def __init__(self, brand_id, extra_cond=None, max_images=15):
print str.format(
'Publishing (brand_id={0}, max_images={1}, extra_cond="{2}")...',
brand_id, max_images, extra_cond)
# 某一单品最大发布的图片数量
self.max_images = max_images
self.brand_id = brand_id
if not extra_cond:
extra_cond = ['1']
elif not iterable(extra_cond):
extra_cond = [extra_cond]
self.extra_cond = extra_cond
self.tot = 0
self.progress = 0
# 国家的展示顺序
self.region_order = {
k: info.region_info()[k]['weight']
for k in info.region_info()
}
self.products_tbl = 'products'
self.prod_mt_tbl = 'products_mfashion_tags'
self.mt_tbl = 'mfashion_tags'
self.prod_ot_tbl = 'products_original_tags'
self.ot_tbl = 'original_tags'
self.price_hist = 'products_price_history'
def __init__(self, name, region, *a, **kw):
self.name = str.format('{0}-{1}', name, '-'.join(region) if region else 'all')
super(MFashionSpider, self).__init__(*a, **kw)
if not region:
self.region_list = self.get_supported_regions()
else:
self.region_list = list((set(region) if iterable(region) else
{region}).intersection(set(self.get_supported_regions())))
def start_requests(self):
for region in self.region_list:
metadata = {'region': region, 'brand_id': getattr(self, 'spider_data')['brand_id'],
'tags_mapping': {}, 'category': []}
tmp = getattr(self, 'spider_data')['home_urls'][region]
start_urls = tmp if iterable(tmp) else [tmp]
for url in start_urls:
m = copy.deepcopy(metadata)
yield Request(url=url, meta={'userdata': m}, callback=self.parse, errback=self.onerr)
def query_match(self,
selects,
table,
matches=None,
extra=None,
tail_str=None,
use_result=False,
distinct=False):
"""
查询：相当于SELECT ... FROM ... WHERE col=val
:param selects: 需要select的字段
:param table: 查询的表名称
:param matches: dict类型，查询条件
:param extra: 其它的查询条件
:param tail_str: 添加在查询语句末尾的字符串
:param use_result:
:return:
"""
if not extra:
extra = ['1']
elif not iterable(extra):
extra = [extra]
if not iterable(selects):
selects = [selects]
def func(arg):
k, v = arg
return unicode.format(u'{0}="{1}"', k,
self.sql_escape(v)) if v else unicode.format(
u'{0} IS NULL', k)
match_str = ' AND '.join(map(func,
matches.items())) if matches else '1'
extra_cond = ' AND '.join(extra)
statement = unicode.format(
u'SELECT {5} {0} FROM {1} WHERE {2} AND {3} {4}',
', '.join(selects), table, match_str, extra_cond,
tail_str if tail_str else '', 'DISTINCT' if distinct else '')
self.db.query(statement.encode('utf-8'))
return self.db.use_result() if use_result else self.db.store_result()
def __init__(self,
src_spec=getattr(glob, 'DATABASE')['DB_SPEC'],
dst_spec=getattr(glob, 'DATABASE')['DB_SPEC'],
cond=None):
self.progress = 0
self.tot = 1
if cond:
if iterable(cond):
self.cond = cond
else:
self.cond = [cond]
else:
self.cond = ['1']
self.src_spec = src_spec
self.dst_spec = dst_spec
def __init__(self, region):
if iterable(region):
HogoBossSpider.spider_data['home_urls'] = {
reg: str.format('http://store-{0}.hugoboss.com', reg)
if reg != 'cn' else 'http://store.hugoboss.cn'
for reg in region
}
else:
k = region
HogoBossSpider.spider_data['home_urls'] = {
k: str.format('http://store-{0}.hugoboss.com', k)
if k != 'cn' else 'http://store.hugoboss.cn'
}
super(HogoBossSpider, self).__init__('Hugo Boss', region)
def start_requests(self):
for region in self.region_list:
metadata = {
'region': region,
'brand_id': self.spider_data['brand_id'],
'tags_mapping': {}
}
tmp = self.spider_data['home_urls']['common']
cookie = {
'DKI_FiftyOneInternationalCookie':
str.format('{0}-{1}', region.upper(),
