]> jfr.im git - yt-dlp.git/blobdiff - yt_dlp/downloader/external.py
[cleanup] Minor fixes (See desc)
[yt-dlp.git] / yt_dlp / downloader / external.py
index c900b3842263b742e754b73d23bfe4144c51eb1b..85c6a6977bfbfa020671aeabc99fa8339b43732d 100644 (file)
@@ -1,5 +1,3 @@
-from __future__ import unicode_literals
-
 import os.path
 import re
 import subprocess
@@ -7,23 +5,22 @@
 import time
 
 from .fragment import FragmentFD
-from ..compat import (
-    compat_setenv,
-    compat_str,
-)
-from ..postprocessor.ffmpeg import FFmpegPostProcessor, EXT_TO_OUT_FORMATS
+from ..compat import compat_setenv, compat_str
+from ..postprocessor.ffmpeg import EXT_TO_OUT_FORMATS, FFmpegPostProcessor
 from ..utils import (
+    Popen,
+    _configuration_args,
+    check_executable,
+    classproperty,
+    cli_bool_option,
     cli_option,
     cli_valueless_option,
-    cli_bool_option,
-    _configuration_args,
     determine_ext,
-    encodeFilename,
     encodeArgument,
+    encodeFilename,
     handle_youtubedl_headers,
-    check_executable,
-    Popen,
     remove_end,
+    traverse_obj,
 )
 
 
@@ -55,7 +52,7 @@ def real_download(self, filename, info_dict):
             }
             if filename != '-':
                 fsize = os.path.getsize(encodeFilename(tmpfilename))
-                self.to_screen('\r[%s] Downloaded %s bytes' % (self.get_basename(), fsize))
+                self.to_screen(f'\r[{self.get_basename()}] Downloaded {fsize} bytes')
                 self.try_rename(tmpfilename, filename)
                 status.update({
                     'downloaded_bytes': fsize,
@@ -73,17 +70,23 @@ def real_download(self, filename, info_dict):
     def get_basename(cls):
         return cls.__name__[:-2].lower()
 
+    @classproperty
+    def EXE_NAME(cls):
+        return cls.get_basename()
+
     @property
     def exe(self):
-        return self.get_basename()
+        return self.EXE_NAME
 
     @classmethod
     def available(cls, path=None):
-        path = check_executable(path or cls.get_basename(), [cls.AVAILABLE_OPT])
-        if path:
-            cls.exe = path
-            return path
-        return False
+        path = check_executable(
+            cls.EXE_NAME if path in (None, cls.get_basename()) else path,
+            [cls.AVAILABLE_OPT])
+        if not path:
+            return False
+        cls.exe = path
+        return path
 
     @classmethod
     def supports(cls, info_dict):
@@ -106,7 +109,7 @@ def _valueless_option(self, command_option, param, expected_value=True):
 
     def _configuration_args(self, keys=None, *args, **kwargs):
         return _configuration_args(
-            self.get_basename(), self.params.get('external_downloader_args'), self.get_basename(),
+            self.get_basename(), self.params.get('external_downloader_args'), self.EXE_NAME,
             keys, *args, **kwargs)
 
     def _call_downloader(self, tmpfilename, info_dict):
@@ -150,7 +153,7 @@ def _call_downloader(self, tmpfilename, info_dict):
             fragment_filename = '%s-Frag%d' % (tmpfilename, frag_index)
             try:
                 src, _ = self.sanitize_open(fragment_filename, 'rb')
-            except IOError as err:
+            except OSError as err:
                 if skip_unavailable_fragments and frag_index > 1:
                     self.report_skip_fragment(frag_index, err)
                     continue
@@ -172,7 +175,7 @@ def _make_cmd(self, tmpfilename, info_dict):
         cmd = [self.exe, '--location', '-o', tmpfilename, '--compressed']
         if info_dict.get('http_headers') is not None:
             for key, val in info_dict['http_headers'].items():
-                cmd += ['--header', '%s: %s' % (key, val)]
+                cmd += ['--header', f'{key}: {val}']
 
         cmd += self._bool_option('--continue-at', 'continuedl', '-', '0')
         cmd += self._valueless_option('--silent', 'noprogress')
@@ -209,7 +212,7 @@ def _make_cmd(self, tmpfilename, info_dict):
         cmd = [self.exe, '-o', tmpfilename]
         if info_dict.get('http_headers') is not None:
             for key, val in info_dict['http_headers'].items():
-                cmd += ['-H', '%s: %s' % (key, val)]
+                cmd += ['-H', f'{key}: {val}']
         cmd += self._configuration_args()
         cmd += ['--', info_dict['url']]
         return cmd
@@ -222,7 +225,7 @@ def _make_cmd(self, tmpfilename, info_dict):
         cmd = [self.exe, '-O', tmpfilename, '-nv', '--no-cookies', '--compression=auto']
         if info_dict.get('http_headers') is not None:
             for key, val in info_dict['http_headers'].items():
-                cmd += ['--header', '%s: %s' % (key, val)]
+                cmd += ['--header', f'{key}: {val}']
         cmd += self._option('--limit-rate', 'ratelimit')
         retry = self._option('--tries', 'retries')
         if len(retry) == 2:
@@ -233,7 +236,7 @@ def _make_cmd(self, tmpfilename, info_dict):
         proxy = self.params.get('proxy')
         if proxy:
             for var in ('http_proxy', 'https_proxy'):
-                cmd += ['--execute', '%s=%s' % (var, proxy)]
+                cmd += ['--execute', f'{var}={proxy}']
         cmd += self._valueless_option('--no-check-certificate', 'nocheckcertificate')
         cmd += self._configuration_args()
         cmd += ['--', info_dict['url']]
@@ -264,7 +267,7 @@ def _make_cmd(self, tmpfilename, info_dict):
 
