diff options
Diffstat (limited to 'python/tests/test_misc.py')
| -rw-r--r-- | python/tests/test_misc.py | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/python/tests/test_misc.py b/python/tests/test_misc.py index dcc1202..7d3e755 100644 --- a/python/tests/test_misc.py +++ b/python/tests/test_misc.py @@ -83,7 +83,7 @@ def test_invalid_cdx(): print("missing warc") raw = "edu,upenn,ldc)/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf 20170828233154 https://www.ldc.upenn.edu/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf application/pdf 200 WL3FEA62TEU4F52Y5DOVQ62VET4QJW7G - - 210251 931661233 -" - assert parse_cdx_line(raw) == None + assert parse_cdx_line(raw) is None print("bad datetime") raw = "edu,upenn,ldc)/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf 2070828233154 https://www.ldc.upenn.edu/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf application/pdf 200 WL3FEA62TEU4F52Y5DOVQ62VET4QJW7G - - 210251 931661233i SEMSCHOLAR-PDF-CRAWL-2017-08-04-20170828231135742-00000-00009-wbgrp-svc284/SEMSCHOLAR-PDF-CRAWL-2017-08-04-20170828232253025-00005-3480~wbgrp-svc284.us.archive.org~8443.warc.gz" |