diff options
| author | Bryan Newbold <bnewbold@archive.org> | 2020-08-14 16:00:39 -0700 |
|---|---|---|
| committer | Bryan Newbold <bnewbold@archive.org> | 2020-08-14 16:00:39 -0700 |
| commit | 2264757db3709620eed4222400327f6f871f8c58 (patch) | |
| tree | 48d97e70513a3fd1b2fa3872dcf27d24064e24bd | |
| parent | e04748c3b0ac1f83d66a217e4d2461f142e1c699 (diff) | |
| download | sandcrawler-2264757db3709620eed4222400327f6f871f8c58.tar.gz sandcrawler-2264757db3709620eed4222400327f6f871f8c58.zip | |
yet more bad PDF sha1
| -rw-r--r-- | python/sandcrawler/pdfextract.py | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/python/sandcrawler/pdfextract.py b/python/sandcrawler/pdfextract.py index 269c5dd..e4150c2 100644 --- a/python/sandcrawler/pdfextract.py +++ b/python/sandcrawler/pdfextract.py @@ -51,6 +51,8 @@ BAD_PDF_SHA1HEX = [ "d17b1e254cce82df5c6eb4fd492cef91e7e11558", "d188762a7e3ab5d4ee8a897204316513e4e636ec", "d6b0f405bf13c23d0e90c54eea527442786d1cd3", + "e01bb7256d77aea258313bb410dfcfc10512f420", + "eb1b39fd7a874896688855a22efddef10272427c", "eb5fffaa590a52bcc3705b888c6ff9c4dc4c45b2", "f68f9a9202a75d2aee35252e104d796f9515001e", ] |