使用Python数据结构表示这个的最佳方法
<specification>
<propertyName> string </propertyName>
<value>
<number>
<value> anyNumberHere </value>
</number>
<text>
<value> anyTextHere </value>
</text>
<URL>
<value> anyURIHere </value>
</URL>
</value>
<!-- ... 1 or more value nodes here ... -->
</specification>
<!-- ... 1 or more specification nodes here ... -->
我需要构建一个API请求，用户会传入这些值。那么，怎样表示这些值才能让使用这个方法的用户更容易传递相应的值呢？
我在想：
列表的列表的字典：
[specification1, specification2, specification3]
这里：
specification1= [value1, value2]
这里：
value1 = {number:anyNumberHere, text:anyTextHere, URL:anyURIHere}
value2 = {number:anyNumberHere, text:anyTextHere, URL:anyURIHere}
但是，我在这里无法包含：<propertyName>。有什么建议吗？
而且，这听起来太复杂了。我们能像在Java中那样使用对象封装吗？我知道可以，但我很好奇，在Python中推荐的方式是什么？
My logic for now, suggestions (incorrect due to propertyName):
#specification is a List of List of Dicts
for spec in specification:
specification_elem = etree.SubElement(root, "specification")
propertyName_elem = etree.SubElement(specification_elem,"propertyName")
propertyName_elem.text = spec_propertyName
for value in spec:
value_elem = etree.SubElement(specification_elem, "value")
for key in value:
key_elem = etree.SubElement(value_elem, key)
keyValue_elem = etree.SubElement(key_elem, "value")
keyValue_elem.text = value[key]
在这里，我会把spec_propertyName作为一个不同的参数传递。所以，用户将传入：specification和spec_propertyName
3 个回答
0
如果属性名称是唯一的，你可以用一个字典来存储字典的列表。如果属性名称不是唯一的，你就可以用一个列表来存储列表的列表，再加上字典。
如果这些列表让你觉得很难管理，你可以创建一个类来存储这些数据：
class Specification:
def __init__(self):
self.propertyName = ""
self.values = []
然后，使用：
spec = Specification()
spec.propertyName = "string"
value = {"URL":"someURI", "text":"someText"}
spec.values.append(value)
0
这里介绍了一种使用 命名元组 的表示方法。你可以随时升级到使用类（这样可以添加一些代码来验证用户输入的内容，并且允许在可选字段中省略 field=None），这样对用户接口没有其他影响，构建 ElementTree 的代码也几乎不需要改动。
# -*- coding: cp1252 -*-
from collections import namedtuple
import xml.etree.cElementTree as etree
Specifications = namedtuple('Specifications', 'specification_list')
Specification = namedtuple('Specification', 'propertyName value_list')
Value = namedtuple('Value', 'number text url')
def make_etree(specifications, encoding):
"""
Convert user's `specifications` to an ElementTree.
`encoding` is encoding of *input* `str` objects.
"""
def ensure_unicode(v):
if isinstance(v, str): return v.decode(encoding)
if isinstance(v, unicode): return v
return unicode(v) # convert numbers etc to unicode strings
root = etree.Element('specifications')
for spec in specifications.specification_list:
specification_elem = etree.SubElement(root, "specification")
propertyName_elem = etree.SubElement(specification_elem, "propertyName")
propertyName_elem.text = ensure_unicode(spec.propertyName)
for value in spec.value_list:
value_elem = etree.SubElement(specification_elem, "value")
for key in value._fields:
kv = getattr(value, key)
if kv is None: continue
key_elem = etree.SubElement(value_elem, key)
keyValue_elem = etree.SubElement(key_elem, "value")
keyValue_elem.text = ensure_unicode(kv)
return etree.ElementTree(root)
# === sample caller code follows ===
specs = Specifications(
specification_list=[
Specification(
propertyName='a prop',
value_list=[
Value(
number=42,
text='universe',
url='http://uww.everywhere',
),
Value(
number=0,
text=None, # optional
url='file:///dev/null',
),
],
),
Specification(
propertyName='b prop',
value_list=[
Value(
number=1,
text='Üñîçøðè', # str object, encoded in cp1252
url=u'Üñîçøðè', # unicode object
),
],
),
],
)
print repr(specs); print
import sys
tree = make_etree(specs, 'cp1252')
import cStringIO
f = cStringIO.StringIO()
tree.write(f, encoding='UTF-8', xml_declaration=True)
print repr(f.getvalue())
print
输出（在第80列处折叠）：
Specifications(specification_list=[Specification(propertyName='a prop', value_li
st=[Value(number=42, text='universe', url='http://uww.everywhere'), Value(number
=0, text=None, url='file:///dev/null')]), Specification(propertyName='b prop', v
alue_list=[Value(number=1, text='\xdc\xf1\xee\xe7\xf8\xf0\xe8', url=u'\xdc\xf1\x
ee\xe7\xf8\xf0\xe8')])])
"<?xml version='1.0' encoding='UTF-8'?>\n<specifications><specification><propert
yName>a prop</propertyName><value><number><value>42</value></number><text><value
>universe</value></text><url><value>http://uww.everywhere</value></url></value><
value><number><value>0</value></number><url><value>file:///dev/null</value></url
></value></specification><specification><propertyName>b prop</propertyName><valu
e><number><value>1</value></number><text><value>\xc3\x9c\xc3\xb1\xc3\xae\xc3\xa7
\xc3\xb8\xc3\xb0\xc3\xa8</value></text><url><value>\xc3\x9c\xc3\xb1\xc3\xae\xc3\
xa7\xc3\xb8\xc3\xb0\xc3\xa8</value></url></value></specification></specification
s>"
6
怎么样，如果我们有一个包含多个规范对象的列表，每个规范对象都有一个 property_name 和一个 value 字典的列表呢？（values 也可以是一个对象的列表，而不是字典。）
举个例子：
class Value(object):
def __init__(self,
number=None,
text=None,
url=None):
self.number = number
self.text = text
self.url = url
class Specification(object):
def __init__(self, propertyName):
self.propertyName = propertyName
self.values = []
spec1 = Specification("Spec1")
spec1.values = [Value(number=3,
url="http://www.google.com"),
Value(text="Hello, World!",
url="http://www.jasonfruit.com")]
spec2 = Specification("Spec2")
spec2.values = [Value(number=27,
text="I can haz cheezburger?",
url="http://stackoverflow.com"),
Value(text="Running out of ideas.",
url="http://news.google.com")]