[extractor/youtube] Support shorter relative time format (#7191)

[yt-dlp.git] / yt_dlp / extractor / youtube.py

diff --git a/yt_dlp/extractor/youtube.py b/yt_dlp/extractor/youtube.py
index ccf97705a18c59f49aa454861599aa247b399410..6e7485c03010a07e4421acf9f05792891900b652 100644 (file)
--- a/yt_dlp/extractor/youtube.py
+++ b/yt_dlp/extractor/youtube.py
@@ -893,9 +893,16 @@ def _extract_thumbnails(data, *path_list):
     def extract_relative_time(relative_time_text):
         """
         Extracts a relative time from string and converts to dt object
-        e.g. 'streamed 6 days ago', '5 seconds ago (edited)', 'updated today'
+        e.g. 'streamed 6 days ago', '5 seconds ago (edited)', 'updated today', '8 yr ago'
         """
-        mobj = re.search(r'(?P<start>today|yesterday|now)|(?P<time>\d+)\s*(?P<unit>microsecond|second|minute|hour|day|week|month|year)s?\s*ago', relative_time_text)
+
+        # XXX: this could be moved to a general function in utils.py
+        # The relative time text strings are roughly the same as what
+        # Javascript's Intl.RelativeTimeFormat function generates.
+        # See: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Intl/RelativeTimeFormat
+        mobj = re.search(
+            r'(?P<start>today|yesterday|now)|(?P<time>\d+)\s*(?P<unit>sec(?:ond)?|s|min(?:ute)?|h(?:our|r)?|d(?:ay)?|w(?:eek|k)?|mo(?:nth)?|y(?:ear|r)?)s?\s*ago',
+            relative_time_text)
         if mobj:
             start = mobj.group('start')
             if start: