Das deutsche Python-Forum

Code: Alles auswählen

from xml.dom.minidom import parse,parseString
from urllib2         import urlopen
from os              import open

def xmlopen(url,enc=None):

	if url[:7].lower() == "http://":
		xmlfd = urlopen(url)
	else:
		xmlfd = open(url,"r")
	if enc == None:
		xmlhl = xmlhead(xmlfd)
		enc = xmlhl['encoding']
		# leider laest sich das bearbeitete xml danach nicht mehr parsen weswegen es neu goeffnet werden muss 
		xmlfd.close()
		xmlfd = urlopen(url)
	xmldoc = parse(xmlfd)
	xmlfd.close()

	return bibxml(xmldoc,enc)

def xmlhead(xmlfd): # Auswerten der Header-Variablen

	xmlc = ""
	xmlhead = ""
	while xmlc != ">":
		xmlhead = xmlhead + xmlc
		xmlc = xmlfd.read(1)
	xmlhead = xmlhead + " />"
	xmlhead = xmlhead.replace("?","")

	attribl = {} 
	head = parseString(xmlhead)
	for attribut in head.childNodes[0].attributes.keys():
		attribl[head.childNodes[0].attributes[attribut].name.encode()] = head.childNodes[0].attributes[attribut].value.encode()
	return attribl

def bibxml(xmldoc,enc):

	name = ""
	bibnodes = {}

	for childnode in xmldoc.childNodes:

		if childnode.nodeType == 1: # Node ist ein Tag
			name = childnode.nodeName.encode(enc)
			# Inhalt
			if childnode.hasChildNodes():
				value = bibxml(childnode,enc)
			elif childnode.nodeValue == None:
				value = childnode.nodeValue
			else:
				value = childnode.nodeValue.encode(enc)

			# Attribute
			if childnode.attributes.keys() != []:
				attribbib = {}
				for attribut in childnode.attributes.keys():
					aname  = childnode.attributes[attribut].name.encode(enc)
					avalue = childnode.attributes[attribut].value.encode(enc)
					attribbib[aname] = avalue
				if value != None:
					value = [{'attribut':attribbib},value]
				else:
					value = {'attribut':attribbib}

			# Eintragen von einen bibliotheks-Eintrag ( zugegeben -> geklaut ;P )
			if value != None:
				if name in bibnodes: 
					if isinstance(bibnodes[name], dict):
						# there is multiple instances of this node - convert to list
						bibnodes[name] = [bibnodes[name]]
					bibnodes[name].append(value)
				else:
					bibnodes[name] = value

		if ((childnode.nodeType == 3) or (childnode.nodeType == 4)): # Node ist ein Text
			if childnode.nodeValue != None:
				value = childnode.nodeValue.encode(enc).replace("\n","")
				value = value.replace("\t","")
				value = value.replace(" ","")
				if value == "":
					continue
			bibnodes = childnode.nodeValue.encode(enc)

	return bibnodes