#!/usr/bin/env python3.1
import datetime, http.cookiejar, optparse, os, sys, shelve, re, urllib.request
parser = optparse.OptionParser()
parser.add_option('-v', '--verbose', action = 'store_true', dest = 'verbose', help = 'turns on verbose status notifications', metavar = 'bool', default = False)
parser.add_option('-d', '--dir', action = 'store', dest = 'dir', help = 'specify directory where the files should be stored', metavar = 'string', default = None)
parser.add_option('-r', '--root', action = 'store', dest = 'root', help = 'specify a site from which data should be mirrored', metavar = 'string', default = None)
parser.add_option('-l', '--log', action = 'store', dest = 'log', help = 'specify a log file to process', metavar = 'string', default = None)
parser.add_option('-e', '--skip-etag', action = 'store_true', dest = 'noetag', help = 'do not process etags', metavar = 'bool', default = False)
(options, args) = parser.parse_args()
assert options.dir, 'Directory not specified'
assert options.root, 'Server not specified'
assert options.log, 'Log file not specified'
assert os.access(options.log, os.R_OK), 'Log file unreadable'
# this is file index - everything is stored in this file
index = shelve.open(options.dir + '/.index')
desc_fields = ('Content-Length', 'Pragma', 'Last-Modified')
ignore_fields = ('Accept-Ranges', 'Age', 'Cache-Control', 'Connection', 'Content-Type', 'Date', 'Expires', 'Server', 'Via', 'X-Cache', 'X-Cache-Lookup', 'X-Powered-By')
if not options.noetag:
desc_fields += 'ETag',
else:
ignore_fields += 'ETag',
block_size = 32768
while True:
unchecked_files = set()
checked_files = 0
# reading log and storing found urls for processing
# check file mtime XXX
with open(options.log, 'r') as log_file:
log_line = re.compile('^[^ ]+ - - \[.*] "(GET|HEAD) (.*?)(\?.*)? HTTP/1.1" (\d+) \d+ "(.*)" "(.*)"$')
for line in log_file:
this_line = log_line.match(line.strip())
if this_line:
unchecked_files.add(this_line.group(2))
for url in unchecked_files:
reload = False
recheck = False
print('Checking file:', url)
# creating empty placeholder in index
if not url in index:
print('This one is new.')
index[url] = {}
reload = True
# creating file name from url
file_name = options.dir + re.compile('%20').sub(' ', url)
# forcibly checking file if no file present
if not reload and not os.access(file_name, os.R_OK):
print('File not found or inaccessible.')
reload = True
# forcibly checking file if file size doesn't match with index data
elif not reload and 'Content-Length' in index[url] and os.stat(file_name).st_size != int(index[url]['Content-Length']):
print('File size is ', os.stat(file_name).st_size, ' and stored file size is ', index[url]['Content-Length'], '.', sep='')
reload = True
# forcibly checking file if index hods Pragma header
if not reload and 'Pragma' in index[url] and index[url]['Pragma'] == 'no-cache':
print('Pragma on: recheck imminent.')
recheck = True
# skipping file processing if there's no need to recheck it and we have checked it at least 4 hours ago
if not recheck and not reload and '__time__' in index[url] and (datetime.datetime.now() - datetime.timedelta(hours = 4) - index[url]['__time__']).days < 0:
continue
try:
with urllib.request.urlopen(options.root + url) as source:
new_headers = {}
headers = source.info()
# stripping unneeded headers (XXX make this inplace?)
for header in headers:
if header in desc_fields:
if header == 'Pragma' and headers[header] != 'no-cache':
print('Pragma:', headers[header])
new_headers[header] = headers[header]
elif not header in ignore_fields:
print('Undefined header "', header, '": ', headers[header], sep='')
# comparing headers with data found in index
# if any header has changed (except Pragma) file is fully downloaded
# same if we get more or less headers
old_keys = set(index[url].keys())
old_keys.discard('__time__')
old_keys.discard('Pragma')
more_keys = set(new_headers.keys()) - old_keys
more_keys.discard('Pragma')
less_keys = old_keys - set(new_headers.keys())
if len(more_keys) > 0:
if not len(old_keys) == 0:
print('More headers appear:', more_keys)
reload = True
elif len(less_keys) > 0:
print('Less headers appear:', less_keys)
else:
for key in index[url].keys():
if key not in ('__time__', 'Pragma') and not index[url][key] == new_headers[key]:
print('Header "', key, '" changed from [', index[url][key], '] to [', new_headers[key], ']', sep='')
reload = True
# downloading file
if reload:
if 'Content-Length' in headers:
print('Downloading', headers['Content-Length'], 'bytes [', end='')
else:
print('Downloading [', end='')
sys.stdout.flush()
# file is created at temporary location and moved in place only when download completes
temp_file = open(options.dir + '/.tmp', 'wb')
buffer = source.read(block_size)
blocks = 0
megs = 0
while len(buffer) > 0:
temp_file.write(buffer)
print('.', end='')
sys.stdout.flush()
buffer = source.read(block_size)
blocks += 1
if blocks > 1024*1024/block_size:
blocks = blocks - 1024*1024/block_size
megs += 1
print('{}Mb'.format(megs), end='')
temp_file.close()
print(']')
os.renames(options.dir + '/.tmp', file_name)
checked_files += 1
# storing new time mark and storing new headers
new_headers['__time__'] = datetime.datetime.now()
index[url] = new_headers
index.sync()
except urllib.error.HTTPError as error:
# in case of error we don't need to do anything actually,
# if file download stalls or fails the file would not be moved to it's location
print(error)
print('[', len(unchecked_files), '/', checked_files, ']')
# checking if there were any files downloaded, if yes - restarting sequence
if checked_files == 0:
break