code/utils/db_utils.py

import os
import sqlite3

from pyserini.search.lucene import LuceneSearcher
import json
from func_timeout import func_set_timeout, FunctionTimedOut
import time
import multiprocessing
from multiprocessing.pool import ThreadPool
import requests


# get the database cursor for a sqlite database path
def get_cursor_from_path(sqlite_path):
    try:
        if not os.path.exists(sqlite_path):
            print("Openning a new connection %s" % sqlite_path)
        connection = sqlite3.connect(sqlite_path, check_same_thread = False)
    except Exception as e:
        print(sqlite_path)
        raise e
    connection.text_factory = lambda b: b.decode(errors="ignore")
    cursor = connection.cursor()
    return cursor

# execute predicted sql with a time limitation
@func_set_timeout(30)
def execute_sql(cursor, sql):
    cursor.execute(sql)

    return cursor.fetchall()

# execute predicted sql with a long time limitation (for buiding content index)
@func_set_timeout(2000)
def execute_sql_long_time_limitation(cursor, sql):
    cursor.execute(sql)

    return cursor.fetchall()

def check_sql_executability(generated_sql, db):
    if not os.path.exists(db):
        raise Exception("Database file not found: %s" % db)
        
    connection = sqlite3.connect(db, check_same_thread = False)
    connection.text_factory = lambda b: b.decode(errors="ignore")
    cursor = connection.cursor()

    if generated_sql.strip() == "":
        return "Error: empty string"
    try:
        execute_sql(cursor, "EXPLAIN QUERY PLAN " + generated_sql)
        execution_error = None
    except FunctionTimedOut as fto:
        print("SQL execution time out error: {}.".format(fto))
        execution_error = "SQL execution times out."
    except Exception as e:
        # print("SQL execution runtime error: {}.".format(e))
        execution_error = str(e)

    cursor.close()
    connection.close()
    
    return execution_error

def is_number(s):
    try:
        float(s)
        return True
    except ValueError:
        return False

def detect_special_char(name):
    for special_char in ['(', '-', ')', ' ', '/']:
        if special_char in name:
            return True

    return False

def add_quotation_mark(s):
    return "`" + s + "`"

def get_column_contents(column_name, table_name, cursor):
    select_column_sql = "SELECT DISTINCT `{}` FROM `{}` WHERE `{}` IS NOT NULL LIMIT 2;".format(column_name, table_name, column_name)
    results = execute_sql_long_time_limitation(cursor, select_column_sql)
    column_contents = [str(result[0]).strip() for result in results]
    # remove empty and extremely-long contents
    column_contents = [content for content in column_contents if len(content) != 0 and len(content) <= 25]

    return column_contents

def get_db_schema_sequence(schema):
    """Build a CHESS-style DDL schema string with inline -- comments.

    Each column line follows the format:
        col_name  TYPE,  -- Example Values: `v1`, `v2` | Column Description: ... | Value Description: ...

    Falls back gracefully when description fields are absent.
    """
    schema_sequence = "database schema:\n"
    for table in schema["schema_items"]:
        table_name_raw = table["table_name"]
        table_name = add_quotation_mark(table_name_raw) if detect_special_char(table_name_raw) else table_name_raw

        column_defs = []
        cols = zip(
            table["column_names"],
            table["column_types"],
            table["column_comments"],
            table["column_contents"],
            table["pk_indicators"],
            table.get("column_descriptions", [""] * len(table["column_names"])),
            table.get("value_descriptions", [""] * len(table["column_names"])),
        )
        for col_name, col_type, col_comment, col_content, pk_indicator, col_desc, val_desc in cols:
            display_name = add_quotation_mark(col_name) if detect_special_char(col_name) else col_name

            type_str = col_type.upper() if col_type else "TEXT"
            suffix = "," if True else ""  # always add comma for DDL style

