import os,re,pickle
from xml.dom import minidom
from xml import xpath
def getDetails():
mainxpath = "//div[@class='left w420']/div[@class='contentpadlr']/"
files = [x for x in os.listdir('.') if x.endswith('.xhtml')]
cars = []
for f in files:
dom = minidom.parse(f)
names = [x.childNodes[1].nodeValue.replace('\n',' ').strip()
for x in xpath.Evaluate(mainxpath+"h4",
dom.documentElement)]
positions = [int(x.childNodes[0].firstChild.nodeValue)
for x in xpath.Evaluate(mainxpath+"h4",
dom.documentElement)]
uls = [x
for x in xpath.Evaluate(mainxpath+"ul[@class='nobull']",
dom.documentElement)]
files = [f] * len(names)
cars += zip(names,positions,uls,files)
carsdict = {}
for n,p,ul,fil in cars:
names = [x.firstChild.nodeValue
for x in xpath.Evaluate("li/div[@class='w240 left']",ul)]
values = [x.getAttribute('alt').split()[0]
for x in xpath.Evaluate("li/div[@class='left']/img",ul)]
scores = dict(zip(names,values))
haspre = re.search(r'\((.*)\)',n)
if haspre:
pre = haspre.groups()[0]
else:
pre = ""
manufacturer, model = re.sub(r'\((.*)\)','',n).split(' ',1)
if manufacturer in ['Alfa','Land']:
m1,model = model.split(' ',1)
manufacturer = manufacturer + ' ' + m1
carsdict[n] = {"position": p, "scores": scores, "manufacturer": manufacturer.encode('utf-8'),
"model": model.encode('utf-8'), "pre": pre, "file": fil}
# dump for faster move later
fp = open('cache.pickle','wb')
pickle.dump(carsdict,fp)
fp.close()
return carsdict
def uniq(s): return dict(zip(s,[1]*len(s))).keys()
# main
if os.path.exists('cache.pickle'):
fp = open('cache.pickle','rb')
carsdict = pickle.load(fp)
fp.close()
else:
carsdict = getDetails()
# write out output
manufacturers = [c["manufacturer"] for c in carsdict.values()]
manufacturers = uniq(manufacturers)
manufacturers.sort()
print '''
Top Gear Survey easy access edition
This is the easy access version of the BBC
Top Gear 2005 car survey. While the survey itself is massively useful, it's
difficult to navigate; you have to look through all 150-odd entries to find the
car you want. Here's a handy index to all the cars in the survey, with links
to let you jump to each.
'''
links = ['%s' % (m.lower().replace(' ','_'),m)
for m in manufacturers]
print ' '.join(links)
print '
'
for m in manufacturers:
cars = [c for c in carsdict.values() if c['manufacturer'] == m]
posns = [c['position'] for c in cars]
avg = sum(posns) / len(posns)
print '- %s (avg. posn. %s) (top)
' % (
m.lower().replace(' ','_'), m, avg
)
print ''
for c in cars:
print '- %s %s: place %s
' % (
'http://www.topgear.com/content/features/stories/2005/11/stories/01/',
c['file'].replace('.xhtml','.html'),
c['manufacturer'], c['model'], c['position']
)
print '
'
print ''