Python学习笔记（十七）内建模块之XML和HTMLParser

参考资料：https://www.liaoxuefeng.com/wiki/001374738125095c955c1e6d8bb493182103fac9270762a000/001407499098340324a6232bfee42348849d53c90576747000

1、操作XML有两种方法：DOM和SAX。DOM会把整个XML读入内存，解析为树，因此占用内存大，解析慢，优点是可以任意遍历树的节点。SAX是流模式，边读边解析，占用内存小，解析快，缺点是我们需要自己处理事件。正常情况下，优先考虑SAX，因为DOM实在太占内存。在Python中使用SAX解析XML非常简洁，通常我们关心的事件是start_element，end_element和char_data，准备好这3个函数，然后就可以解析xml了。下面是我的学习代码：

from xml.parsers.expat import ParserCreate

class mySaxHandler(object):
    def start_element(self, name, attrs):
        print 'sax:start_element %s, attrs %s' % (name, attrs)

    def end_element(self, name):
        print 'sax:end_element %s' % name

    def char_data(self, text):
        print 'sax:char_data %s' % text

def Test():
    handler = mySaxHandler()
    parser = ParserCreate()
    parser.returns_unicode = True
    parser.StartElementHandler = handler.start_element
    parser.EndElementHandler = handler.end_element
    parser.CharacterDataHandler = handler.char_data
    xml = r'<?xml version="1.0"?>'
    xml = xml + r'<root>'
    xml = xml + 'data'
    xml = xml + r'</root>'
    parser.Parse(xml)

2、HTML本质上是XML的子集，但是HTML的语法没有XML那么严格，所以不能用标准的DOM或SAX来解析HTML。Python提供了HTMLParser来非常方便地解析HTML。下面是我的学习代码：

from HTMLParser import HTMLParser
from htmlentitydefs import name2codepoint
from collections import namedtuple

#会议类
class Meeting(object):
    def __init__(self, name, time, address):
        #会议名称
        self.name = name
        #会议时间
        self.time = time
        #会议地址
        self.address = address

#解析标签元组类
Tag = namedtuple("Tag", ["name", "property", "value"])

#用于解析网站https://www.python.org/events/python-events/中包含的会议信息
class PythonMeetingParser(HTMLParser):
    def myInit(self): 
        #会议列表
        self.mList = []
        #当前会议对象
        self.m = None
        #解析标签对象列表
        self.tags = []
        #标题标签对象
        self.tags.append(Tag('h3', 'class', 'event-title')) 
        #时间标签对象
        self.tags.append(Tag('time', None, None))
        #地址标签对象
        self.tags.append(Tag('span', 'class', 'event-location'))
        #标题内容标签对象
        self.tags.append(Tag('a', None, None))
        #当前标签类型 0-标题 1-时间 2-地址 3-标题内容
        self.iTag = None
        self.iTag1 = None

    #判断当前标签类型
    def getTagIndex(self, tag, attrs):
        result = -1
        for t in self.tags:
            result = result + 1
            if t.name == tag.lower():
                if t.property != None and t.value != None:
                    for att in attrs:
                        if att[0].lower() == t.property and att[1].lower() == t.value:
                            return result
                else:
                    return result
        return None

    def handle_starttag(self, tag, attrs):
        self.iTag1 = self.getTagIndex(tag, attrs)
        if self.iTag1 != 3:
            self.iTag = self.iTag1

    def handle_endtag(self, tag):
        pass

    def handle_startendtag(self, tag, attrs):
        pass

    def handle_data(self, data):
        if self.iTag != None and data != None:
            if self.iTag == 0 and self.iTag1 == 3:
                if self.m == None:
                    self.m = Meeting(None, None, None)
                if data.replace('\n', '').replace(' ', '') != '':
                    self.m.name = data.replace('\n', ' ')
            if self.iTag == 1:
                if self.m == None:
                    self.m = Meeting(None, None, None)
                if data.replace('\n', '').replace(' ', '') != '':
                    self.m.time = data.replace('\n', ' ').replace('\xe2', '-').replace('\x80', '-').replace('\x93', '-')
            if self.iTag == 2:
                if self.m == None:
                    self.m = Meeting(None, None, None)
                if data.replace('\n', '').replace(' ', '') != '':
                    self.m.address = data.replace('\n', ' ')
            if self.m != None and self.m.name != None and self.m.time != None and self.m.address != None:
                self.mList.append(self.m)
                self.m = None

    def handle_comment(self, data):
        pass

    def handle_entityref(self, name):
        pass

    def handle_charref(self, name):
        pass

#测试入口
def Test():
    parser = PythonMeetingParser()
    #接收一个待解析的文件名（手动复制上述网站的HTML源码保存到本地即可）
    s = raw_input("input a filename:")
    try:
        with open(s, 'r') as f:
            ss = f.read()
            parser.myInit()
            parser.feed(ss)
            c = len(parser.mList)
            if c > 0:
                print 'parser meeting count:', c
                print 'No.\tName\tTime\tAddress'
                i = 1
                for mm in parser.mList:
                    print '%d\t%s\t%s\t%s' % (i, mm.name, mm.time, mm.address)
                    i = i + 1
            else:
                print 'no metting'
    except BaseException, e:
        print 'Error on read file %s: %s' % (s, e.message)

注：HTMLParser.feed可连续调用多次，也就是不一定一次把整个HTML字符串都塞进去，可以一部分一部分塞进去。

今天就学习到这里啦，下一节从常用第三方模块学起。