What Gets Extracted — Tableau Prep

RapidDox parses .tfl and .tflx flow files and extracts every step in execution order — inputs, transformations, outputs — plus connections, parameters, and the flow diagram.

.tfl.tflxInput stepsTransform stepsOutput stepsFlow diagrams

Open a Tableau Prep flow in RapidDox and you can generate documentation straight away — or customise it before export. This page lists every section of a flow that RapidDox documents.

How RapidDox Reads a Prep Flow

RapidDox opens the flow file directly — no Tableau Prep installation, no Tableau Server, and no Tableau Cloud connection required. Packaged flows (.tflx) are unpacked first; the inner flow definition and display settings are then parsed, every step is walked in execution order, and a flow diagram is rendered alongside the per-step details.

Connections

For every connection in the flow, RapidDox extracts:

AttributeDescription
NameThe connection name shown in Prep.
Connection TypeFile or database type (CSV, Excel, SQL Server, Snowflake, etc.).
Is PackagedWhether the source is bundled inside a .tflx archive.
File NameSource file path or database identifier.

Parameters

Each flow parameter is documented with its name, data type, current value, description, whether it is required to run the flow, and the values it accepts. Numeric types are displayed as Number (whole) or Number (decimal); allowed values are listed in full when the parameter is restricted to a fixed set.

Input Steps

Every input step is documented with:

  • Step name, type, and any description captured in Prep
  • The connection it draws from, and whether the source includes a header row
  • The full field list, with each field's data type
  • File filters applied to wildcard / union inputs — regex pattern, value range, top / bottom N rank, and relative-date filters (e.g. "Last 7 days", "Today", "Quarter to date")
  • Database union filters — inclusion or exclusion patterns and the tables they match

Transform Steps

Cleaning Operations

Cleaning steps are broken out into the individual operations they perform. RapidDox documents these operation types:

OperationWhat's documented
Remove column / Remove columnsThe column or columns removed at this step.
Rename columnOriginal name and new name.
Bulk rename columnsThe selection rule, explicitly included or exempted columns, and the rename operation applied.
Add column (calculated)The new column's name and the expression that builds it.
Change column typeThe column, its new data type, and any conversion calculation.
Quick calculation columnThe column, the expression, and the calculation type.
Remap valuesThe column being remapped and the old → new value map.
Range filterThe numeric or date ranges kept.
Filter operationThe filter expression.
Rich null filterThe column, the null-handling rule, and any description.

Joins and Unions

Joins are documented with their source steps, join type (inner, left, right, outer), and join conditions. Unions are documented with their source steps and union type.

Pivots and Unpivots

Pivot steps are documented with the fields being pivoted and the value fields being aggregated. Unpivot steps are documented with the column groups being reshaped and whether Prep's smart-default detection is in use.

Aggregations

Aggregate steps are documented with the group-by fields, the aggregated fields, and the aggregation function applied to each (SUM, AVG, COUNT, MAX, MIN, and so on).

New Rows

New-rows steps are documented with the row values generated and the structure of the rows added to the flow.

Output Steps

RapidDox documents the seven Prep output destinations:

DestinationWhat's documented
Write to HyperOutput table name, output file path, and any TDS file produced alongside it.
Write to DatabaseTarget table, schema, connection, and write-back mode.
Write to ExcelOutput file path and sheet name.
Write to CSVOutput file path.
Write to Data CloudTarget table and connection.
Write to CRM AnalyticsDataset name and connection.
Publish ExtractPublished data source name, whether the sheet is shown, and the include-all flag.

Scripts and Predictions

Script (extensibility) steps are documented with the script function name, the language (Python or R), and the script body. Prediction steps are documented with the model name and type, the fields feeding the model, and the prediction output fields.

Groups (Nested Containers)

Group containers are documented as a single step that lists the nested steps inside and the inputs and outputs that cross the group boundary, preserving the structure the flow author created.

Flow Diagrams

Each output step produces a node-by-node diagram of the path leading to it, rendered into the exported document as an image. The diagram uses the same step layout that is shown in Tableau Prep.

Large Flows — Standard and Optimized Modes

For large or complex flows, RapidDox offers two ways to package the Word output. When the flow is large enough that a single document might fail to render, RapidDox prompts you to choose:

  • Standard Mode — packages everything into a single .docx file.
  • Optimized Mode — splits the documentation into multiple linked .docx files organised by output step, saved into a folder you choose.
Which mode should I pick?
On smaller flows RapidDox doesn't show the prompt — Standard Mode is used. On large flows, Optimized Mode is the safer choice: the prompt itself recommends it, because very large single-file exports can fail to render. You can still choose Standard Mode if you prefer one consolidated file.
If RapidDox encounters an unfamiliar step
Tableau Prep adds new step types over time. If your flow uses a step RapidDox doesn't yet recognise, the app stops processing and asks you to email the file to support@rapiddox.com so support can extend coverage in a future release.

Next Steps