diff options
| author | Bryan Newbold <bnewbold@archive.org> | 2022-09-12 10:03:35 -0700 |
|---|---|---|
| committer | Bryan Newbold <bnewbold@archive.org> | 2022-09-12 10:03:35 -0700 |
| commit | a283b054dc98620046dff28cbb16663564b8320b (patch) | |
| tree | 935b570b1cb62d0053e1f3b257929f35b11a229b | |
| parent | bbaa351a109c001a46e10f52549b43bcef48d348 (diff) | |
| download | sandcrawler-a283b054dc98620046dff28cbb16663564b8320b.tar.gz sandcrawler-a283b054dc98620046dff28cbb16663564b8320b.zip | |
bad PDF sha1
| -rw-r--r-- | python/sandcrawler/pdfextract.py | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/python/sandcrawler/pdfextract.py b/python/sandcrawler/pdfextract.py index 3eb7581..2441891 100644 --- a/python/sandcrawler/pdfextract.py +++ b/python/sandcrawler/pdfextract.py @@ -87,10 +87,12 @@ BAD_PDF_SHA1HEX: List[str] = [ "781dafda896a9f5c30f3d0a011f79a3b79b574c4", "788672c7c2bcdecf6e2f6a2177c01e60f04d9cfb", "79d6cba3c6e577a0f3a3a9fe575680d38454938d", + "7b8b7e8e4b789579a7d2fda329db52528383a652", "7c5c925cfb7c5a861b5c0a1d923308f9bedd335e", "7cfc0739be9c49d94272110a0a748256bdde9be6", "7daf61526ec825151f384cc1db510ca5237d5d80", "7e9d846f3bf9ce15cdb991b78cc870ab8a2bed76", + "800e47a7ed214f7acac85cc29aa7b0f9c0e218ae", "8398b211a5ec4da1195a4ba1bc29ca8c0ac40f67", "859d7ec532a0bf3b52b17c7f2d8ecc58410c0aad", "88edcbab1cac2d70af5870422974afc253f4f0c6", @@ -125,9 +127,11 @@ BAD_PDF_SHA1HEX: List[str] = [ "b8b427e5b3d650ba9e03197f9c3917e25b878930", "bad48b89b639b5b7df2c6a2d5288181fcb8b0e35", "be0cda7642e9247b3ee41cd2017fa709aab4f344", + "beff1b0c24aa99989be73c66dfb1d1e7578e370b", "c1b583fbd052572f08158d39ffe4d7510dadbebb", "c2526f75a013dc67b14ce1e2d0e4fc80bb93c6e1", "c4abbb284f4acaca9e8ceb88f842901984e84d33", + "c58e028269c8dfd3a442f6745c81b4c0e8610c43", "c7220d1bf1e71fb755d9f26bbdd4c539dc162960", "c7687fa6f637c7d32a25be0e772867d87536d35c", "c7d8b37ec99cf0d987e60667f05299f200e18a5d", |