OLD | NEW |
| (Empty) |
1 # coding: utf-8 | |
2 | |
3 # This file is part of the Adblock Plus web scripts, | |
4 # Copyright (C) 2006-2016 Eyeo GmbH | |
5 # | |
6 # Adblock Plus is free software: you can redistribute it and/or modify | |
7 # it under the terms of the GNU General Public License version 3 as | |
8 # published by the Free Software Foundation. | |
9 # | |
10 # Adblock Plus is distributed in the hope that it will be useful, | |
11 # but WITHOUT ANY WARRANTY; without even the implied warranty of | |
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
13 # GNU General Public License for more details. | |
14 # | |
15 # You should have received a copy of the GNU General Public License | |
16 # along with Adblock Plus. If not, see <http://www.gnu.org/licenses/>. | |
17 | |
18 import json | |
19 import os | |
20 import tempfile | |
21 import time | |
22 import traceback | |
23 from datetime import datetime | |
24 from errno import EEXIST | |
25 | |
26 import MySQLdb | |
27 | |
28 from sitescripts.web import url_handler | |
29 from sitescripts.utils import get_config, setupStderr | |
30 from sitescripts.filterhits import db, geometrical_mean | |
31 from sitescripts.filterhits.web import common | |
32 | |
33 | |
34 def log_filterhits(data, basepath, query_string): | |
35 """ | |
36 This logs the provided filterhits data as JSON to a file named after | |
37 the current timestamp in a directory named after the current date. | |
38 """ | |
39 now = time.gmtime() | |
40 | |
41 dir_name = time.strftime("%Y-%m-%d", now) | |
42 path = os.path.join(basepath, dir_name) | |
43 try: | |
44 os.makedirs(path) | |
45 except OSError as e: | |
46 if e.errno != EEXIST: | |
47 raise | |
48 | |
49 with tempfile.NamedTemporaryFile( | |
50 prefix=str(int(time.mktime(now))) + "-", | |
51 suffix=".log", | |
52 dir=path, | |
53 delete=False | |
54 ) as f: | |
55 print >> f, "[%s] %s" % (time.strftime("%d/%b/%Y:%H:%M:%S", now), query_
string) | |
56 json.dump(data, f) | |
57 return f.name | |
58 | |
59 | |
60 @url_handler("/submit") | |
61 def submit(environ, start_response): | |
62 setupStderr(environ["wsgi.errors"]) | |
63 config = get_config() | |
64 | |
65 # Check that this is a POST request | |
66 if environ["REQUEST_METHOD"] != "POST": | |
67 return common.show_error("Unsupported request method", start_response) | |
68 | |
69 # Parse the submitted JSON | |
70 try: | |
71 data = json.loads(environ["wsgi.input"].read(int(environ["CONTENT_LENGTH
"]))) | |
72 except (KeyError, IOError, ValueError): | |
73 return common.show_error("Error while parsing JSON data.", start_respons
e) | |
74 | |
75 # Make sure the submitted data was contained within an object at least | |
76 if not isinstance(data, dict): | |
77 return common.show_error("Error, data must be contained within an object
.", start_response) | |
78 | |
79 # Log the data to a file | |
80 log_dir = config.get("filterhitstats", "log_dir") | |
81 try: | |
82 log_file = log_filterhits(data, log_dir, environ.get("QUERY_STRING", "")
) | |
83 except (OSError, IOError): | |
84 traceback.print_exc() | |
85 return common.show_error("Failed to write data to log file!", start_resp
onse, | |
86 "500 Logging error") | |
87 | |
88 # Update the geometrical_mean aggregations in the database | |
89 interval = config.get("filterhitstats", "interval") | |
90 try: | |
91 db_connection = db.connect() | |
92 try: | |
93 db.write(db_connection, geometrical_mean.update(interval, data)) | |
94 finally: | |
95 db_connection.close() | |
96 except: | |
97 # Updating the aggregations in the database failed for whatever reason, | |
98 # log the details but continue to return 204 to the client to avoid the | |
99 # re-transmission of data. | |
100 processing_error_log = os.path.join(config.get("filterhitstats", "log_di
r"), | |
101 "processing-errors.log") | |
102 with open(processing_error_log, "a+") as f: | |
103 message = "Problem processing data file %s:\n%s" % ( | |
104 log_file, traceback.format_exc() | |
105 ) | |
106 print >> f, "[%s] %s" % (datetime.now().strftime("%d/%b/%Y:%H:%M:%S
%z"), message) | |
107 | |
108 # Send back a 204 No Content | |
109 start_response("204 No Content", []) | |
110 return [] | |
OLD | NEW |