You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
217 lines
5.5 KiB
217 lines
5.5 KiB
##############################################################################
|
|
#
|
|
# Copyright (c) 2004 Zope Foundation and Contributors.
|
|
# All Rights Reserved.
|
|
#
|
|
# This software is subject to the provisions of the Zope Public License,
|
|
# Version 2.1 (ZPL). A copy of the ZPL should accompany this distribution.
|
|
# THIS SOFTWARE IS PROVIDED "AS IS" AND ANY AND ALL EXPRESS OR IMPLIED
|
|
# WARRANTIES ARE DISCLAIMED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
|
|
# WARRANTIES OF TITLE, MERCHANTABILITY, AGAINST INFRINGEMENT, AND FITNESS
|
|
# FOR A PARTICULAR PURPOSE.
|
|
#
|
|
##############################################################################
|
|
"""Utility functions
|
|
"""
|
|
|
|
import asyncore
|
|
import errno
|
|
import logging
|
|
import os
|
|
import re
|
|
import stat
|
|
import time
|
|
import calendar
|
|
|
|
logger = logging.getLogger('waitress')
|
|
|
|
def find_double_newline(s):
|
|
"""Returns the position just after a double newline in the given string."""
|
|
pos1 = s.find(b'\n\r\n') # One kind of double newline
|
|
if pos1 >= 0:
|
|
pos1 += 3
|
|
pos2 = s.find(b'\n\n') # Another kind of double newline
|
|
if pos2 >= 0:
|
|
pos2 += 2
|
|
|
|
if pos1 >= 0:
|
|
if pos2 >= 0:
|
|
return min(pos1, pos2)
|
|
else:
|
|
return pos1
|
|
else:
|
|
return pos2
|
|
|
|
def concat(*args):
|
|
return ''.join(args)
|
|
|
|
def join(seq, field=' '):
|
|
return field.join(seq)
|
|
|
|
def group(s):
|
|
return '(' + s + ')'
|
|
|
|
short_days = ['sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat']
|
|
long_days = ['sunday', 'monday', 'tuesday', 'wednesday',
|
|
'thursday', 'friday', 'saturday']
|
|
|
|
short_day_reg = group(join(short_days, '|'))
|
|
long_day_reg = group(join(long_days, '|'))
|
|
|
|
daymap = {}
|
|
for i in range(7):
|
|
daymap[short_days[i]] = i
|
|
daymap[long_days[i]] = i
|
|
|
|
hms_reg = join(3 * [group('[0-9][0-9]')], ':')
|
|
|
|
months = ['jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul',
|
|
'aug', 'sep', 'oct', 'nov', 'dec']
|
|
|
|
monmap = {}
|
|
for i in range(12):
|
|
monmap[months[i]] = i + 1
|
|
|
|
months_reg = group(join(months, '|'))
|
|
|
|
# From draft-ietf-http-v11-spec-07.txt/3.3.1
|
|
# Sun, 06 Nov 1994 08:49:37 GMT ; RFC 822, updated by RFC 1123
|
|
# Sunday, 06-Nov-94 08:49:37 GMT ; RFC 850, obsoleted by RFC 1036
|
|
# Sun Nov 6 08:49:37 1994 ; ANSI C's asctime() format
|
|
|
|
# rfc822 format
|
|
rfc822_date = join(
|
|
[concat(short_day_reg, ','), # day
|
|
group('[0-9][0-9]?'), # date
|
|
months_reg, # month
|
|
group('[0-9]+'), # year
|
|
hms_reg, # hour minute second
|
|
'gmt'
|
|
],
|
|
' '
|
|
)
|
|
|
|
rfc822_reg = re.compile(rfc822_date)
|
|
|
|
def unpack_rfc822(m):
|
|
g = m.group
|
|
return (
|
|
int(g(4)), # year
|
|
monmap[g(3)], # month
|
|
int(g(2)), # day
|
|
int(g(5)), # hour
|
|
int(g(6)), # minute
|
|
int(g(7)), # second
|
|
0,
|
|
0,
|
|
0,
|
|
)
|
|
|
|
# rfc850 format
|
|
rfc850_date = join(
|
|
[concat(long_day_reg, ','),
|
|
join(
|
|
[group('[0-9][0-9]?'),
|
|
months_reg,
|
|
group('[0-9]+')
|
|
],
|
|
'-'
|
|
),
|
|
hms_reg,
|
|
'gmt'
|
|
],
|
|
' '
|
|
)
|
|
|
|
rfc850_reg = re.compile(rfc850_date)
|
|
# they actually unpack the same way
|
|
def unpack_rfc850(m):
|
|
g = m.group
|
|
yr = g(4)
|
|
if len(yr) == 2:
|
|
yr = '19' + yr
|
|
return (
|
|
int(yr), # year
|
|
monmap[g(3)], # month
|
|
int(g(2)), # day
|
|
int(g(5)), # hour
|
|
int(g(6)), # minute
|
|
int(g(7)), # second
|
|
0,
|
|
0,
|
|
0
|
|
)
|
|
|
|
# parsdate.parsedate - ~700/sec.
|
|
# parse_http_date - ~1333/sec.
|
|
|
|
weekdayname = ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun']
|
|
monthname = [None, 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
|
|
'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
|
|
|
|
def build_http_date(when):
|
|
year, month, day, hh, mm, ss, wd, y, z = time.gmtime(when)
|
|
return "%s, %02d %3s %4d %02d:%02d:%02d GMT" % (
|
|
weekdayname[wd],
|
|
day, monthname[month], year,
|
|
hh, mm, ss)
|
|
|
|
def parse_http_date(d):
|
|
d = d.lower()
|
|
m = rfc850_reg.match(d)
|
|
if m and m.end() == len(d):
|
|
retval = int(calendar.timegm(unpack_rfc850(m)))
|
|
else:
|
|
m = rfc822_reg.match(d)
|
|
if m and m.end() == len(d):
|
|
retval = int(calendar.timegm(unpack_rfc822(m)))
|
|
else:
|
|
return 0
|
|
return retval
|
|
|
|
class logging_dispatcher(asyncore.dispatcher):
|
|
logger = logger
|
|
|
|
def log_info(self, message, type='info'):
|
|
severity = {
|
|
'info': logging.INFO,
|
|
'warning': logging.WARN,
|
|
'error': logging.ERROR,
|
|
}
|
|
self.logger.log(severity.get(type, logging.INFO), message)
|
|
|
|
def cleanup_unix_socket(path):
|
|
try:
|
|
st = os.stat(path)
|
|
except OSError as exc:
|
|
if exc.errno != errno.ENOENT:
|
|
raise # pragma: no cover
|
|
else:
|
|
if stat.S_ISSOCK(st.st_mode):
|
|
try:
|
|
os.remove(path)
|
|
except OSError: # pragma: no cover
|
|
# avoid race condition error during tests
|
|
pass
|
|
|
|
class Error(object):
|
|
|
|
def __init__(self, body):
|
|
self.body = body
|
|
|
|
class BadRequest(Error):
|
|
code = 400
|
|
reason = 'Bad Request'
|
|
|
|
class RequestHeaderFieldsTooLarge(BadRequest):
|
|
code = 431
|
|
reason = 'Request Header Fields Too Large'
|
|
|
|
class RequestEntityTooLarge(BadRequest):
|
|
code = 413
|
|
reason = 'Request Entity Too Large'
|
|
|
|
class InternalServerError(Error):
|
|
code = 500
|
|
reason = 'Internal Server Error'
|