Lines of
samesite.py
from check-in 38b25713eb
that are changed by the sequence of edits moving toward
check-in 827033dd7e:
1: #!/usr/bin/env python3.1
2:
3: import datetime, http.cookiejar, optparse, os, sys, shelve, re, urllib.request
4:
5: parser = optparse.OptionParser()
6: parser.add_option('-v', '--verbose', action = 'store_true', dest = 'verbose', help = 'turns on verbose status notifications', metavar = 'bool', default = False)
7: parser.add_option('-d', '--dir', action = 'store', dest = 'dir', help = 'specify directory where the files should be stored', metavar = 'string', default = None)
8: parser.add_option('-r', '--root', action = 'store', dest = 'root', help = 'specify a site from which data should be mirrored', metavar = 'string', default = None)
9: parser.add_option('-l', '--log', action = 'store', dest = 'log', help = 'specify a log file to process', metavar = 'string', default = None)
10: parser.add_option('-e', '--skip-etag', action = 'store_true', dest = 'noetag', help = 'do not process etags', metavar = 'bool', default = False)
11: (options, args) = parser.parse_args()
12:
13: assert options.dir, 'Directory not specified'
14: assert options.root, 'Server not specified'
15: assert options.log, 'Log file not specified'
16: assert os.access(options.log, os.R_OK), 'Log file unreadable'
17:
18: # this is file index - everything is stored in this file
19: index = shelve.open(options.dir + '/.index')
20: desc_fields = ('Content-Length', 'Pragma', 'Last-Modified')
21: ignore_fields = ('Accept-Ranges', 'Age', 'Cache-Control', 'Connection', 'Content-Type', 'Date', 'Expires', 'Server', 'Via', 'X-Cache', 'X-Cache-Lookup', 'X-Powered-By')
22:
23: if not options.noetag:
24: desc_fields += 'ETag',
25: else:
26: ignore_fields += 'ETag',
27:
28: block_size = 32768
29:
30: while True:
31: unchecked_files = set()
32: checked_files = 0
33:
34: # reading log and storing found urls for processing
35: # check file mtime XXX
36: with open(options.log, 'r') as log_file:
37: log_line = re.compile('^[^ ]+ - - \[.*] "(GET|HEAD) (.*?)(\?.*)? HTTP/1.1" (\d+) \d+ "(.*)" "(.*)"$')
38: for line in log_file:
39: this_line = log_line.match(line.strip())
40: if this_line:
41: unchecked_files.add(this_line.group(2))
42:
43: for url in unchecked_files:
44: reload = False
45: recheck = False
38b25713eb 2010-07-26 46: print('Checking file:', url)
47:
48: # creating empty placeholder in index
49: if not url in index:
38b25713eb 2010-07-26 50: print('This one is new.')
51: index[url] = {}
52: reload = True
53:
54: # creating file name from url
55: file_name = options.dir + re.compile('%20').sub(' ', url)
56:
57: # forcibly checking file if no file present
58: if not reload and not os.access(file_name, os.R_OK):
38b25713eb 2010-07-26 59: print('File not found or inaccessible.')
60: reload = True
61:
62: # forcibly checking file if file size doesn't match with index data
63: elif not reload and 'Content-Length' in index[url] and os.stat(file_name).st_size != int(index[url]['Content-Length']):
38b25713eb 2010-07-26 64: print('File size is ', os.stat(file_name).st_size, ' and stored file size is ', index[url]['Content-Length'], '.', sep='')
65: reload = True
66:
67: # forcibly checking file if index hods Pragma header
68: if not reload and 'Pragma' in index[url] and index[url]['Pragma'] == 'no-cache':
38b25713eb 2010-07-26 69: print('Pragma on: recheck imminent.')
70: recheck = True
71:
72: # skipping file processing if there's no need to recheck it and we have checked it at least 4 hours ago
73: if not recheck and not reload and '__time__' in index[url] and (datetime.datetime.now() - datetime.timedelta(hours = 4) - index[url]['__time__']).days < 0:
74: continue
75:
76: try:
77: with urllib.request.urlopen(options.root + url) as source:
78: new_headers = {}
79: headers = source.info()
80:
81: # stripping unneeded headers (XXX make this inplace?)
82: for header in headers:
83: if header in desc_fields:
84: if header == 'Pragma' and headers[header] != 'no-cache':
85: print('Pragma:', headers[header])
86: new_headers[header] = headers[header]
87: elif not header in ignore_fields:
88: print('Undefined header "', header, '": ', headers[header], sep='')
89:
90: # comparing headers with data found in index
91: # if any header has changed (except Pragma) file is fully downloaded
92: # same if we get more or less headers
93: old_keys = set(index[url].keys())
94: old_keys.discard('__time__')
95: old_keys.discard('Pragma')
96: more_keys = set(new_headers.keys()) - old_keys
97: more_keys.discard('Pragma')
98: less_keys = old_keys - set(new_headers.keys())
99: if len(more_keys) > 0:
100: if not len(old_keys) == 0:
101: print('More headers appear:', more_keys)
102: reload = True
103: elif len(less_keys) > 0:
104: print('Less headers appear:', less_keys)
105: else:
106: for key in index[url].keys():
107: if key not in ('__time__', 'Pragma') and not index[url][key] == new_headers[key]:
108: print('Header "', key, '" changed from [', index[url][key], '] to [', new_headers[key], ']', sep='')
109: reload = True
110:
111: # downloading file
112: if reload:
113: if 'Content-Length' in headers:
114: print('Downloading', headers['Content-Length'], 'bytes [', end='')
115: else:
116: print('Downloading [', end='')
117: sys.stdout.flush()
118:
119: # file is created at temporary location and moved in place only when download completes
120: temp_file = open(options.dir + '/.tmp', 'wb')
121: buffer = source.read(block_size)
122: blocks = 0
123: megs = 0
124: while len(buffer) > 0:
125: temp_file.write(buffer)
126: print('.', end='')
127: sys.stdout.flush()
128: buffer = source.read(block_size)
129: blocks += 1
130: if blocks > 1024*1024/block_size:
131: blocks = blocks - 1024*1024/block_size
132: megs += 1
133: print('{}Mb'.format(megs), end='')
134: temp_file.close()
135: print(']')
136: os.renames(options.dir + '/.tmp', file_name)
137:
138: checked_files += 1
139:
140: # storing new time mark and storing new headers
141: new_headers['__time__'] = datetime.datetime.now()
142: index[url] = new_headers
143: index.sync()
144:
145: except urllib.error.HTTPError as error:
146: # in case of error we don't need to do anything actually,
147: # if file download stalls or fails the file would not be moved to it's location
148: print(error)
149:
38b25713eb 2010-07-26 150: print('[', len(unchecked_files), '/', checked_files, ']')
151:
152: # checking if there were any files downloaded, if yes - restarting sequence
153: if checked_files == 0:
154: break