include_tables to exclude some specific column in SQLDatabase

[nishantpandey107]

Hi,

I know there is an option of custom table, but having said that, is there an option of excluding any specific column name in table while doing this.

db = SQLDatabase(engine, include_tables=[], sample_rows_in_table_info=5)

column where there is some sensitive info like PPI info[SSN/HICN] or password hash etc.

[dosubot[bot]]

🤖

Hi,

Thank you for your question. Currently, the LangChain framework does not provide a direct way to exclude specific columns when using the include_tables option in SQLDatabase. The include_tables option allows you to specify which tables to include, but it does not provide column-level control.

However, there are a couple of workarounds you could consider:

1. Creating views in your database: You can create views in your database that exclude the sensitive columns, and then include these views instead of the original tables. This would require direct manipulation of your database and is not handled by the LangChain framework itself.

2. Modifying the LangChain source code: You could add a new parameter to the SQLDatabase class that specifies which columns to ignore, and then modify the get_table_info and _get_sample_rows methods to exclude these columns. However, this would require a good understanding of the LangChain source code and the SQLAlchemy library.

Here is the relevant part of the SQLDatabase class for reference:

class SQLDatabase:
    """SQLAlchemy wrapper around a database."""

    def __init__(
        self,
        engine: Engine,
        schema: Optional[str] = None,
        metadata: Optional[MetaData] = None,
        ignore_tables: Optional[List[str]] = None,
        include_tables: Optional[List[str]] = None,
        sample_rows_in_table_info: int = 3,
        indexes_in_table_info: bool = False,
        custom_table_info: Optional[dict] = None,
        view_support: bool = False,
        max_string_length: int = 300,
    ):
        """Create engine from database URI."""
        ...

    def get_usable_table_names(self) -> Iterable[str]:
        """Get names of tables available."""
        ...

    def get_table_info(self, table_names: Optional[List[str]] = None) -> str:
        """Get information about specified tables."""
        ...

    def run(self, command: str, fetch: str = "all") -> str:
        """Execute a SQL command and return a string representing the results."""
        ...

You can find the full source code for the SQLDatabase class in the LangChain repository here.

I hope this helps! Let me know if you have any other questions.

Sources

• libs/langchain/langchain/utilities/sql_database.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.