def fetch_versions(self):
if self.versions:
return
result = get(self.url)
tables = result('.tabel95')
self.title = tables.find('.titulo').contents()[0].strip()
for i, table in enumerate(tables[2:-1:2]):
trs = query(table)('tr')
release = trs.find('.NewsTitle').text().partition(',')[0]
release = re.sub('version ', '', release, 0, re.I)
infos = trs.next().find('.newsDate').eq(0).text()
infos = re.sub('(?:should)? works? with ', '', infos, 0, re.I)
for tr in trs[2:]:
tr = query(tr)
language = tr('.language')
if not language:
continue
completeness = language.next().text().partition(' ')[0]
language = language.text()
download = tr('a[href*=updated]') or tr('a[href*=original]')
if not download:
continue
hearing_impaired = \
bool(tr.next().find('img[title="Hearing Impaired"]'))
download = download.attr.href
self.add_version(download, language, release, infos,
completeness, hearing_impaired)
def run(self):
while 1:
print('开始爬取...')
page_goods = []
url = 'http://list.showjoy.com/search/?page={page}'.format(page=self.page)
html = query(Sp(url).get('text'))
containers = html('.brick-cover')
# 获取最大页
if self.max_count + self.max_page == 0:
self.max_count = int(html('.highlight').text())
self.max_page = int(self.max_count / 20) + 1
print('获取最大页：' + str(self.max_page))
for g in containers:
g = query(g)
url = g('.brick-pic').eq(0).attr('href')
img = g('.brick-pic').eq(0)('img').eq(0).attr('src')
brand = {
'name': g('.brand').eq(0)('img').eq(0).attr('alt'),
'img': g('.brand').eq(0)('img').eq(0).attr('src')
}
title = g('.brick-title').text().strip()
# 价格转换为数字格式
price = g('.price').text().strip()
price = float(re.findall('¥(\d+\.\d+)', price)[0])
# 销量转换为数字格式
sales = g('.sales').text().strip()
sales = int(re.findall('最近成交(\d+)笔', sales)[0])
page_goods.append({
'url': url,
'img': img,
'brand': brand,
'title': title,
'price': price,
'sales': sales
})
print('{title}，价格：{price}，销量：{sales}'.format(title=title, price=price, sales=sales))
print('第 {page} 页：商品总数：{count}'.format(page=self.page, count=len(page_goods)))
self.goods += page_goods
self.page += 1
if self.page > self.max_page:
self.end_time = time.time()
break
print(
'完成所有商品获取，总页数：{max_page}，商品总数:{count}，耗费时间：{times}'.format(max_page=self.max_page,
count=len(self.goods),
times=self.end_time-self.start_time,
))
def __extractWeight(html):
weight = {}
try:
query_site_info = query(html)('div.siteinfo')
html_font = query_site_info('font')
keys = ['baidu_weight', 'key_count', 'net_flow', 'place']
for i in range(html_font.size()):
weight[keys[i]] = query(html_font[i]).text()
except Exception, e:
print e
def parse_next_page_href(html):
next_page_list = []
try:
query_page_a_s = query(html)('#nav a.fl')
for item in query_page_a_s:
next_page_list.append(
'https://www.google.com.hk' + query(item).attr('href'))
except Exception, e:
print 'DataParser: parse next page info failed'
print e
def __extract_img_info(item):
item_json = {}
try:
query_img_div = query(item)
query_a_img = query_img_div('.r a')
query_summary_span = query_img_div('.s')
#print 'query_summary_span:', query_summary_span
#print 'query_summary_span', query_summary_span.html().decode('iso-8859-1')
if(query_a_img != None):
img_title = query_a_img.text()
img_web_url = query_a_img.attr('href')
img_thumnail_src = query_img_div(
'img').attr('src')
# print query_a_img('div.s')
img_original_src = query_img_div(
'div.s')('a').attr('href')
img_original_src = urlparse.parse_qs(
urlparse.urlparse(img_original_src).query, True)['imgurl'][0]
item_json['img_web_url'] = img_web_url
item_json['img_title'] = img_title
item_json['img_thumnail_src'] = img_thumnail_src
item_json['img_original_src'] = img_original_src
item_json['img_weight'] = 1
item_json['img_datetime'] = '2015.05.01'
# print img_title
# print img_web_url
# print img_original_src
# print img_thumnail_src
except Exception, e:
# print 'extract item json failed'
print e
pass
def process_goods(self, html):
"""
清理获取到的商品数据，返回商品列表
