below).
|
||
|
if (range_overlaps_request && !range_contains(in_range, front_request, info.piece_length()))
|
||
|
{
|
||
|
// the start of the next block to receive is stored
|
||
|
// in m_piece. We need to append the rest of that
|
||
|
// block from the http receive buffer and then
|
||
|
// (if it completed) call incoming_piece() with
|
||
|
// m_piece as buffer.
|
||
|
|
||
|
int piece_size = int(m_piece.size());
|
||
|
int copy_size = (std::min)((std::min)(front_request.length - piece_size
|
||
|
, recv_buffer.left()), int(range_end - range_start - m_received_body));
|
||
|
if (copy_size > m_chunk_pos && m_chunk_pos > 0) copy_size = m_chunk_pos;
|
||
|
if (copy_size > 0)
|
||
|
{
|
||
|
TORRENT_ASSERT(m_piece.size() == m_received_in_piece);
|
||
|
m_piece.resize(piece_size + copy_size);
|
||
|
std::memcpy(&m_piece[0] + piece_size, recv_buffer.begin, copy_size);
|
||
|
TORRENT_ASSERT(int(m_piece.size()) <= front_request.length);
|
||
|
recv_buffer.begin += copy_size;
|
||
|
m_received_body += copy_size;
|
||
|
m_body_start += copy_size;
|
||
|
if (m_chunk_pos > 0)
|
||
|
{
|
||
|
TORRENT_ASSERT(m_chunk_pos >= copy_size);
|
||
|
m_chunk_pos -= copy_size;
|
||
|
}
|
||
|
TORRENT_ASSERT(m_received_body <= range_end - range_start);
|
||
|
TORRENT_ASSERT(int(m_piece.size()) <= front_request.length);
|
||
|
incoming_piece_fragment(copy_size);
|
||
|
TORRENT_ASSERT(m_piece.size() == m_received_in_piece);
|
||
|
}
|
||
|
|
||
|
if (maybe_harvest_block())
|
||
|
recv_buffer = receive_buffer();
|
||
|
if (associated_torrent().expired()) return;
|
||
|
}
|
||
|
|
||
|
// report all received blocks to the bittorrent engine
|
||
|
while (!m_requests.empty()
|
||
|
&& range_contains(in_range, m_requests.front(), info.piece_length())
|
||
|
&& m_block_pos >= m_requests.front().length)
|
||
|
{
|
||
|
peer_request r = m_requests.front();
|
||
|
TORRENT_ASSERT(recv_buffer.left() >= r.length);
|
||
|
|
||
|
incoming_piece_fragment(r.length);
|
||
|
incoming_piece(r, recv_buffer.begin);
|
||
|
m_requests.pop_front();
|
||
|
if (associated_torrent().expired()) return;
|
||
|
TORRENT_ASSERT(m_block_pos >= r.length);
|
||
|
m_block_pos -= r.length;
|
||
|
m_received_body += r.length;
|
||
|
TORRENT_ASSERT(receive_buffer().begin + m_body_start == recv_buffer.begin);
|
||
|
TORRENT_ASSERT(m_received_body <= range_end - range_start);
|
||
|
cut_receive_buffer(m_body_start + r.length, t->block_size() + request_size_overhead);
|
||
|
if (m_chunk_pos > 0)
|
||
|
{
|
||
|
TORRENT_ASSERT(m_chunk_pos >= r.length);
|
||
|
m_chunk_pos -= r.length;
|
||
|
}
|
||
|
m_body_start = 0;
|
||
|
recv_buffer = receive_buffer();
|
||
|
}
|
||
|
|
||
|
if (!m_requests.empty())
|
||
|
{
|
||
|
if (in_range.start + in_range.length < m_requests.front().start + m_requests.front().length
|
||
|
&& (m_received_body + recv_buffer.left() >= range_end - range_start))
|
||
|
{
|
||
|
int piece_size = int(m_piece.size());
|
||
|
int copy_size = (std::min)((std::min)(m_requests.front().length - piece_size
|
||
|
, recv_buffer.left()), int(range_end - range_start - m_received_body));
|
||
|
TORRENT_ASSERT(copy_size >= 0);
|
||
|
if (copy_size > 0)
|
||
|
{
|
||
|
TORRENT_ASSERT(m_piece.size() == m_received_in_piece);
|
||
|
m_piece.resize(piece_size + copy_size);
|
||
|
std::memcpy(&m_piece[0] + piece_size, recv_buffer.begin, copy_size);
