feat: Add Tool mode Support to AstraCQL component

[edwinjosechittilappilly]

This pull request includes significant changes to the AstraDBCQLToolComponent class in the src/backend/base/langflow/components/tools/astradb_cql.py file. The changes involve refactoring the component class, modifying input types, and adding new methods for output handling.

Refactoring and input modifications:

• Replaced StrInput with MessageTextInput for several input fields, such as keyspace, table_name, api_endpoint, and projection_fields. [1] [2]
• Added the tool_mode attribute to DictInput fields partition_keys and clustering_keys.

Output handling and method additions:

• Introduced outputs attribute with Output definitions for api_run_model and api_as_dataframe.
• Added as_dataframe method to convert results into a DataFrame.

Code removal:

• Removed methods create_args_schema and build_tool, which were responsible for schema creation and tool building.

[Copilot]

Pull Request Overview

This PR refactors the AstraDBCQLToolComponent to support tool mode, enhancing input and output handling for Astra DB CQL operations. Key changes include replacing StrInput with MessageTextInput, adding the tool_mode flag to key inputs, and introducing an as_dataframe method for converting REST output into a DataFrame while removing deprecated schema creation and tool building methods.

[Copilot]

The as_dataframe method calls self.astra_rest(args) twice, which appears to be a duplicate call. Consider removing one of these calls to avoid unnecessary execution.

Suggested change

results = self.astra_rest(args)

[edwinjosechittilappilly]

@carlosrcoelho we might need a code owner to test this tool