:param ret:
:return:
"""
goods = []
d_goods = html('.gl-item')
# 解析商品数据
for d_good in d_goods:
d_good = query(d_good)
good = {
'id': d_good.attr('data-sku'),
'title': d_good('.p-name > a').text(),
'price': d_good('.p-price > strong > i').text(),
'url': 'https:' + d_good('div.p-img > a').attr('href')
}
# 商品封面
icon = d_good('.p-img > a > img').attr('source-data-lazy-img')
if not icon or icon == 'done':
icon = d_good('.p-img > a > img').attr('src')
good.setdefault('icon', 'https:' + icon)
goods.append(good)
return goods
def parse_item(i):
d = {}
q = query(i)
d['lat'] = float(q.attr('data-latitude')) if len(q.attr('data-latitude')) > 1 else ''
d['lng'] = float(q.attr('data-longitude')) if len(q.attr('data-longitude')) > 1 else ''
d['postdata'] = q.find('.itemdate').text()
d['price'] = q.find('.itemph').text().split(' ')[0]
d['title'] = q.find('a').text()
d['link'] = q.find('a').attr('href')
# print d['title']
d['district'] = ''
district = q.find('.itempn').find('font').text()
if district is not None and district is not "":
for dis in districts.keys():
if dis in district:
d['district'] = districts[dis]
break
for k in validationDistrict.keys():
r = validationDistrict[k]
# print '%s > %s > %s' % (r[0], d['lat'], r[2])
if r[0] > d['lat'] and d['lat'] > r[2]:
# print ' %s > %s > %s' % (r[1], d['lng'], r[3])
if r[1] < d['lng'] and d['lng'] < r[3]:
d['district'] = districts[k]
# print '%s %s - %s [%s]' % (k, d['price'], d['title'], d['link'])
return d if d['district'] is not '' else None
def fetch_versions(self):
if self.versions:
return
result = self._page or session.get(self.url)
tables = result('.tabel95')
self.title = tables.find('.titulo').contents()[0].strip()
for i, table in enumerate(tables[2:-1:2]):
trs = query(table)('tr')
release = encode(trs.find('.NewsTitle').text().partition(',')[0])
release = re.sub('version ', '', release, 0, re.I)
infos = encode(trs.next().find('.newsDate').eq(0).text())
infos = re.sub('(?:should)? works? with ', '', infos, 0, re.I)
for tr in trs[2:]:
tr = query(tr)
language = tr('.language')
if not language:
continue
completeness = encode(language.next().text().partition(' ')[0])
language = encode(language.text())
download = tr('a[href*=updated]') or tr('a[href*=original]')
if not download:
continue
hearing_impaired = \
bool(tr.next().find('img[title="Hearing Impaired"]'))
url = encode(download.attr.href)
favorite = tr('a[href*="saveFavorite"]')[0].attrib['href']
id, language_id, version = \
re.search(r'(\d+),(\d+),(\d+)', favorite).groups()
self.add_version(
id=id,
language_id=language_id,
version=version,
url=url,
language=language,
release=release,
infos=infos,
completeness=completeness,
hearing_impaired=hearing_impaired,
)
def get_max_page(self):
"""
获取商品搜索最大页
:return:
"""
url = 'http://search.jd.com/Search?keyword=' + self.keyword
max_page = query(spider(url).get('text')).html('#J_topPage > span > i').text()
print('获取最大页：' + max_page)
return max_page
def get_cards():
cards = []
page = 1
start_time = time()
while 1:
params = {
'resourceType': 1, # 0是软件排行，1是游戏排行
'page': page
}
ret = spider('http://www.wandoujia.com/api/top/more', params=params, fields='json')
json_data = ret.get('json')
content = unpack_dict(json_data, 'data.content')
# 到最大页则 content 为空
if not content:
end_time = time()
print('爬取完成，总页数：{page}，游戏总数：{count}，耗时：{times}'.format(page=page - 1, count=len(cards), times=end_time-start_time))
return cards
document = query(content)
cards_dom = document('.card')
for card_dom in cards_dom:
# 游戏名称、下载量、图标、下载地址
# 下载地址需安装豌豆荚
card_dom = query(card_dom)
download_btn = card_dom('a.i-source.install-btn')
name = download_btn.attr('data-name')
downloads = download_btn.attr('data-install')
icon = download_btn.attr('data-app-icon')
url = download_btn.attr('href')
cards.append({
'name': name,
'downloads': downloads,
'icon': icon,
'url': url
})
print('完成第 {page} 页爬取，当前总数：{count}'.format(page=page, count=len(cards)))
page += 1
