X-Git-Url: https://code.delx.au/youtube-cgi/blobdiff_plain/50ddeee3bf353c6bfd4130a7ad6948277c386e4b..a34ea48b292a73519fa06c4bfdd11d7f0730d42f:/youtube.cgi diff --git a/youtube.cgi b/youtube.cgi index 293db8a..a54fa91 100755 --- a/youtube.cgi +++ b/youtube.cgi @@ -1,8 +1,9 @@ #!/usr/bin/env python +from __future__ import division + import cookielib import cgi -import itertools import json from lxml import html import os @@ -11,6 +12,7 @@ import resource import shutil import subprocess import sys +import time import urllib import urllib2 import urlparse @@ -26,6 +28,8 @@ MIMETYPES = { } QUALITIES = { + "hd1080": 5, + "hd720": 4, "large": 3, "medium": 2, "small": 1, @@ -74,14 +78,27 @@ cookiejar = cookielib.CookieJar() urlopener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cookiejar)) referrer = "" -def urlopen(url): +def urlopen(url, offset=None): global referrer req = urllib2.Request(url) if referrer: req.add_header("Referer", referrer) referrer = url + req.add_header("User-Agent", USER_AGENT) - return urlopener.open(req) + + if offset: + req.add_header("Range", "bytes=%d-" % offset) + + res = urlopener.open(req) + + content_range = res.info().getheader("Content-Range") + if content_range: + tokens = content_range.split() + assert tokens[0] == "bytes" + start = int(tokens[1].split("-")[0]) + assert start == offset + return res def parse_url(url): f = urlopen(url) @@ -124,22 +141,26 @@ def get_player_config(doc): return player_config def get_best_video(player_config): - url_data = urlparse.parse_qs(player_config["args"]["url_encoded_fmt_stream_map"]) - url_data = itertools.izip_longest( - url_data["url"], - url_data["type"], - url_data["quality"], - url_data.get("sig", []), - ) + url_data_list = player_config["args"]["url_encoded_fmt_stream_map"].split(",") + best_url = None best_quality = None best_extension = None - for video_url, mimetype, quality, signature in url_data: - mimetype = mimetype.split(";")[0] + for url_data in url_data_list: + url_data = urlparse.parse_qs(url_data) + video_url = url_data["url"][0] + mimetype = url_data["type"][0].split(";")[0] + quality = url_data["quality"][0] + signature = url_data["sig"][0] + + if quality not in QUALITIES: + continue if mimetype not in MIMETYPES: continue + extension = MIMETYPES[mimetype] - quality = QUALITIES.get(quality.split(",")[0], -1) + quality = QUALITIES.get(quality, -1) + if best_quality is None or quality > best_quality: if signature: video_url = append_to_qs(video_url, {"signature": signature}) @@ -210,18 +231,94 @@ def cgimain(): ) return +def pp_size(size): + suffixes = ["", "KiB", "MiB", "GiB"] + for i, suffix in enumerate(suffixes): + if size < 1024: + break + size /= 1024 + return "%.2f %s" % (size, suffix) + +def copy_with_progress(content_length, infile, outfile): + def print_status(): + rate = 0 + if now != last_ts: + rate = last_bytes_read / (now - last_ts) + sys.stdout.write("\33[2K\r") + sys.stdout.write("%s / %s (%s/sec)" % ( + pp_size(bytes_read), + pp_size(content_length), + pp_size(rate), + )) + sys.stdout.flush() + + last_ts = 0 + last_bytes_read = 0 + bytes_read = 0 + while True: + now = time.time() + if now - last_ts > 0.5: + print_status() + last_ts = now + last_bytes_read = 0 + + buf = infile.read(32768) + if not buf: + break + outfile.write(buf) + last_bytes_read += len(buf) + bytes_read += len(buf) + + # Newline at the end + print_status() + print + def main(): try: url = sys.argv[1] except: print >>sys.stderr, "Usage: %s http://youtube.com/watch?v=FOOBAR" % sys.argv[0] sys.exit(1) + doc = parse_url(url) video_url, filename = get_video_url(doc) - data = urlopen(video_url) - outfile = open(filename, "w") - shutil.copyfileobj(data, outfile) - data.close() + print "Downloading", filename.encode("utf-8") + + outfile = open(filename, "a") + offset = outfile.tell() + if offset > 0: + print "Resuming download from", pp_size(offset) + total_size = None + + while True: + try: + video_data = urlopen(video_url, offset) + except urllib2.HTTPError, e: + if e.code == 416: + print "File is complete!" + break + else: + raise + + content_length = int(video_data.info().getheader("Content-Length")) + if total_size is None: + total_size = content_length + + try: + copy_with_progress(content_length, video_data, outfile) + except IOError, e: + print + + video_data.close() + if outfile.tell() != total_size: + old_offset = offset + offset = outfile.tell() + if old_offset == offset: + time.sleep(1) + print "Restarting download from", pp_size(offset) + else: + break + outfile.close() @@ -230,5 +327,9 @@ if __name__ == "__main__": if os.environ.has_key("SCRIPT_NAME"): cgimain() else: - main() + try: + main() + except KeyboardInterrupt: + print "\nExiting..." + sys.exit(1)