08ae38b6ce 2010-06-25 1: #!/usr/bin/env python3.1
08ae38b6ce 2010-06-25 2:
e7b837a681 2010-08-25 3: import datetime, http.cookiejar, os, sys, shelve, spacemap, re, urllib.request
e7b837a681 2010-08-25 4:
e7b837a681 2010-08-25 5: class Config:
e7b837a681 2010-08-25 6: __slots__ = frozenset(['_config', '_default', '_section', 'options', 'root'])
e7b837a681 2010-08-25 7: _default = {
e7b837a681 2010-08-25 8: 'general': {
e7b837a681 2010-08-25 9: 'port': '8008',
e7b837a681 2010-08-25 10: },
e7b837a681 2010-08-25 11: '_other': {
e7b837a681 2010-08-25 12: 'verbose': 'no',
e7b837a681 2010-08-25 13: 'noetag': 'no',
e7b837a681 2010-08-25 14: 'noparts': 'no',
e7b837a681 2010-08-25 15: },}
e7b837a681 2010-08-25 16:
e7b837a681 2010-08-25 17: # function to read in config file
e7b837a681 2010-08-25 18: def __init__(self):
e7b837a681 2010-08-25 19: import configparser, optparse
e7b837a681 2010-08-25 20:
e7b837a681 2010-08-25 21: parser = optparse.OptionParser()
e7b837a681 2010-08-25 22: parser.add_option('-c', '--config', dest = 'config', help = 'config file location', metavar = 'FILE', default = 'samesite.conf')
e7b837a681 2010-08-25 23: (self.options, args) = parser.parse_args()
e7b837a681 2010-08-25 24:
e7b837a681 2010-08-25 25: assert os.access(self.options.config, os.R_OK), "Fatal error: can't read {}".format(self.options.config)
e7b837a681 2010-08-25 26:
e7b837a681 2010-08-25 27: configDir = re.compile('^(.*)/[^/]+$').match(self.options.config)
e7b837a681 2010-08-25 28: if configDir:
e7b837a681 2010-08-25 29: self.root = configDir.group(1)
e7b837a681 2010-08-25 30: else:
e7b837a681 2010-08-25 31: self.root = os.getcwd()
e7b837a681 2010-08-25 32:
e7b837a681 2010-08-25 33: self._config = configparser.ConfigParser()
e7b837a681 2010-08-25 34: self._config.readfp(open(self.options.config))
e7b837a681 2010-08-25 35:
e7b837a681 2010-08-25 36: for section in self._config.sections():
e7b837a681 2010-08-25 37: if section != 'general':
e7b837a681 2010-08-25 38: if self._config.has_option(section, 'dir'):
e7b837a681 2010-08-25 39: if re.compile('^/$').match(self._config.get(section, 'dir')):
e7b837a681 2010-08-25 40: self._config.set(section, 'dir', self.root + os.sep + section)
e7b837a681 2010-08-25 41: thisDir = re.compile('^(.*)/$').match(self._config.get(section, 'dir'))
e7b837a681 2010-08-25 42: if thisDir:
e7b837a681 2010-08-25 43: self._config.set(section, 'dir', thisDir.group(1))
e7b837a681 2010-08-25 44: if not re.compile('^/(.*)$').match(self._config.get(section, 'dir')):
e7b837a681 2010-08-25 45: self._config.set(section, 'dir', self.root + os.sep + self._config.get(section, 'dir'))
e7b837a681 2010-08-25 46: else:
e7b837a681 2010-08-25 47: self._config.set(section, 'dir', self.root + os.sep + section)
e7b837a681 2010-08-25 48:
e7b837a681 2010-08-25 49: if not self._config.has_option(section, 'root'):
e7b837a681 2010-08-25 50: self._config.set(section, 'root', section)
e7b837a681 2010-08-25 51:
e7b837a681 2010-08-25 52: # function to select config file section or create one
e7b837a681 2010-08-25 53: def section(self, section):
e7b837a681 2010-08-25 54: if not self._config.has_section(section):
e7b837a681 2010-08-25 55: self._config.add_section(section)
e7b837a681 2010-08-25 56: self._section = section
e7b837a681 2010-08-25 57:
e7b837a681 2010-08-25 58: # function to get config parameter, if parameter doesn't exists the default
e7b837a681 2010-08-25 59: # value or None is substituted
e7b837a681 2010-08-25 60: def __getitem__(self, name):
e7b837a681 2010-08-25 61: if not self._config.has_option(self._section, name):
e7b837a681 2010-08-25 62: if self._section in self._default:
e7b837a681 2010-08-25 63: if name in self._default[self._section]:
e7b837a681 2010-08-25 64: self._config.set(self._section, name, self._default[self._section][name])
e7b837a681 2010-08-25 65: else:
e7b837a681 2010-08-25 66: self._config.set(self._section, name, None)
e7b837a681 2010-08-25 67: elif name in self._default['_other']:
e7b837a681 2010-08-25 68: self._config.set(self._section, name, self._default['_other'][name])
e7b837a681 2010-08-25 69: else:
e7b837a681 2010-08-25 70: self._config.set(self._section, name, None)
e7b837a681 2010-08-25 71: return(self._config.get(self._section, name))
e7b837a681 2010-08-25 72:
e7b837a681 2010-08-25 73: config = Config()
e7b837a681 2010-08-25 74:
e7b837a681 2010-08-25 75: #assert options.port or os.access(options.log, os.R_OK), 'Log file unreadable'
e7b837a681 2010-08-25 76:
e7b837a681 2010-08-25 77: const_desc_fields = set(['Content-Length', 'Pragma', 'Last-Modified'])