def parse_img_info(html):
print 'parse_img_info()'
result_list = []
try:
query_img_div_s = query(html)('.rc')
if query_img_div_s == None:
return result_list
for item in query_img_div_s:
result_list.append(DataParser.__extract_img_info(item))
except Exception, e:
print 'DataParser: parse img info failed'
print e
def main(sc):
q = query(findapturl)
print 'starting: %s' % (q('p').eq(0).find('.itemph').text())
geo = q('*[data-latitude]').filter(lambda i: query(this).attr('data-latitude') != '')
district = q('p').filter(lambda i: query(this).find('.itempn').each(in_district))
results = []
resultNodes = geo + district
for item in resultNodes:
obj = parse_item(item)
if obj:
results.append( obj )
f = open(''.join([os.path.expanduser('~'),'/.craigslist.json']), 'r+')
for item in results:
if len(item['link']) > 1 and item['link'] not in data:
print item['price']
try:
cost = int(item['price'].replace('$',''))
except:
continue
if cost < maxSMSRent:
shortened = requests.get('http://is.gd/create.php?format=simple&url=%s' % (item['link']))
# notify( '%s [%s] %s %s' % (item['price'], item['district'], shortened.text, item['title']) )
page = requests.get(item['link'])
email(page.text, item['link'], '*****@*****.**')
# email(page.text, item['link'], '*****@*****.**')
item['added'] = datetime.datetime.now()
data[item['link']] = item
f = open(''.join(crapy_JSON_db), 'w')
f.write(json.dumps(data, indent=4, default=date_encoder))
f.close()
sc.enter(sleeptime, 1, main, (sc,))
def get_page_goods(self, page):
page = page * 2 - 1 # 京东商品搜索页每页请求两次
url = 'http://search.jd.com/s_new.php'
""" 获取上半部分商品 """
params = {
"keyword": self.keyword,
"page": page,
"click": "0",
"enc": "utf-8",
"qrst": "1",
"rt": "1",
"s": "110",
"stop": "1",
"vt": "2"
}
html = query(spider(url, params=params).get('text'))
goods = self.process_goods(html)
""" 获取下半部分商品 """
params = {
"keyword": self.keyword,
"show_items": ','.join([g.get('id') for g in goods]),
"enc": "utf-8",
"page": "2",
"log_id": "1510505434.63851",
"qrst": "1",
"rt": "1",
"s": "28",
"scrolling": "y",
"stop": "1",
"tpl": "2_M",
"vt": "2"
}
html = query(spider(url, params=params).get('text'))
goods.extend(self.process_goods(html))
print('第 {page} 页商品数量：{count}'.format(page=page, count=len(goods)))
return goods
def get_recommend_goods(self, page, keyword):
"""
获取（左侧）商品精选
:param page:
:param keyword:
:return:
"""
goods = []
url = 'http://x.jd.com/Search'
params = {
"page": page,
"keyword": keyword,
"_": time.time() * 1000,
"adType": "7",
"ad_ids": "291:20",
"area": "1",
"callback": "jQuery5618052",
"enc": "utf-8",
"xtest": "new_search"
}
ret = spider(url, fields='json', params=params, debug=True)
if ret.get('code') == 200:
# 解析商品数据，格式化保存
json_data = ret.get('json')
json_goods = json_data.get('291', [])
for json_good in json_goods:
good = {
'id': json_good.get('sku_id'), # 通过 ID 取商品评价
'title': query(json_good.get('ad_title')).text(),
'icon': 'http://img1.360buyimg.com/n1/' + json_good.get('image_url'),
'price': json_good.get('pc_price'),
'url': json_good.get('click_url')
}
# 获取商品评价
good.setdefault('comments', self.get_comments(good.get('id')))
goods.append(good)
print('已获取商品：' + good.get('title'))
else:
print('第 {page} 页商品获取完成'.format(page=page))
else:
print('获取商品出错：' + str(ret.get('err')))
return goods
def get_one_page(url: str):
"""
将 $url 对应的文章保存为 html 和 md
虽然最后有清洗的步骤，最后格式还是需要手动调整（比如强迫症的我）
"""
assert url.strip(), "Invalid url!"
headers = {
"user-agent": Faker().chrome(),
}
if url[-1] == "/":
url = url[:-1]
page_id = url.split("/")[-1]
if not os.path.exists(f"files/{page_id}.html"):
print(f"Retrieving {url}...")
r = requests.get(url, headers=headers)
doc = query(r.text)
print(f"Saving the article to files/{page_id}.html...")
with open(f"files/{page_id}.html", "w", encoding="utf-8") as f:
f.write(doc("article").html())
print("Converting html to md...")
os.system(f"pandoc -o files/{page_id}.md files/{page_id}.html")
print("Cleaning the md...")
