[downloader/(common/http)] Changes calculation of the rate-limit. (Fix #2297, fix...

[youtube-dl] / youtube_dl / downloader / http.py

diff --git a/youtube_dl/downloader/http.py b/youtube_dl/downloader/http.py
index cc8b9c9a7f5cac66a4b59d460c7678917aa98cc0..462be2739143e19ea27406731c069ca461787b41 100644 (file)
--- a/youtube_dl/downloader/http.py
+++ b/youtube_dl/downloader/http.py
@@ -14,6 +14,8 @@ from ..utils import (
 
 
 class HttpFD(FileDownloader):
+    _TEST_FILE_SIZE = 10241
+
     def real_download(self, filename, info_dict):
         url = info_dict['url']
         tmpfilename = self.temp_name(filename)
@@ -28,8 +30,10 @@ class HttpFD(FileDownloader):
         basic_request = compat_urllib_request.Request(url, None, headers)
         request = compat_urllib_request.Request(url, None, headers)
 
-        if self.params.get('test', False):
-            request.add_header('Range', 'bytes=0-10240')
+        is_test = self.params.get('test', False)
+
+        if is_test:
+            request.add_header('Range', 'bytes=0-%s' % str(self._TEST_FILE_SIZE - 1))
 
         # Establish possible resume length
         if os.path.isfile(encodeFilename(tmpfilename)):
@@ -100,6 +104,15 @@ class HttpFD(FileDownloader):
             return False
 
         data_len = data.info().get('Content-length', None)
+
+        # Range HTTP header may be ignored/unsupported by a webserver
+        # (e.g. extractor/scivee.py, extractor/bambuser.py).
+        # However, for a test we still would like to download just a piece of a file.
+        # To achieve this we limit data_len to _TEST_FILE_SIZE and manually control
+        # block size when downloading a file.
+        if is_test and (data_len is None or int(data_len) > self._TEST_FILE_SIZE):
+            data_len = self._TEST_FILE_SIZE
+
         if data_len is not None:
             data_len = int(data_len) + resume_len
             min_data_len = self.params.get("min_filesize", None)
@@ -115,16 +128,21 @@ class HttpFD(FileDownloader):
         byte_counter = 0 + resume_len
         block_size = self.params.get('buffersize', 1024)
         start = time.time()
+
+        # measure time over whole while-loop, so slow_down() and best_block_size() work together properly
+        now = None  # needed for slow_down() in the first loop run
+        before = start  # start measuring
         while True:
+
             # Download and write
-            before = time.time()
-            data_block = data.read(block_size)
-            after = time.time()
+            data_block = data.read(block_size if not is_test else min(block_size, data_len - byte_counter))
+            byte_counter += len(data_block)
+
+            # exit loop when download is finished
             if len(data_block) == 0:
                 break
-            byte_counter += len(data_block)
 
-            # Open file just in time
+            # Open destination file just in time
             if stream is None:
                 try:
                     (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
@@ -140,11 +158,22 @@ class HttpFD(FileDownloader):
                 self.to_stderr(u"\n")
                 self.report_error(u'unable to write data: %s' % str(err))
                 return False
+
+            # Apply rate limit
+            self.slow_down(start, now, byte_counter - resume_len)
+
+            # end measuring of one loop run
+            now = time.time()
+            after = now
+
+            # Adjust block size
             if not self.params.get('noresizebuffer', False):
                 block_size = self.best_block_size(after - before, len(data_block))
 
+            before = after
+
             # Progress message
-            speed = self.calc_speed(start, time.time(), byte_counter - resume_len)
+            speed = self.calc_speed(start, now, byte_counter - resume_len)
             if data_len is None:
                 eta = percent = None
             else:
@@ -162,8 +191,8 @@ class HttpFD(FileDownloader):
                 'speed': speed,
             })
 
-            # Apply rate limit
-            self.slow_down(start, byte_counter - resume_len)
+            if is_test and byte_counter == data_len:
+                break
 
         if stream is None:
             self.to_stderr(u"\n")