e7b837a681 2010-08-25 78: const_ignore_fields = set(['Accept-Ranges', 'Age', 'Cache-Control', 'Connection', 'Content-Type', 'Date', 'Expires', 'Server', 'Via', 'X-Cache', 'X-Cache-Lookup', 'X-Powered-By'])
80f8e3804a 2010-08-20 79:
80f8e3804a 2010-08-20 80: block_size = 4096
80f8e3804a 2010-08-20 81:
80f8e3804a 2010-08-20 82: '''
80f8e3804a 2010-08-20 83: # later, kqueue would be good but later
80f8e3804a 2010-08-20 84: class Connection:
80f8e3804a 2010-08-20 85: __slots__ = frozenset(('__address', '__input', '__socket', '__status', 'error', 'method', 'url', 'http_version'))
80f8e3804a 2010-08-20 86:
80f8e3804a 2010-08-20 87: def __init__(self, socket, address):
80f8e3804a 2010-08-20 88: self.__address = address
80f8e3804a 2010-08-20 89: self.__input = b''
80f8e3804a 2010-08-20 90: self.__socket = socket
80f8e3804a 2010-08-20 91: self.__status = 0
80f8e3804a 2010-08-20 92:
80f8e3804a 2010-08-20 93: def read(self, kev):
80f8e3804a 2010-08-20 94: buffer = self.__socket.recv(kev.data)
80f8e3804a 2010-08-20 95: exhausted = False
80f8e3804a 2010-08-20 96: if len(buffer) == 0:
80f8e3804a 2010-08-20 97: eof = True
80f8e3804a 2010-08-20 98: else:
80f8e3804a 2010-08-20 99: self.__input += buffer
80f8e3804a 2010-08-20 100: while not exhausted:
80f8e3804a 2010-08-20 101: if self.__status == -1:
80f8e3804a 2010-08-20 102: exhausted = True
80f8e3804a 2010-08-20 103: elif self.__status == 0:
80f8e3804a 2010-08-20 104: endstring = self.__input.find(b'\n')
80f8e3804a 2010-08-20 105: if endstring > 0:
80f8e3804a 2010-08-20 106: print('Processing request line.')
80f8e3804a 2010-08-20 107: line = self.__input[:endstring].decode('ascii')
80f8e3804a 2010-08-20 108: self.__input = self.__input[endstring + 1:]
80f8e3804a 2010-08-20 109: isRequest = re.compile('(GET) ([^ ]+) HTTP/(1\.0)').match(line)
80f8e3804a 2010-08-20 110: if not isRequest:
80f8e3804a 2010-08-20 111: self.error = 'Not a HTTP connection.'
80f8e3804a 2010-08-20 112: self.__status = -1
80f8e3804a 2010-08-20 113: else:
80f8e3804a 2010-08-20 114: self.method = isRequest.group(1)
80f8e3804a 2010-08-20 115: self.url = isRequest.group(2)
80f8e3804a 2010-08-20 116: self.http_version = isRequest.group(3)
80f8e3804a 2010-08-20 117: self.__status = 1
80f8e3804a 2010-08-20 118: else:
80f8e3804a 2010-08-20 119: exhausted = True
80f8e3804a 2010-08-20 120: elif self.__status == 1:
80f8e3804a 2010-08-20 121: endstring = self.__input.find(b'\n')
80f8e3804a 2010-08-20 122: if endstring > 0:
80f8e3804a 2010-08-20 123: print('Processing header line.' + repr(self.__input))
80f8e3804a 2010-08-20 124: line = self.__input[:endstring].decode('ascii')
80f8e3804a 2010-08-20 125: self.__input = self.__input[endstring + 1:]
80f8e3804a 2010-08-20 126: isHeader = re.compile('([^:]*): +(.*)').match(line)
80f8e3804a 2010-08-20 127: if not isHeader:
80f8e3804a 2010-08-20 128: self.error = 'Bad header.'
80f8e3804a 2010-08-20 129: return(False)
80f8e3804a 2010-08-20 130: # process header here
80f8e3804a 2010-08-20 131: elif endstring == 0:
80f8e3804a 2010-08-20 132: self.__status = 2
80f8e3804a 2010-08-20 133: else:
80f8e3804a 2010-08-20 134: exhausted = True
80f8e3804a 2010-08-20 135:
80f8e3804a 2010-08-20 136: def write(self, kev):
80f8e3804a 2010-08-20 137: pass
80f8e3804a 2010-08-20 138:
80f8e3804a 2010-08-20 139: if options.port:
80f8e3804a 2010-08-20 140: import select, socket
80f8e3804a 2010-08-20 141:
80f8e3804a 2010-08-20 142: sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
80f8e3804a 2010-08-20 143: try:
80f8e3804a 2010-08-20 144: sock.bind(('127.0.0.1', int(options.port)))
80f8e3804a 2010-08-20 145: sock.listen(-1)
80f8e3804a 2010-08-20 146:
80f8e3804a 2010-08-20 147: kq = select.kqueue()
80f8e3804a 2010-08-20 148: assert kq.fileno() != -1, "Fatal error: can't initialise kqueue."
80f8e3804a 2010-08-20 149:
80f8e3804a 2010-08-20 150: kq.control([select.kevent(sock, select.KQ_FILTER_READ, select.KQ_EV_ADD)], 0)
80f8e3804a 2010-08-20 151: timeout = None
80f8e3804a 2010-08-20 152:
80f8e3804a 2010-08-20 153: connections = {sock.fileno(): None}
80f8e3804a 2010-08-20 154:
80f8e3804a 2010-08-20 155: while True:
80f8e3804a 2010-08-20 156: kevs = kq.control(None, 1, timeout)
80f8e3804a 2010-08-20 157:
80f8e3804a 2010-08-20 158: for kev in kevs:
80f8e3804a 2010-08-20 159: if type(connections[kev.ident]) == Connection:
80f8e3804a 2010-08-20 160: print(kev.ident, kev.data, kev.filter, kev.flags)
80f8e3804a 2010-08-20 161: assert kev.data != 0, 'No data available.'
