change database schema, auto detect max page of node

Author: oldshensheep

requirements-analysis.txt

@@ -0,0 +1,2 @@
+pandas==2.0.1
+plotly==5.14.1

requirements.txt

@@ -1,5 +1,4 @@
 arrow==1.2.3
-pandas==2.0.1
-plotly==5.14.1
+httpx==0.24.1
 Scrapy==2.9.0
 SQLAlchemy==2.0.17

v2ex_scrapy/DB.py

@@ -26,9 +26,9 @@ def __new__(cls):
             cls._instance = super().__new__(cls)
         return cls._instance
 
-    def __init__(self):
+    def __init__(self, database_name="v2ex.sqlite"):
         self.engine = create_engine(
-            "sqlite:///v2ex.sqlite",
+            f"sqlite:///{database_name}",
             echo=False,
             json_serializer=lambda x: json.dumps(x, ensure_ascii=False),
         )
@@ -61,7 +61,7 @@ def get_max_topic_id(self) -> int:
 
     def get_topic_comment_count(self, topic_id) -> int:
         result = self.session.execute(
-            text("select count(*) from comment where topic_id = :q"), {"q": topic_id}
+            text("select reply_count from topic where id = :q"), {"q": topic_id}
         ).fetchone()
         if result is None or result[0] is None:
             return 0

v2ex_scrapy/items.py

@@ -26,14 +26,15 @@ class TopicItem(Base):
     id_: Mapped[int] = mapped_column(name="id", primary_key=True)
     author: Mapped[str] = mapped_column(nullable=False)
     title: Mapped[str] = mapped_column(nullable=False)
-    content: Mapped[str]
+    content: Mapped[str] = mapped_column()
     node: Mapped[str] = mapped_column(nullable=False)
     tag: Mapped[list[str]] = mapped_column(nullable=False)
     clicks: Mapped[int] = mapped_column(nullable=False)
     votes: Mapped[int] = mapped_column(nullable=False)
     create_at: Mapped[int] = mapped_column(nullable=False)
     thank_count: Mapped[int] = mapped_column(nullable=False)
     favorite_count: Mapped[int] = mapped_column(nullable=False)
+    reply_count: Mapped[int] = mapped_column(nullable=False)
 
     @staticmethod
     def err_topic(topic_id: int):
@@ -49,6 +50,7 @@ def err_topic(topic_id: int):
             votes=-1,
             thank_count=-1,
             favorite_count=-1,
+            reply_count=-1,
         )
 
 
@@ -72,6 +74,7 @@ class CommentItem(Base):
     content: Mapped[str] = mapped_column(nullable=False)
     thank_count: Mapped[int] = mapped_column(nullable=False)
     create_at: Mapped[int] = mapped_column(nullable=False)
+    no: Mapped[int] = mapped_column(nullable=False)
 
 
 @dataclass(kw_only=True)

v2ex_scrapy/middlewares.py

@@ -7,13 +7,13 @@
 
 import random
 import time
-from http.cookies import SimpleCookie
 
 import scrapy
 import scrapy.http.response.html
 from scrapy import signals
-
+from scrapy.exceptions import IgnoreRequest
 from v2ex_scrapy.DB import DB, LogItem
+from v2ex_scrapy import utils
 
 
 class TutorialScrapySpiderMiddleware:
@@ -95,10 +95,14 @@ def process_request(self, request: scrapy.Request, spider):
         return None
 
     def process_response(
-        self, request, response: scrapy.http.response.html.HtmlResponse, spider
+        self,
+        request: scrapy.Request,
+        response: scrapy.http.response.html.HtmlResponse,
+        spider: scrapy.Spider,
     ):
         # Called with the response returned from the downloader.
-
+        if response.status == 403:
+            raise IgnoreRequest(f"403 url {response.url}")
         # Must either;
         # - return a Response object
         # - return a Request object
@@ -118,10 +122,8 @@ def process_exception(self, request, exception, spider):
     def spider_opened(self, spider: scrapy.Spider):
         self.proxies = spider.settings.get("PROXIES", [])  # type: ignore
 
-        if type(cookie_str := spider.settings.get("COOKIES", "")) == str:
-            simple_cookie = SimpleCookie()
-            simple_cookie.load(cookie_str)  # type: ignore
-            self.cookies = {k: v.value for k, v in simple_cookie.items()}
+        cookie_str = spider.settings.get("COOKIES", "")
+        self.cookies = utils.cookie_str2cookie_dict(cookie_str) # type: ignore
 
         spider.logger.info("Spider opened: %s" % spider.name)
 

v2ex_scrapy/pipelines.py

@@ -11,12 +11,7 @@
 # don't remove
 import v2ex_scrapy.insert_ignore
 from v2ex_scrapy.DB import DB
-from v2ex_scrapy.items import (
-    CommentItem,
-    MemberItem,
-    TopicItem,
-    TopicSupplementItem,
-)
+from v2ex_scrapy.items import CommentItem, MemberItem, TopicItem, TopicSupplementItem
 
 ItemsType = Union[TopicItem, CommentItem, MemberItem, TopicSupplementItem]
 
@@ -48,7 +43,11 @@ def process_item(
                 self.db.session.commit()
         return item
 
