diff options
author | Martin Czygan <martin.czygan@gmail.com> | 2020-11-24 15:02:56 +0100 |
---|---|---|
committer | Martin Czygan <martin.czygan@gmail.com> | 2020-11-24 15:02:56 +0100 |
commit | bfd6e08c3b1ffb16c98321d15dd5da6e8db400de (patch) | |
tree | 42ba05e01e3eb68dcc79d2df693136b1d785ec02 /fuzzycat | |
parent | 77d9c26819ab59a183c86a4b3ec3b5b80e85dd14 (diff) | |
download | fuzzycat-bfd6e08c3b1ffb16c98321d15dd5da6e8db400de.tar.gz fuzzycat-bfd6e08c3b1ffb16c98321d15dd5da6e8db400de.zip |
wip: hl
Diffstat (limited to 'fuzzycat')
-rw-r--r-- | fuzzycat/cluster.py | 2 | ||||
-rw-r--r-- | fuzzycat/verify.py | 1 |
2 files changed, 2 insertions, 1 deletions
diff --git a/fuzzycat/cluster.py b/fuzzycat/cluster.py index 723d0fa..dfc08b7 100644 --- a/fuzzycat/cluster.py +++ b/fuzzycat/cluster.py @@ -258,7 +258,7 @@ def test_sandcrawler_slugify() -> None: (r"👾 🙇 💁 🙅 🙆 🙋 🙎 🙍", "👾🙇💁🙅🙆🙋🙎🙍"), (r"2️⃣ 3️⃣ 4️⃣ 5️⃣", "2345"), (r"﷽ ", "﷽"), - (r"̗̺͖̹̯͓Ṯ̤͍̥͇͈h̲́e͏͓̼̗̙̼̣͔ ͇̜̱̠͓͍ͅN͕͠e̗̱z̘̝̜̺͙p̤̺̹͍̯͚e̠̻̠͜r̨̤͍̺̖͔̖̖d̠̟̭̬̝͟i̦͖̩͓͔̤a̠̗̬͉̙n͚͜ ̻̞̰͚ͅh̵͉i̳̞v̢͇ḙ͎͟-҉̭̩̼͔m̤̭̫i͕͇̝̦n̗͙ḍ̟ ̯̲͕͞ǫ̟̯̰̲͙̻̝f ̪̰̰̗̖̭̘͘c̦͍̲̞͍̩̙ḥ͚a̮͎̟̙͜ơ̩̹͎s̤.̝̝ ҉Z̡̖̜͖̰̣͉̜a͖̰͙̬͡l̲̫̳͍̩g̡̟̼̱͚̞̬ͅo̗͜.̟", + (r"\"̗̺͖̹̯͓Ṯ̤͍̥͇͈h̲́e͏͓̼̗̙̼̣͔ ͇̜̱̠͓͍ͅN͕͠e̗̱z̘̝̜̺͙p̤̺̹͍̯͚e̠̻̠͜r̨̤͍̺̖͔̖̖d̠̟̭̬̝͟i̦͖̩͓͔̤a̠̗̬͉̙n͚͜ ̻̞̰͚ͅh̵͉i̳̞v̢͇ḙ͎͟-҉̭̩̼͔m̤̭̫i͕͇̝̦n̗͙ḍ̟ ̯̲͕͞ǫ̟̯̰̲͙̻̝f ̪̰̰̗̖̭̘͘c̦͍̲̞͍̩̙ḥ͚a̮͎̟̙͜ơ̩̹͎s̤.̝̝ ҉Z̡̖̜͖̰̣͉̜a͖̰͙̬͡l̲̫̳͍̩g̡̟̼̱͚̞̬ͅo̗͜.̟", "thenezperdianhivemindofchaoszalgo"), (r"The quick brown fox jumps over the lazy dog", "thequickbrownfoxjumpsoverthelazydog"), (r"The quick brown fox jumps over the lazy dog", "thequickbrownfoxjumpsoverthelazydog"), diff --git a/fuzzycat/verify.py b/fuzzycat/verify.py index fc298c1..e6310cb 100644 --- a/fuzzycat/verify.py +++ b/fuzzycat/verify.py @@ -52,6 +52,7 @@ WIPv1 (10m) "total": 9481815 } + TODO: allow to pass in a DOI blacklist, e.g. a list of DOI which are not valid any more; example: https://fatcat.wiki/release/azbcyqjnmrdofigpgk24ck4rpq, https://fatcat.wiki/release/eb2uf5ae7bedxe22jasf2l3faa |