Spaces:
Sleeping
Sleeping
File size: 9,219 Bytes
8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea 8d465b0 cc98bea |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 |
import os
import re
import gradio as gr
os.environ["TRANSFORMERS_NO_TF"] = "1"
os.environ["TRANSFORMERS_NO_FLAX"] = "1"
os.environ["HF_HUB_DISABLE_TELEMETRY"] = "1"
CRUD_VERB = {
"INSERT": "load",
"UPDATE": "changed",
"DELETE": "deleted",
"SELECT": "specified",
}
# ---------- helpers ----------
def plural(n: int, singular: str = "record", plural_word: str = "records") -> str:
return f"{n} {singular if n == 1 else plural_word}"
def detect_command(sql_text: str) -> str:
m = re.search(r"\b(INSERT|UPDATE|DELETE|SELECT)\b", sql_text, flags=re.IGNORECASE)
return m.group(1).upper() if m else "OTHER"
def parse_table_name(sql: str) -> str:
for pat in [
r"INSERT\s+INTO\s+([A-Za-z0-9\.\[\]_]+)",
r"UPDATE\s+([A-Za-z0-9\.\[\]_]+)",
r"DELETE\s+FROM\s+([A-Za-z0-9\.\[\]_]+)",
r"\bFROM\s+([A-Za-z0-9\.\[\]_]+)",
]:
m = re.search(pat, sql, flags=re.IGNORECASE)
if m:
return m.group(1)
return ""
def clean_statement(text: str) -> str:
text = re.sub(r"^(What|Which|How|Give|Provide)[^:]*:\s*", "", text, flags=re.IGNORECASE).strip()
if text and text[-1] not in ".!?":
text += "."
return text[0].upper() + text[1:] if text else text
def infer_in_list_count(where_clause: str) -> None:
if not where_clause:
return None
m = re.search(r"\bIN\s*\(\s*([^)]+?)\s*\)", where_clause, flags=re.IGNORECASE | re.DOTALL)
if not m:
return None
items = [x.strip() for x in m.group(1).split(",") if x.strip()]
return len(items) if items else None
# ---------- deterministic CRUD summaries ----------
def summarize_insert(sql: str) -> str:
table = parse_table_name(sql)
cols_match = re.search(r"\(\s*([^)]+?)\s*\)\s*VALUES", sql, flags=re.IGNORECASE | re.DOTALL)
cols = []
if cols_match:
cols = [c.strip().strip("[]") for c in cols_match.group(1).split(",")]
try:
uid_idx = cols.index("user_id")
except ValueError:
uid_idx = None
tuples = re.findall(r"VALUES\s*\(\s*([^)]+?)\s*\)", sql, flags=re.IGNORECASE | re.DOTALL)
if not tuples:
tuples = re.findall(r"\(\s*([^)]+?)\s*\)", sql, flags=re.IGNORECASE)
user_ids = []
for tup in tuples:
parts = [p.strip() for p in tup.split(",")]
if uid_idx is not None and uid_idx < len(parts):
uid_raw = parts[uid_idx].strip().strip("'").strip('"')
if re.fullmatch(r"-?\d+", uid_raw):
user_ids.append(int(uid_raw))
count = len(tuples)
verb = CRUD_VERB["INSERT"]
if user_ids:
groups = {}
for u in user_ids:
groups[u] = groups.get(u, 0) + 1
if len(groups) == 1:
uid = next(iter(groups))
return f"{count} record(s) {verb} into table {table} (column user_id {uid})."
else:
parts = [f"{n} with column user_id {uid}" for uid, n in sorted(groups.items())]
return f"{count} record(s) {verb} into table {table} ({', '.join(parts)})."
else:
return f"{count} record(s) {verb} into table {table}."
def summarize_update(sql: str) -> str:
table = parse_table_name(sql)
set_match = re.search(r"\bSET\b\s+(.+?)(\bWHERE\b|;|$)", sql, flags=re.IGNORECASE | re.DOTALL)
changed_cols = []
if set_match:
assigns = [a.strip() for a in set_match.group(1).split(",")]
for a in assigns:
col = a.split("=")[0].strip().strip("[]")
if col:
changed_cols.append(col)
where = ""
w = re.search(r"\bWHERE\b\s+(.+?)(;|$)", sql, flags=re.IGNORECASE | re.DOTALL)
if w:
where = re.sub(r"\s+", " ", w.group(1)).strip()
verb = CRUD_VERB["UPDATE"]
base = f"Record(s) {verb} in table {table}"
if changed_cols:
cols_txt = ", ".join(changed_cols)
base += f" (changed: column(s) {cols_txt})"
if where:
base += f" with {where}"
return base + "."
def summarize_delete(sql: str) -> str:
table = parse_table_name(sql)
where = ""
w = re.search(r"\bWHERE\b\s+(.+?)(;|$)", sql, flags=re.IGNORECASE | re.DOTALL)
if w:
where = re.sub(r"\s+", " ", w.group(1)).strip()
n_targets = infer_in_list_count(where)
verb = CRUD_VERB["DELETE"]
base = (f"{plural(n_targets, 'record', 'records')} {verb} from {table}"
if n_targets else f"Records {verb} from {table}")
if where:
base += f" that match {where}"
return base + "."
