Don't crash on a string decoding exception

atwardowski · atwardowski · commit d3ddb49d9eaf · 2013-10-16T13:50:01.000-04:00
diff --git a/README.rst b/README.rst
@@ -466,6 +466,8 @@ When using ``copytruncate`` style log rotation, two race conditions can occur:
    watch which may be truncated are generally going to be large enough
    and slow-filling enough that this won't crop up in the wild.
 
+----
+
 When you get an error similar to ``ImportError: No module named
 _sqlite3`` your python seems to miss the sqlite3-module. This can be the
 case on FreeBSD and probably other systems. If so, use the local package
@@ -474,6 +476,13 @@ manager or port system to build that module. On FreeBSD::
     cd /usr/ports/databases/py-sqlite3
     sudo make install clean
 
+----
+
+Binary data in your logs will be converted to escape sequences or ?'s depending on the encoding settings to prevent decoding exceptions from crashing beaver.
+
+   | http://docs.python.org/2/library/codecs.html#codecs.replace_errors
+   | malformed data is replaced with a suitable replacement character such as '?' in bytestrings and '\ufffd' in Unicode  strings.
+
 Credits
 =======
 
diff --git a/beaver/worker/tail.py b/beaver/worker/tail.py
@@ -82,11 +82,11 @@ def open(self, encoding=None):
                 _file = gzip.open(self._filename, 'rb')
             else:
                 if encoding:
-                    _file = io.open(self._filename, 'r', encoding=encoding)
+                    _file = io.open(self._filename, 'r', encoding=encoding, errors='replace')
                 elif self._encoding:
-                    _file = io.open(self._filename, 'r', encoding=self._encoding)
+                    _file = io.open(self._filename, 'r', encoding=self._encoding, errors='replace')
                 else:
-                    _file = io.open(self._filename, 'r')
+                    _file = io.open(self._filename, 'r', errors='replace')
         except IOError, e:
             self._log_warning(str(e))
             _file = None
diff --git a/beaver/worker/worker.py b/beaver/worker/worker.py
@@ -582,11 +582,11 @@ def open(self, filename, encoding=None):
             else:
                 file_encoding = self._beaver_config.get_field('encoding', filename)
                 if encoding:
-                    _file = io.open(filename, "r", encoding=encoding)
+                    _file = io.open(filename, "r", encoding=encoding, errors='replace')
                 elif file_encoding:
-                    _file = io.open(filename, "r", encoding=file_encoding)
+                    _file = io.open(filename, "r", encoding=file_encoding, errors='replace')
                 else:
-                    _file = io.open(filename, "r")
+                    _file = io.open(filename, "r", errors='replace')
         except IOError, e:
             self._logger.warning(str(e))
             _file = None