80f8e3804a 2010-08-20 162: if kev.filter == select.KQ_FILTER_READ:
80f8e3804a 2010-08-20 163: connections[kev.ident].read(kev)
80f8e3804a 2010-08-20 164: elif kev.filter == select.KQ_FILTER_WRITE:
80f8e3804a 2010-08-20 165: connections[kev.ident].write(kev)
80f8e3804a 2010-08-20 166: else:
80f8e3804a 2010-08-20 167: assert kev.filter in (select.KQ_FILTER_READ, select.KQ_FILTER_WRITE), 'Do we support other filters?'
80f8e3804a 2010-08-20 168: else:
80f8e3804a 2010-08-20 169: (conn, addr) = sock.accept()
80f8e3804a 2010-08-20 170: print('Connection from ' + repr(addr))
80f8e3804a 2010-08-20 171: kq.control([select.kevent(conn, select.KQ_FILTER_READ, select.KQ_EV_ADD)], 0)
80f8e3804a 2010-08-20 172: connections[conn.fileno()] = Connection(conn, addr)
80f8e3804a 2010-08-20 173:
80f8e3804a 2010-08-20 174: if kev.flags >> 15 == 1:
80f8e3804a 2010-08-20 175: kq.control([select.kevent(kev.ident, select.KQ_FILTER_READ, select.KQ_EV_DELETE)], 0)
80f8e3804a 2010-08-20 176: kq.control([select.kevent(kev.ident, select.KQ_FILTER_WRITE, select.KQ_EV_DELETE)], 0)
80f8e3804a 2010-08-20 177: del(connections[kev.ident])
80f8e3804a 2010-08-20 178: finally:
80f8e3804a 2010-08-20 179: sock.close()
80f8e3804a 2010-08-20 180: '''
80f8e3804a 2010-08-20 181:
e7b837a681 2010-08-25 182: # XXX how about rechecking files?
e7b837a681 2010-08-25 183: if True:
80f8e3804a 2010-08-20 184: import http.server
80f8e3804a 2010-08-20 185:
80f8e3804a 2010-08-20 186: class MyRequestHandler(http.server.BaseHTTPRequestHandler):
80f8e3804a 2010-08-20 187: def __process(self):
80f8e3804a 2010-08-20 188: # reload means file needs to be reloaded to serve request
80f8e3804a 2010-08-20 189: reload = False
80f8e3804a 2010-08-20 190: # recheck means file needs to be checked, this also means that if file hav been modified we can serve older copy
80f8e3804a 2010-08-20 191: recheck = False
80f8e3804a 2010-08-20 192: # file_stat means file definitely exists
80f8e3804a 2010-08-20 193: file_stat = None
80f8e3804a 2010-08-20 194: # requested_ranges holds data about any range requested
80f8e3804a 2010-08-20 195: requested_ranges = None
80f8e3804a 2010-08-20 196: # records holds data from index locally, should be written back upon successfull completion
80f8e3804a 2010-08-20 197: record = None
80f8e3804a 2010-08-20 198: info = 'Checking file: ' + self.path
80f8e3804a 2010-08-20 199:
d0071bdbc7 2010-08-20 200: myPath = re.compile('^(.*?)(\?.*)$').match(self.path)
d0071bdbc7 2010-08-20 201: if myPath:
d0071bdbc7 2010-08-20 202: my_path = myPath.group(1)
d0071bdbc7 2010-08-20 203: else:
d0071bdbc7 2010-08-20 204: my_path = self.path
d0071bdbc7 2010-08-20 205:
e7b837a681 2010-08-25 206: config.section(self.headers['Host'])
e7b837a681 2010-08-25 207:
e7b837a681 2010-08-25 208: if not os.access(config['dir'], os.X_OK):
e7b837a681 2010-08-25 209: os.mkdir(config['dir'])
e7b837a681 2010-08-25 210: # this is file index - everything is stored in this file
e7b837a681 2010-08-25 211: # _parts - list of stored parts of file
e7b837a681 2010-08-25 212: # _time - last time the file was checked
e7b837a681 2010-08-25 213: # everything else is just the headers
e7b837a681 2010-08-25 214: index = shelve.open(config['dir'] + os.sep + '.index')
e7b837a681 2010-08-25 215:
e7b837a681 2010-08-25 216: desc_fields = const_desc_fields.copy()
e7b837a681 2010-08-25 217: ignore_fields = const_ignore_fields.copy()
e7b837a681 2010-08-25 218: if not config['noetag']:
e7b837a681 2010-08-25 219: desc_fields.add('ETag')
e7b837a681 2010-08-25 220: else:
e7b837a681 2010-08-25 221: ignore_fields.add('ETag')
e7b837a681 2010-08-25 222:
80f8e3804a 2010-08-20 223: proxy_ignored = ('Accept', 'Accept-Encoding',
80f8e3804a 2010-08-20 224: 'Cache-Control', 'Connection',
80f8e3804a 2010-08-20 225: 'Host',
fb10031536 2010-08-21 226: 'If-Modified-Since', 'If-Unmodified-Since',
80f8e3804a 2010-08-20 227: 'User-Agent',
80f8e3804a 2010-08-20 228: 'Via',
80f8e3804a 2010-08-20 229: 'X-Forwarded-For',
80f8e3804a 2010-08-20 230: )
80f8e3804a 2010-08-20 231:
fb10031536 2010-08-21 232: print('===============[ {} request ]==='.format(self.command))
80f8e3804a 2010-08-20 233:
80f8e3804a 2010-08-20 234: for header in self.headers:
80f8e3804a 2010-08-20 235: if header in proxy_ignored:
80f8e3804a 2010-08-20 236: pass
80f8e3804a 2010-08-20 237: elif header in ('Range'):
80f8e3804a 2010-08-20 238: isRange = re.compile('bytes=(\d+)-(\d+)').match(self.headers[header])
80f8e3804a 2010-08-20 239: if isRange:
e7b837a681 2010-08-25 240: requested_ranges = spacemap.SpaceMap({int(isRange.group(1)): int(isRange.group(2)) + 1})
80f8e3804a 2010-08-20 241: else:
80f8e3804a 2010-08-20 242: return()
e7b837a681 2010-08-25 243: elif header in ('Pragma'):
e7b837a681 2010-08-25 244: if my_path in index:
e7b837a681 2010-08-25 245: index[my_path][header] = self.headers[header]
80f8e3804a 2010-08-20 246: else:
80f8e3804a 2010-08-20 247: print('Unknown header - ', header, ': ', self.headers[header], sep='')
80f8e3804a 2010-08-20 248: return()
80f8e3804a 2010-08-20 249: print(header, self.headers[header])
80f8e3804a 2010-08-20 250:
80f8e3804a 2010-08-20 251: # creating empty placeholder in index
80f8e3804a 2010-08-20 252: # if there's no space map and there's no file in real directory - we have no file
80f8e3804a 2010-08-20 253: # if there's an empty space map - file is full
80f8e3804a 2010-08-20 254: # space map generally covers every bit of file we don't posess currently
d0071bdbc7 2010-08-20 255: if not my_path in index:
80f8e3804a 2010-08-20 256: info += '\nThis one is new.'
