creating torrents

Author: Arvid Norberg, arvid@rasterbar.com
Version: 1.0.0

overview

This section describes the functions and classes that are used to create torrent files. It is a layered API with low level classes and higher level convenience functions. A torrent is created in 4 steps:

  1. first the files that will be part of the torrent are determined.
  2. the torrent properties are set, such as tracker url, web seeds, DHT nodes etc.
  3. Read through all the files in the torrent, SHA-1 all the data and set the piece hashes.
  4. The torrent is bencoded into a file or buffer.

If there are a lot of files and or deep directoy hierarchies to traverse, step one can be time consuming.

Typically step 3 is by far the most time consuming step, since it requires to read all the bytes from all the files in the torrent.

All of these classes and functions are declared by including libtorrent/create_torrent.hpp.

high level example

file_storage fs;

// recursively adds files in directories
add_files(fs, "./my_torrent");

create_torrent t(fs);
t.add_tracker("http://my.tracker.com/announce");
t.set_creator("libtorrent example");

// reads the files and calculates the hashes
set_piece_hashes(t, ".");

ofstream out("my_torrent.torrent", std::ios_base::binary);
bencode(std::ostream_iterator<char>(out), t.generate());

add_files

template <class Pred>
void add_files(file_storage& fs, std::string const& path, Pred p
        , boost::uint32_t flags = 0);
template <class Pred>
void add_files(file_storage& fs, std::wstring const& path, Pred p
        , boost::uint32_t flags = 0);

void add_files(file_storage& fs, std::string const& path
        , boost::uint32_t flags = 0);
void add_files(file_storage& fs, std::wstring const& path
        , boost::uint32_t flags = 0);

Adds the file specified by path to the file_storage object. In case path refers to a diretory, files will be added recursively from the directory.

If specified, the predicate p is called once for every file and directory that is encountered. files for which p returns true are added, and directories for which p returns true are traversed. p must have the following signature:

bool Pred(std::string const& p);

and for the wide string version:

bool Pred(std::wstring const& p);

The path that is passed in to the predicate is the full path of the file or directory. If no predicate is specified, all files are added, and all directories are traveresed.

The ".." directory is never traversed.

The flags argument should be the same as the flags passed to the create_torrent constructor.

set_piece_hashes()

template <class Fun>
void set_piece_hashes(create_torrent& t, std::string const& p, Fun f);
template <class Fun>
void set_piece_hashes(create_torrent& t, std::wstring const& p, Fun f);
template <class Fun>
void set_piece_hashes(create_torrent& t, std::string const& p, Fun f
        , error_code& ec);
template <class Fun>
void set_piece_hashes(create_torrent& t, std::wstring const& p, Fun f
        , error_code& ec);

void set_piece_hashes(create_torrent& t, std::string const& p);
void set_piece_hashes(create_torrent& t, std::wstring const& p);
void set_piece_hashes(create_torrent& t, std::string const& p
        , error_code& ec);
void set_piece_hashes(create_torrent& t, std::wstring const& p
        , error_code& ec);

This function will assume that the files added to the torrent file exists at path p, read those files and hash the content and set the hashes in the create_torrent object. The optional function f is called in between every hash that is set. f must have the following signature:

void Fun(int);

The overloads that don't take an error_code& may throw an exception in case of a file error, the other overloads sets the error code to reflect the error, if any.

file_storage

The file_storage class represents a file list and the piece size. Everything necessary to interpret a regular bittorrent storage file structure. Its synopsis:

class file_storage
{
public:

        bool is_valid() const;

        enum flags_t
        {
                pad_file = 1,
                attribute_hidden = 2,
                attribute_executable = 4
        };

        void add_file(file_entry const& e);
        void add_file(std::string const& p, size_type size, int flags = 0);
        void add_file(std::wstring const& p, size_type size, int flags = 0);
        void rename_file(int index, std::string const& new_filename);
        void rename_file(int index, std::wstring const& new_filename);

        std::vector<file_slice> map_block(int piece, size_type offset
                , int size) const;
        peer_request map_file(int file, size_type offset, int size) const;

        typedef std::vector<internal_file_entry>::const_iterator iterator;
        typedef std::vector<internal_file_entry>::const_reverse_iterator reverse_iterator;

        iterator begin() const;
        iterator end() const;
        reverse_iterator rbegin();
        reverse_iterator rend() const;
        int num_files() const;

        file_entry at(int index) const;

        size_type total_size() const;
        void set_num_pieces(int n);
        int num_pieces() const;
        void set_piece_length(int l);
        int piece_length() const;
        int piece_size(int index) const;

