def main(output_file, input_files):
combined = {}
for filename in input_files:
with open(filename, "r") as fid:
for line in fid:
dr = DetectorResult.from_json(line.strip())
if dr.filename in combined:
if dr.dialect == combined[dr.filename].dialect:
# allow it if the dialect is the same
continue
else:
raise KeyError("Duplicate result for file: %s" %
dr.filename)
combined[dr.filename] = dr
with open(output_file, "w") as fid:
for filename in sorted(combined.keys()):
dr = combined[filename]
dr.original_detector = dr.detector
dr.detector = "reference"
fid.write(dr.to_json() + "\n")
def load_detector_results(result_file):
"""
Load the results from a given detector result file. Verify each record in
the process.
"""
detector_names = set()
results = {}
with open(result_file, "r") as fid:
for idx, line in enumerate(fid.readlines()):
try:
record = DetectorResult.from_json(line.strip())
except json.JSONDecodeError:
print(
"\nError parsing the following record in file (line %i): "
"%s\n---\n%s" % (idx + 1, result_file, line.strip()))
raise SystemExit(1)
detector_names.add(record.detector)
fname = record.filename
if not os.path.isabs(fname):
fname = os.path.abspath(fname)
record.filename = fname
if fname in results:
raise ValueError(
"Duplicate result for file %s in detector file %s" %
(record.filename, result_file))
record.validate()
results[fname] = record
if len(detector_names) > 1:
raise ValueError("More than one detector name in file: %s" %
result_file)
detector = detector_names.pop()
return detector, results