code.delx.au - webdl/blob - common.py

   1 import hashlib
   2 import io
   3 import json
   4 import logging
   5 import lxml.etree
   6 import lxml.html
   7 import os
   8 import re
   9 import requests
  10 import requests_cache
  11 import shutil
  12 import signal
  13 import subprocess
  14 import time
  15 import urllib.parse
  16
  17
  18 try:
  19     import autosocks
  20     autosocks.try_autosocks()
  21 except ImportError:
  22     pass
  23
  24
  25 logging.basicConfig(
  26     format = "%(levelname)s %(message)s",
  27     level = logging.INFO if os.environ.get("DEBUG", None) is None else logging.DEBUG,
  28 )
  29
  30 CACHE_FILE = os.path.join(
  31     os.environ.get("XDG_CACHE_HOME", os.path.expanduser("~/.cache")),
  32     "webdl",
  33     "requests_cache"
  34 )
  35 if not os.path.isdir(os.path.dirname(CACHE_FILE)):
  36     os.makedirs(os.path.dirname(CACHE_FILE))
  37
  38 requests_cache.install_cache(CACHE_FILE, backend='sqlite', expire_after=3600)
  39
  40
  41 class Node(object):
  42     def __init__(self, title, parent=None):
  43         self.title = title
  44         if parent:
  45             parent.children.append(self)
  46         self.parent = parent
  47         self.children = []
  48         self.can_download = False
  49
  50     def get_children(self):
  51         if not self.children:
  52             self.fill_children()
  53         return self.children
  54
  55     def fill_children(self):
  56         pass
  57
  58     def download(self):
  59         raise NotImplemented
  60
  61
  62 def load_root_node():
  63     root_node = Node("Root")
  64
  65     import iview
  66     iview.fill_nodes(root_node)
  67
  68     import sbs
  69     sbs.fill_nodes(root_node)
  70
  71     import brightcove
  72     brightcove.fill_nodes(root_node)
  73
  74     return root_node
  75
  76 valid_chars = frozenset("-_.()!@#%^ abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789")
  77 def sanify_filename(filename):
  78     filename = "".join(c for c in filename if c in valid_chars)
  79     assert len(filename) > 0
  80     return filename
  81
  82 def ensure_scheme(url):
  83     parts = urllib.parse.urlparse(url)
  84     if parts.scheme:
  85         return url
  86     parts = list(parts)
  87     parts[0] = "http"
  88     return urllib.parse.urlunparse(parts)
  89
  90 http_session = requests.Session()
  91 http_session.headers["User-Agent"] = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:21.0) Gecko/20100101 Firefox/21.0"
  92
  93 def grab_text(url):
  94     logging.debug("grab_text(%r)", url)
  95     request = http_session.prepare_request(requests.Request("GET", url))
  96     response = http_session.send(request)
  97     return response.text
  98
  99 def grab_html(url):
 100     logging.debug("grab_html(%r)", url)
 101     request = http_session.prepare_request(requests.Request("GET", url))
 102     response = http_session.send(request, stream=True)
 103     doc = lxml.html.parse(io.BytesIO(response.content), lxml.html.HTMLParser(encoding="utf-8", recover=True))
 104     response.close()
 105     return doc
 106
 107 def grab_xml(url):
 108     logging.debug("grab_xml(%r)", url)
 109     request = http_session.prepare_request(requests.Request("GET", url))
 110     response = http_session.send(request, stream=True)
 111     doc = lxml.etree.parse(io.BytesIO(response.content), lxml.etree.XMLParser(encoding="utf-8", recover=True))
 112     response.close()
 113     return doc
 114
 115 def grab_json(url):
 116     logging.debug("grab_json(%r)", url)
 117     request = http_session.prepare_request(requests.Request("GET", url))
 118     response = http_session.send(request)
 119     return response.json()
 120
 121 def exec_subprocess(cmd):
 122     logging.debug("Executing: %s", cmd)
 123     try:
 124         p = subprocess.Popen(cmd)
 125         ret = p.wait()
 126         if ret != 0:
 127             logging.error("%s exited with error code: %s", cmd[0], ret)
 128             return False
 129         else:
 130             return True
 131     except OSError as e:
 132         logging.error("Failed to run: %s -- %s", cmd[0], e)
 133     except KeyboardInterrupt:
 134         logging.info("Cancelled: %s", cmd)
 135         try:
 136             p.terminate()
 137             p.wait()
 138         except KeyboardInterrupt:
 139             p.send_signal(signal.SIGKILL)
 140             p.wait()
 141     return False
 142
 143
 144 def check_command_exists(cmd):
 145     try:
 146         subprocess.check_output(cmd, stderr=subprocess.STDOUT)
 147         return True
 148     except Exception:
 149         return False
 150
 151 def find_ffmpeg():
 152     for ffmpeg in ["avconv", "ffmpeg"]:
 153         if check_command_exists([ffmpeg, "--help"]):
 154             return ffmpeg
 155
 156     raise Exception("You must install ffmpeg or libav-tools")
