Changeset Navigation

Changeset 1319

Timestamp:

09/02/06 13:39:10

Author:

fumanchu

Message:

Fix for #535 (Apache re-processes Range headers). _cpmodpy now uses a new "ignore_headers(['Range'])" Tool by default.

Files:

r1275	r1319
19	19	func()
20	20
21		cherrypy.config.update({'log.screen': False})
	21	cherrypy.config.update({'log.screen': False,
	22	"tools.ignore_headers.on": True,
	23	"tools.ignore_headers.headers": ['Range'],
	24	})
22	25
23	26	if cherrypy.engine.state == cherrypy._cpengine.STOPPED:

r1311	r1319
313	313	default_toolbox.tidy = Tool('before_finalize', tidy.tidy)
314	314	default_toolbox.nsgmls = Tool('before_finalize', tidy.nsgmls)
	315	default_toolbox.ignore_headers = Tool('before_request_body', cptools.ignore_headers)
315	316
316	317

r1303	r1319
121	121	xff = xff.split(',')[-1].strip()
122	122	request.remote.ip = xff
	123
	124
	125	def ignore_headers(headers=('Range',)):
	126	"""Delete request headers whose field names are included in 'headers'.
	127
	128	This is a useful tool for working behind certain HTTP servers;
	129	for example, Apache duplicates the work that CP does for 'Range'
	130	headers, and will doubly-truncate the response.
	131	"""
	132	request = cherrypy.request
	133	for name in headers:
	134	if name in request.headers:
	135	del request.headers[name]
123	136
124	137

r1317	r1319
111	111	import cherrypy
112	112	cherrypy.config.update({
113		"log.error.file": os.path.join(curdir, "test.log"),
	113	"log.error_file": os.path.join(curdir, "test.log"),
114	114	"environment": "production",
115	115	})