用到的包:xml.dom.minidom 文档
代码托管位置 github-pytools
有一个表,里面数据量比较大,每天一更新,其字段可以通过xml配置文件进行配置,即,可能每次建表的字段不一样。
上游跑时会根据配置从源文件中提取,到入库这一步需要根据配置进行建表。
写了一个简单的xml,配置需要字段及类型
上游读取到对应的数据
入库这一步,先把原表删除,根据配置建新表
<?xml version="1.0" encoding="UTF-8"?>
<!-- 表名 ,数据库名 可灵活配置插入哪个库哪个表 --> <table name="top_query" db_name="evaluting_sys"> <!-- 非业务主键,自增长,可配名,其他 INTEGER UNSIGNED AUTO_INCREMENT --> <primary_key> <name>id</name> </primary_key> <!-- 字段开始 --> <field> <name>query</name> <type>varchar(200)</type> <is_index>false</is_index> <description>query</description> </field> <field> <name>pv</name> <type>integer</type> <is_index>false</is_index> <description>pv</description> </field> <field> <name>avg_money</name> <type>integer</type> <is_index>false</is_index> <description></description> </field> <!-- 字段配置结束 --> </table>
#!/usr/bin/python
# -*- coding:utf-8 -*- #author: wklken #desc: use to read db xml config. #----------------------- #2012-02-18 created #---------------------- import sys,os from xml.dom import minidom, Node def read_dbconfig_xml(xml_file_path): content = {} root = minidom.parse(xml_file_path) table = root.getElementsByTagName("table")[0] #read dbname and table name. table_name = table.getAttribute("name") db_name = table.getAttribute("db_name") if len(table_name) > 0 and len(db_name) > 0: db_sql = "create database if not exists `" + db_name +"`; use " + db_name + ";" table_drop_sql = "drop " + table_name + " if exists " + table_name + ";" content.update({"db_sql" : db_sql}) content.update({"table_sql" : table_drop_sql }) else: print "Error:attribute is not define well! db_name=" + db_name + " ;table_name=" + table_name sys.exit(1) #print table_name, db_name table_create_sql = "create table " + table_name +"(" #read primary cell primary_key = table.getElementsByTagName("primary_key")[0] primary_key_name = primary_key.getElementsByTagName("name")[0].childNodes[0].nodeValue table_create_sql += primary_key_name + " INTEGER NOT NULL AUTO_INCREMENT PRIMARY KEY," #print primary_key.toxml() #read ordernary field fields = table.getElementsByTagName("field") f_index = 0 for field in fields: f_index += 1 name = field.getElementsByTagName("name")[0].childNodes[0].nodeValue type = field.getElementsByTagName("type")[0].childNodes[0].nodeValue table_create_sql += name + " " + type if f_index != len(fields): table_create_sql += "," is_index = field