with open(f"files/{page_id}.md") as f:
lines = f.readlines()
trash_dots_lines = []
for idx, line in enumerate(lines):
if line.startswith(":::"):
trash_dots_lines.append(idx)
if re.search(r"{#.*?}", line):
lines[idx] = re.sub(r"{#.*?}", "", line)
for x in trash_dots_lines[::-1]:
lines.pop(x)
text = re.sub(r"{\..*?}", "", "".join(lines), flags=re.S)
while re.search(r"{\..*?}", text, flags=re.S):
text = re.sub(r"{\..*?}", "", text, flags=re.S)
text = re.sub(r"<div>.*?</div>", "", text, flags=re.S)
text = text.replace("\\\n", "").replace("\n\n\n",
"\n\n").replace("\n\n\n", "\n\n")
with open(f"files/{page_id}.md", "w", encoding="utf-8") as f:
f.write(text)
class Users(pymodel.BaseModel):
def __init__(self):
self.name = UserInfo()
class Employees(pymodel.BaseModel):
def __init__(self):
import datetime
self.set_model_name("employees")
self.firstname = str
self.lastname = str
self.bithdate = datetime.datetime
employees = Employees()
qr = pyquery.query(employees.get_model_name())
qr.project(employees.firstname, employees.lastname)
print qr.pipeline
# user=Users()
# c=pydoc.document.nam + pydoc.document.x>("1+{0}+3",15)
# print c
# pyaggregatebuilders.Match(pydoc.FilterFiels)
# c=pyaggregatebuilders.Project(
# pydoc.document.fullName<<("concat(firstName,'{0}',lastName)","xxx")
#
# )
# print c
# c=(pyfuncs.regex(pydoc.document.name,"12"))
# print c
import pymongo
from pymongo import MongoClient
import pyfuncs
import pydocs
Fields = pydocs.Fields()
# print pyfuncs.toDouble(X.name)
# fields=pydoc.Fields()
# x=pyfuncs.cmp(fields.amount,fields.name)==0
# print isinstance(x,pydoc.Fields)
# # c=x.__owner__
# print x.__tree__
cnn = MongoClient(host="localhost", port=27017)
db = cnn.get_database("hrm")
db.authenticate(name="root", password="******")
qr = pyquery.query(db, "test.coll001")
qr = qr.where(pyfuncs.regex(Fields.fx, "^312313$"))
# qr.project({
# Fields.Users.username:1,
# Fields.Users.fullName:pyfuncs.concat(Fields.Users.firstName, " ",Fields.Users.lastname)
# })
# qr=qr+2
# #.set(x=1,y=2)
# import pprint
# items=list(qr.objects)
import pprint
x = list(qr.objects)
pprint.pprint(list(qr.items))
def get_page_content(page):
items = []
domain = 'https://www.qiushibaike.com'
# 获取原始链接，内容详情页
url = 'https://www.qiushibaike.com/8hr/page/{0}/'.format(page)
ret = spider(url)
text = ret.get('text')
index_document = query(text)
articles = index_document('.article')
for article in articles:
href = query(article)('a.contentHerf').attr('href')
items.append({'url': domain + href})
print('第 {page} 页获取链接数：{count}'.format(page=page, count=len(items)))
if len(items) == 0:
print(ret)
for index, item in enumerate(items):
for i in range(0, 2):
# 访问详情页，获取内容 + 评论
text = spider(item['url']).get('text')
document = query(text)
# 内容
content = document('#single-next-link > div').text()
if not content:
print('获取失败，重试获取，进度：{index}/{maxlength}'.format(
index=index + 1, maxlength=len(items)))
continue
# 内容配图
img_href = document('#single-next-link > div.thumb > img').attr(
'src') or ''
if img_href:
img_href = 'https:' + img_href
# 评论
comments = []
comments_dom = document('.comment-block > div.replay > span.body')
for span in comments_dom:
comments.append(query(span).text())
item.update({
'content': content,
'img_href': img_href,
'comments': comments
})
print('获取第 {page} 页，进度：{index}/{maxlength}'.format(
page=page, index=index + 1, maxlength=len(items)))
break
print('第 {page} 页获取完成'.format(page=page))
if page == 1:
max_page = int(
index_document(
'#content-left > ul > li:nth-child(7) > a > span').text())
print('最大页：' + str(max_page))
return max_page, items
return items
def get(url, raw=False, **params):
global last_url
url = urljoin(last_url, url)
request = requests.get(url, headers={'Referer': last_url}, params=params)
last_url = url
return request.content if raw else query(request.content)
