]> jfr.im git - yt-dlp.git/blobdiff - yt_dlp/postprocessor/modify_chapters.py
[ie/EuroParlWebstream] Support new URL format (#9647)
[yt-dlp.git] / yt_dlp / postprocessor / modify_chapters.py
index dca87620037ef44cd1d0d801cd9ad2dd66a801a9..f5219868c842f7b8aa2fa3023e47a346df45928a 100644 (file)
@@ -3,17 +3,9 @@
 import os
 
 from .common import PostProcessor
-from .ffmpeg import (
-    FFmpegPostProcessor,
-    FFmpegSubtitlesConvertorPP
-)
+from .ffmpeg import FFmpegPostProcessor, FFmpegSubtitlesConvertorPP
 from .sponsorblock import SponsorBlockPP
-from ..utils import (
-    orderedSet,
-    PostProcessingError,
-    prepend_extension,
-)
-
+from ..utils import PostProcessingError, orderedSet, prepend_extension
 
 _TINY_CHAPTER_DURATION = 1
 DEFAULT_SPONSORBLOCK_CHAPTER_TITLE = '[SponsorBlock]: %(category_names)l'
@@ -24,13 +16,14 @@ def __init__(self, downloader, remove_chapters_patterns=None, remove_sponsor_seg
                  *, sponsorblock_chapter_title=DEFAULT_SPONSORBLOCK_CHAPTER_TITLE, force_keyframes=False):
         FFmpegPostProcessor.__init__(self, downloader)
         self._remove_chapters_patterns = set(remove_chapters_patterns or [])
-        self._remove_sponsor_segments = set(remove_sponsor_segments or [])
+        self._remove_sponsor_segments = set(remove_sponsor_segments or []) - set(SponsorBlockPP.NON_SKIPPABLE_CATEGORIES.keys())
         self._ranges_to_remove = set(remove_ranges or [])
         self._sponsorblock_chapter_title = sponsorblock_chapter_title
         self._force_keyframes = force_keyframes
 
     @PostProcessor._restrict_to(images=False)
     def run(self, info):
+        self._fixup_chapters(info)
         # Chapters must be preserved intact when downloading multiple formats of the same video.
         chapters, sponsor_chapters = self._mark_chapters_to_remove(
             copy.deepcopy(info.get('chapters')) or [],
@@ -38,16 +31,20 @@ def run(self, info):
         if not chapters and not sponsor_chapters:
             return [], info
 
-        real_duration = self._get_real_video_duration(info)
+        real_duration = self._get_real_video_duration(info['filepath'])
         if not chapters:
-            chapters = [{'start_time': 0, 'end_time': real_duration, 'title': info['title']}]
+            chapters = [{'start_time': 0, 'end_time': info.get('duration') or real_duration, 'title': info['title']}]
 
         info['chapters'], cuts = self._remove_marked_arrange_sponsors(chapters + sponsor_chapters)
         if not cuts:
             return [], info
+        elif not info['chapters']:
+            self.report_warning('You have requested to remove the entire video, which is not possible')
+            return [], info
 
-        if self._duration_mismatch(real_duration, info.get('duration')):
-            if not self._duration_mismatch(real_duration, info['chapters'][-1]['end_time']):
+        original_duration, info['duration'] = info.get('duration'), info['chapters'][-1]['end_time']
+        if self._duration_mismatch(real_duration, original_duration, 1):
+            if not self._duration_mismatch(real_duration, info['duration']):
                 self.to_screen(f'Skipping {self.pp_key()} since the video appears to be already cut')
                 return [], info
             if not info.get('__real_download'):
@@ -57,6 +54,7 @@ def run(self, info):
                 self.write_debug('Expected and actual durations mismatch')
 
         concat_opts = self._make_concat_opts(cuts, real_duration)
+        self.write_debug('Concat spec = %s' % ', '.join(f'{c.get("inpoint", 0.0)}-{c.get("outpoint", "inf")}' for c in concat_opts))
 
         def remove_chapters(file, is_sub):
             return file, self.remove_chapters(file, cuts, concat_opts, self._force_keyframes and not is_sub)
@@ -67,12 +65,13 @@ def remove_chapters(file, is_sub):
         # Renaming should only happen after all files are processed
         files_to_remove = []
         for in_file, out_file in in_out_files:
+            mtime = os.stat(in_file).st_mtime
             uncut_file = prepend_extension(in_file, 'uncut')
             os.replace(in_file, uncut_file)
             os.replace(out_file, in_file)
+            self.try_utime(in_file, mtime, mtime)
             files_to_remove.append(uncut_file)
 
-        info['_real_duration'] = info['chapters'][-1]['end_time']
         return files_to_remove, info
 
     def _mark_chapters_to_remove(self, chapters, sponsor_chapters):
@@ -104,7 +103,7 @@ def _mark_chapters_to_remove(self, chapters, sponsor_chapters):
             'start_time': start,
             'end_time': end,
             'category': 'manually_removed',
-            '_categories': [('manually_removed', start, end)],
+            '_categories': [('manually_removed', start, end, 'Manually removed')],
             'remove': True,
         } for start, end in self._ranges_to_remove)
 
@@ -295,15 +294,14 @@ def _remove_tiny_rename_sponsors(self, chapters):
             c.pop('_was_cut', None)
             cats = c.pop('_categories', None)
             if cats:
-                category = min(cats, key=lambda c: c[2] - c[1])[0]
-                cats = orderedSet(x[0] for x in cats)
+                category, _, _, category_name = min(cats, key=lambda c: c[2] - c[1])
                 c.update({
                     'category': category,
-                    'categories': cats,
-                    'name': SponsorBlockPP.CATEGORIES[category],
-                    'category_names': [SponsorBlockPP.CATEGORIES[c] for c in cats]
+                    'categories': orderedSet(x[0] for x in cats),
+                    'name': category_name,
+                    'category_names': orderedSet(x[3] for x in cats),
                 })
-                c['title'] = self._downloader.evaluate_outtmpl(self._sponsorblock_chapter_title, c)
+                c['title'] = self._downloader.evaluate_outtmpl(self._sponsorblock_chapter_title, c.copy())
                 # Merge identically named sponsors.
                 if (new_chapters and 'categories' in new_chapters[-1]
                         and new_chapters[-1]['title'] == c['title']):
@@ -320,7 +318,7 @@ def remove_chapters(self, filename, ranges_to_cut, concat_opts, force_keyframes=
         self.to_screen(f'Removing chapters from {filename}')
         self.concat_files([in_file] * len(concat_opts), out_file, concat_opts)
         if in_file != filename:
-            os.remove(in_file)
+            self._delete_downloaded_files(in_file, msg=None)
         return out_file
 
     @staticmethod
@@ -333,6 +331,6 @@ def _make_concat_opts(chapters_to_remove, duration):
                 continue
             opts[-1]['outpoint'] = f'{s["start_time"]:.6f}'
             # Do not create 0 duration chunk at the end.
-            if s['end_time'] != duration:
+            if s['end_time'] < duration:
                 opts.append({'inpoint': f'{s["end_time"]:.6f}'})
         return opts