80f8e3804a 2010-08-20 257: reload = True
e7b837a681 2010-08-25 258: record = {}
80f8e3804a 2010-08-20 259: else:
d0071bdbc7 2010-08-20 260: record = index[my_path]
e7b837a681 2010-08-25 261:
e7b837a681 2010-08-25 262: if not '_parts' in record:
e7b837a681 2010-08-25 263: record['_parts'] = None
d0071bdbc7 2010-08-20 264:
d0071bdbc7 2010-08-20 265: # creating file name from my_path
e7b837a681 2010-08-25 266: file_name = config['dir'] + os.sep + re.compile('%20').sub(' ', my_path)
80f8e3804a 2010-08-20 267: # partial file or unfinished download
e7b837a681 2010-08-25 268: temp_name = config['dir'] + os.sep + '.parts' + re.compile('%20').sub(' ', my_path)
80f8e3804a 2010-08-20 269:
80f8e3804a 2010-08-20 270: # forcibly checking file if no file present
80f8e3804a 2010-08-20 271: if os.access(file_name, os.R_OK):
80f8e3804a 2010-08-20 272: file_stat = os.stat(file_name)
80f8e3804a 2010-08-20 273: elif '_parts' in record and os.access(temp_name, os.R_OK):
80f8e3804a 2010-08-20 274: file_stat = os.stat(temp_name)
80f8e3804a 2010-08-20 275: elif not reload:
e7b837a681 2010-08-25 276: print(record)
80f8e3804a 2010-08-20 277: info += '\nFile not found or inaccessible.'
e7b837a681 2010-08-25 278: record['_parts'] = None
80f8e3804a 2010-08-20 279: reload = True
80f8e3804a 2010-08-20 280:
80f8e3804a 2010-08-20 281: # forcibly checking file if file size doesn't match with index data
80f8e3804a 2010-08-20 282: if not reload:
e7b837a681 2010-08-25 283: if '_parts' in record and record['_parts'] == spacemap.SpaceMap():
80f8e3804a 2010-08-20 284: if 'Content-Length' in record and file_stat and file_stat.st_size != int(record['Content-Length']):
80f8e3804a 2010-08-20 285: info += '\nFile size is {} and stored file size is {}.'.format(file_stat.st_size, record['Content-Length'])
80f8e3804a 2010-08-20 286: reload = True
80f8e3804a 2010-08-20 287:
80f8e3804a 2010-08-20 288: # forcibly checking file if index holds Pragma header
80f8e3804a 2010-08-20 289: if not reload and 'Pragma' in record and record['Pragma'] == 'no-cache':
80f8e3804a 2010-08-20 290: info +='\nPragma on: recheck imminent.'
80f8e3804a 2010-08-20 291: recheck = True
80f8e3804a 2010-08-20 292:
80f8e3804a 2010-08-20 293: # skipping file processing if there's no need to recheck it and we have checked it at least 4 hours ago
80f8e3804a 2010-08-20 294: if not recheck and not reload and '_time' in record and (datetime.datetime.now() - datetime.timedelta(hours = 4) - record['_time']).days < 0:
80f8e3804a 2010-08-20 295: recheck = True
80f8e3804a 2010-08-20 296:
80f8e3804a 2010-08-20 297: print(info)
80f8e3804a 2010-08-20 298: if reload or recheck:
80f8e3804a 2010-08-20 299:
80f8e3804a 2010-08-20 300: try:
e7b837a681 2010-08-25 301: request = 'http://' + config['root'] + my_path
fb10031536 2010-08-21 302: needed = None
e7b837a681 2010-08-25 303: # XXX and if we specify full file we don't go partial?
80f8e3804a 2010-08-20 304: if requested_ranges != None:
80f8e3804a 2010-08-20 305: if '_parts' in record and record['_parts'] != None:
80f8e3804a 2010-08-20 306: needed = record['_parts'] & requested_ranges
e7b837a681 2010-08-25 307: elif config['noparts']:
e7b837a681 2010-08-25 308: needed = record['_parts']
80f8e3804a 2010-08-20 309: else:
80f8e3804a 2010-08-20 310: needed = requested_ranges
80f8e3804a 2010-08-20 311: ranges = ()
fb10031536 2010-08-21 312: print('Missing ranges: {}, requested ranges: {}, needed ranges: {}.'.format(record['_parts'], requested_ranges, needed))
fb10031536 2010-08-21 313: if len(needed) > 0:
fb10031536 2010-08-21 314: needed.rewind()
fb10031536 2010-08-21 315: while True:
fb10031536 2010-08-21 316: range = needed.pop()
fb10031536 2010-08-21 317: if range[0] == None:
fb10031536 2010-08-21 318: break
fb10031536 2010-08-21 319: ranges += '{}-{}'.format(range[0], range[1] - 1),
fb10031536 2010-08-21 320: request = urllib.request.Request(request, headers = {'Range': 'bytes=' + ','.join(ranges)})
80f8e3804a 2010-08-20 321:
80f8e3804a 2010-08-20 322: with urllib.request.urlopen(request) as source:
80f8e3804a 2010-08-20 323: new_record = {}
80f8e3804a 2010-08-20 324: new_record['_parts'] = record['_parts']
80f8e3804a 2010-08-20 325: headers = source.info()
80f8e3804a 2010-08-20 326:
80f8e3804a 2010-08-20 327: # stripping unneeded headers (XXX make this inplace?)
