def _split_configuration(self, projectfile, temp_dir):
num_pieces = multiprocessing.cpu_count()
tree = ET(file=projectfile)
num_files = len(tree.findall('./files/file'))
splitfiles = []
files_per_job = int(math.ceil(float(num_files)/num_pieces))
for idx in xrange(num_pieces):
tree = ET(file=projectfile)
root = tree.getroot()
start = idx*files_per_job
end = start + files_per_job
if end > num_files:
end = None
for elem in ('files', 'images', 'pages',
'file-name-disambiguation'):
elem_root = root.find(elem)
to_keep = elem_root.getchildren()[start:end]
to_remove = [x for x in elem_root.getchildren()
if not x in to_keep]
for node in to_remove:
elem_root.remove(node)
out_file = os.path.join(temp_dir,
"{0}-{1}.ScanTailor".format(
os.path.splitext(os.path.basename(
projectfile))[0], idx))
tree.write(out_file)
splitfiles.append(out_file)
return splitfiles
def dump(self, stream):
if self.prettyprint:
self.indent(self.xml)
document = ET(self.xml)
header = '<?xml version="1.0" encoding="%s"?>' % self.encoding
stream.write(header.encode(self.encoding))
document.write(stream, encoding=self.encoding)
`assin-eval.py` script.
"""
import argparse
from xml.etree.cElementTree import ElementTree as ET
import numpy as np
from collections import Counter
if __name__ == '__main__':
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument('train', help='XML file with training data')
parser.add_argument('test', help='XML file with test data')
parser.add_argument('output', help='Output tagged XML file')
args = parser.parse_args()
tree = ET()
root_train = tree.parse(args.train)
similarities_train = np.array(
[float(pair.get('similarity')) for pair in root_train])
similarity_avg = similarities_train.mean()
entailments_train = [pair.get('entailment') for pair in root_train]
entailment_counter = Counter(entailments_train)
majority_entailment, _ = entailment_counter.most_common(1)[0]
root_test = tree.parse(args.test)
for pair in root_test:
pair.set('similarity', str(similarity_avg))
pair.set('entailment', majority_entailment)
tree.write(args.output, 'utf-8')
def __call__(self, stream):
self.xml = ET(file=stream)
(keys, defaults) = self.find_graphml_keys(self.xml)
for g in self.xml.findall("{%s}graph" % self.NS_GRAPHML):
yield self.make_graph(g, keys, defaults)
# convenience script for interactive use:
# myougiden $ python3.2
# >>> from qjm import *
from myougiden import config
from xml.etree.cElementTree import ElementTree as ET
import gzip
et = ET()
jm = et.parse(gzip.open(config['paths']['jmdictgz'], 'r'))
from xml.etree.cElementTree import tostring
def tos(element):
return tostring(element).decode()