|
||
|
recv_buffer.begin += copy_size;
|
||
|
m_received_body += copy_size;
|
||
|
m_body_start += copy_size;
|
||
|
incoming_piece_fragment(copy_size);
|
||
|
TORRENT_ASSERT(m_piece.size() == m_received_in_piece);
|
||
|
}
|
||
|
TORRENT_ASSERT(m_received_body == range_end - range_start);
|
||
|
}
|
||
|
}
|
||
|
|
||
|
TORRENT_ASSERT(m_received_body <= range_end - range_start);
|
||
|
// if we're in chunked encoding mode, we have to wait for the complete
|
||
|
// tail header before we can consider have received the block, otherwise
|
||
|
// we'll get out of sync with the next http response. m_chunk_pos is set
|
||
|
// to -1 when the tail header has been received
|
||
|
if (m_received_body == range_end - range_start
|
||
|
&& (!m_parser.chunked_encoding() || m_chunk_pos == -1))
|
||
|
{
|
||
|
int size_to_cut = recv_buffer.begin - receive_buffer().begin;
|
||
|
|
||
|
TORRENT_ASSERT(receive_buffer().left() < size_to_cut + 1
|
||
|
|| receive_buffer()[size_to_cut] == 'H');
|
||
|
|
||
|
cut_receive_buffer(size_to_cut, t->block_size() + request_size_overhead);
|
||
|
if (m_chunk_pos > 0)
|
||
|
{
|
||
|
TORRENT_ASSERT(m_chunk_pos >= size_to_cut);
|
||
|
m_chunk_pos -= size_to_cut;
|
||
|
}
|
||
|
recv_buffer = receive_buffer();
|
||
|
m_file_requests.pop_front();
|
||
|
m_parser.reset();
|
||
|
m_body_start = 0;
|
||
|
m_received_body = 0;
|
||
|
m_chunk_pos = 0;
|
||
|
m_partial_chunk_header = 0;
|
||
|
|
||
|
torrent_info const& info = t->torrent_file();
|
||
|
while (!m_file_requests.empty()
|
||
|
&& info.orig_files().internal_at(m_file_requests.front()).pad_file)
|
||
|
{
|
||
|
// the next file is a pad file. We didn't actually send
|
||
|
// a request for this since it most likely doesn't exist on
|
||
|
// the web server anyway. Just pretend that we received a
|
||
|
// bunch of zeroes here and pop it again
|
||
|
int file_index = m_file_requests.front();
|
||
|
m_file_requests.pop_front();
|
||
|
size_type file_size = info.orig_files().file_size(info.orig_files().internal_at(file_index));
|
||
|
TORRENT_ASSERT(m_block_pos < front_request.length);
|
||
|
int pad_size = (std::min)(file_size, size_type(front_request.length - m_block_pos));
|
||
|
|
||
|
// insert zeroes to represent the pad file
|
||
|
m_piece.resize(m_piece.size() + pad_size, 0);
|
||
|
m_block_pos += pad_size;
|
||
|
incoming_piece_fragment(pad_size);
|
||
|
|
||
|
if (maybe_harvest_block())
|
||
|
recv_buffer = receive_buffer();
|
||
|
if (associated_torrent().expired()) return;
|
||
|
}
|
||
|
continue;
|
||
|
}
|
||
|
if (bytes_transferred == 0 || payload_transferred == 0)
|
||
|
{
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded()
|
||
|
== dl_target);
|
||
|
#endif
|
||
|
break;
|
||
|
}
|
||
|
TORRENT_ASSERT(payload_transferred > 0);
|
||
|
}
|
||
|
TORRENT_ASSERT(bytes_transferred == 0);
|
||
|
#ifdef TORRENT_DEBUG
|
||
|
TORRENT_ASSERT(m_statistics.last_payload_downloaded()
|
||
|
+ m_statistics.last_protocol_downloaded() == dl_target);
|
||
|
#endif
|
||
|
}
|
||
|
|
||
|
void web_peer_connection::get_specific_peer_info(peer_info& p) const
|
||
|
{
|
||
|
web_connection_base::get_specific_peer_info(p);
|
||
|
p.flags |= peer_info::local_connection;
|
||
|
p.connection_type = peer_info::web_seed;
|
||
|
}
|
||
|
|
||
|
}
|
||
|
|