Simplified SpiderMiddlewareManager by making it inherit from MiddlewareManager

This commit is contained in:
Pablo Hoffman 2010-09-06 10:40:33 -03:00
parent cc72f03e10
commit 5f58af2005
2 changed files with 20 additions and 45 deletions

View File

@ -61,7 +61,7 @@ class Scraper(object):
def __init__(self, engine, settings):
self.sites = {}
self.spidermw = SpiderMiddlewareManager()
self.spidermw = SpiderMiddlewareManager.from_settings(settings)
itemproc_cls = load_object(settings['ITEM_PROCESSOR'])
self.itemproc = itemproc_cls.from_settings(settings)
self.concurrent_items = settings.getint('CONCURRENT_ITEMS')

View File

@ -1,65 +1,40 @@
"""
This module implements the Spider Middleware manager. For more information see
the Spider Middleware doc in:
docs/topics/spider-middleware.rst
Spider Middleware manager
See documentation in docs/topics/spider-middleware.rst
"""
from scrapy import log
from twisted.python.failure import Failure
from scrapy.exceptions import NotConfigured
from scrapy.utils.misc import load_object
from scrapy.utils.conf import build_component_list
from scrapy.middleware import MiddlewareManager
from scrapy.utils.defer import mustbe_deferred
from scrapy.conf import settings
from scrapy.utils.conf import build_component_list
def _isiterable(possible_iterator):
return hasattr(possible_iterator, '__iter__')
class SpiderMiddlewareManager(object):
def __init__(self):
self.loaded = False
self.enabled = {}
self.disabled = {}
self.spider_middleware = []
self.result_middleware = []
self.exception_middleware = []
self.load()
class SpiderMiddlewareManager(MiddlewareManager):
component_name = 'spider middleware'
@classmethod
def _get_mwlist_from_settings(cls, settings):
return build_component_list(settings['SPIDER_MIDDLEWARES_BASE'], \
settings['SPIDER_MIDDLEWARES'])
def _add_middleware(self, mw):
super(SpiderMiddlewareManager, self)._add_middleware(mw)
if hasattr(mw, 'process_spider_input'):
self.spider_middleware.append(mw.process_spider_input)
self.methods['process_spider_input'].append(mw.process_spider_input)
if hasattr(mw, 'process_spider_output'):
self.result_middleware.insert(0, mw.process_spider_output)
self.methods['process_spider_output'].append(mw.process_spider_output)
if hasattr(mw, 'process_spider_exception'):
self.exception_middleware.insert(0, mw.process_spider_exception)
def load(self):
"""Load middleware defined in settings module"""
mwlist = build_component_list(settings['SPIDER_MIDDLEWARES_BASE'], \
settings['SPIDER_MIDDLEWARES'])
self.enabled.clear()
self.disabled.clear()
for mwpath in mwlist:
try:
cls = load_object(mwpath)
mw = cls()
self.enabled[cls.__name__] = mw
self._add_middleware(mw)
except NotConfigured, e:
self.disabled[cls.__name__] = mwpath
if e.args:
log.msg(e)
log.msg("Enabled spider middlewares: %s" % ", ".join(self.enabled.keys()), \
level=log.DEBUG)
self.loaded = True
self.methods['process_spider_exception'].append(mw.process_spider_exception)
def scrape_response(self, scrape_func, response, request, spider):
fname = lambda f:'%s.%s' % (f.im_self.__class__.__name__, f.im_func.__name__)
def process_spider_input(response):
for method in self.spider_middleware:
for method in self.methods['process_spider_input']:
try:
result = method(response=response, spider=spider)
assert result is None, \
@ -72,7 +47,7 @@ class SpiderMiddlewareManager(object):
def process_spider_exception(_failure):
exception = _failure.value
for method in self.exception_middleware:
for method in self.methods['process_spider_exception']:
result = method(response=response, exception=exception, spider=spider)
assert result is None or _isiterable(result), \
'Middleware %s must returns None, or an iterable object, got %s ' % \
@ -82,7 +57,7 @@ class SpiderMiddlewareManager(object):
return _failure
def process_spider_output(result):
for method in self.result_middleware:
for method in self.methods['process_spider_output']:
result = method(response=response, result=result, spider=spider)
assert _isiterable(result), \
'Middleware %s must returns an iterable object, got %s ' % \