80f8e3804a 2010-08-20 328: for header in headers:
80f8e3804a 2010-08-20 329: if header in desc_fields:
80f8e3804a 2010-08-20 330: #if header == 'Pragma' and headers[header] != 'no-cache':
80f8e3804a 2010-08-20 331: if header == 'Content-Length':
80f8e3804a 2010-08-20 332: if 'Content-Range' not in headers:
fb10031536 2010-08-21 333: new_record[header] = int(headers[header])
80f8e3804a 2010-08-20 334: else:
80f8e3804a 2010-08-20 335: new_record[header] = headers[header]
80f8e3804a 2010-08-20 336: elif header == 'Content-Range':
80f8e3804a 2010-08-20 337: range = re.compile('^bytes (\d+)-(\d+)/(\d+)$').match(headers[header])
80f8e3804a 2010-08-20 338: if range:
fb10031536 2010-08-21 339: new_record['Content-Length'] = int(range.group(3))
80f8e3804a 2010-08-20 340: else:
80f8e3804a 2010-08-20 341: assert False, 'Content-Range unrecognized.'
80f8e3804a 2010-08-20 342: elif not header in ignore_fields:
80f8e3804a 2010-08-20 343: print('Undefined header "', header, '": ', headers[header], sep='')
80f8e3804a 2010-08-20 344:
80f8e3804a 2010-08-20 345: # comparing headers with data found in index
80f8e3804a 2010-08-20 346: # if any header has changed (except Pragma) file is fully downloaded
80f8e3804a 2010-08-20 347: # same if we get more or less headers
80f8e3804a 2010-08-20 348: old_keys = set(record.keys())
80f8e3804a 2010-08-20 349: old_keys.discard('_time')
80f8e3804a 2010-08-20 350: old_keys.discard('Pragma')
80f8e3804a 2010-08-20 351: more_keys = set(new_record.keys()) - old_keys
80f8e3804a 2010-08-20 352: more_keys.discard('Pragma')
80f8e3804a 2010-08-20 353: less_keys = old_keys - set(new_record.keys())
80f8e3804a 2010-08-20 354: if len(more_keys) > 0:
80f8e3804a 2010-08-20 355: if not len(old_keys) == 0:
80f8e3804a 2010-08-20 356: print('More headers appear:', more_keys)
80f8e3804a 2010-08-20 357: reload = True
80f8e3804a 2010-08-20 358: elif len(less_keys) > 0:
80f8e3804a 2010-08-20 359: print('Less headers appear:', less_keys)
80f8e3804a 2010-08-20 360: else:
80f8e3804a 2010-08-20 361: for key in record.keys():
80f8e3804a 2010-08-20 362: if key[0] != '_' and key != 'Pragma' and not record[key] == new_record[key]:
80f8e3804a 2010-08-20 363: print('Header "', key, '" changed from [', record[key], '] to [', new_record[key], ']', sep='')
fb10031536 2010-08-21 364: print(type(record[key]), type(new_record[key]))
80f8e3804a 2010-08-20 365: reload = True
80f8e3804a 2010-08-20 366:
80f8e3804a 2010-08-20 367: if reload:
80f8e3804a 2010-08-20 368: print('Reloading.')
80f8e3804a 2010-08-20 369: if os.access(temp_name, os.R_OK):
80f8e3804a 2010-08-20 370: os.unlink(temp_name)
80f8e3804a 2010-08-20 371: if os.access(file_name, os.R_OK):
80f8e3804a 2010-08-20 372: os.unlink(file_name)
fb10031536 2010-08-21 373: if new_record['_parts'] == None or reload:
e7b837a681 2010-08-25 374: new_record['_parts'] = spacemap.SpaceMap({0: int(new_record['Content-Length'])})
fb10031536 2010-08-21 375: print(new_record)
80f8e3804a 2010-08-20 376:
80f8e3804a 2010-08-20 377: # downloading file or segment
80f8e3804a 2010-08-20 378: if 'Content-Length' in new_record:
fb10031536 2010-08-21 379: if needed == None:
fb10031536 2010-08-21 380: needed = new_record['_parts']
80f8e3804a 2010-08-20 381: else:
fb10031536 2010-08-21 382: if len(needed) > 1:
80f8e3804a 2010-08-20 383: print("Multipart requests currently not supported.")
80f8e3804a 2010-08-20 384: assert False, 'Skip this one for now.'
80f8e3804a 2010-08-20 385: else:
80f8e3804a 2010-08-20 386: assert False, 'No Content-Length or Content-Range header.'
