Add field `webpage_url_domain`

[yt-dlp.git] / yt_dlp / YoutubeDL.py

diff --git a/yt_dlp/YoutubeDL.py b/yt_dlp/YoutubeDL.py
index 6341694700e642b133944bf8cee6668c8629407b..39fd2e1fd7ab527241806607b1b31dba03fb879d 100644 (file)
--- a/yt_dlp/YoutubeDL.py
+++ b/yt_dlp/YoutubeDL.py
@@ -70,6 +70,7 @@
     format_field,
     formatSeconds,
     GeoRestrictedError,
+    get_domain,
     HEADRequest,
     int_or_none,
     iri_to_uri,
@@ -96,6 +97,7 @@
     ReExtractInfo,
     register_socks_protocols,
     RejectedVideoReached,
+    remove_terminal_sequences,
     render_table,
     replace_extension,
     SameFileError,
@@ -327,7 +329,7 @@ class YoutubeDL(object):
     bidi_workaround:   Work around buggy terminals without bidirectional text
                        support, using fridibi
     debug_printtraffic:Print out sent and received HTTP traffic
-    include_ads:       Download ads as well
+    include_ads:       Download ads as well (deprecated)
     default_search:    Prepend this string if an input url is not valid.
                        'auto' for elaborate guessing
     encoding:          Use this encoding instead of the system-specified.
@@ -633,13 +635,6 @@ def check_deprecated(param, option, suggestion):
                 self.print_debug_header()
             self.add_default_info_extractors()
 
-        for pp_def_raw in self.params.get('postprocessors', []):
-            pp_def = dict(pp_def_raw)
-            when = pp_def.pop('when', 'post_process')
-            pp_class = get_postprocessor(pp_def.pop('key'))
-            pp = pp_class(self, **compat_kwargs(pp_def))
-            self.add_post_processor(pp, when=when)
-
         hooks = {
             'post_hooks': self.add_post_hook,
             'progress_hooks': self.add_progress_hook,
@@ -649,6 +644,13 @@ def check_deprecated(param, option, suggestion):
             for ph in self.params.get(opt, []):
                 fn(ph)
 
+        for pp_def_raw in self.params.get('postprocessors', []):
+            pp_def = dict(pp_def_raw)
+            when = pp_def.pop('when', 'post_process')
+            self.add_post_processor(
+                get_postprocessor(pp_def.pop('key'))(self, **compat_kwargs(pp_def)),
+                when=when)
+
         register_socks_protocols()
 
         def preload_download_archive(fn):
@@ -735,6 +737,9 @@ def add_progress_hook(self, ph):
     def add_postprocessor_hook(self, ph):
         """Add the postprocessing progress hook"""
         self._postprocessor_hooks.append(ph)
+        for pps in self._pps.values():
+            for pp in pps:
+                pp.add_progress_hook(ph)
 
     def _bidi_workaround(self, message):
         if not hasattr(self, '_output_channel'):
@@ -776,6 +781,7 @@ def to_stderr(self, message, only_once=False):
     def to_console_title(self, message):
         if not self.params.get('consoletitle', False):
             return
+        message = remove_terminal_sequences(message)
         if compat_os_name == 'nt':
             if ctypes.windll.kernel32.GetConsoleWindow():
                 # c_wchar_p() might not be necessary if `message` is
@@ -1369,11 +1375,11 @@ def progress(msg):
         min_wait, max_wait = self.params.get('wait_for_video')
         diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
         if diff is None and ie_result.get('live_status') == 'is_upcoming':
-            diff = random.randrange(min_wait or 0, max_wait) if max_wait else min_wait
+            diff = random.randrange(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait)
             self.report_warning('Release time of video is not known')
         elif (diff or 0) <= 0:
             self.report_warning('Video should already be available according to extracted info')
-        diff = min(max(diff, min_wait or 0), max_wait or float('inf'))
+        diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
         self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
 
         wait_till = time.time() + diff
@@ -1419,6 +1425,7 @@ def add_default_extra_info(self, ie_result, ie, url):
                 'webpage_url': url,
                 'original_url': url,
                 'webpage_url_basename': url_basename(url),
+                'webpage_url_domain': get_domain(url),
             })
         if ie is not None:
             self.add_extra_info(ie_result, {
@@ -1452,6 +1459,7 @@ def process_ie_result(self, ie_result, download=True, extra_info=None):
                     info_copy['id'] = ie.get_temp_id(ie_result['url'])
                 self.add_default_extra_info(info_copy, ie, ie_result['url'])
                 self.add_extra_info(info_copy, extra_info)
+                info_copy, _ = self.pre_process(info_copy)
                 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
                 if self.params.get('force_write_download_archive', False):
                     self.record_download_archive(info_copy)
@@ -1542,6 +1550,7 @@ def _fixup(r):
                     'extractor': ie_result['extractor'],
                     'webpage_url': ie_result['webpage_url'],
                     'webpage_url_basename': url_basename(ie_result['webpage_url']),
+                    'webpage_url_domain': get_domain(ie_result['webpage_url']),
                     'extractor_key': ie_result['extractor_key'],
                 })
                 return r
@@ -1703,6 +1712,7 @@ def get_entry(i):
                 'extractor': ie_result['extractor'],
                 'webpage_url': ie_result['webpage_url'],
                 'webpage_url_basename': url_basename(ie_result['webpage_url']),
+                'webpage_url_domain': get_domain(ie_result['webpage_url']),
                 'extractor_key': ie_result['extractor_key'],
             }
 
@@ -3391,7 +3401,8 @@ def print_debug_header(self):
         def get_encoding(stream):
             ret = getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__)
             if not supports_terminal_sequences(stream):
-                ret += ' (No ANSI)'
+                from .compat import WINDOWS_VT_MODE
+                ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
             return ret
 
         encoding_str = 'Encodings: locale %s, fs %s, out %s, err %s, pref %s' % (