         if info_dict.get('http_headers') is not None:
             for key, val in info_dict['http_headers'].items():
-                cmd += ['--header', '%s: %s' % (key, val)]
+                cmd += ['--header', f'{key}: {val}']
         cmd += self._option('--max-overall-download-limit', 'ratelimit')
         cmd += self._option('--interface', 'source_address')
         cmd += self._option('--all-proxy', 'proxy')
@@ -282,10 +285,10 @@ def _make_cmd(self, tmpfilename, info_dict):
         dn = os.path.dirname(tmpfilename)
         if dn:
             if not os.path.isabs(dn):
-                dn = '.%s%s' % (os.path.sep, dn)
+                dn = f'.{os.path.sep}{dn}'
             cmd += ['--dir', dn + os.path.sep]
         if 'fragments' not in info_dict:
-            cmd += ['--out', '.%s%s' % (os.path.sep, os.path.basename(tmpfilename))]
+            cmd += ['--out', f'.{os.path.sep}{os.path.basename(tmpfilename)}']
         cmd += ['--auto-file-renaming=false']
 
         if 'fragments' in info_dict:
@@ -296,7 +299,7 @@ def _make_cmd(self, tmpfilename, info_dict):
                 fragment_filename = '%s-Frag%d' % (os.path.basename(tmpfilename), frag_index)
                 url_list.append('%s\n\tout=%s' % (fragment['url'], fragment_filename))
             stream, _ = self.sanitize_open(url_list_file, 'wb')
-            stream.write('\n'.join(url_list).encode('utf-8'))
+            stream.write('\n'.join(url_list).encode())
             stream.close()
             cmd += ['-i', url_list_file]
         else:
@@ -306,17 +309,14 @@ def _make_cmd(self, tmpfilename, info_dict):
 
 class HttpieFD(ExternalFD):
     AVAILABLE_OPT = '--version'
-
-    @classmethod
-    def available(cls, path=None):
-        return super().available(path or 'http')
+    EXE_NAME = 'http'
 
     def _make_cmd(self, tmpfilename, info_dict):
         cmd = ['http', '--download', '--output', tmpfilename, info_dict['url']]
 
         if info_dict.get('http_headers') is not None:
             for key, val in info_dict['http_headers'].items():
-                cmd += ['%s:%s' % (key, val)]
+                cmd += [f'{key}:{val}']
         return cmd
 
 
@@ -364,9 +364,11 @@ def _call_downloader(self, tmpfilename, info_dict):
         if not self.params.get('verbose'):
             args += ['-hide_banner']
 
-        args += info_dict.get('_ffmpeg_args', [])
+        args += traverse_obj(info_dict, ('downloader_options', 'ffmpeg_args'), default=[])
 
-        # This option exists only for compatibility. Extractors should use `_ffmpeg_args` instead
+        # These exists only for compatibility. Extractors should use
+        # info_dict['downloader_options']['ffmpeg_args'] instead
+        args += info_dict.get('_ffmpeg_args') or []
         seekable = info_dict.get('_seekable')
         if seekable is not None:
             # setting -seekable prevents ffmpeg from guessing if the server
@@ -383,13 +385,15 @@ def _call_downloader(self, tmpfilename, info_dict):
         # if end_time:
         #     args += ['-t', compat_str(end_time - start_time)]
 
-        if info_dict.get('http_headers') is not None and re.match(r'^https?://', urls[0]):
-            # Trailing \r\n after each HTTP header is important to prevent warning from ffmpeg/avconv:
-            # [http @ 00000000003d2fa0] No trailing CRLF found in HTTP header.
-            headers = handle_youtubedl_headers(info_dict['http_headers'])
-            args += [
+        http_headers = None
+        if info_dict.get('http_headers'):
+            youtubedl_headers = handle_youtubedl_headers(info_dict['http_headers'])
+            http_headers = [
+                # Trailing \r\n after each HTTP header is important to prevent warning from ffmpeg/avconv:
+                # [http @ 00000000003d2fa0] No trailing CRLF found in HTTP header.
                 '-headers',
-                ''.join('%s: %s\r\n' % (key, val) for key, val in headers.items())]
+                ''.join(f'{key}: {val}\r\n' for key, val in youtubedl_headers.items())
+            ]
 
         env = None
         proxy = self.params.get('proxy')
@@ -442,6 +446,11 @@ def _call_downloader(self, tmpfilename, info_dict):
                 args += ['-rtmp_conn', conn]
 
         for i, url in enumerate(urls):
+            # We need to specify headers for each http input stream
+            # otherwise, it will only be applied to the first.
+            # https://github.com/yt-dlp/yt-dlp/issues/2696
+            if http_headers is not None and re.match(r'^https?://', url):
+                args += http_headers
             args += self._configuration_args((f'_i{i + 1}', '_i')) + ['-i', url]
 
         args += ['-c', 'copy']
@@ -510,11 +519,13 @@ class AVconvFD(FFmpegFD):
     pass
 
 
-_BY_NAME = dict(
-    (klass.get_basename(), klass)
+_BY_NAME = {
+    klass.get_basename(): klass
     for name, klass in globals().items()
     if name.endswith('FD') and name not in ('ExternalFD', 'FragmentFD')
-)
+}
+
+_BY_EXE = {klass.EXE_NAME: klass for klass in _BY_NAME.values()}
 
 
 def list_external_downloaders():
@@ -526,4 +537,4 @@ def get_external_downloader(external_downloader):
         downloader . """
     # Drop .exe extension on Windows
     bn = os.path.splitext(os.path.basename(external_downloader))[0]
-    return _BY_NAME.get(bn)
+    return _BY_NAME.get(bn, _BY_EXE.get(bn))