self.spider_data['curreny'][region])
}
start_urls = tmp if iterable(tmp) else [tmp]
for url in start_urls:
m = copy.deepcopy(metadata)
yield Request(url=url,
meta={'userdata': m},
callback=self.parse,
errback=self.onerr,
cookies=cookie,
dont_filter=True)
def run(self):
last_update = self.last_update
extra_cond = self.extra_cond
if not extra_cond:
extra_cond = []
elif not iterable(extra_cond):
extra_cond = [extra_cond]
if last_update:
extra_cond.append(
unicode.format(u'update_time > "{0}"', last_update))
extra_cond.append('mapping_list IS NOT NULL')
# MFashion标签的缓存记录
cached_mfashion = {}
# 标签更新原理：original_tags存放原始标签。根据update_time字段可以得到最近更新过的标签。由于标签系统具备一定传染性，所以
# 该标签对应brand/region下的所有标签都必须重做
rs = self.db.query_match(['brand_id', 'region'],
self.original_tags_tbl, {},
extra=extra_cond,
distinct=True)
# 需要处理的标签
tag_dict = {}
for i in xrange(rs.num_rows()):
brand_id, region = rs.fetch_row()[0]
for val in self.db.query_match(
['idmappings', 'mapping_list'],
self.original_tags_tbl, {
'brand_id': brand_id,
'region': region
},
extra='mapping_list IS NOT NULL').fetch_row(maxrows=0):
tag_dict[val[0]] = json.loads(val[1].replace("'", '"'))
# 删除旧单品/标签关系
self.db.execute(
str.format(
'DELETE FROM p2 USING {0} AS p1, {1} AS p2 WHERE p1.idproducts=p2.idproducts '
'AND p1.brand_id={2} AND region="{3}"', self.products,
self.prod_mt_tbl, brand_id, region))
self.tot = len(tag_dict)
self.progress = 0
for tid, rule in tag_dict.items():
self.progress += 1
self.db.start_transaction()
try:
# 所有相关的单品
pid_list = [
int(val[0]) for val in self.db.query_match(
['idproducts'], self.prod_tag_tbl, {
'id_original_tags': tid
}).fetch_row(maxrows=0)
]
# 添加MFashion标签
for tag in rule:
if tag not in cached_mfashion:
self.db.insert({'tag': tag},
self.mfashion_tags_tbl,
ignore=True)
tid = int(
self.db.query_match(['idmfashion_tags'],
self.mfashion_tags_tbl, {
'tag': tag
}).fetch_row()[0][0])
cached_mfashion[tag] = tid
self.db.insert([{
'idproducts': pid,
'id_mfashion_tags': cached_mfashion[tag]
} for pid in pid_list],
self.prod_mt_tbl,
ignore=True)
self.db.commit()
except ValueError:
self.db.rollback()
except:
self.db.rollback()
raise
def set_up_spider(spider_class, data, spider_type='default'):
"""
设置爬虫对象
@param spider_type: 爬虫类型，是update还是monitor，还是普通爬虫
@param spider_class:
@param data: 爬虫的配置参数
@return:
"""
crawler = Crawler(Settings())
crawler.settings.values['BOT_NAME'] = 'mstore_bot'
if spider_type == 'update':
crawler.settings.values['ITEM_PIPELINES'] = {'scrapper.pipelines.UpdatePipeline': 800}
brand_list = [int(tmp) for tmp in (data['brand'] if 'brand' in data else [])]
if 'region' in data:
region_list = data['region']
elif 'r' in data:
region_list = data['r']
else:
region_list = None
spider = spider_class(brand_list, region_list, getattr(glob, 'DATABASE')['DB_SPEC'])
welcome_msg = str.format('Updating started, processing the following brands: {0}',
', '.join(str(tmp) for tmp in brand_list))
# TODO update类型的spider可以指定多个品牌，导致user agent很不好处理。这是一种不太好的架构，以后考虑改成：一个UpdateSpider对应一个特定的品牌
major_brand = brand_list[0]
elif spider_type == 'monitor':
crawler.settings.values['ITEM_PIPELINES'] = {'scrapper.pipelines.MonitorPipeline': 800}
brand = int(data['brand'][0])
region = data['region'][0]
idmonitor = int(data['idmonitor'][0])
parameter = {'brand_id': brand, 'region': region}
spider = spider_class(idmonitor, parameter, getattr(glob, 'DATABASE')['DB_SPEC'])
welcome_msg = str.format('STARTING MONITORING, idmonitory={0}, brand={1}, region={2}', idmonitor, brand,
