File size: 9,688 Bytes
778d47d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 | import os
import argparse
import json
import re
from utils.db_utils import check_sql_executability
def extract_sql_in_code_block(pred_sql_text):
"""
Extracts the SQL query from a text block that contains code block marked by triple backticks (```sql ... ```).
Args:
pred_sql_text (str): The input text that may contain a SQL code block.
Returns:
str: The extracted SQL query or an empty string if no SQL code block is found.
"""
# Use regex to search for the SQL code block enclosed in triple backticks
sql_block_match = re.search(r"```sql\s+(.+?)\s+```", pred_sql_text, re.DOTALL)
if sql_block_match:
# Extract the SQL query from the matched block
sql_query = sql_block_match.group(1).strip()
return sql_query
else:
return pred_sql_text
parser = argparse.ArgumentParser()
parser.add_argument('--pred_file', type=str, default='data/evaluate/phi-3-planner_spider_dev.jsonl')
parser.add_argument('--planner_only', action='store_true')
args = parser.parse_args()
if "spider_dev" in args.pred_file:
args.dev_file = "data/sft_data_collections/spider/dev.json"
elif "spider_dk" in args.pred_file:
args.dev_file = 'data/sft_spider_dk_text2sql.json'
elif "spider_realistic" in args.pred_file:
args.dev_file = 'data/sft_spider_realistic_text2sql.json'
elif "spider_syn" in args.pred_file:
args.dev_file = 'data/sft_spider_syn_text2sql.json'
elif "dr_spider" in args.pred_file:
args.dev_file = 'data/sft_dr_spider_text2sql.json'
elif "bird" in args.pred_file and 'dev' in args.pred_file:
args.dev_file = "data/full_value_matching_schema_insight_bird_062024_with_evidence_dev_text2sql.json"
elif "bird_train" in args.pred_file:
args.dev_file = "data/full_value_matching_schema_insight_bird_062024_with_evidence_train_text2sql.json"
# Load dev_file to get the correct order of (db_id, question) keys
with open(args.dev_file) as dev_fp:
dev_data = json.load(dev_fp)
# Create a list of keys from dev_file to sort by
dev_keys = [f"{sample['db_id']} {sample['question']}" for sample in dev_data]
def get_final_sql(predict_sql, fixed_sql, db_path):
sqls_priority = [fixed_sql, predict_sql]
sqls_priority = [sql for sql in sqls_priority if sql is not None]
for sql in sqls_priority:
# check if the sql is executable
execution_error = check_sql_executability(sql, db_path)
if execution_error is None:
return sql
return predict_sql
def get_predict_sqls():
# Load predictions and store them using the same (db_id, question) key
predicted_sqls_dict = {}
with open(args.pred_file) as pred_fp:
for line in pred_fp:
sample = json.loads(line)
db_id = sample.get('db_id')
question = sample.get('question')
key = f"{db_id} {question}"
predict_sqls = []
if not args.planner_only:
predict_sqls = [x for x in sample.get('fixed_sqls', []) if x is not None]
predict_sqls = [extract_sql_in_code_block(x) for x in predict_sqls]
predict_sqls.extend(sample['predict_sqls'])
if 'final_sql' in sample:
predict_sqls = [sample['final_sql']]
elif 'fixed_sqls' in sample:
pair_sqls = [(x, y) for x, y in zip(sample['predict_sqls'], sample['fixed_sqls'])]
predict_sqls = [get_final_sql(x, y, sample['db_path']) for x, y in pair_sqls]
predict_sqls = [re.sub(r'\border\b(?!\s+BY)', r'`order`', query).replace("``", "`") for query in predict_sqls]
pred_sql = None
for sql in predict_sqls:
execution_error = check_sql_executability(sql, sample["db_path"])
if execution_error is not None:
if "syntax error" in execution_error:
print(sql)
continue
pred_sql = sql
break
if pred_sql is None:
pred_sql = predict_sqls[0]
pred_sql = pred_sql.replace("\n", " ").strip()
if len(pred_sql.strip()) == 0:
pred_sql = "SELECT;"
# Store the prediction in the dictionary using the key
predicted_sqls_dict[key] = (pred_sql, sample['db_id'])
return predicted_sqls_dict
if 'spider' in args.pred_file:
predicted_sqls_dict = get_predict_sqls()
predicted_sqls = [predicted_sqls_dict.get(key, ("SELECT", ""))[0] for key in dev_keys]
elif 'bird' in args.pred_file:
predicted_sqls_dict = get_predict_sqls()
bird_results_dict = {}
for idx, key in enumerate(dev_keys):
pred_sql, db_id = predicted_sqls_dict.get(key, ("SELECT", ""))
bird_results_dict[idx] = pred_sql + "\t----- bird -----\t" + db_id
if "bird" in args.pred_file:
