Merge branch 'develop'

Conflicts:
	CHANGELOG
	pafy/pafy.py
	setup.py

CHANGELOG

@@ -1,3 +1,11 @@
+19 June 2014
+Version 0.3.50
+
+[Feature] - Resume partially downloaded files (issue #32)
+[Bugfix] - Fix javascript function extraction
+
+-------------------------------------------------------------------------------
+
 12 June 2014
 Version 0.3.48
 

pafy/pafy.py

@@ -25,7 +25,7 @@
 
 from __future__ import unicode_literals
 
-__version__ = "0.3.48"
+__version__ = "0.3.50"
 __author__ = "nagev"
 __license__ = "GPLv3"
 
@@ -275,7 +275,7 @@ def _extract_function_from_js(name, js):
 def _get_mainfunc_from_js(js):
     """ Return main signature decryption function from javascript as dict. """
     dbg("Scanning js for main function.")
-    m = re.search(r'\w\.sig\|\|(\w+)\(\w+\.\w+\)', js)
+    m = re.search(r'\w\.sig\|\|([$\w]+)\(\w+\.\w+\)', js)
     funcname = m.group(1)
     dbg("Found main function: %s", funcname)
     function = _extract_function_from_js(funcname, js)
@@ -685,6 +685,15 @@ def download(self, filepath="", quiet=False, callback=None):
         """ Download.  Use quiet=True to supress output. Return filename. """
         # pylint: disable=R0914
         # Too many local variables - who cares?
+
+        def safe_filename(ff):
+            """ Remove troublesome characters in basename. """
+            d, b = os.path.split(ff)
+            ok = re.compile(r'[^\\/?*$\'"%&:<>|]')
+            b = "".join(x if ok.match(x) else "_" for x in b)
+            ff = os.path.join(d, b)
+            return ff.encode("utf8", "ignore")
+
         status_string = ('  {:,} Bytes [{:.2%}] received. Rate: [{:4.0f} '
                          'KB/s].  ETA: [{:.0f} secs]')
 
@@ -696,21 +705,49 @@ def download(self, filepath="", quiet=False, callback=None):
         total = int(response.info()['Content-Length'].strip())
         chunksize, bytesdone, t0 = 16384, 0, time.time()
         fname = filepath or self.filename
+        dirname = os.path.dirname(fname)
+
+        if dirname and not os.path.isdir(dirname):
+            raise IOError("Invalid directory: %s" % dirname)
+
+        elif os.path.isdir(fname):
+            fname = os.path.join(fname, self.filename)
+
+        temp = os.path.join(os.path.dirname(fname), self._parent.title)
+        tempfname = "%s_%s_%s.part" % (temp, self._parent.videoid, self.itag)
+        safename, fmode, offset = safe_filename(tempfname), "wb", 0
+
+        if os.path.exists(tempfname) and os.stat(tempfname).st_size < total:
+            offset = os.stat(tempfname).st_size
+            fmode = "ab"
+
+        elif os.path.exists(safename) and os.stat(safename).st_size < total:
+            offset = os.stat(safename).st_size
+            fmode = "ab"
 
         try:
-            outfh = open(fname, 'wb')
+            outfh = open(tempfname, fmode)
 
         except IOError:
-            ok = re.compile(r'[^\\/?*$\'"%&:<>|]')
-            fname = "".join(x if ok.match(x) else "_" for x in self.filename)
-            outfh = open(fname.encode("utf8", "ignore"), 'wb')
+            # remove special chars from filename
+            fname = safe_filename(fname)
+            tempfname = safename
+            outfh = open(tempfname, fmode)
+
+        if offset:
+            # partial file exists, resume download
+            resuming_opener = build_opener()
+            resuming_opener.addheaders = [('User-Agent', g.ua),
+                                          ("Range", "bytes=%s-" % offset)]
+            response = resuming_opener.open(self.url)
+            bytesdone = offset
 
         while True:
             chunk = response.read(chunksize)
             outfh.write(chunk)
             elapsed = time.time() - t0
             bytesdone += len(chunk)
-            rate = (bytesdone / 1024) / elapsed
+            rate = ((bytesdone - offset) / 1024) / elapsed
             eta = (total - bytesdone) / (rate * 1024)
             progress_stats = (bytesdone, bytesdone * 1.0 / total, rate, eta)
 
@@ -726,6 +763,7 @@ def download(self, filepath="", quiet=False, callback=None):
             if callback:
                 callback(total, *progress_stats)
 
+        os.rename(tempfname, fname)
         return fname
 
 

setup.py

@@ -19,7 +19,7 @@
     name='Pafy',
     packages=['pafy'],
     scripts=['scripts/ytdl'],
-    version='0.3.48',
+    version='0.3.50',
     description="Python API for YouTube, query and download YouTube content",
     keywords=["Pafy", "API", "YouTube", "youtube", "download", "video"],
     author="nagev",

tests/test.py

@@ -71,13 +71,35 @@ def get_all_funcs(self):
     @stdout_to_null
     def test_pafy_download(self):
         """ Test downloading. """
-
         callback = lambda a, b, c, d, e: 0
         vid = pafy.new("DsAn_n6O5Ns", gdata=True)
         vstream = vid.audiostreams[-1]
-        name = vstream.download(filepath="file/", callback=callback)
+        name = vstream.download(callback=callback)
         self.assertEqual(name[0:5], "WASTE")
 
+    @stdout_to_null
+    def test_pafy_download_resume(self):
+        """ Test resuming a partial download. """
+        tempname = "WASTE  2 SECONDS OF YOUR LIFE_DsAn_n6O5Ns_171.part"
+        open(tempname, "w").write("abc")
+        vid = pafy.new("DsAn_n6O5Ns", gdata=True)
+        vstream = vid.audiostreams[-1].download()
+        name = "WASTE  2 SECONDS OF YOUR LIFE.ogg"
+        self.assertEqual(22675, os.stat(name).st_size)
+
+    def test_pafy_download_invalid_dirname(self):
+        """ Test user specified invalid path. """
+        vid = pafy.new("DsAn_n6O5Ns", gdata=True)
+        self.assertRaises(IOError, vid.audiostreams[-1].download, "/bad/h/")
+
+    @stdout_to_null
+    def test_pafy_download_to_dir(self):
+        """ Test user specified path. """
+        vid = pafy.new("DsAn_n6O5Ns", gdata=True)
+        vstream = vid.audiostreams[-1].download("/tmp")
+        name = "/tmp/WASTE  2 SECONDS OF YOUR LIFE.ogg"
+        self.assertEqual(22675, os.stat(name).st_size)
+
     def test_lazy_pafy(self):
         """ Test create pafy object without fetching data. """
 
@@ -276,15 +298,15 @@ def test_misc_tests(self):
         'identifier': 'https://youtu.be/watch?v=07FYdnEawAQ',
         'videoid': '07FYdnEawAQ',
         'title': 'Justin Timberlake - Tunnel Vision (Explicit)',
-        'length': 420,
-        'duration': '00:07:00',
+        'length': 419,
+        'duration': '00:06:59',
         'author': 'justintimberlakeVEVO',
         'username': 'justintimberlakeVEVO',
         'published': '2013-07-03 22:00:16',
         'thumb': 'http://i1.ytimg.com/vi/07FYdnEawAQ/default.jpg',
         'category': 'Music',
         'description': '55e8e6e2b219712bf94d67c2434530474a503265',
-        'bestsize': 80664244,
+        'bestsize': 79885533,
         'all streams': 21,
         'normal streams': 6,
         'video streams': 13,