Parsing XML using SAX

Fredrik Lundh fredrik at pythonware.com
Mon Oct 23 16:58:22 CEST 2006


Nathan Harmston wrote:

> However it is quite messy at the moment. The main reason is that Python
> doesnt have a switch statement.

a switch statement wouldn't save you a single line, so I find it a bit hard
to believe that it's the main reason...

> def startElement(self,name,attributes):
>                if name == "sbml":
>                        s = Sbml(attributes['xmlns'], attributes['version'],
> attributes['level'])
>                        self.sbmlDict['sbml'] = s
>                elif name == "model":
>                        m = Model(attributes['id'], attributes['name'])
>                        self.sbmlDict['model'] = m
>                elif name == "listOfCompartments":
>                        self.inListOfCompartments = bool(1)
>                elif name == "compartment" and self.inListOfCompartments:
>                        c = Compartment(attributes['id'],
> attributes['name'])
>                        self.tempList.append(c)
> .......................................snip
>
> I would use a dictionary for this, but this would require the use of many
> extra methods for each tag name, and this would lead to clutter aswell. Does
> anyone have any suggestions for reducing the number of lines and making my
> code look neater than a large amount of methods or elif statements.

forget about SAX and silly state machines, and use a modern XML library;
here's some ElementTree inspiration:

    http://effbot.org/zone/element-iterparse.htm#incremental-decoding

there are several other newer libraries that provide built-in data binding support,
including lxml.objectify and Amara.

</F> 






More information about the Python-list mailing list