在 Python 中解析 XML 但忽略实体

Question

I have to pase an XML file and store data into a database. The problem is this XML have some entity that I don't wanto to import but, instead, I want the raw entity tag. To clarify better I have the following schema:

<!ENTITY exa "example">
.....
<mytag>&exa;</mytag>

If I try to parse the above code reading the tag "mytag" using the folliwng code:

import xml.etree.ElementTree as ET

tree = ET.parse(xmlfile)
root = tree.getroot()

for item in root:
        if item.tag == "mytag":

I read the string "example". Instead I want to have the tag "exa". I guess is possible but cause I'm new t python delelompent I can not find the right way to get this result. Some suggestions? Thank you

Answer 1

Here is an example to start:

import os
import re
from lxml import etree

xmlfile = 'testfile.xml'
xml_path = '%s/%s' % (os.path.dirname(os.path.realpath(__file__)), xmlfile)

parser = etree.XMLParser(resolve_entities=False)
tree = etree.parse(xml_path, parser)
# root = tree.getroot()

root = tree.xpath('/mytag')

for item in root:
    entity = etree.tostring(item, pretty_print=True).decode('utf-8')
    print('ENTITY     : ', entity)
    entity_value = re.findall(r'&(.*?);', entity)
    print('Parsed str : ', entity_value)

But there may be a simpler way to recover the value.

Answer 2

You can modify each of the ENTITY tags in the xml file so that they have the values you want in them and then modify them back at the end.

You could create a class that clones your xml file:

import os
import re

class NoEntities:
    """
    Creates a clone of the target xml file such that the <!ENTITY x "y"> tags
    become <!ENTITY x "x">.
    """

    def __init__(self, xmlFile):
        self.targetName = xmlFile
        self.tmpName = 'temp.xml'

    def __enter__(self):
        match = r'<!ENTITY\s+(\S+)\s+"[^"]+"\s*>'
        replace = r'<!ENTITY \1 "\1">'

        with open(self.targetName) as target:
            with open(self.tmpName, 'w') as tmp:
                tmp.writelines(
                    re.sub(match, replace, line)
                    for line in target
                )

        return self.tmpName

    def __exit__(self, *exec_info):
        os.remove(self.tmpName)

And then use it inside a with block:

import xml.etree.ElementTree as ET

with NoEntities(pathToOriginalXml) as noEntityXml:
    tree = ET.parse(noEntityXml)
    # Do what you like...