This visualization helped me understand what was going on better than the lines of code in the existing example. Perhaps it's also worth mentioning in the readme that `OrderedDict` is what's returned from from `parse` and we're serializing to json because `pprint` does not work with OrderedDict.
If false, it suppresses the XML declaration, as required by #55.
xmltodict becomes slow when you have an XML file with large texts. Enabling the parser.buffer_text option dramatically increases performance. Code used to benchmark: import xmltodict import time xml = "<root>" + ("a"*70+"\n")*10000 + "</root>" s=time.time() x=xmltodict.parse(xml) print(time.time() - s) # 19.9860811234 seconds without buffer_text # 0.059289932251 seconds with buffer_text # So 300 times faster