            comment_parts = []
            if col_content:
                examples = ", ".join(f"`{v}`" for v in col_content[:3])
                comment_parts.append(f"Example Values: {examples}")
            if col_desc:
                comment_parts.append(f"Column Description: {col_desc}")
            elif col_comment:
                comment_parts.append(f"Column Description: {col_comment}")
            if val_desc:
                comment_parts.append(f"Value Description: {val_desc}")
            if pk_indicator != 0:
                comment_parts.append("Primary Key")

            if comment_parts:
                column_defs.append(
                    f"    {display_name} {type_str},  -- {' | '.join(comment_parts)}"
                )
            else:
                column_defs.append(f"    {display_name} {type_str},")

        col_block = "\n".join(column_defs)
        schema_sequence += f"CREATE TABLE {table_name}\n(\n{col_block}\n);\n"

    if len(schema["foreign_keys"]) != 0:
        schema_sequence += "-- Foreign keys:\n"
        for foreign_key in schema["foreign_keys"]:
            fk = [add_quotation_mark(p) if detect_special_char(p) else p for p in foreign_key]
            schema_sequence += f"-- {fk[0]}.{fk[1]} = {fk[2]}.{fk[3]}\n"

    return schema_sequence.strip()

def retrieve_most_similar_column_content(question, db_id, table_name, column_name):
    # requests to retrieval api to get most similar column content
    pass


def get_db_schema_sequence_with_matched_examples(schema, question):
    schema_sequence = "database schema:\n"
    for table in schema["schema_items"]:
        table_name, table_comment = table["table_name"], table["table_comment"]
        if detect_special_char(table_name):
            table_name = add_quotation_mark(table_name)
        
        # if table_comment != "":
        #     table_name += " ( comment : " + table_comment + " )"

        column_info_list = []
        for column_name, column_type, column_comment, pk_indicator in \
            zip(table["column_names"], table["column_types"], table["column_comments"], table["pk_indicators"]):
            if detect_special_char(column_name):
                column_name = add_quotation_mark(column_name)
            additional_column_info = []
            # column type
            
            # pk indicator
            if pk_indicator != 0:
                additional_column_info.append("primary key")

            additional_column_info.append(f"type: {column_type}")
            # column comment
            if column_comment != "":
                additional_column_info.append("meaning: " + column_comment)
            # representive column values
            if len(column_content) != 0:
                additional_column_info.append("values: " + " , ".join(column_content))
            
            column_info_list.append(column_name + " | " + " ; ".join(additional_column_info))
        
        schema_sequence += "table "+ table_name + " , columns = [\n  " + "\n  ".join(column_info_list) + "\n]\n"

    if len(schema["foreign_keys"]) != 0:
        schema_sequence += "foreign keys:\n"
        for foreign_key in schema["foreign_keys"]:
            for i in range(len(foreign_key)):
                if detect_special_char(foreign_key[i]):
                    foreign_key[i] = add_quotation_mark(foreign_key[i])
            schema_sequence += "{}.{} = {}.{}\n".format(foreign_key[0], foreign_key[1], foreign_key[2], foreign_key[3])
    else:
        schema_sequence += "foreign keys: None\n"
    return schema_sequence.strip()

def get_matched_content_sequence(matched_contents):
    content_sequence = ""
    if len(matched_contents) != 0:
        content_sequence += "matched contents:\n"
        for tc_name, contents in matched_contents.items():
            table_name = tc_name.split(".")[0]
            column_name = tc_name.split(".")[1]
            if detect_special_char(table_name):
                table_name = add_quotation_mark(table_name)
            if detect_special_char(column_name):
                column_name = add_quotation_mark(column_name)
            
            content_sequence += table_name + "." + column_name + " ( " + " , ".join(contents) + " )\n"
    else:
        content_sequence = "matched contents: None"
    return content_sequence.strip()

def get_most_similar_column_contents(args):
    base_url, source, question, db_id, table_name, column_name = args
    # base_url = "http://localhost:8005"
    url = f"{base_url}/search_column_content"
    payload = {
        "source": source,
        "db_id": db_id,
        "table": table_name,
        "column": column_name,
        "query": question,
        "k": 2
    }

    response = requests.post(url, json=payload)
    if response.status_code == 200:
        return response.json()["results"]
    else:
        print("No results: ", source, db_id, table_name, column_name)
        return []

def get_db_schema(api_url, source, question, db_path, db_comments, db_id,
                  db_descriptions=None):
    """Build the schema dict for a database.

