def harmonization_extract_worker(args):
filename = args[0]
name_queue = args[1]
try:
tree = extract.parse_xml(filename)
harmonized = {}
harmonized['spl_set_id'] = extract.extract_set_id(tree)
harmonized['id'] = extract.extract_id(tree)
harmonized['spl_version'] = extract.extract_version_number(tree)
harmonized['is_original_packager'] = extract.is_original_packager(tree)
harmonized['ProductNDCs'] = extract.extract_product_ndcs(tree)
harmonized['OriginalPackagerProductNDSs'] = \
extract.extract_original_packager_product_ndcs(tree)
harmonized['PackageNDCs'] = extract.extract_package_ndcs(tree)
name_queue.put(harmonized)
except Exception as inst:
print filename + 'has a problem'
print inst
def _extract(self, filename):
''' Moving this code from the `spl_harmonization` file, since it is the
only part of that file that is needed now that we have converted to a
map reduction.`
'''
try:
tree = extract.parse_xml(filename)
harmonized = {}
harmonized['spl_set_id'] = spl.extract.extract_set_id(tree)
harmonized['id'] = spl.extract.extract_id(tree)
harmonized['spl_version'] = spl.extract.extract_version_number(tree)
harmonized['is_original_packager'] = \
spl.extract.is_original_packager(tree)
harmonized['spl_product_ndc'] = spl.extract.extract_product_ndcs(tree)
harmonized['original_packager_product_ndc'] = \
spl.extract.extract_original_packager_product_ndcs(tree)
harmonized['package_ndc'] = spl.extract.extract_package_ndcs(tree)
return harmonized
except:
logging.warn('ERROR processing SPL data: %s', filename)
return None
def _extract(self, filename):
''' Moving this code from the `spl_harmonization` file, since it is the
only part of that file that is needed now that we have converted to a
map reduction.`
'''
try:
tree = extract.parse_xml(filename)
harmonized = {}
harmonized['spl_set_id'] = spl.extract.extract_set_id(tree)
harmonized['id'] = spl.extract.extract_id(tree)
harmonized['spl_version'] = spl.extract.extract_version_number(tree)
harmonized['is_original_packager'] = \
spl.extract.is_original_packager(tree)
harmonized['spl_product_ndc'] = spl.extract.extract_product_ndcs(tree)
harmonized['original_packager_product_ndc'] = \
spl.extract.extract_original_packager_product_ndcs(tree)
harmonized['package_ndc'] = spl.extract.extract_package_ndcs(tree)
return harmonized
except:
logging.warn('ERROR processing SPL data: %s', filename)
return None
def harmonization_extract_worker(args):
filename = args[0]
name_queue = args[1]
try:
tree = extract.parse_xml(filename)
harmonized = {}
unii_list = []
intermediate = []
harmonized['unii'] = extract_unii.extract_unii(tree)
harmonized['set_id'] = extract_unii.extract_set_id(tree)
harmonized['name'] = extract_unii.extract_unii_name(tree)
# zipping together two arrays, since they came from the same xpath locations
# these are the NUI codes and their respective names
# we might be able to get the names from somewhere else and avoid the zip
intermediate = zip(extract_unii.extract_unii_other_code(tree),
extract_unii.extract_unii_other_name(tree))
header = ['number', 'name']
harmonized['va'] = [dict(zip(header, s)) for s in intermediate]
unii_list.append(harmonized)
name_queue.put(unii_list)
except Exception as inst:
print filename + 'has a problem'
print inst
def setUp(self):
self.tree = extract_spl.parse_xml(os.path.dirname(__file__) + "/data/METHOXSALEN.xml")
def setUp(self):
self.tree = extract_spl.parse_xml(os.path.dirname(__file__) +
'/data/METHOXSALEN.xml')
def setUp(self):
self.tree = extract.parse_xml(data_filename('lipitor-orig.xml'))
def setUp(self):
self.tree = extract.parse_xml(data_filename('no-title.xml'))
def setUp(self):
self.tree = extract.parse_xml(data_filename('cough.xml'))
def setUp(self):
self.tree = extract.parse_xml(data_filename('humira.xml'))