[tor-commits] [ooni-probe/master] Escape strings before writing them to the log.

[art at torproject.org]

commit 68cac6eabf474dd32f317e1671f3e9dc2b5fd697
Author: Arturo Filastò <art at fuffa.org>
Date:   Wed Apr 23 16:25:49 2014 +0200

    Escape strings before writing them to the log.
    
    This fixes: https://github.com/TheTorProject/ooni-probe/issues/302 for ooni-probe
---
 ooni/tests/test_utils.py |   12 ++++++++++-
 ooni/utils/log.py        |   52 +++++++++++++++++++++++++++++++++++++---------
 2 files changed, 53 insertions(+), 11 deletions(-)

diff --git a/ooni/tests/test_utils.py b/ooni/tests/test_utils.py
index 10aa0cc..e21ab89 100644
--- a/ooni/tests/test_utils.py
+++ b/ooni/tests/test_utils.py
@@ -1,7 +1,9 @@
 import os
 from twisted.trial import unittest
 
-from ooni.utils import pushFilenameStack
+from ooni.utils import pushFilenameStack, log
+
+
 class TestUtils(unittest.TestCase):
     def test_pushFilenameStack(self):
         basefilename = os.path.join(os.getcwd(), 'dummyfile')
@@ -20,3 +22,11 @@ class TestUtils(unittest.TestCase):
             self.assertEqual(str(i-1), str(c))
             f.close()
 
+    def test_log_encode(self):
+        logmsgs = (
+            (r"spam\x07\x08", "spam\a\b"),
+            (r"spam\x07\x08", u"spam\a\b"),
+            (r"ham\u237e", u"ham"+u"\u237e")
+        )
+        for encoded_logmsg, logmsg in logmsgs:
+            self.assertEqual(log.log_encode(logmsg), encoded_logmsg)
diff --git a/ooni/utils/log.py b/ooni/utils/log.py
index 7fafcfb..beba78f 100644
--- a/ooni/utils/log.py
+++ b/ooni/utils/log.py
@@ -1,5 +1,6 @@
-import sys
 import os
+import sys
+import codecs
 import logging
 import traceback
 
@@ -10,10 +11,33 @@ from twisted.python.logfile import DailyLogFile
 
 from ooni import otime
 
-## Get rid of the annoying "No route found for
-## IPv6 destination warnings":
+# Get rid of the annoying "No route found for
+# IPv6 destination warnings":
 logging.getLogger("scapy.runtime").setLevel(logging.ERROR)
 
+
+def log_encode(logmsg):
+    """
+    I encode logmsg (a str or unicode) as printable ASCII. Each case
+    gets a distinct prefix, so that people differentiate a unicode
+    from a utf-8-encoded-byte-string or binary gunk that would
+    otherwise result in the same final output.
+    """
+    if isinstance(logmsg, unicode):
+        return codecs.encode(logmsg, 'unicode_escape')
+    elif isinstance(logmsg, str):
+        try:
+            unicodelogmsg = logmsg.decode('utf-8')
+        except UnicodeDecodeError:
+            return codecs.encode(logmsg, 'string_escape')
+        else:
+            return codecs.encode(unicodelogmsg, 'unicode_escape')
+    else:
+        raise Exception("I accept only a unicode object or a string, "
+                        "not a %s object like %r" % (type(logmsg),
+                                                     repr(logmsg)))
+
+
 class LogWithNoPrefix(txlog.FileLogObserver):
     def emit(self, eventDict):
         text = txlog.textFromEventDict(eventDict)
@@ -23,6 +47,7 @@ class LogWithNoPrefix(txlog.FileLogObserver):
         util.untilConcludes(self.write, "%s\n" % text)
         util.untilConcludes(self.flush)  # Hoorj!
 
+
 class OONILogger(object):
     def start(self, logfile=None, application_name="ooniprobe"):
         from ooni.settings import config
@@ -30,16 +55,17 @@ class OONILogger(object):
         daily_logfile = None
 
         if not logfile:
-            logfile = config.basic.logfile
+            logfile = os.path.expanduser(config.basic.logfile)
 
         log_folder = os.path.dirname(logfile)
         log_filename = os.path.basename(logfile)
 
         daily_logfile = DailyLogFile(log_filename, log_folder)
 
-        txlog.msg("Starting %s on %s (%s UTC)" %  (application_name, otime.prettyDateNow(),
-                                                     otime.utcPrettyDateNow()))
- 
+        txlog.msg("Starting %s on %s (%s UTC)" % (application_name,
+                                                  otime.prettyDateNow(),
+                                                  otime.utcPrettyDateNow()))
+
         self.fileObserver = txlog.FileLogObserver(daily_logfile)
         self.stdoutObserver = LogWithNoPrefix(sys.stdout)
 
@@ -52,28 +78,34 @@ class OONILogger(object):
 
 oonilogger = OONILogger()
 
+
 def start(logfile=None, application_name="ooniprobe"):
     oonilogger.start(logfile, application_name)
 
+
 def stop():
     oonilogger.stop()
 
+
 def msg(msg, *arg, **kw):
     from ooni.settings import config
     if config.logging:
-        print "%s" % msg
+        print "%s" % log_encode(msg)
+
 
 def debug(msg, *arg, **kw):
     from ooni.settings import config
     if config.advanced.debug and config.logging:
-        print "[D] %s" % msg
+        print "[D] %s" % log_encode(msg)
+
 
 def err(msg, *arg, **kw):
     from ooni.settings import config
     if config.logging:
         if isinstance(msg, Exception):
             msg = "%s: %s" % (msg.__class__.__name__, msg)
-        print "[!] %s" % msg
+        print "[!] %s" % log_encode(msg)
+
 
 def exception(error):
     """