class MyItem(Item):
"""
定义爬虫的目标字段
"""
target_item = HtmlField(css_select='html')
title = TextField(css_select='head title')
article = HtmlField(css_select='article')
def test_html_field_with_many():
field = HtmlField(css_select="a.test_link", many=True)
values = field.extract(html_etree=html_etree)
assert len(values) == 5
assert values[0] == '<a class="test_link" href="https://github.com/howie6879/">hello1 github.</a>\n'
assert values[4] == '<a class="test_link" href="https://github.com/howie6879/">hello5 github.</a>\n' \
' Some text outside.\n'
def test_html_field():
field_en = HtmlField(css_select="div.brand a")
field_zh = HtmlField(css_select="div.brand p")
assert (
field_en.extract(html_etree=html_etree)
== '<a href="https://github.com">Github</a>'
)
assert field_zh.extract(html_etree=html_etree) == "<p>你好</p>\n"
class RankingItem(Item):
target_item = TextField(css_select='.rank-list')
ranking_title = TextField(css_select='h3.wrap-title')
more = AttrField(css_select='h3>a.more', attr='href')
book_list = HtmlField(css_select='div.book-list>ul>li', many=True)
async def clean_ranking_title(self, ranking_title):
if isinstance(ranking_title, list):
return ranking_title[0].text
else:
return str(ranking_title).split('榜')[0] + '榜'
async def clean_more(self, more):
return "https:" + more
class ChinaNewsItem(Item):
"""
定义目标字段抓取规则
"""
#target_item =
title = TextField(css_select='h1')
content = HtmlField(css_select='div.left_zw')
async def clean_title(self, value):
"""
清洗目标数据
:param value: 初始目标数据
:return:
"""
return value
def test_html_field():
field = HtmlField(css_select="div.brand a")
assert (field.extract(
html_etree=html_etree) == '<a href="https://github.com">Github</a>')
class RankingItem(Item):
target_item = TextField(css_select='div.rank_i_p_list')
ranking_title = TextField(css_select='div.rank_i_p_tit')
more = AttrField(css_select='div.rank_i_more a', attr='href')
book_list = HtmlField(css_select='div.rank_i_p_list>div.rank_i_li',
many=True)
class WechatItem(Item):
"""
基于 Ruia 的微信页面 Item 提取类
示例：https://mp.weixin.qq.com/s/NKnTiLixjB9h8fSd7Gq8lw
"""
# 文章标题
# doc_name = AttrField(css_select='meta[property="og:title"]', attr="content")
doc_name = AttrField(css_select='meta[property="og:title"]',
attr="content",
default="")
# 描述
doc_des = AttrField(css_select='meta[property="og:description"]',
attr="content",
default="")
# 文章作者
doc_author = AttrField(css_select='meta[property="og:article:author"]',
attr="content",
default="")
# 文章链接，这里的链接有过期时间，但是在微信体系内打开并不会过期，所以可以用
doc_link = AttrField(css_select='meta[property="og:url"]',
attr="content",
default="")
# 文章类型
doc_type = AttrField(css_select='meta[property="og:type"]',
attr="content",
default="")
# 文章发布时间戳
doc_ts = RegexField(
re_select=r"var ct = \"(\d{1,10})\"\;",
default=time.time(),
)
# 文章发布日期
doc_date = RegexField(
re_select=r"var ct = \"(\d{1,10})\"\;",
default=ts_to_str_date(time.time()),
)
# doc_date_f1 = TextField(css_select="em#publish_time", default="")
# doc_date_f2 = RegexField(
# re_select=r"o=\"(20\d.*)\"\;",
# default=ts_to_str_date(time.time(), "%Y-%m-%d %H:%M"),
# )
# doc_ts_f1 = TextField(css_select="em#publish_time", default="")
# doc_ts_f2 = RegexField(
# re_select=r"o=\"(20\d.*)\"\;",
# default=ts_to_str_date(time.time(), "%Y-%m-%d %H:%M"),
# )
# 文章图
doc_image = AttrField(css_select='meta[property="og:image"]',
attr="content",
default="")
# 公众号名称
doc_source_name = TextField(
css_select="div.profile_inner>strong.profile_nickname", default="")
# 公众号元数据
doc_source_meta_list = TextField(
css_select="p.profile_meta>span.profile_meta_value",
many=True,
default=["", ""])
# 核心html
doc_core_html = HtmlField(css_select="div#js_content", default="")
# 公众号昵称
doc_source_account_nick = ""
# 公众号介绍
doc_source_account_intro = ""
# 文本内容，兼容
doc_content = ""
# 常量
# 信息来源
doc_source = "liuli_wechat"
async def clean_doc_source_meta_list(self, value: list):
"""从doc_source_meta_list提取公众号昵称和介绍"""
self.doc_source_account_nick = value[0]
self.doc_source_account_intro = value[1]
return value
async def clean_doc_core_html(self, value: str):
"""清洗核心html"""
return text_compress(
str(value).strip().replace("visibility: visible;", "").replace(
"<br>", "").replace("data-src", "src"))
async def clean_doc_date(self, value):
"""
清洗时间，数据格式 2021-12-17 08:48
"""
try:
value = ts_to_str_date(value)
except Exception as _:
value = ts_to_str_date(time.time())
return value
async def clean_doc_ts(self, value):
"""
清洗时间戳，数据格式1620567960
"""
try:
value = int(value)
except Exception as _:
value = int(time.time())
return value
class HackerNewsItem(Item):
target_item = TextField(css_select='tr.athing')
title = TextField(css_select='a.storylink')
url = AttrField(css_select='a.storylink', attr='href')
content = HtmlField(css_select='a.storylink')
