search

arthurherbout / crypto_code_detection

Automatic Detection of Custom Cryptographic C Code
8 stars 4 forks source link

Issue with file name in aggregated files #31

Closed arnaudstiegler closed 4 years ago

arnaudstiegler commented 4 years ago

@corentinllorca, @redouane-dziri, Not sure who's in charge for that, but file names for crypto-library are root paths, and it's not very convenient for QA'ing the results

Hadrien-Cornier commented 4 years ago

Indeed test=pd.read_json('C:/Users/hadri/Documents/GitHub/crypto_code_detection/data/full_data.json') print(test.iloc[0]['file_name']) '/Users/red/Documents/CP/crypto_code_detection/data/crypto-library/files/nettle/eddsa-hash.c'

arnaudstiegler commented 4 years ago

I fixed the file name for the crypto-library on my branch, will merge soon with newly generated full_data.json. However, couldn't run the train_test .py file for now so the train.json and test.json are still the old version on my branch

redouane-dziri commented 4 years ago

The only issue is with files from crypto-library right? @arnaudstiegler you're saying this is fixed locally for you then? :) If yes let us know when this is merged and fixed on master

redouane-dziri commented 4 years ago

Ok so need to look at generations of test and train

redouane-dziri commented 4 years ago

PR #38 should've fixed this, and I re-generated all the necessary jsons in it. Close this when you want.