        // index accessors
        sha1_hash const& hash(int index) const;
        std::string const& symlink(int index) const;
        time_t mtime(int index) const;
        size_type file_base(int index) const;
        void set_file_base(int index, size_type off);
        std::string file_path(int index) const;
        std::string file_name(int index) const;
        size_type file_size(int index) const;
        bool pad_file_at(int index) const;
        size_type file_offset(int index) const;

        // iterator accessors
        sha1_hash hash(internal_file_entry const& fe) const;
        std::string const& symlink(internal_file_entry const& fe) const;
        time_t mtime(internal_file_entry const& fe) const;
        int file_index(internal_file_entry const& fe) const;
        size_type file_base(internal_file_entry const& fe) const;
        void set_file_base(internal_file_entry const& fe, size_type off);
        std::string file_path(internal_file_entry const& fe) const;
        std::string file_name(internal_file_entry const& fe) const;
        size_type file_size(internal_file_entry const& fe) const;
        bool pad_file_at(internal_file_entry const& fe) const;
        size_type file_offset(internal_file_entry const& fe) const;

        void set_name(std::string const& n);
        void set_name(std::wstring const& n);
        const std::string& name() const;

        void swap(file_storage& ti);
}

add_file()

void add_file(file_entry const& e);
void add_file(std::string const& p, size_type size, int flags = 0);
void add_file(std::wstring const& p, size_type size, int flags = 0);

Adds a file to the file storage. The flags argument sets attributes on the file. The file attributes is an extension and may not work in all bittorrent clients. The possible arreibutes are:

pad_file
attribute_hidden
attribute_executable

If more files than one are added, certain restrictions to their paths apply. In a multi-file file storage (torrent), all files must share the same root directory.

That is, the first path element of all files must be the same. This shared path element is also set to the name of the torrent. It can be changed by calling set_name.

The built in functions to traverse a directory to add files will make sure this requirement is fulfilled.

file_base() set_file_base()

size_type file_base(int index) const;
void set_file_base(int index, size_type off);

The file base of a file is the offset within the file on the filsystem where it starts to write. For the most part, this is always 0. It's possible to map several files (in the torrent) into a single file on the filesystem by making them all point to the same filename, but with different file bases, so that they don't overlap. torrent_info::remap_files can be used to use a new file layout.

create_torrent

The create_torrent class has the following synopsis:

struct create_torrent
{
        enum {
                optimize = 1
                , merkle = 2
                , modification_time = 4
                , symlinks = 8
                , calculate_file_hashes = 16
        };
        create_torrent(file_storage& fs, int piece_size = 0, int pad_size_limit = -1
                , int flags = optimize, int alignment = 0x4000);
        create_torrent(torrent_info const& ti);

        entry generate() const;

        file_storage const& files() const;

        void set_comment(char const* str);
        void set_creator(char const* str);
        void set_hash(int index, sha1_hash const& h);
        void set_file_hash(int index, sha1_hash const& h);
        void add_url_seed(std::string const& url);
        void add_http_seed(std::string const& url);
        void add_node(std::pair<std::string, int> const& node);
        void add_tracker(std::string const& url, int tier = 0);
        void set_root_cert(std::string const& pem);
        void set_priv(bool p);

        int num_pieces() const;
        int piece_length() const;
        int piece_size(int i) const;
        bool priv() const;
};

create_torrent()

enum {
        optimize = 1
        , merkle = 2
        , modification_time = 4
        , symlinks = 8
        , calculate_file_hashes = 16
};
create_torrent(file_storage& fs, int piece_size = 0, int pad_size_limit = -1
        , int flags = optimize, int alignment = 0x4000);
create_torrent(torrent_info const& ti);

The piece_size is the size of each piece in bytes. It must be a multiple of 16 kiB. If a piece size of 0 is specified, a piece_size will be calculated such that the torrent file is roughly 40 kB.

If a pad_size_limit is specified (other than -1), any file larger than the specified number of bytes will be preceeded by a pad file to align it with the start of a piece. The pad_file_limit is ignored unless the optimize flag is passed. Typically it doesn't make sense to set this any lower than 4kiB.

The overload that takes a torrent_info object will make a verbatim copy of its info dictionary (to preserve the info-hash). The copy of the info dictionary will be used by generate(). This means that none of the member functions of create_torrent that affects the content of the info dictionary (such as set_hash()), will have any affect.

