spending some time studying pycurl and libcurl documentation, i still can't find a (simple) way, how to get HTTP status message (reason-phrase) in pycurl.
status code is easy:
import pycurl
import cStringIO
curl = pycurl.Curl()
buff = cStringIO.StringIO()
curl.setopt(pycurl.URL, 'http://example.org')
curl.setopt(pycurl.WRITEFUNCTION, buff.write)
curl.perform()
print "status code: %s" % curl.getinfo(pycurl.HTTP_CODE)
# -> 200
# print "status message: %s" % ???
# -> "OK"
i've found a solution myself, which does what i need, but could be more robust (works for HTTP).
it's based on a fact that captured headers obtained by pycurl.HEADERFUNCTION
include the status line.
import pycurl
import cStringIO
import re
curl = pycurl.Curl()
buff = cStringIO.StringIO()
hdr = cStringIO.StringIO()
curl.setopt(pycurl.URL, 'http://example.org')
curl.setopt(pycurl.WRITEFUNCTION, buff.write)
curl.setopt(pycurl.HEADERFUNCTION, hdr.write)
curl.perform()
print "status code: %s" % curl.getinfo(pycurl.HTTP_CODE)
# -> 200
status_line = hdr.getvalue().splitlines()[0]
m = re.match(r'HTTP\/\S*\s*\d+\s*(.*?)\s*$', status_line)
if m:
status_message = m.groups(1)
else:
status_message = ''
print "status message: %s" % status_message
# -> "OK"