expat3.py

python web programming 部分

#!/usr/bin/python
import sys
from xml.parsers import expat

class GetElements:
    def __init__(self,name):
        self.name     = name
        self.parser   = expat.ParserCreate()
        self.parser.StartElementHandler = self.start_element
        self.parser.EndElementHandler = self.end_element
        self.parser.CharacterDataHandler = self.character_data

    def parse(self,file):
        self.results = [ ]
        self.currentname = None
        self.parser.ParseFile(file)
        return self.results
    def start_element(self,name,attrs):
        if name == self.name:
            self.currentattrs = attrs
            self.currentname = name
            self.currentdata = [ ]
            
    def character_data(self,data):
        if self.currentname == self.name:
            self.currentdata.append(data)
            
    def end_element(self,name):
        if name == self.currentname:
            self.results.append(("".join(self.currentdata),self.currentattrs))
            self.currentname = None

p = GetElements("item")
items = p.parse(open(sys.argv[1]))

for text,attr in items:
    num = attr.get("num",1)
    unit = attr.get("units","none")
    if unit == "none":
        unit = ""

    print "%4s %-8s  %s" % (num, unit, text)