[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 6afc1b73070cd410992733a59c0cd8bc9b9bff0e..18b1443b887ca323e87348636c0febbb878705cc 100755 (executable)
--- a/youtube_dl/YoutubeDL.py
+++ b/youtube_dl/YoutubeDL.py
@@ -241,7 +241,7 @@ class YoutubeDL(object):
                        - "detect_or_warn": check whether we can do anything
                                            about it, warn otherwise (default)
     source_address:    (Experimental) Client-side IP address to bind to.
-    call_home:         Boolean, true if we are allowed to contact the
+    call_home:         Boolean, true iff we are allowed to contact the
                        youtube-dl servers for debugging.
     sleep_interval:    Number of seconds to sleep before each download.
     listformats:       Print an overview of available video formats and exit.
@@ -590,7 +590,7 @@ class YoutubeDL(object):
             return None
 
     def _match_entry(self, info_dict, incomplete):
-        """ Returns None if the file should be downloaded """
+        """ Returns None iff the file should be downloaded """
 
         video_title = info_dict.get('title', info_dict.get('id', 'video'))
         if 'title' in info_dict:
@@ -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():
@@ -1986,8 +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