def test_queries(test_label, query_input_file, query_output_file):
"""Test checker for query inputs (from query_input_file).
Write the output to query_output_file.
For sentences, capitalize the first word and add a period at the
end to the suggestions.
Arguments:
- `test_label`:
- `query_input_file`:
- `query_output_file`:
"""
query_list = get_inputs(test_label, query_input_file)
checker = spell_checker.SpellChecker()
if sys.argv[1] == 'dummy-test':
checker.get_posterior_fn = dummy_posterior_fn
checker.run_spell_check(query_list)
suggestion_dict = checker.get_suggestion_dict()
# print 'suggestion_dict', suggestion_dict
# TODO: Remove this
save_memoize_table()
# Note: The query_list must be sent and not suggestion_dict.keys()
# cos the original order of queries must be maintained
write_outputs(test_label, query_list, suggestion_dict,
query_output_file)
EF1: {5}\tNum queries: {2}\n'.format(
str(datetime.now()), test_label, len(query_list), *stats)
print 'stats_str', stats_str
f.write(stats_str)
f.close()
if __name__ == '__main__':
commandline_args_str = 'Format: ' + sys.argv[0] + ' arg\n' + 'arg = run-test: run all tests and write to results file.\n' + 'arg = calc-stats: calculate stats from results in file.\n'
# test_labels = ['phrases']
test_labels = ['words', 'phrases', 'sentences']
if len (sys.argv) != 2:
print 'commandline_args_str', commandline_args_str
exit (0)
elif sys.argv[1] in ['run-test', 'dummy-test']:
for test_label in test_labels:
test_queries(test_label,
'../data/' + test_label + '.input',
'../data/' + test_label + '.output')
save_memoize_table()
elif sys.argv[1] == 'calc-stats':
for test_label in test_labels:
calc_stats(test_label,
'../data/' + test_label + '.output',
'../data/' + test_label + '.tsv',
'../data/' + test_label + '.stats')
save_memoize_table()
else:
print 'commandline_args_str', commandline_args_str
exit (0)