my_xml

Help module to parse a simple XML buffer and store it as a read-only (mostly) dictionary-type object (MyXml). This dictionary can hold other dictionaries, nodes-lists, or leaf nodes. Access to the nodes is by using attributes.

xml = parse("Val") xml.Foo.Bar == "Val" True xml.Foo.Bar Val

I don't like to use the built in Python DOM parsers for simple XML data, but this module is good only for simple XML! No name-spaces, CDATA and other fancy features are supported.

There are three factory functions, "parse", "parse_file" and "parse_object".

• parse takes an XML string and builds MyXml object from it.

• parse_file takes a file name reads it and do the same.

Both functions take an optional list of tags names from the beginning of the XML data, to ignore.

• parse_object takes a complex python object (of dictionaries, sequences and scalars) and creates MyXml object from it.

It is possible, but not convenient, to construct an XML trees using this module.

Usage Examples:

xml = parse(''' ... <?xml bla bla bla> ... ...

... One Two & Three ... ... ... ... ... Bla Bla Bla ... ... No ... Value ... ... ''')

• An XML node is an attribute of the MyXml object

xml.Main.Text One Two & Three

• And also

xml.Main.Text == "One Two & Three" True

xml.Main.Text.value == "One Two & Three" True

There is also a way to access a node with "nd_" prefix (so we can access python reserved words), this will also return EMPY_NODE if the node doesn't exists.

xml.nd_Main.nd_Text One Two & Three

• A node can be looked at as a list with one item

xml.Main.Double.Double[0] is xml.Main.Double.Double True

• Nodes Lists are regular lists

len(xml.Main.List.Item) 3 unicode(xml.Main.List.Item[2]) u'Bla Bla Bla'

• MyXml object is a dictionary

xml["Main"]["Text"] == xml.Main["Text"] True xml.Main.get("Text") == xml["Main"].Text True

• There is also a very simple XPath-like method

xml.xpath("Main/List/Item")[2] Bla Bla Bla

• Attributes can be accessed with an "at_" prefix

xml.Main.List.Item[1].at_ccc u'ab+c'

• Access the attributes dictionary with "at_dict"

xml.Main.List.Item[0].at_dict["aaa"] u'bbb'

• Every value can be looked at as a number and a boolean

xml.Main.BoolNum.boolean False

• Also attribute can be looked at as booleans or numbers

xml.Main.BoolNum.at_num.number * 2 7.0 xml.xpath("Main/BoolNum").at_bool.boolean True

• But if the value is not a number or boolean (yes, no, true, false, 1, 0) the
• return value is None

xml.Main.List.Item[0].at_aaa.number

• "get" and "xpath" return an empty node by default, so we can still use the
• number/boolean attributes.

bool(xml.get("foo").boolean) False

xml.xpath("Main/foo").number is None True

• Printing MyXml objects keeps the original order and adds indentation.
• The indentation is not thread safe though.

print xml.Main.List

Bla Bla Bla

• Constructing MyXml object from a python complex object:

xml = parse_object({ ... "foo1": "bar", ... "foo2": ["bar1", "bar2", "bar3"], ... "foo3": {"bar": "foo"}, ... "foo4": 5 ... }, "Main") # "Main" is the name of the top most node

xml.xpath("Main/foo4").number 5

• The names of the nodes that hold a sequence items, are the type name of the
• sequence (list, tuple, set, generator).

xml.xpath("Main/foo2/list")[1] == "bar2" True

• Finally - not very useful - but you can modify MyXml object

add_returns_self = xml.add(MyNode("bar5", "foo5")) # MyNode(value, name) xml.foo5.at_dict["attr"] = "attr value" xml.xpath("Main/foo5").at_attr == "attr value" True

One can also use the other built in dictionary and list methods, but this is not recommended

xml # Here the order is not preserved because of the python dictionary

5 bar bar1 bar2 bar3 foo bar5

Please note that this module is not efficient in parsing large XML buffers. It uses string slicing heavily.

Erez Bibi

Please send comments and questions to erezbibi AT users DOT sourceforge DOT net