with open("predict_dev.json", "w", encoding = 'utf-8') as f:
f.write(json.dumps(bird_results_dict, indent = 2, ensure_ascii = False))
os.system("sh bird_evaluation/run_evaluation.sh predict_dev.json")
elif "spider_dev" in args.pred_file:
with open("pred_sqls.txt", "w+", encoding = 'utf-8') as f:
for sql in predicted_sqls:
f.write(sql + "\n")
print("Execution accuracy:")
os.system('python -u test_suite_sql_eval/evaluation.py --gold ./data/sft_data_collections/spider/dev_gold.sql --pred pred_sqls.txt --db ./data/sft_data_collections/spider/database --etype exec')
with open("pred_sqls.txt", "w+", encoding = 'utf-8') as f:
for sql in predicted_sqls:
f.write(sql + "\n")
print("Test suit execution accuracy:")
os.system('python -u test_suite_sql_eval/evaluation.py --gold ./data/sft_data_collections/spider/dev_gold.sql --pred pred_sqls.txt --db test_suite_sql_eval/test_suite_database --etype exec')
elif "spider_dk" in args.pred_file:
with open("pred_sqls.txt", "w", encoding = 'utf-8') as f:
for sql in predicted_sqls:
f.write(sql + "\n")
print("Execution accuracy:")
os.system('python -u test_suite_sql_eval/evaluation.py --gold ./data/sft_data_collections/Spider-DK/dk_gold.sql --pred pred_sqls.txt --db ./data/sft_data_collections/spider/database --etype exec')
elif "spider_realistic" in args.pred_file:
with open("pred_sqls.txt", "w", encoding = 'utf-8') as f:
for sql in predicted_sqls:
f.write(sql + "\n")
print("Execution accuracy:")
os.system('python -u test_suite_sql_eval/evaluation.py --gold ./data/sft_data_collections/spider-realistic/realistic_gold.sql --pred pred_sqls.txt --db ./data/sft_data_collections/spider/database --etype exec')
with open("pred_sqls.txt", "w", encoding = 'utf-8') as f:
for sql in predicted_sqls:
f.write(sql + "\n")
print("Test suit execution accuracy:")
os.system('python -u test_suite_sql_eval/evaluation.py --gold ./data/sft_data_collections/spider-realistic/realistic_gold.sql --pred pred_sqls.txt --db test_suite_sql_eval/test_suite_database --etype exec')
elif "spider_syn" in args.pred_file:
with open("pred_sqls.txt", "w", encoding = 'utf-8') as f:
for sql in predicted_sqls:
f.write(sql + "\n")
print("Execution accuracy:")
os.system('python -u test_suite_sql_eval/evaluation.py --gold ./data/sft_data_collections/Spider-Syn/Spider-Syn/syn_dev_gold.sql --pred pred_sqls.txt --db ./data/sft_data_collections/spider/database --etype exec')
with open("pred_sqls.txt", "w", encoding = 'utf-8') as f:
for sql in predicted_sqls:
f.write(sql + "\n")
print("Test suit execution accuracy:")
os.system('python -u test_suite_sql_eval/evaluation.py --gold ./data/sft_data_collections/Spider-Syn/Spider-Syn/syn_dev_gold.sql --pred pred_sqls.txt --db test_suite_sql_eval/test_suite_database --etype exec')
elif "dr_spider" in args.pred_file:
test_set_names = [
# DB Schema related
'DB_schema_synonym',
'DB_schema_abbreviation',
'DB_DBcontent_equivalence',
# NLQ related
'NLQ_keyword_synonym',
'NLQ_keyword_carrier',
'NLQ_column_synonym',
'NLQ_column_carrier',
'NLQ_column_attribute',
'NLQ_column_value',
'NLQ_value_synonym',
'NLQ_multitype',
'NLQ_others',
# SQL related
'SQL_comparison',
'SQL_sort_order',
'SQL_NonDB_number',
'SQL_DB_text',
'SQL_DB_number',
]
#test_set_names = ['DB_schema_synonym', 'DB_schema_abbreviation', 'DB_DBcontent_equivalence']
raw_dataset = json.load(open(args.dev_file))
for test_set_name in test_set_names:
if test_set_name not in args.pred_file:
continue
print(test_set_name)
test_set_predicted_sqls = [predicted_sql for predicted_sql, raw_data in zip(predicted_sqls, raw_dataset) if raw_data["source"] == "dr.spider-" + test_set_name]
database_file_path = "database_post_perturbation" if test_set_name.startswith("DB_") else "databases"
db_path = os.path.join("./data/sft_data_collections/diagnostic-robustness-text-to-sql/data/", test_set_name, database_file_path)
gold_path = os.path.join("./data/sft_data_collections/diagnostic-robustness-text-to-sql/data/", test_set_name, "gold_post_perturbation.sql")
with open("pred_sqls.txt", "w", encoding = 'utf-8') as f:
for sql in test_set_predicted_sqls:
f.write(sql + "\n")
print("Execution accuracy:")
os.system('python -u test_suite_sql_eval/evaluation.py --gold {} --pred pred_sqls.txt --db {} --etype exec'.format(gold_path, db_path))
|