Fast convert big xml to cvs.

fast_xml2cvs.py

from lxml import etree
from time import time

st = time()
c = 0

order_pattern = [
    '1st_column',
    '2nd_column',
    '3rd_column',
    ...
    'n_column',
]


el_holder = ['', ] * len(order_pattern)

with open('out.csv', 'w') as f:
    for event, element in etree.iterparse('in.xml', tag='T'):
        # out_str = ','.join([child.tag for child in element]) + '\n'
        # print(out_str)
        h = el_holder[:]
        for child in element:

            try:
                if child.text:
                    if h[order_pattern.index(child.tag)]:
                        h[order_pattern.index(child.tag)] += '||' + child.text.strip()
                    else:
                        h[order_pattern.index(child.tag)] = child.text.strip()
            except (ValueError, TypeError):
                continue

        f.write(','.join(h) + '\n')
        element.clear()
        while element.getprevious() is not None:
            del element.getparent()[0]
        c += 1

print('Processed: {0}\tTime: {1}'.format(c, time() - st))