The flags arguments specifies options for the torrent creation. It can be any combination of the following flags:

optimize
This will insert pad files to align the files to piece boundaries, for optimized disk-I/O.
merkle
This will create a merkle hash tree torrent. A merkle torrent cannot be opened in clients that don't specifically support merkle torrents. The benefit is that the resulting torrent file will be much smaller and not grow with more pieces. When this option is specified, it is recommended to have a fairly small piece size, say 64 kiB. When creating merkle torrents, the full hash tree is also generated and should be saved off separately. It is accessed through the merkle_tree() function.
modification_time
This will include the file modification time as part of the torrent. This is not enabled by default, as it might cause problems when you create a torrent from separate files with the same content, hoping to yield the same info-hash. If the files have different modification times, with this option enabled, you would get different info-hashes for the files.
symlink
If this flag is set, files that are symlinks get a symlink attribute set on them and their data will not be included in the torrent. This is useful if you need to reconstruct a file hierarchy which contains symlinks.
calculate_file_hashes
If this is set, the set_piece_hashes() function will, as it calculates the piece hashes, also calculate the file hashes and add those associated with each file. Note that unless you use the set_piece_hashes() function, this flag will have no effect.

alignment is used when pad files are enabled. This is the size eligible files are aligned to. The default is the default bittorrent block size of 16 kiB. It is common to align to the piece size of the torrent.

generate()

entry generate() const;

This function will generate the .torrent file as a bencode tree. In order to generate the flat file, use the bencode() function.

It may be useful to add custom entries to the torrent file before bencoding it and saving it to disk.

If anything goes wrong during torrent generation, this function will return an empty entry structure. You can test for this condition by querying the type of the entry:

file_storage fs;
// add file ...
create_torrent t(fs);
// add trackers and piece hashes ...
e = t.generate();

if (e.type() == entry::undefined_t)
{
        // something went wrong
}

For instance, you cannot generate a torrent with 0 files in it. If you don't add any files to the file_storage, torrent generation will fail.

set_comment()

void set_comment(char const* str);

Sets the comment for the torrent. The string str should be utf-8 encoded. The comment in a torrent file is optional.

set_creator()

void set_creator(char const* str);

Sets the creator of the torrent. The string str should be utf-8 encoded. This is optional.

set_hash()

void set_hash(int index, sha1_hash const& h);

This sets the SHA-1 hash for the specified piece (index). You are required to set the hash for every piece in the torrent before generating it. If you have the files on disk, you can use the high level convenience function to do this. See set_piece_hashes().

set_file_hash()

void set_file_hash(int index, sha1_hash const& h);

This sets the sha1 hash for this file. This hash will end up under the key sha1 associated with this file (for multi-file torrents) or in the root info dictionary for single-file torrents.

add_url_seed() add_http_seed()

void add_url_seed(std::string const& url);
void add_http_seed(std::string const& url);

This adds a url seed to the torrent. You can have any number of url seeds. For a single file torrent, this should be an HTTP url, pointing to a file with identical content as the file of the torrent. For a multi-file torrent, it should point to a directory containing a directory with the same name as this torrent, and all the files of the torrent in it.

The second function, add_http_seed() adds an HTTP seed instead.

add_node()

void add_node(std::pair<std::string, int> const& node);

This adds a DHT node to the torrent. This especially useful if you're creating a tracker less torrent. It can be used by clients to bootstrap their DHT node from. The node is a hostname and a port number where there is a DHT node running. You can have any number of DHT nodes in a torrent.

add_tracker()

void add_tracker(std::string const& url, int tier = 0);

Adds a tracker to the torrent. This is not strictly required, but most torrents use a tracker as their main source of peers. The url should be an http:// or udp:// url to a machine running a bittorrent tracker that accepts announces for this torrent's info-hash. The tier is the fallback priority of the tracker. All trackers with tier 0 are tried first (in any order). If all fail, trackers with tier 1 are tried. If all of those fail, trackers with tier 2 are tried, and so on.

set_root_cert()

void set_root_cert(std::string const& pem);

This function sets an X.509 certificate in PEM format to the torrent. This makes the torrent an SSL torrent. An SSL torrent requires that each peer has a valid certificate signed by this root certificate. For SSL torrents, all peers are connecting over SSL connections. For more information on SSL torrents, see the manual.

The string is not the path to the cert, it's the actual content of the certificate, loaded into a std::string.

set_priv() priv()

void set_priv(bool p);
bool priv() const;

Sets and queries the private flag of the torrent.

merkle_tree()

std::vector<sha1_hash> const& merkle_tree() const;

This function returns the merkle hash tree, if the torrent was created as a merkle torrent. The tree is created by generate() and won't be valid until that function has been called. When creating a merkle tree torrent, the actual tree itself has to be saved off separately and fed into libtorrent the first time you start seeding it, through the torrent_info::set_merkle_tree() function. From that point onwards, the tree will be saved in the resume data.