80f8e3804a 2010-08-20 387:
fb10031536 2010-08-21 388: new_record['_time'] = datetime.datetime.now()
fb10031536 2010-08-21 389: if self.command not in ('HEAD'):
fb10031536 2010-08-21 390: # file is created at temporary location and moved in place only when download completes
fb10031536 2010-08-21 391: if not os.access(temp_name, os.R_OK):
e7b837a681 2010-08-25 392: empty_name = config['dir'] + os.sep + '.tmp'
fb10031536 2010-08-21 393: with open(empty_name, 'w+b') as some_file:
fb10031536 2010-08-21 394: pass
fb10031536 2010-08-21 395: os.renames(empty_name, temp_name)
fb10031536 2010-08-21 396: temp_file = open(temp_name, 'r+b')
fb10031536 2010-08-21 397: needed.rewind()
fb10031536 2010-08-21 398: while True:
fb10031536 2010-08-21 399: (start, end) = needed.pop()
fb10031536 2010-08-21 400: if start == None:
fb10031536 2010-08-21 401: break
fb10031536 2010-08-21 402: stream_last = start
fb10031536 2010-08-21 403: old_record = new_record
fb10031536 2010-08-21 404: if end - start < block_size:
fb10031536 2010-08-21 405: req_block_size = end - start
fb10031536 2010-08-21 406: else:
fb10031536 2010-08-21 407: req_block_size = block_size
fb10031536 2010-08-21 408: buffer = source.read(req_block_size)
fb10031536 2010-08-21 409: length = len(buffer)
fb10031536 2010-08-21 410: while length > 0 and stream_last < end:
fb10031536 2010-08-21 411: stream_pos = stream_last + length
fb10031536 2010-08-21 412: assert not stream_pos > end, 'Received more data then requested: pos:{} start:{} end:{}.'.format(stream_pos, start, end)
fb10031536 2010-08-21 413: temp_file.seek(stream_last)
fb10031536 2010-08-21 414: temp_file.write(buffer)
e7b837a681 2010-08-25 415: new_record['_parts'] = new_record['_parts'] - spacemap.SpaceMap({stream_last: stream_pos})
fb10031536 2010-08-21 416: index[my_path] = old_record
fb10031536 2010-08-21 417: index.sync()
fb10031536 2010-08-21 418: old_record = new_record
fb10031536 2010-08-21 419: stream_last = stream_pos
fb10031536 2010-08-21 420: if end - stream_last < block_size:
fb10031536 2010-08-21 421: req_block_size = end - stream_last
fb10031536 2010-08-21 422: buffer = source.read(req_block_size)
fb10031536 2010-08-21 423: length = len(buffer)
fb10031536 2010-08-21 424: # moving downloaded data to real file
fb10031536 2010-08-21 425: temp_file.close()
fb10031536 2010-08-21 426:
fb10031536 2010-08-21 427: print(new_record)
fb10031536 2010-08-21 428: index[my_path] = new_record
fb10031536 2010-08-21 429: index.sync()
80f8e3804a 2010-08-20 430:
80f8e3804a 2010-08-20 431: except urllib.error.HTTPError as error:
80f8e3804a 2010-08-20 432: # in case of error we don't need to do anything actually,
80f8e3804a 2010-08-20 433: # if file download stalls or fails the file would not be moved to it's location
80f8e3804a 2010-08-20 434: print(error)
e7b837a681 2010-08-25 435:
e7b837a681 2010-08-25 436: if '_parts' in index[my_path] and index[my_path]['_parts'] == spacemap.SpaceMap():
e7b837a681 2010-08-25 437: # just moving
e7b837a681 2010-08-25 438: # drop old dirs XXX
e7b837a681 2010-08-25 439: print('Moving temporary file to new destination.')
e7b837a681 2010-08-25 440: os.renames(temp_name, file_name)
fb10031536 2010-08-21 441:
80f8e3804a 2010-08-20 442: if self.command == 'HEAD':
80f8e3804a 2010-08-20 443: self.send_response(200)
d0071bdbc7 2010-08-20 444: if 'Content-Length' in index[my_path]:
d0071bdbc7 2010-08-20 445: self.send_header('Content-Length', index[my_path]['Content-Length'])
80f8e3804a 2010-08-20 446: self.send_header('Accept-Ranges', 'bytes')
80f8e3804a 2010-08-20 447: self.send_header('Content-Type', 'application/octet-stream')
d0071bdbc7 2010-08-20 448: if 'Last-Modified' in index[my_path]:
d0071bdbc7 2010-08-20 449: self.send_header('Last-Modified', index[my_path]['Last-Modified'])
80f8e3804a 2010-08-20 450: self.end_headers()
80f8e3804a 2010-08-20 451: else:
e7b837a681 2010-08-25 452: if ('_parts' in index[my_path] and index[my_path]['_parts'] != spacemap.SpaceMap()) or not os.access(file_name, os.R_OK):
80f8e3804a 2010-08-20 453: file_name = temp_name
80f8e3804a 2010-08-20 454:
80f8e3804a 2010-08-20 455: with open(file_name, 'rb') as real_file:
80f8e3804a 2010-08-20 456: file_stat = os.stat(file_name)
fb10031536 2010-08-21 457: if 'Range' in self.headers:
fb10031536 2010-08-21 458: self.send_response(206)
80f8e3804a 2010-08-20 459: ranges = ()
80f8e3804a 2010-08-20 460: requested_ranges.rewind()
80f8e3804a 2010-08-20 461: while True:
80f8e3804a 2010-08-20 462: pair = requested_ranges.pop()
80f8e3804a 2010-08-20 463: if pair[0] == None:
80f8e3804a 2010-08-20 464: break
80f8e3804a 2010-08-20 465: ranges += '{}-{}'.format(pair[0], str(pair[1] - 1)),
fb10031536 2010-08-21 466: self.send_header('Content-Range', 'bytes {}/{}'.format(','.join(ranges), index[my_path]['Content-Length']))
80f8e3804a 2010-08-20 467: else:
fb10031536 2010-08-21 468: self.send_response(200)
80f8e3804a 2010-08-20 469: self.send_header('Content-Length', str(file_stat.st_size))
e7b837a681 2010-08-25 470: requested_ranges = spacemap.SpaceMap({0: file_stat.st_size})
fb10031536 2010-08-21 471: self.send_header('Last-Modified', index[my_path]['Last-Modified'])
