Ticket #637: signals_in_engine.patch

_cpconfig.py

@@ -106 +106 @@
-        'request.show_tracebacks': False,
-        'log.screen': False,
-        },
+    "embedded": {
+        # For use with CherryPy embedded in another deployment stack.
+        'engine.autoreload_on': False,
+        'checker.on': False,
+        'tools.log_headers.on': False,
+        'request.show_tracebacks': False,
+        'log.screen': False,
+        'engine.SIGHUP': None,
+        'engine.SIGTERM': None,
+        },
-    "test_suite": {
-        'engine.autoreload_on': False,
-        'checker.on': False,

_cpengine.py

@@ -16 +16 @@
-STARTING = None
-STARTED = 1
-
-try:
-    if hasattr(signal, "SIGHUP"):
-        def SIGHUP(signum=None, frame=None):
-            cherrypy.engine.reexec()
-        signal.signal(signal.SIGHUP, SIGHUP)
-
-    if hasattr(signal, "SIGTERM"):
-        def SIGTERM(signum=None, frame=None):
-            cherrypy.server.stop()
-            cherrypy.engine.stop()
-        signal.signal(signal.SIGTERM, SIGTERM)
-except ValueError, _signal_exc:
-    if _signal_exc.args[0] != "signal only works in main thread":
-        raise
-
-
-class PerpetualTimer(threading._Timer):
-    
-    def run(self):
@@ -43 +28 @@
-
-
-class Engine(object):
-
+
-    
+    Servers and gateways should not instantiate Request objects directly.
+    Instead, they should ask an Engine object for a request via the
+    Engine.request method.
+    
+    Blocking is completely optional! The Engine's blocking, signal and
+    interrupt handling, privilege dropping, and autoreload features are
+    not a good idea when driving CherryPy applications from another
+    deployment tool (but an Engine is a great deployment tool itself).
+    By calling start(blocking=False), you avoid blocking and interrupt-
+    handling issues. By setting Engine.SIGHUP and Engine.SIGTERM to None,
+    you can completely disable the signal handling (and therefore disable
+    autoreloads triggered by SIGHUP). Set Engine.autoreload_on to False
+    to disable autoreload entirely.
+    """
+    
-    # Configurable attributes
-    request_class = _cprequest.Request
-    response_class = _cprequest.Response
@@ -81 +81 @@
-        
-        self.state = STARTED
-        
+        self._set_signals()
+        
-        freq = self.deadlock_poll_freq
-        if freq > 0:
-            self.monitor_thread = PerpetualTimer(freq, self.monitor)
@@ -263 +265 @@
-        
-        self.start()
-    
+    
+    #                           Signal handling                           #
+    
+    SIGHUP = None
+    SIGTERM = None
+    
+    if hasattr(signal, "SIGHUP"):
+        def SIGHUP(self, signum=None, frame=None):
+            self.reexec()
+    
+    if hasattr(signal, "SIGTERM"):
+        def SIGTERM(signum=None, frame=None):
+            cherrypy.server.stop()
+            self.stop()
+    
+    def _set_signals(self):
+        if self.SIGHUP:
+            signal.signal(signal.SIGHUP, self.SIGHUP)
+        if self.SIGTERM:
+            signal.signal(signal.SIGTERM, self.SIGTERM)
+    
+    
+    #                           Drop privileges                           #
+    
-    # Special thanks to Gavin Baker: http://antonym.org/node/100.
-    try:
-        import pwd, grp