cwprep

Generate a Tableau Prep data flow (.tfl/.tflx) file from a declarative flow definition.

Args: flow_name: Display name for the flow in Tableau Prep. connection: Connection settings. Must include a "type" key: For type="database": Required keys: host. Conditionally required: username (required for mysql/postgres, optional for sqlserver with sspi authentication), dbname (required for mysql/postgres, often empty at connection-level for sqlserver). Optional keys: port (default depends on db_class), db_class ("mysql"|"sqlserver"|"postgres"|"adb_mysql", default "mysql"), authentication ("sspi"|"sqlserver"|"", default ""), schema (e.g. "dbo" for SQL Server input_table nodes). For type="file": No additional keys needed at this level. File connections are created automatically from input_excel/input_csv nodes. nodes: Ordered list of node definitions. Each node is a dict with a "type" key and a "name" key (used to reference this node from downstream nodes). Additional keys depend on the node type:

• input_sql: sql (str)
• input_table: table (str)
• input_excel: filename, sheet (str), fields? (list)
• input_csv: filename, fields? (list), separator?, locale?, charset?, contains_headers?
• input_csv_union: file_names (list of str), fields? (list), separator?, locale?, charset?, contains_headers?
• join: left, right (node names), left_col, right_col, join_type?
• union: parents (list of node names)
• filter: parent (node name), expression (str)
• value_filter: parent, field, values (list), exclude? (bool)
• calculation: parent, column_name, formula
• aggregate: parent, group_by (list), aggregations? (list of {field, function, output_name?})
• keep_only: parent, columns (list)
• remove_columns: parent, columns (list)
• rename: parent, renames ({old: new})
• pivot: parent, pivot_column, aggregate_column, new_columns, group_by?, aggregation?
• unpivot: parent, columns_to_unpivot, name_column?, value_column?
• quick_calc: parent, column_name, calc_type (lowercase|uppercase|titlecase|trim_spaces|remove_extra_spaces|remove_all_spaces|remove_letters|remove_punctuation)
• change_type: parent, fields ({column: target_type})
• duplicate_column: parent, source_column, new_column_name?
• output_server: parent, datasource_name, project_name?, server_url? output_path: File path for the generated flow file. Use .tfl extension for standard flows. Use .tflx extension for packaged flows with embedded data files. data_files: Optional dict mapping filenames to list of absolute source file paths. Required when output_path ends with .tflx. Example: {"orders.xlsx": ["C:/data/orders.xlsx"]}

Returns: A summary string with the output file path and node count.

List all supported node types and their required/optional parameters.

Returns: A JSON string describing every supported operation type.

Translate a Tableau Prep flow to equivalent ANSI SQL (CTE format).

Use this to preview the logical equivalence of a data flow as SQL, for review and verification purposes. Supports two input modes:

Mode 1 — From declarative definition (same format as generate_tfl): Provide flow_name, connection, and nodes.

Mode 2 — From existing .tfl file: Provide tfl_path pointing to a .tfl file on disk.

If both tfl_path and (connection + nodes) are provided, tfl_path takes priority.

Args: flow_name: Display name for the flow (used in Mode 1 and header comments). connection: Connection settings (same format as generate_tfl). Required for Mode 1. nodes: Ordered list of node definitions (same format as generate_tfl). Required for Mode 1. tfl_path: Path to an existing .tfl file. Required for Mode 2.

Returns: ANSI SQL string with CTEs representing the flow logic, including a flow summary header and step-by-step comments.

Validate a flow definition without generating a file.

Checks for:

• Required connection fields
• Valid node types
• Node reference integrity (parent/left/right references exist)
• Required fields for each node type

Args: flow_name: Flow display name. connection: Database connection settings (same format as generate_tfl). nodes: Ordered list of node definitions (same format as generate_tfl).

Returns: A JSON string with "valid" (bool) and "errors" (list of error strings).