Extract metrics from ocrevalUAtion HTML reports

ocrevaluation_extract.py

from bs4 import BeautifulSoup                                                                                           
import glob                                                                                                             
                                                                                                                        
                                                                                                                        
METRICS = ["CER", "WER", "WER (order independent)"]                                                                     
FILES = glob.glob("*.html")                                                                                             
                                                                                                                        
print("file;" + ";".join(METRICS))                                                                                      
                                                                                                                        
for f in FILES:                                                                                                         
    soup = BeautifulSoup(open(f), "html.parser")                                                                        
                                                                                                                        
    print(f, end="")                                                                                                    
    for metric in METRICS:                                                                                              
        label = soup.find(text=metric).parent                                                                           
        value = label.nextSibling.text.strip()                                                                          
        print(";" + value, end="")                                                                                      
    print()