80f8e3804a 2010-08-20 472: self.send_header('Content-Type', 'application/octet-stream')
80f8e3804a 2010-08-20 473: self.end_headers()
80f8e3804a 2010-08-20 474: if self.command in ('GET'):
fb10031536 2010-08-21 475: if len(requested_ranges) > 0:
fb10031536 2010-08-21 476: requested_ranges.rewind()
fb10031536 2010-08-21 477: (start, end) = requested_ranges.pop()
fb10031536 2010-08-21 478: else:
fb10031536 2010-08-21 479: start = 0
fb10031536 2010-08-21 480: end = index[my_path]['Content-Length']
80f8e3804a 2010-08-20 481: real_file.seek(start)
80f8e3804a 2010-08-20 482: if block_size > end - start:
80f8e3804a 2010-08-20 483: req_block_size = end - start
80f8e3804a 2010-08-20 484: else:
80f8e3804a 2010-08-20 485: req_block_size = block_size
80f8e3804a 2010-08-20 486: buffer = real_file.read(req_block_size)
80f8e3804a 2010-08-20 487: length = len(buffer)
80f8e3804a 2010-08-20 488: while length > 0:
80f8e3804a 2010-08-20 489: self.wfile.write(buffer)
80f8e3804a 2010-08-20 490: start += len(buffer)
80f8e3804a 2010-08-20 491: if req_block_size > end - start:
80f8e3804a 2010-08-20 492: req_block_size = end - start
80f8e3804a 2010-08-20 493: if req_block_size == 0:
80f8e3804a 2010-08-20 494: break
80f8e3804a 2010-08-20 495: buffer = real_file.read(req_block_size)
80f8e3804a 2010-08-20 496: length = len(buffer)
80f8e3804a 2010-08-20 497:
80f8e3804a 2010-08-20 498: def do_HEAD(self):
80f8e3804a 2010-08-20 499: return self.__process()
80f8e3804a 2010-08-20 500: def do_GET(self):
80f8e3804a 2010-08-20 501: return self.__process()
80f8e3804a 2010-08-20 502:
e7b837a681 2010-08-25 503: config.section('general')
e7b837a681 2010-08-25 504: server = http.server.HTTPServer(('127.0.0.1', int(config['port'])), MyRequestHandler)
80f8e3804a 2010-08-20 505: server.serve_forever()
80f8e3804a 2010-08-20 506:
80f8e3804a 2010-08-20 507: else:
80f8e3804a 2010-08-20 508: while True:
80f8e3804a 2010-08-20 509: unchecked_files = set()
80f8e3804a 2010-08-20 510: checked_files = 0
80f8e3804a 2010-08-20 511:
80f8e3804a 2010-08-20 512: # reading log and storing found urls for processing
80f8e3804a 2010-08-20 513: # check file mtime XXX
80f8e3804a 2010-08-20 514: with open(options.log, 'r') as log_file:
80f8e3804a 2010-08-20 515: log_line = re.compile('^[^ ]+ - - \[.*] "(GET|HEAD) (.*?)(\?.*)? HTTP/1.1" (\d+) \d+ "(.*)" "(.*)"$')
80f8e3804a 2010-08-20 516: for line in log_file:
80f8e3804a 2010-08-20 517: this_line = log_line.match(line.strip())
80f8e3804a 2010-08-20 518: if this_line:
80f8e3804a 2010-08-20 519: unchecked_files.add(this_line.group(2))
80f8e3804a 2010-08-20 520:
80f8e3804a 2010-08-20 521: for url in unchecked_files:
80f8e3804a 2010-08-20 522: reload = False
80f8e3804a 2010-08-20 523: recheck = False
80f8e3804a 2010-08-20 524: info = 'Checking file: ' + url
80f8e3804a 2010-08-20 525:
80f8e3804a 2010-08-20 526: # creating empty placeholder in index
80f8e3804a 2010-08-20 527: if not url in index:
80f8e3804a 2010-08-20 528: info += '\nThis one is new.'
80f8e3804a 2010-08-20 529: index[url] = {}
80f8e3804a 2010-08-20 530: reload = True
80f8e3804a 2010-08-20 531:
80f8e3804a 2010-08-20 532: # creating file name from url
80f8e3804a 2010-08-20 533: file_name = options.dir + re.compile('%20').sub(' ', url)
80f8e3804a 2010-08-20 534:
80f8e3804a 2010-08-20 535: # forcibly checking file if no file present
80f8e3804a 2010-08-20 536: if not reload and not os.access(file_name, os.R_OK):
80f8e3804a 2010-08-20 537: info += '\nFile not found or inaccessible.'
80f8e3804a 2010-08-20 538: reload = True
80f8e3804a 2010-08-20 539:
80f8e3804a 2010-08-20 540: # forcibly checking file if file size doesn't match with index data
80f8e3804a 2010-08-20 541: elif not reload and 'Content-Length' in index[url] and os.stat(file_name).st_size != int(index[url]['Content-Length']):
80f8e3804a 2010-08-20 542: info += '\nFile size is ' + os.stat(file_name).st_size + ' and stored file size is ' + index[url]['Content-Length'] + '.'
80f8e3804a 2010-08-20 543: reload = True
80f8e3804a 2010-08-20 544:
80f8e3804a 2010-08-20 545: # forcibly checking file if index hods Pragma header
80f8e3804a 2010-08-20 546: if not reload and 'Pragma' in index[url] and index[url]['Pragma'] == 'no-cache':
80f8e3804a 2010-08-20 547: info +='\nPragma on: recheck imminent.'
80f8e3804a 2010-08-20 548: recheck = True
80f8e3804a 2010-08-20 549:
80f8e3804a 2010-08-20 550: # skipping file processing if there's no need to recheck it and we have checked it at least 4 hours ago
80f8e3804a 2010-08-20 551: if not recheck and not reload and (options.noupdate or ('_time' in index[url] and (datetime.datetime.now() - datetime.timedelta(hours = 4) - index[url]['_time']).days < 0)):
80f8e3804a 2010-08-20 552: if options.verbose:
80f8e3804a 2010-08-20 553: print(info)
80f8e3804a 2010-08-20 554: continue
80f8e3804a 2010-08-20 555: else:
80f8e3804a 2010-08-20 556: print(info)
80f8e3804a 2010-08-20 557:
80f8e3804a 2010-08-20 558: try:
80f8e3804a 2010-08-20 559: with urllib.request.urlopen(options.root + url) as source:
80f8e3804a 2010-08-20 560: new_headers = {}
80f8e3804a 2010-08-20 561: headers = source.info()
80f8e3804a 2010-08-20 562:
80f8e3804a 2010-08-20 563: # stripping unneeded headers (XXX make this inplace?)
