[YoutubeDL] Introduce CSS3 like string operators

[youtube-dl] / youtube_dl / YoutubeDL.py

diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py
index 1795097ae8ba5faebbbf9efff91b5b70e0132c86..18b1443b887ca323e87348636c0febbb878705cc 100755 (executable)
--- a/youtube_dl/YoutubeDL.py
+++ b/youtube_dl/YoutubeDL.py
@@ -49,7 +49,7 @@ from .utils import (
     DownloadError,
     encode_compat_str,
     encodeFilename,
-    error_to_str,
+    error_to_compat_str,
     ExtractorError,
     format_bytes,
     formatSeconds,
@@ -677,13 +677,13 @@ class YoutubeDL(object):
                 else:
                     return ie_result
             except ExtractorError as e:  # An error we somewhat expected
-                self.report_error(error_to_str(e), e.format_traceback())
+                self.report_error(compat_str(e), e.format_traceback())
                 break
             except MaxDownloadsReached:
                 raise
             except Exception as e:
                 if self.params.get('ignoreerrors', False):
-                    self.report_error(error_to_str(e), tb=encode_compat_str(traceback.format_exc()))
+                    self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
                     break
                 else:
                     raise
@@ -898,6 +898,9 @@ class YoutubeDL(object):
             STR_OPERATORS = {
                 '=': operator.eq,
                 '!=': operator.ne,
+                '^=': lambda attr, value: attr.startswith(value),
+                '$=': lambda attr, value: attr.endswith(value),
+                '*=': lambda attr, value: value in attr,
             }
             str_operator_rex = re.compile(r'''(?x)
                 \s*(?P<key>ext|acodec|vcodec|container|protocol)
@@ -1244,6 +1247,12 @@ class YoutubeDL(object):
             except (ValueError, OverflowError, OSError):
                 pass
 
+        # Auto generate title fields corresponding to the *_number fields when missing
+        # in order to always have clean titles. This is very common for TV series.
+        for field in ('chapter', 'season', 'episode'):
+            if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
+                info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
+
         subtitles = info_dict.get('subtitles')
         if subtitles:
             for _, subtitle in subtitles.items():
@@ -1312,7 +1321,7 @@ class YoutubeDL(object):
             # only set the 'formats' fields if the original info_dict list them
             # otherwise we end up with a circular reference, the first (and unique)
             # element in the 'formats' field in info_dict is info_dict itself,
-            # wich can't be exported to json
+            # which can't be exported to json
             info_dict['formats'] = formats
         if self.params.get('listformats'):
             self.list_formats(info_dict)
@@ -1461,7 +1470,7 @@ class YoutubeDL(object):
             if dn and not os.path.exists(dn):
                 os.makedirs(dn)
         except (OSError, IOError) as err:
-            self.report_error('unable to create directory ' + error_to_str(err))
+            self.report_error('unable to create directory ' + error_to_compat_str(err))
             return
 
         if self.params.get('writedescription', False):
@@ -1512,7 +1521,7 @@ class YoutubeDL(object):
                             sub_info['url'], info_dict['id'], note=False)
                     except ExtractorError as err:
                         self.report_warning('Unable to download subtitle for "%s": %s' %
-                                            (sub_lang, error_to_str(err.cause)))
+                                            (sub_lang, error_to_compat_str(err.cause)))
                         continue
                 try:
                     sub_filename = subtitles_filename(filename, sub_lang, sub_format)
@@ -1791,6 +1800,10 @@ class YoutubeDL(object):
         res = ''
         if fdict.get('ext') in ['f4f', 'f4m']:
             res += '(unsupported) '
+        if fdict.get('language'):
+            if res:
+                res += ' '
+            res += '[%s]' % fdict['language']
         if fdict.get('format_note') is not None:
             res += fdict['format_note'] + ' '
         if fdict.get('tbr') is not None:
@@ -1982,8 +1995,19 @@ class YoutubeDL(object):
         https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
         ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
         data_handler = compat_urllib_request_DataHandler()
+
+        # When passing our own FileHandler instance, build_opener won't add the
+        # default FileHandler and allows us to disable the file protocol, which
+        # can be used for malicious purposes (see
+        # https://github.com/rg3/youtube-dl/issues/8227)
+        file_handler = compat_urllib_request.FileHandler()
+
+        def file_open(*args, **kwargs):
+            raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
+        file_handler.file_open = file_open
+
         opener = compat_urllib_request.build_opener(
-            proxy_handler, https_handler, cookie_processor, ydlh, data_handler)
+            proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler)
 
         # Delete the default user-agent header, which would otherwise apply in
         # cases where our custom HTTP handler doesn't come into play
@@ -2041,4 +2065,4 @@ class YoutubeDL(object):
                                    (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
                 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
                     self.report_warning('Unable to download thumbnail "%s": %s' %
-                                        (t['url'], error_to_str(err)))
+                                        (t['url'], error_to_compat_str(err)))