python處理xml檔案_執行個體Python處理XML檔案的方法

需求

有一個表，裡面資料量比較大，每天一更新，其字段可以通過xml配置檔案進行配置，即，可能每次建表的字段不一樣。

上遊跑時會根據配置從源檔案中提取，到入庫這一步需要根據配置進行建表。

解決

寫了一個簡單的xml，配置需要字段及類型

上遊讀取到對應的資料

入庫這一步，先把原表删除，根據配置建新表

XML檔案

<?xml version="1.0" encoding="UTF-8"?>

id

query

varchar(200)

false

query

pv

integer

false

pv

avg_money

integer

false

處理腳本

#!/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.getElementsByTagName("is_index")[0].childNodes[0].nodeValue

table_create_sql += ");"

content.update({"table_create_sql" : table_create_sql})

#character set latin1 collate latin1_danish_ci;

print content

if __name__ == "__main__":

read_dbconfig_xml(sys.argv[1])

PYTHON解析XML大檔案[SAX]需求

讀取xml資料檔案，檔案較大，需要實時處理插入到資料庫

xml文檔

100000

男

北京，海澱區

437

1989

333

242

null

大學

1

2月14日

處理

sax處理時并不會像dom一樣可以以類似節點的次元進行讀取,它隻有 開始标簽 内容 結束标簽 之分

處理思想是：通過一個handler，對開始标簽，内容，結束标簽各有一個處理函數

代碼及注解

person 處理類

from xml.sax import handler,parseString

class PersonHandler(handler.ContentHandler):

def __init__(self, db_ops):

#db op obj

self.db_ops = db_ops

#存儲一個person的map

self.person = {}

#目前的tag

self.current_tag = ""

#是否是tag之間的内容 ，目的拿到tag間内容，不受空白的幹擾

self.in_quote = 0

#開始，清空map

def startElement(self, name, attr):

#以person，清空map

if name == "person":

self.person = {}

#記錄 狀态

self.current_tag = name

self.in_quote = 1

#結束，插入資料庫

def endElement(self, name):

#以person結尾 代表讀取一個person的資訊結束

if name == "person":

#do something

in_fields = tuple([ ('"' + self.person.get(i,"") + '"') for i in fields ])

print in_sql % in_fields

db_ops.insert( in_sql%(in_fields))

#處理

self.in_quote = 0

def characters(self, content):

#若是在tag之間的内容，更新到map中

if self.in_quote:

self.person.update({self.current_tag: content})

加上入庫的完整代碼

#!/usr/bin/python

# -*- coding:utf-8 -*-

#parse_person.py

#version : 0.1

#author : [email protected]

#desc : parse person.xml and out sql

import sys,os

import MySQLdb

reload(sys)

sys.setdefaultencoding('utf-8')

in_sql = "insert into person(id,sex,address,fansNum,summary,wbNum,gzNum,blog,edu,work,renZh,brithday) values(%s, %s, %s, %s, %s, %s,

%s, %s, %s, %s, %s, %s)"

fields = ("id","sex","address","fansNum","summary","wbNum","gzNum","blog","edu","work","renZh","brithday")

#資料庫方法

class Db_Connect:

def __init__(self, db_host, user, pwd, db_name, charset="utf8", use_unicode = True):

print "init begin"

print db_host, user, pwd, db_name, charset , use_unicode

self.conn = MySQLdb.Connection(db_host, user, pwd, db_name, charset=charset , use_unicode=use_unicode)

print "init end"

def insert(self, sql):

try:

n = self.conn.cursor().execute(sql)

return n

except MySQLdb.Warning, e:

print "Error: execute sql '",sql,"' failed"

def close(self):

self.conn.close()

#person 處理類

from xml.sax import handler,parseString

class PersonHandler(handler.ContentHandler):

def __init__(self, db_ops):

#db op obj

self.db_ops = db_ops

#存儲一個person的map

self.person = {}

#目前的tag

self.current_tag = ""

#是否是tag之間的内容

self.in_quote = 0

#開始，清空map

def startElement(self, name, attr):

#以person，清空map

if name == "person":

self.person = {}

#記錄 狀态

self.current_tag = name

self.in_quote = 1

#結束，插入資料庫

def endElement(self, name):

#以person結尾 代表讀取一個person的資訊結束

if name == "person":

#do something

in_fields = tuple([ ('"' + self.person.get(i,"") + '"') for i in fields ])

print in_sql % in_fields

db_ops.insert( in_sql%(in_fields))

#處理

self.in_quote = 0

def characters(self, content):

#若是在tag之間的内容，更新到map中

if self.in_quote:

self.person.update({self.current_tag: content})

if __name__ == "__main__":

f = open("./person.xml")

#如果源檔案gbk 轉碼 若是utf-8，去掉decode.encode

db_ops = Db_Connect("127.0.0.1", "root", "root", "test")

parseString(f.read().decode("gbk").encode("utf-8"), PersonHandler(db_ops))

f.close()

db_ops.close()