output_dir = args.output
use_resources_dump = args.continue_generation
time = datetime.datetime.today()
logging.basicConfig(filename='{}/generator_{:%Y-%m-%d-%H-%M}.log'.format(output_dir, time), level=logging.DEBUG)
resource_dump_file = output_dir + '/resource_dump.json'
resource_dump_exists = os.path.exists(resource_dump_file)
if (resource_dump_exists and not use_resources_dump):
warning_message = 'Warning: The file {} exists which indicates an error. Remove file or continue generation after fixing with --continue'.format(
resource_dump_file)
print warning_message
sys.exit(1)
reload(sys)
sys.setdefaultencoding("utf-8")
not_instanced_templates = collections.Counter()
used_resources = collections.Counter(json.loads(open(resource_dump_file).read())) if use_resources_dump else collections.Counter()
file_mode = 'a' if use_resources_dump else 'w'
templates = read_template_file(template_file)
try:
generate_dataset(templates, output_dir, file_mode)
except:
print 'exception occured, look for error in log file'
save_cache(resource_dump_file, used_resources)
else:
save_cache('{}/used_resources_{:%Y-%m-%d-%H-%M}.json'.format(output_dir, time), used_resources)
finally:
log_statistics(used_resources, SPECIAL_CLASSES, not_instanced_templates)
sys.setdefaultencoding("utf-8")
dataset_root, _ = os.path.splitext(dataset_file)
used_resources_root, _ = os.path.splitext(used_resources_file)
filtered_sparql_file = '{}_filtered_{:d}_{}.sparql'.format(
dataset_root, MINIMUM, COMP.__name__)
filtered_en_file = '{}_filtered_{:d}_{}.en'.format(dataset_root, MINIMUM,
COMP.__name__)
used_resources = collections.Counter(
json.loads(open(used_resources_file).read()))
filtered_resources = [
elem_cnt for elem_cnt in list(used_resources.items())
if elem_cnt[1] >= MINIMUM
]
save_cache('{}_filter_{:d}.json'.format(used_resources_root, MINIMUM),
collections.Counter(dict(filtered_resources)))
valid_encoded_resources = [
encode(elem_cnt1[0]) for elem_cnt1 in filtered_resources
]
check = lambda encoded_entity: encoded_entity in valid_encoded_resources
valid_lines = []
filtered_queries = []
with open(dataset_root + '.sparql', 'r') as sparql_file:
for linenumber, line in enumerate(sparql_file):
entities = extract_encoded_entities(line)
valid = COMP(list(map(check, entities)))
if valid:
filtered_queries.append(line)
valid_lines.append(linenumber)