-    def close_spider(self, spider):
+    def save_all(self):
         for _, v in self.data.items():
             self.db.session.add_all(v)
+            self.db.session.commit()
+
+    def close_spider(self, spider):
+        self.save_all()
         self.db.close()

v2ex_scrapy/spiders/CommonSpider.py

@@ -9,13 +9,10 @@
 
 
 class CommonSpider:
-    def __init__(
-        self, logger, update_topic=False, update_member=False, update_comment=False
-    ):
+    def __init__(self, logger, update_member=False, update_comment=False):
         self.db = DB()
         self.logger = logger
         self.UPDATE_MEMBER = update_member
-        # only work when UPDATE_TOPIC
         self.UPDATE_COMMENT = update_comment
 
     def parse_topic_err(self, failure):

v2ex_scrapy/spiders/V2exNodeTopicSpider.py

@@ -1,16 +1,26 @@
+import httpx
 import scrapy
 import scrapy.http.response.html
+from parsel import Selector
+from scrapy.utils.project import get_project_settings
 
 from v2ex_scrapy.DB import DB
 from v2ex_scrapy.items import TopicItem
 from v2ex_scrapy.spiders.CommonSpider import CommonSpider
+from v2ex_scrapy import utils
 
 
 class V2exTopicSpider(scrapy.Spider):
     name = "v2ex-node"
 
     UPDATE_TOPIC_WHEN_REPLY_CHANGE = True
     UPDATE_COMMENT = True  # only work when UPDATE_TOPIC_WHEN_REPLY_CHANGE = True
+    URL = "https://www.v2ex.com/go/"
+
+    """
+    现存在的几个问题，因为节点的排序是动态的，如果爬完一页后未爬的主题跑到爬完的页数里那就爬不到了。
+    解决方法1，开始爬取时先获取全部帖子ID再开始爬，获取ID的速度比较快所以排序改变的幅度不会很大。
+    """
 
     def __init__(self, node="flamewar", *args, **kwargs):
         super().__init__(*args, **kwargs)
@@ -19,11 +29,25 @@ def __init__(self, node="flamewar", *args, **kwargs):
         self.common_spider = CommonSpider(
             self.logger, update_comment=self.UPDATE_COMMENT
         )
+        settings = get_project_settings()
+        resp = httpx.get(
+            f"{self.URL}{self.node}",
+            timeout=10,
+            follow_redirects=True,
+            cookies=utils.cookie_str2cookie_dict(settings.get("COOKIES", "")),  # type: ignore
+            headers={"User-Agent": settings.get("USER_AGENT", "")},  # type: ignore
+        ).text
+        max_page = (
+            Selector(text=resp)
+            .xpath('//tr/td[@align="left" and @width="92%"]/a[last()]/text()')
+            .get("1")
+        )
+        self.max_page = int(max_page)
 
     def start_requests(self):
-        for i in range(552, 0, -1):
+        for i in range(self.max_page, 0, -1):
             yield scrapy.Request(
-                url=f"https://www.v2ex.com/go/{self.node}?p={i}",
+                url=f"{self.URL}{self.node}?p={i}",
                 callback=self.parse,
                 cb_kwargs={"page": i},
             )
@@ -33,6 +57,7 @@ def parse(self, response: scrapy.http.response.html.HtmlResponse, page: int):
             (int(x), int(y))
             for x, y in zip(
                 response.xpath('//span[@class="item_title"]/a/@id').re(r"\d+"),
+                # not correct when some comments are deleted, fuck
                 response.xpath('//span[@class="item_title"]/a/@href').re(r"reply(\d+)"),
             )
         ]

v2ex_scrapy/utils.py

@@ -1,3 +1,4 @@
+from http.cookies import SimpleCookie
 import json
 from typing import Union
 
@@ -27,6 +28,12 @@ def json_to_str(j):
     return json.dumps(j, ensure_ascii=False)
 
 
+def cookie_str2cookie_dict(cookie_str: str):
+    simple_cookie = SimpleCookie()
+    simple_cookie.load(cookie_str)
+    return {k: v.value for k, v in simple_cookie.items()}
+
+
 if __name__ == "__main__":
     a = ["2022-04-28 13:24:38 +08:00", "287 天前", "1 小时前"]
 

v2ex_scrapy/v2ex_parser.py

@@ -42,8 +42,10 @@ def parse_comment(response: scrapy.http.response.html.HtmlResponse, topic_id):
         reply_content = cbox.xpath('.//div[@class="reply_content"]').get("")
         reply_time = cbox.css(".ago::attr(title)").get("")
         thank_count = cbox.css(".fade::text").get("0").strip()
+        no = cbox.css(".no::text").get("-1").strip()
         yield CommentItem(
             id_=int(comment_id),
+            no=int(no),
             commenter=author_name,
             topic_id=topic_id,
             content=reply_content,
@@ -74,7 +76,9 @@ def parse_topic(response: scrapy.http.response.html.HtmlResponse, topic_id):
         )
 
     topic_content = response.css(".cell .topic_content").get("")
-
+    topic_reply_count = response.css(".box > .cell > .gray::text").re_first(
+        r"(\d+) 条回复", "0"
+    )
     yield TopicItem(
         id_=topic_id,
         author=topic_author,
@@ -87,6 +91,7 @@ def parse_topic(response: scrapy.http.response.html.HtmlResponse, topic_id):
         votes=int(topic_vote),
         thank_count=int(topic_thank_count),
         favorite_count=int(topic_favorite_count),
+        reply_count=int(topic_reply_count),
     )