# Script-Generated Question

## **Overview**

{% hint style="info" %}
This feature only available in **row labeling** project and disabled by default. Please reach out to [**support@datasaur.ai**](mailto:support@datasaur.ai) if your team needs this feature, and we'll assist you!
{% endhint %}

**Script-generated questions** is an advanced question type that allows dynamic question generation based on each row’s data. Using [TypeScript](https://www.typescriptlang.org/), users can configure logic that determines which questions should appear for a given row. This flexibility makes it ideal for cases where static question sets are insufficient.

### When to Use Script-Generated Questions

* You need **dynamic questions** that change based on row data.
* You want to **automate question creation** instead of manually defining them.
* You require **custom logic** to determine which questions appear.
* You need **dropdown options that are generated based on the data in each row.**

## Configuration

### **Defining Question Identifiers**

An **identifier** is a unique representation of a question. It serves as a reference to the question type, label, and configuration, allowing the script to access and manipulate the question during execution. Before generating questions dynamically, all possible question identifiers **must be defined during project creation**.

This is configured in the <mark style="color:red;">**`identifiers`**</mark> section of the script.

![Identifiers](/files/wrZcd4ObKw0lr8FZ5ZFR)

Each identifier should have the following attributes:

* <mark style="color:red;">**`id`**</mark> — A unique, zero-indexed integer, signifying the relative order of the questions.
* <mark style="color:red;">**`label`**</mark> — A human-readable string that represents the question.
* <mark style="color:red;">**`type`**</mark> — The question type (i.e. text, multiple choice, date).
  * Currently, all question types are supported **except** **Grouped Attributes** and **Script-Generated Questions**, with some caveats detailed [here](#limitation).
  * See [this](/data-studio-projects/lets-get-labeling/label-sets.md#question-types) for more details on each question type.
* <mark style="color:red;">**`config.multiple`**</mark> — Configuration that allows a question to have multiple answers.

<details>

<summary>Example Configuration</summary>

Here’s an example of an identifier for a **text** question and a **multiple-choice** question.

```tsx
identifiers: Record<string, QuestionIdentifier> = {
  ["Text Question"]: {
    id: 0, 
    label: "Text Question", 
    config: { multiple: false }, 
    type: QuestionType.TEXT
  },
  ["Multiple Choice Question"]: {
    id: 1, 
    label: "Multiple Choice Question", 
    config: { multiple: true }, 
    type: QuestionType.MULTIPLE_CHOICE
  },
};
```

</details>

### Dynamically Generating Questions

The core function that determines which questions appear is <mark style="color:red;">`getQuestionsForRow`</mark>. This function runs for each row in the dataset, evaluating its data and returning the appropriate questions.

{% hint style="info" %}
**Tip:** You can view available <mark style="color:red;">`row`</mark> and <mark style="color:red;">`columns`</mark> attributes in the editor by pressing **Ctrl / Cmd + Space**.
{% endhint %}

<details>

<summary>Example Usage</summary>

In the following example:

* A **text question** appears if the column <mark style="color:red;">`DATA`</mark> contains the word <mark style="color:red;">`"text"`</mark>.
* Otherwise, a **multiple-choice question** is shown, with options generated from the values in <mark style="color:red;">`DATA`</mark>.

```tsx
getQuestionsForRow({ row, columns }: RowContext): Question[] {
  if (getCellValueByColumnLabel("DATA").includes("text")) {
    return [{ ...this.identifiers["Text Question"], required: false }];
  }

  const options = getCellValueByColumnLabel("DATA").split(",");
  return [
    {
      ...this.identifiers["Multiple Choice Question"],
      required: false,
      config: {
        ...this.identifiers["Multiple Choice Question"],
        options: options.map((o) => ({ id: o, label: o })),
      },
    },
  ];
```

</details>

Here’s how it will look once the project is created.

![](/files/vT0DzTicrosKNaVBufzn)

![](/files/lA8oWdrS3dN90nvgIKo0)

You can find a sample CSV, as well as the full script to recreate the project below.

{% file src="/files/242pkAHNwlNkpZlnUhKZ" %}

{% file src="/files/29qV2UhkOvkjUWWDhxkN" %}

## Limitation

### Dropdown and Hierarchical Dropdown

We currently **don’t support** configuring dropdown or hierarchical dropdown question types in script-generated questions.

We recommend using **multiple choice** or **single choice** question types, as these also allow users to select from a set of options.

### Question Set Management

1. Any script-generated questions **will not be saved** to the library when the question set is saved.
2. When configuring the question set in **Label set management**, these questions **will not appear** in the question type dropdown.

### Question Logic

Since question logic relies on a predefined order, it’s not currently possible to configure logic for questions that reference a script-generated one. This is because script-generated questions are created dynamically during the project, which makes it difficult to set conditions for subsequent questions based on them.

What this means is that questions following a script-generated one can still have logic, but they can only reference other question types, not the script-generated question itself.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.datasaur.ai/advanced/script-generated-question.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.