region)
major_brand = brand
else:
crawler.settings.values['ITEM_PIPELINES'] = {'scrapper.pipelines.ProductImagePipeline': 800,
'scrapper.pipelines.ProductPipeline': 300} \
if getattr(glob, 'DATABASE')['WRITE_DATABASE'] else {}
if 'job' in data:
job_path = get_job_path(spider_class.spider_data['brand_id']) + '-1'
if 'rst-job' in data:
shutil.rmtree(job_path, ignore_errors=True)
crawler.settings.values['JOBDIR'] = job_path
# Telnet支持
# crawler.settings.values['TELNETCONSOLE_HOST'] = '127.0.0.1'
# if 'telnet' in data and data['telnet']:
# start_port = int(data['telnet'][0])
# else:
# start_port = spider_class.spider_data['brand_id']
# crawler.settings.values['TELNETCONSOLE_PORT'] = [start_port, start_port + 8]
# 图像数据存储
crawler.settings.values['IMAGES_STORE'] = get_images_store(spider_class.spider_data['brand_id'])
# crawler.settings.values['IMAGES_THUMBS'] = {'small': (480, 480), 'medium': (1200, 1200)}
crawler.settings.values['IMAGES_MIN_HEIGHT'] = 128
crawler.settings.values['IMAGES_MIN_WIDTH'] = 128
# 获取爬虫区域
region_list = data['r']
if not region_list:
region_list = spider_class.get_supported_regions()
elif not iterable(region_list):
region_list = [region_list]
region_list = filter(lambda val: info.region_info()[val]['status'], region_list)
if 'exclude-region' in data:
for r in data['exclude-region']:
if r in region_list:
region_list.pop(region_list.index(r))
spider = spider_class(region_list)
welcome_msg = str.format('Spider started, processing the following regions: {0}', ', '.join(region_list))
major_brand = spider_class.spider_data['brand_id']
crawler.settings.values['AUTOTHROTTLE_ENABLED'] = False
# 设置spider的user agent
# 优先级如下：
# 1. 命令行中--user-agent参数指定
# 2. 配置文件指定，参见global_settings['USER_AGENT']项目
# 3. spider_class中，spider_data['USER_AGENT']指定
# 4. 默认为chrome
# TODO 以上第2项暂未实现
ua_map = {
'chrome': 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/30.0.1599.69 Safari/537.36',
'iphone': 'Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_3_2 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8H7 Safari/6533.18.5',
'ipad': 'Mozilla/5.0 (iPad; U; CPU OS 3_2 like Mac OS X; en-us) AppleWebKit/531.21.10 (KHTML, like Gecko) Version/4.0.4 Mobile/7B334b Safari/531.21.10'}
if 'user-agent' in data:
ua = data['user-agent'][0]
else:
spider_spec = info.spider_info()[major_brand]['spider_class']
if 'user_agent' in spider_spec.spider_data:
ua = spider_spec.spider_data['user_agent']
else:
ua = 'chrome'
crawler.settings.values['USER_AGENT'] = ua_map[ua.lower()] if ua.lower() in ua_map else ua
# 设置spider的proxy信息
crawler.settings.values['DOWNLOADER_MIDDLEWARES'] = {
'scrapy.contrib.downloadermiddleware.httpproxy.HttpProxyMiddleware': 1}
if 'proxy' in data:
try:
crawler.settings.values['PROXY_ENABLED'] = True
except AttributeError:
crawler.settings.values['PROXY_ENABLED'] = False
else:
crawler.settings.values['PROXY_ENABLED'] = False
# TODO deal with cookies
# cookie_flag = getattr(glob, 'COOKIES_ENABLED', False)
# try:
# cookie_flag = (data['cookie'][0].lower() == 'true')
# except (IndexError, KeyError):
# pass
# crawler.settings.values['COOKIES_ENABLED'] = cookie_flag
#
# try:
# crawler.settings.values['COOKIES_DEBUG'] = getattr(glob, 'DEBUG')['COOKIES_DEBUG']
# except (AttributeError, KeyError):
# crawler.settings.values['COOKIES_DEBUG'] = False
crawler.signals.connect(reactor.stop, signal=signals.spider_closed)
crawler.configure()
spider.log(welcome_msg, log.INFO)
crawler.crawl(spider)
crawler.start()
return spider
