script.module.gorrent2http/README.md

109 lines
4.6 KiB
Markdown

script.module.torrent2http
==========================
This add-on is binding to [torrent2http](https://github.com/anteo/torrent2http) client.
It is bundled with torrent2http binaries for Android ARM, Linux x86/x64/ARM, Windows x86/x64 and Darwin/OSX x64 platforms.
You can download it from my [repository](http://bit.ly/184XKjm)
This add-on can be used to stream media files from torrents without need to download entire files.
Internally, it runs pre-compiled torrent2http client binary, which starts local HTTP server, presenting contents of torrent.
Next, request to HTTP server can be sent to receive list of files or to start streaming of needed file inside of torrent.
Usage
-----
### Get list of files inside torrent ###
Getting list of files inside torrent is straightforward:
import xbmc
from torrent2http import State, Engine, MediaType
from contextlib import closing
# Create instance of Engine
engine = Engine(uri="...")
files = []
# Ensure we'll close engine on exception
with closing(engine):
# Start engine
engine.start()
# Wait until files received
while not files and not xbmc.abortRequested:
# Will list only video files in torrent
files = engine.list(media_types=[MediaType.VIDEO])
# Check if there is loading torrent error and raise exception
engine.check_torrent_error()
xbmc.sleep(200)
### Start streaming ###
import xbmc
from torrent2http import State, Engine, MediaType
from contextlib import closing
# XBMC addon handle
handle = ...
# Playable list item
listitem = ...
# We can know file_id of needed video file on this step, if no, we'll try to detect one.
file_id = None
# Flag will set to True when engine is ready to resolve URL to XBMC
ready = False
# Set pre-buffer size to 15Mb. This is a size of file that need to be downloaded before we resolve URL to XMBC
pre_buffer_bytes = 15*1024*1024
engine = Engine(uri="...")
with closing(engine):
# Start engine and instruct torrent2http to begin download first file,
# so it can start searching and connecting to peers
engine.start(file_id or 0)
while not xbmc.abortRequested and not ready:
xbmc.sleep(500)
status = engine.status()
# Check if there is loading torrent error and raise exception
engine.check_torrent_error(status)
# Trying to detect file_id
if file_id is None:
# Get torrent files list, filtered by video file type only
files = engine.list(media_types=[MediaType.VIDEO])
# If torrent metadata is not loaded yet then continue
if files is None:
continue
# Torrent has no video files
if not files:
break
# Select first matching file
file_id = files[0].index
file_status = files[0]
else:
# If we've got file_id already, get file status
file_status = engine.file_status(file_id)
# If torrent metadata is not loaded yet then continue
if not file_status:
continue
if status.state == State.DOWNLOADING:
# Wait until minimum pre_buffer_bytes downloaded before we resolve URL to XBMC
if file_status.download >= pre_buffer_bytes:
ready = True
break
elif status.state in [State.FINISHED, State.SEEDING]:
# We have already downloaded file
ready = True
break
# Here you can update pre-buffer progress dialog, for example.
# Note that State.CHECKING also need waiting until fully finished, so it better to use resume_file option
# for engine to avoid CHECKING state if possible.
# ...
if ready:
# Resolve URL to XBMC
listitem.SetPath(file_status.url)
xbmcplugin.SetResolvedUrl(handle, True, listitem)
# Wait until playing finished or abort requested
while not xbmc.abortRequested and xbmc.Player().isPlaying():
xbmc.sleep(500)
### Fully working example ###
You can look into fully working [example](https://github.com/anteo/plugin.video.okino/blob/master/resources/lib/okino/torrent/stream/t2h_stream.py). This is
excerpt from my plugin using script.module.torrent2http. It rather complicated but it includes error handling, showing progress dialog and other features.