def __init__(self):
model=DataBaseModel()
engine = model.db_connect()
model.create_drawing_table(engine)
self._session = sessionmaker(bind=engine)
self._handlers=[DrawingItemDataBaseHanlder(self._session),
DrawingTagItemDataBaseHanlder(self._session),
DrawingIntroItemDataBaseHanlder(self._session)]
class JdBaseSpider(scrapy.Spider):
allowed_domains = ["http://www.jd.com/", "item.jd.com", "list.jd.com", "d.3.cn", "search.jd.com"]
def __init__(self):
self._model = DataBaseModel()
self._engine = self._model.db_connect()
self._session = sessionmaker(bind=self._engine)()
def _get_drawing_by_unique_key(self, key):
drawings = self._session.query(DrawingTable).filter(DrawingTable.isbn == key)
if drawings.count() == 0:
return None
else:
return drawings.first()
def _is_empty(self, intro):
return intro == '' or intro == 'n/a'
def _is_content_or_author_intro_empty(self, drawing):
if not drawing:
return True
return self._is_empty(drawing.content_intro) or self._is_empty(drawing.author_intro)
def _handle_drawing_page(self, response):
item = DrawingItemBuilder(response).build()
# DRAWING_UNIQUE_KEY absent
if not item[DRAWING_UNIQUE_KEY]:
logger.warn("drop the item found in %s since no %s found" % (response.url, DRAWING_UNIQUE_KEY))
return
drawing = self._get_drawing_by_unique_key(item[DRAWING_UNIQUE_KEY])
# drawing has not been in DataBase
if self._create_new_drawing_if_not_exist and not drawing:
yield item
# unique tags for a single DRAWING_UNIQUE_KEY
if item['press']:
yield DrawingTagItemBuilder(item[DRAWING_UNIQUE_KEY], u"出版社", item['press']).build()
if item['author']:
yield DrawingTagItemBuilder(item[DRAWING_UNIQUE_KEY], u"作者/绘者/译者", item['author']).build()
if item['drawer']:
yield DrawingTagItemBuilder(item[DRAWING_UNIQUE_KEY], u"作者/绘者/译者", item['drawer']).build()
if item['translator']:
yield DrawingTagItemBuilder(item[DRAWING_UNIQUE_KEY], u"作者/绘者/译者", item['translator']).build()
# add one more tag
if self._tag_type and self._tag_value:
yield DrawingTagItemBuilder(item[DRAWING_UNIQUE_KEY], self._tag_type, self._tag_value).build()
# crawl for content introduction and/or author introduction
if self._is_content_or_author_intro_empty(drawing) and item['product_code']:
drawing_detail_page = u"http://d.3.cn/desc/" + item['product_code'] + u"?cdn=1&callback=showdesc"
yield scrapy.Request(drawing_detail_page,
callback=lambda rsp, key=item[DRAWING_UNIQUE_KEY]: self._handle_drawing_detail_page(
rsp, key))
def _handle_drawing_detail_page(self, response, key):
yield DrawingIntroItemBuilder(response, key).build()
def mark_crawling_task_to_done(crawl_task):
model = DataBaseModel()
engine = model.db_connect()
session = sessionmaker(bind=engine)()
for row in session.query(CrawlListTable).filter(CrawlListTable.id == crawl_task.tag_id):
logger.info("crawl done %d %s %s %s" % (row.id, row.tag_type, row.tag_value, row.start_url))
row.done = True
session.commit()
session.close()
def get_jd_drawings_to_crawl(by_search=False):
model = DataBaseModel()
engine = model.db_connect()
session = sessionmaker(bind=engine)()
to_crawl_list = []
for row in session.query(CrawlListTable).filter(CrawlListTable.done == False).filter(
CrawlListTable.by_search == by_search).all():
to_crawl_list.append(DrawingsToCrawl(row.id, row.tag_type, row.tag_value, row.start_url))
session.close()
return to_crawl_list
class DuplicateInDataBasePipeline(object):
def __init__(self):
self._model = DataBaseModel()
self._engine = self._model.db_connect()
self._session = sessionmaker(bind=self._engine)
def _exist_in_db(self,item):
return self._session().query(DrawingTable).filter(
DrawingTable.isbn == item[DRAWING_UNIQUE_KEY]).count() > 0
def process_item(self, item, spider):
if type(item) == DrawingItem:
if self._exist_in_db(item):
raise DropItem("Duplicate item found in DB: %s" % item)
return item
# -*- coding: utf-8 -*-
import logging
import time
import requests
import string
from drawing.util.log import setup_logger
from drawing.models.tables.drawing import DrawingTable
from drawing.models.model import DataBaseModel
from sqlalchemy.orm import sessionmaker
logger = setup_logger(loggername=__name__, console=logging.DEBUG)
DATABASE_MODEL = DataBaseModel()
DATABASE_ENGINE = DATABASE_MODEL.db_connect()
class CrawlListGenerator(object):
def _valid_isbn(self, isbn):
return len(isbn) == 13 or len(isbn) == 10
def _book_to_crawl(self, row):
return self._valid_isbn(row.isbn) and (not row.tried_in_douban) and row.single_book
def generate(self):
db_session = sessionmaker(bind=DATABASE_ENGINE)
return [row.isbn for row
in db_session().query(DrawingTable.isbn, DrawingTable.tried_in_douban, DrawingTable.single_book).all()
if self._book_to_crawl(row)]