Python自用代码（某方标准类网页源代码清洗）

#coding=utf-8
from pymongo import MongoClient
from lxml import etree
import requests

s = [u'标准编号：',u'发布单位：',u'发布日期：',u'状态：',u'实施日期：',u'开本页数：',u'采用关系：',
    u'中图分类号：',u'中国标准分类号：',u'国际标准分类号：',u'国别：',u'关键词：',u'摘要：']

# 获取数据库
def get_db():
    client = MongoClient('IP', 27017)
    db = client.wanfang
    db.authenticate("用户名","密码") 
    return db

# 获取第num条数据
def get_data(table, num):
    i = 1
    for item in table.find({}, {"content":1,"_id":0}):
        if i==num:
            if item.has_key('content') and item['content']:
                return item['content']
        else:
            i+=1
            continue

# 列表转字符串
def list_str(list):
    if len(list)!=0:
        return list[0]
    else:
        return ""

# 提取分类号
def code_ls(list):
    if len(list)!=0:
        ls = list[0].split()
        shanchu = []
        for i in ls:
            if ("("in i) or (")"in i) or ("（"in i) or("）"in i):
                shanchu.append(i)
        for i in shanchu:
            ls.remove(i)
        return ls
    else:
        return ""

# 构造关键词列表
def keywords_ls(list):
    if len(list)!=0:
        return list
    else:
        return ""

# 替代标准
def replace_str(replace):
    if replace!="":
        ls = [i.strip().replace("\r\n", "") for i in replace]
        if len(ls)!=0:
            return ls[0][5:]
        else:
            return ""
    else:
        return ""

# 提取摘要
def summary_str(list):
    if len(list)!=0:
        if list[0][0]!="<":
            return list[0]
        else:
            return ""
    else:
        return ""

# 调整日期格式
def date_str(list):
    if len(list)!=0:
        year = list[0].find(u'年')
        month = list[0].find(u'月')
        day = list[0].find(u'日')
        if month-year==2:
            list[0] = list[0].replace(u"年",u"年0")
        if day-mOnth==2:
            list[0] = list[0].replace(u"月",u"月0")
        return list[0].replace(u"日","").replace(u"月","-").replace(u"年","-")
    else:
        return ""

# 调整采标格式
def adopted_ls(string, ls):
    dc = {}
    loc = string.find(',')
    if loc==-1:
        return ls
    else:
        dc["code"] = string[:loc].strip()
        dc["type"] = string[loc+1:loc+4]
        ls.append(dc)
        return adopted_ls(string[loc+4:],ls)

# 构造标准入库字典
def standard_dict(html):
    dc = {}
    tree = etree.HTML(html)
    # 标准名称
    dc["title"] = list_str(tree.xpath("//h1/text()"))
    # 外文名称
    dc["title_eng"] = list_str(tree.xpath("//h2/text()"))
    # 标准编号
    dc["standard_number"] = list_str(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[0])))
    # 发布单位
    dc["publishing_department"] = list_str(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[1])))
    # 发布日期
    dc["release_date"] = date_str(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[2])))
    # 状态
    dc["state"] = list_str(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[3])))
    # 实施日期
    dc["enforcement_date"] = date_str(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[4])))
    # 开本页数
    dc["pages"] = list_str(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[5])))
    # 采用关系
    dc["adopted"] = adopted_ls(list_str(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[6]))), [])
    # 中图分类号
    dc["clc"] = code_ls(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[7])))
    # 中国标准分类号
    dc["ccs"] = code_ls(tree.xpath("//span[text()='%s']/following-sibling::*/child::*/text()"%(s[8])))
    # 国际标准分类号
    dc["ics"] = code_ls(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[9])))
    # 国别
    dc["country"] = list_str(tree.xpath("//span[text()='%s']/following-sibling::*/text()"%(s[10])))
    # 关键词
    dc["keywords"] = keywords_ls(tree.xpath("//span[text()='%s']/following-sibling::*/child::*/text()"%(s[11])))
    # 摘要
    dc["summary"] = summary_str(tree.xpath("//span[text()='%s']/parent::*/following-sibling::*/text()"%(s[12])))
    # 替代标准
    dc["replace_for"] = replace_str(tree.xpath("//div[@id='replaceStandard']//child::*//text()"))
    return dc

# 主函数
def main():
    db = get_db()
    collection=db.standard
    collection2 = db.standard_cleaned
    for item in collection.find({}, {"content":1,"_id":0}):
        if item.has_key('content') and item['content']:
            dc = standard_dict(item['content'])
            collection2.insert(dc)

if __name__ == '__main__':
    main()
    
    # 以下代码用于测试清洗特定一条数据
    # db = get_db()
    # collection=db.standard
    # collection2 = db.standard_cleaned
    # data = get_data(collection, 8)
    # dc = standard_dict(data)
    # collection2.insert(dc)
    # for k,v in dc.items():
    #     print k,v

    # # 以下代码用于测试提取摘要
    # data = requests.get('http://d.wanfangdata.com.cn/Standard/ISO%208528-5-2013')
    # dc = standard_dict(data.text)
    # for k,v in dc.items():
    #     print k,v

    # # 以下代码用于测试修改日期格式
    # l1 = [u"2017年6月28日"]
    # l2 = [u"2017年10月27日"]
    # l3 = [u"2017年12月1日"]
    # l4 = [u"2017年7月1日"]
    # print date_str(l1)
    # print date_str(l2)
    # print date_str(l3)
    # print date_str(l4)