    Args:
        api_url: URL of the BM25 column-content retrieval service.
        source: Dataset identifier ('bird', 'spider', …).
        question: Natural-language question (used for BM25 retrieval).
        db_path: Path to the SQLite file.
        db_comments: Legacy tables.json comment dict {db_id: {table: …}}.
        db_id: Database identifier string.
        db_descriptions: Optional BIRD CSV descriptions loaded via
            bird_csv_utils.load_db_descriptions().  When provided, each
            schema item gets ``column_descriptions`` and
            ``value_descriptions`` lists for CHESS-style DDL rendering.
    """
    if db_id in db_comments:
        db_comment = db_comments[db_id]
    else:
        db_comment = None

    cursor = get_cursor_from_path(db_path)
    
    results = execute_sql(cursor, "SELECT name FROM sqlite_master WHERE type='table';")
    table_names = [result[0].lower() for result in results]

    schema = dict()
    schema["schema_items"] = []
    foreign_keys = []

    for table_name in table_names:
        if table_name == "sqlite_sequence":
            continue

        results = execute_sql(cursor, "SELECT name, type, pk FROM PRAGMA_TABLE_INFO('{}')".format(table_name))
        column_names_in_one_table = [result[0].lower() for result in results]
        column_types_in_one_table = [result[1].lower() for result in results]
        pk_indicators_in_one_table = [result[2] for result in results]

        with ThreadPool(processes=16) as pool:
            column_contents = pool.map(
                get_most_similar_column_contents,
                [(api_url, source, question, db_id, table_name, col) for col in column_names_in_one_table],
            )

        results = execute_sql(cursor, "SELECT * FROM pragma_foreign_key_list('{}');".format(table_name))
        for result in results:
            if None not in [result[3], result[2], result[4]]:
                foreign_keys.append([table_name.lower(), result[3].lower(), result[2].lower(), result[4].lower()])

        if db_comment is not None:
            if table_name in db_comment:
                table_comment = db_comment[table_name]["table_comment"]
                column_comments = [
                    db_comment[table_name]["column_comments"].get(col, "")
                    for col in column_names_in_one_table
                ]
            else:
                table_comment = ""
                column_comments = ["" for _ in column_names_in_one_table]
        else:
            table_comment = ""
            column_comments = ["" for _ in column_names_in_one_table]

        has_none_indicators = []
        for col in column_names_in_one_table:
            cursor.execute(f"SELECT COUNT(*) FROM `{table_name}` WHERE `{col}` IS NULL")
            count = cursor.fetchone()[0]
            has_none_indicators.append(1 if count > 0 else 0)

        # Enrich with BIRD CSV descriptions when available
        table_csv = {}
        if db_descriptions:
            table_csv = db_descriptions.get(table_name, {})

        column_descriptions = []
        value_descriptions = []
        for col in column_names_in_one_table:
            col_info = table_csv.get(col, {})
            column_descriptions.append(col_info.get("column_description", ""))
            value_descriptions.append(col_info.get("value_description", ""))

        schema["schema_items"].append({
            "table_name": table_name,
            "table_comment": table_comment,
            "column_names": column_names_in_one_table,
            "column_types": column_types_in_one_table,
            "column_comments": column_comments,
            "column_contents": column_contents,
            "pk_indicators": pk_indicators_in_one_table,
            "has_none_indicators": has_none_indicators,
            "column_descriptions": column_descriptions,
            "value_descriptions": value_descriptions,
        })

    schema["foreign_keys"] = foreign_keys
    return schema