 157
 158 def find_ffprobe():
 159     for ffprobe in ["avprobe", "ffprobe"]:
 160         if check_command_exists([ffprobe, "--help"]):
 161             return ffprobe
 162
 163     raise Exception("You must install ffmpeg or libav-tools")
 164
 165 def get_duration(filename):
 166     ffprobe = find_ffprobe()
 167
 168     cmd = [
 169         ffprobe,
 170         filename,
 171         "-show_format_entry", "duration",
 172         "-v", "quiet",
 173     ]
 174     output = subprocess.check_output(cmd).decode("utf-8")
 175     for line in output.split("\n"):
 176         if line.startswith("duration="):
 177             return float(line.split("=")[1]) # ffprobe
 178         if re.match(R'^[0-9.]*$', line):
 179             return float(line) # avprobe
 180
 181     raise Exception("Unable to determine video duration of " + filename)
 182
 183 def check_video_durations(flv_filename, mp4_filename):
 184     flv_duration = get_duration(flv_filename)
 185     mp4_duration = get_duration(mp4_filename)
 186
 187     if abs(flv_duration - mp4_duration) > 1:
 188         logging.error(
 189             "The duration of %s is suspicious, did the remux fail? Expected %s == %s",
 190             mp4_filename, flv_duration, mp4_duration
 191         )
 192         return False
 193
 194     return True
 195
 196 def remux(infile, outfile):
 197     logging.info("Converting %s to mp4", infile)
 198
 199     ffmpeg = find_ffmpeg()
 200     cmd = [
 201         ffmpeg,
 202         "-i", infile,
 203         "-bsf:a", "aac_adtstoasc",
 204         "-acodec", "copy",
 205         "-vcodec", "copy",
 206         "-y",
 207         outfile,
 208     ]
 209     if not exec_subprocess(cmd):
 210         return False
 211
 212     if not check_video_durations(infile, outfile):
 213         return False
 214
 215     os.unlink(infile)
 216     return True
 217
 218 def convert_to_mp4(filename):
 219     with open(filename, "rb") as f:
 220         fourcc = f.read(4)
 221     basename, ext = os.path.splitext(filename)
 222
 223     if ext == ".mp4" and fourcc == b"FLV\x01":
 224         os.rename(filename, basename + ".flv")
 225         ext = ".flv"
 226         filename = basename + ext
 227
 228     if ext in (".flv", ".ts"):
 229         filename_mp4 = basename + ".mp4"
 230         return remux(filename, filename_mp4)
 231
 232     return ext == ".mp4"
 233
 234
 235 def download_hds(filename, video_url, pvswf=None):
 236     filename = sanify_filename(filename)
 237     logging.info("Downloading: %s", filename)
 238
 239     video_url = "hds://" + video_url
 240     if pvswf:
 241         param = "%s pvswf=%s" % (video_url, pvswf)
 242     else:
 243         param = video_url
 244
 245     cmd = [
 246         "livestreamer",
 247         "-f",
 248         "-o", filename,
 249         param,
 250         "best",
 251     ]
 252     if exec_subprocess(cmd):
 253         return convert_to_mp4(filename)
 254     else:
 255         return False
 256
 257 def download_hls(filename, video_url):
 258     filename = sanify_filename(filename)
 259     video_url = "hlsvariant://" + video_url
 260     logging.info("Downloading: %s", filename)
 261
 262     cmd = [
 263         "livestreamer",
 264         "-f",
 265         "-o", filename,
 266         video_url,
 267         "best",
 268     ]
 269     if exec_subprocess(cmd):
 270         return convert_to_mp4(filename)
 271     else:
 272         return False
 273
 274 def download_http(filename, video_url):
 275     filename = sanify_filename(filename)
 276     logging.info("Downloading: %s", filename)
 277
 278     cmd = [
 279         "curl",
 280         "--fail", "--retry", "3",
 281         "-o", filename,
 282         video_url,
 283     ]
 284     if exec_subprocess(cmd):
 285         return convert_to_mp4(filename)
 286     else:
 287         return False
 288
 289 def natural_sort(l, key=None):
 290     ignore_list = ["a", "the"]
 291     def key_func(k):
 292         if key is not None:
 293             k = key(k)
 294         k = k.lower()
 295         newk = []
 296         for c in re.split("([0-9]+)", k):
 297             c = c.strip()
 298             if c.isdigit():
 299                 newk.append(c.zfill(5))
 300             else:
 301                 for subc in c.split():
 302                     if subc not in ignore_list:
 303                         newk.append(subc)
 304         return newk
 305
 306     return sorted(l, key=key_func)
 307
 308 def append_to_qs(url, params):
 309     r = list(urllib.parse.urlsplit(url))
 310     qs = urllib.parse.parse_qs(r[3])
 311     for k, v in params.items():
 312         if v is not None:
 313             qs[k] = v
 314         elif k in qs:
 315             del qs[k]
 316     r[3] = urllib.parse.urlencode(sorted(qs.items()), True)
 317     url = urllib.parse.urlunsplit(r)
 318     return url
 319