def summarize_select(sql: str) -> str:
table = parse_table_name(sql)
cols = "data"
cm = re.search(r"\bSELECT\b\s+(.+?)\bFROM\b", sql, flags=re.IGNORECASE | re.DOTALL)
if cm:
cols_raw = cm.group(1).strip()
cols = "all columns" if cols_raw == "*" else re.sub(r"\s+", " ", cols_raw)
where = ""
w = re.search(r"\bWHERE\b\s+(.+?)(;|$)", sql, flags=re.IGNORECASE | re.DOTALL)
if w:
where = re.sub(r"\s+", " ", w.group(1)).strip()
verb = CRUD_VERB["SELECT"]
base = f"{cols} will be {verb} from {table}"
if where:
base += f" that match {where}"
return base + "."
def deterministic_summary(sql_text: str) -> str:
cmd = detect_command(sql_text)
if cmd == "INSERT":
return summarize_insert(sql_text)
if cmd == "UPDATE":
return summarize_update(sql_text)
if cmd == "DELETE":
return summarize_delete(sql_text)
if cmd == "SELECT":
return summarize_select(sql_text)
return "Unrecognized SQL command."
# ---------- optional T5 rephrase ----------
_HAS_T5 = False
try:
from transformers import T5Tokenizer, T5ForConditionalGeneration
_HAS_T5 = True
except Exception:
_HAS_T5 = False
_T5_TOKENIZER = None
_T5_MODEL = None
CRUD_PROMPT = {
"INSERT": "Rewrite as a clear statement that new records will be added. Keep numbers the same.",
"UPDATE": "Rewrite as a clear statement that existing records will be updated. Keep names and conditions.",
"DELETE": "Rewrite as a clear statement that records will be deleted. Keep conditions if present.",
"SELECT": "Rewrite as a clear statement that data will be retrieved. Keep table/filters.",
"OTHER": "Rewrite as a short, clear statement for non-technical users.",
}
def load_t5():
global _T5_TOKENIZER, _T5_MODEL
if _T5_TOKENIZER is None or _T5_MODEL is None:
_T5_TOKENIZER = T5Tokenizer.from_pretrained("mrm8488/t5-base-finetuned-wikiSQL-sql-to-en")
_T5_MODEL = T5ForConditionalGeneration.from_pretrained("mrm8488/t5-base-finetuned-wikiSQL-sql-to-en")
def rephrase_with_t5(summary: str, cmd: str) -> str:
if not _HAS_T5:
return summary
load_t5()
instruction = CRUD_PROMPT.get(cmd, CRUD_PROMPT["OTHER"])
input_text = f"explain sql in plain english statement: {instruction} {summary}"
feats = _T5_TOKENIZER([input_text], return_tensors="pt")
out = _T5_MODEL.generate(
input_ids=feats["input_ids"],
attention_mask=feats["attention_mask"],
max_new_tokens=64,
do_sample=False,
)
decoded = _T5_TOKENIZER.decode(out[0], skip_special_tokens=True)
return clean_statement(decoded)
def _bad_rephrase(text: str) -> bool:
if text is None:
return True
t = str(text).strip()
if len(t) >= 2 and ((t[0] == t[-1] == '"') or (t[0] == t[-1] == "'")):
t = t[1:-1].strip()
t_norm = re.sub(r"[\s\.\!\?]+$", "", t).strip().lower()
# added "true" here to force fallback for True./False. answers
if t_norm in ("false", "true", "none", "null", "n/a", "na", ""):
return True
if re.match(r"^(what|which|how)\b", t_norm, re.IGNORECASE):
return True
return len(t_norm) < 3
def explain(sql_text: str):
sql_text = (sql_text or "").strip()
cmd = detect_command(sql_text)
deterministic = deterministic_summary(sql_text)
final = None
if _HAS_T5 and cmd != "OTHER":
try:
final = rephrase_with_t5(deterministic, cmd)
except Exception as e:
print(f"T5 rephrase failed: {e}")
final = None
if _bad_rephrase(final):
final = deterministic
return final
# ---------- UI ----------
EXAMPLES = [
# INSERT
"""INSERT INTO demo_database..user_records (record_id, person_id, created_at)
VALUES (101, 5, GETDATE()), (102, 5, GETDATE()), (103, 5, GETDATE());""",
# UPDATE
"""UPDATE users
SET status = 'active', last_login = GETDATE()
WHERE user_id IN (101, 102, 103);""",
# DELETE
"""DELETE FROM orders
WHERE order_date < '2024-01-01' AND status = 'cancelled';""",
# SELECT
"""SELECT user_id, email, created_at
FROM accounts
WHERE email LIKE '%@example.com' AND created_at >= '2025-01-01';""",
]
with gr.Blocks(theme=gr.themes.Glass()) as demo:
gr.Markdown("## π CRUD-SQL2Text")
sql_in = gr.Textbox(label="Enter SQL Query", lines=8, placeholder="Paste your SQL statement here...")
final_out = gr.Textbox(label="Natural Language Output", lines=3)
btn = gr.Button("Explain SQL")
btn.click(explain, inputs=[sql_in], outputs=[final_out])
gr.Examples(examples=EXAMPLES, inputs=[sql_in], outputs=[final_out], fn=explain, cache_examples=False)
demo.launch(share=True)
|