graph-by-quality.py

from bokeh.io import show, output_file
from bokeh.models import ColumnDataSource
from bokeh.palettes import viridis
from bokeh.models.mappers import LinearColorMapper
from bokeh.models import Range1d, LinearAxis
from bokeh.plotting import figure,save
import sys
import re

#bits liberally borrowed from bokeh example code

# need:
# pip install bokeh

# takes pot file or other output file, so hash:password or pwdump style uname:uid:foo:hash:password and makes a graph
# as long as

def big_palette(size, palette_func):
    if size < 256:
        return palette_func(size)
    p = palette_func(256)
    out = []
    for i in range(size):
        idx = int(i * 256.0 / size)
        out.append(p[idx])
    return out

output_file(sys.argv[1]+"q.html")

maxh=0
maxlen=0


freq=[]
pos=[]
tick=0
total=0
maximv=0
total=0
imv=0
itv=0

#parse our input file to get max
with open(sys.argv[1], encoding="utf8") as inf:
    for line in inf:
        ans=line.rstrip()
        if re.match(r'Recovered\.\.\.\.\.\.\.\.: (\d+)/(\d+)',ans):                    
            m = re.search(r'Recovered\.\.\.\.\.\.\.\.: (\d+)/(\d+)',ans)                   
            if m.group(1):
                mv=m.group(1)                            
                imv = int(mv)

            if m.group(2):
                tv=m.group(2)                            
                itv = int(tv)

        if maximv<imv:
            maximv=imv
                
        if total<itv:
            total=itv                

maximv=total

maxp=round(total / (maximv*10))*10 # max percentage from 10 to 100
if maxp<1:
    maxp=1
sf=100/maxp
imv=0

#parse our input file ( #cracked as list, total first )
with open(sys.argv[1], encoding="utf8") as inf:
    for line in inf:
        ans=line.rstrip()
        if re.match(r'Recovered\.\.\.\.\.\.\.\.: (\d+)/(\d+)',ans):                    
            m = re.search(r'Recovered\.\.\.\.\.\.\.\.: (\d+)/(\d+)',ans)                   
            if m.group(1):
                mv=m.group(1)                            
                imv = int(mv)

                freq.append(imv)
                pos.append(tick)
                tick=tick+1

#get an appropriate palette, and reverse the order
cm = big_palette(tick, viridis)
mcm = cm[::-1]


msource = ColumnDataSource(data=dict(ticks=pos, cracked=freq, color=mcm ))

p = figure(x_range=(0,tick), y_range=(0,100.0*total/sf), height=600, title="Passwords by Time To Crack (number on left, proportion of total on right)", toolbar_location=None, tools="")

p.xaxis.axis_label='Ticks (10s each)'
p.yaxis.axis_label='Quantity'

p.extra_y_ranges = {"Percentage": Range1d(start=0, end=maxp)}
p.add_layout(LinearAxis(y_range_name="Percentage"), 'right')

p.vbar(x='ticks', top='cracked', width=1, color='color', source=msource)

p.xgrid.grid_line_color = None
p.ygrid.grid_line_color = None

show(p)
#save(p)