80f8e3804a 2010-08-20 564: for header in headers:
80f8e3804a 2010-08-20 565: if header in desc_fields:
80f8e3804a 2010-08-20 566: if header == 'Pragma' and headers[header] != 'no-cache':
80f8e3804a 2010-08-20 567: print('Pragma:', headers[header])
80f8e3804a 2010-08-20 568: new_headers[header] = headers[header]
80f8e3804a 2010-08-20 569: elif not header in ignore_fields:
80f8e3804a 2010-08-20 570: print('Undefined header "', header, '": ', headers[header], sep='')
80f8e3804a 2010-08-20 571:
80f8e3804a 2010-08-20 572: # comparing headers with data found in index
80f8e3804a 2010-08-20 573: # if any header has changed (except Pragma) file is fully downloaded
80f8e3804a 2010-08-20 574: # same if we get more or less headers
80f8e3804a 2010-08-20 575: old_keys = set(index[url].keys())
80f8e3804a 2010-08-20 576: old_keys.discard('_time')
80f8e3804a 2010-08-20 577: old_keys.discard('Pragma')
80f8e3804a 2010-08-20 578: more_keys = set(new_headers.keys()) - old_keys
80f8e3804a 2010-08-20 579: more_keys.discard('Pragma')
80f8e3804a 2010-08-20 580: less_keys = old_keys - set(new_headers.keys())
80f8e3804a 2010-08-20 581: if len(more_keys) > 0:
80f8e3804a 2010-08-20 582: if not len(old_keys) == 0:
80f8e3804a 2010-08-20 583: print('More headers appear:', more_keys)
80f8e3804a 2010-08-20 584: reload = True
80f8e3804a 2010-08-20 585: elif len(less_keys) > 0:
80f8e3804a 2010-08-20 586: print('Less headers appear:', less_keys)
80f8e3804a 2010-08-20 587: else:
80f8e3804a 2010-08-20 588: for key in index[url].keys():
80f8e3804a 2010-08-20 589: if key[0] != '_' and key != 'Pragma' and not index[url][key] == new_headers[key]:
80f8e3804a 2010-08-20 590: print('Header "', key, '" changed from [', index[url][key], '] to [', new_headers[key], ']', sep='')
80f8e3804a 2010-08-20 591: reload = True
80f8e3804a 2010-08-20 592:
80f8e3804a 2010-08-20 593: # downloading file
80f8e3804a 2010-08-20 594: if reload:
80f8e3804a 2010-08-20 595: if 'Content-Length' in headers:
80f8e3804a 2010-08-20 596: print('Downloading', headers['Content-Length'], 'bytes [', end='')
80f8e3804a 2010-08-20 597: else:
80f8e3804a 2010-08-20 598: print('Downloading [', end='')
80f8e3804a 2010-08-20 599: sys.stdout.flush()
80f8e3804a 2010-08-20 600:
80f8e3804a 2010-08-20 601: # file is created at temporary location and moved in place only when download completes
80f8e3804a 2010-08-20 602: temp_file = open(options.dir + os.sep + '.tmp', 'wb')
80f8e3804a 2010-08-20 603: buffer = source.read(block_size)
80f8e3804a 2010-08-20 604: megablocks = 0
80f8e3804a 2010-08-20 605: blocks = 0
80f8e3804a 2010-08-20 606: megs = 0
80f8e3804a 2010-08-20 607: while len(buffer) > 0:
80f8e3804a 2010-08-20 608: temp_file.write(buffer)
80f8e3804a 2010-08-20 609: buffer = source.read(block_size)
80f8e3804a 2010-08-20 610: blocks += 1
80f8e3804a 2010-08-20 611: if blocks > 102400/block_size:
80f8e3804a 2010-08-20 612: megablocks += 1
80f8e3804a 2010-08-20 613: if megablocks > 10:
80f8e3804a 2010-08-20 614: megablocks = megablocks - 10
80f8e3804a 2010-08-20 615: megs += 1
80f8e3804a 2010-08-20 616: print('{}Mb'.format(megs), end='')
80f8e3804a 2010-08-20 617: else:
80f8e3804a 2010-08-20 618: print('.', end='')
80f8e3804a 2010-08-20 619: blocks = blocks - 102400/block_size
80f8e3804a 2010-08-20 620: sys.stdout.flush()
80f8e3804a 2010-08-20 621: temp_file.close()
80f8e3804a 2010-08-20 622: print(']')
80f8e3804a 2010-08-20 623: os.renames(options.dir + os.sep + '.tmp', file_name)
80f8e3804a 2010-08-20 624:
80f8e3804a 2010-08-20 625: checked_files += 1
80f8e3804a 2010-08-20 626:
80f8e3804a 2010-08-20 627: # storing new time mark and storing new headers
80f8e3804a 2010-08-20 628: new_headers['_time'] = datetime.datetime.now()
80f8e3804a 2010-08-20 629: index[url] = new_headers
80f8e3804a 2010-08-20 630: index.sync()
80f8e3804a 2010-08-20 631:
80f8e3804a 2010-08-20 632: except urllib.error.HTTPError as error:
80f8e3804a 2010-08-20 633: # in case of error we don't need to do anything actually,
80f8e3804a 2010-08-20 634: # if file download stalls or fails the file would not be moved to it's location
80f8e3804a 2010-08-20 635: print(error)
80f8e3804a 2010-08-20 636:
80f8e3804a 2010-08-20 637: if options.verbose:
80f8e3804a 2010-08-20 638: print('[', len(unchecked_files), '/', checked_files, ']')
80f8e3804a 2010-08-20 639:
80f8e3804a 2010-08-20 640: # checking if there were any files downloaded, if yes - restarting sequence
80f8e3804a 2010-08-20 641: if checked_files == 0:
80f8e3804a 2010-08-20 642: break