Skip to contents

Dynamic Field Configuration

dynamic_field_config.Rmd

This package provides dynamic field capabilities with flexible types and dependency chain support.

The package supports three main types of dynamic configurations:

  1. Choice Configuration: Populate dropdown and radio button choices
  2. Parameter Configuration: Validate and use data in URL query parameters
  3. Unique Value Configuration: Validate field values against existing database records

All types include automatic saving to the survey database and error handling. The dynamic configuration is specified through the dynamic_config parameter in survey(). It accepts a list of configuration objects.

Choice Configuration

Choice Configuration provides:

  • Dependent fields with dynamic updates based on parent selections
  • Flexible handling of dropdown and radio button options
  • Support for both single and multi-level dependencies

Example

Imagine I want to develop an R package feedback survey that allows users to select one of my packages and a dependent version number populated from a database table:

dynamic_config = list(
  list(
    config_type = "choice",                   # Type of configuration
    table_name = "config_packages",           # Database table to populate choices from
    config_col = "package"                    # Column with the choices
  ),
  list(
    config_type = "choice",                   # Type of configuration 
    table_name = "config_packages_versions",  # Database table for dependent choices
    parent_table_name = "config_packages",    # Parent table for dependency
    parent_id_col = "package_id",             # Links to parent table
    config_col = "version"                    # Column for dependent choices
  )
)

This configuration creates two dependent fields where the second field’s choices update based on the selection in the first field.

Parameter Configuration

Parameter Configuration provides:

  • URL parameter validation against database tables
  • Optional display text mapping with hidden field integration
  • Accept values from URL parameters
  • Store values in hidden fields
  • Look up display text from database tables

Example

Imagine I want to publish the link to my R package feedback form on my github profile, but I also want post it on bluesky and possibly present it at posit::conf(2025). I can use the URL query to track where I’m receiving feedback from.

# URL: http://127.0.0.1:3838/?source=github

dynamic_config = list(
  list(
    config_type = "param",              # Type of configuration
    table_name = "config_source",       # Database table with valid parameters 
    config_col = "source",              # Matches URL parameter name
    display_col = "display_text"        # Optional: Show display text
  )
)

This feature is useful for tracking individuals, groups, or referral sources. It also allows you to pipe data into the survey UI using a hidden text field with the same name.

Unique Value Configuration

Unique Value Configuration provides:

  • Text normalization for comparison
  • Configurable validation responses (warn or stop)
  • Custom warning or error messages
  • Integration with form submission logic

Example

Imagine I want to create a different survey to submit issues relate to my package but to prevent users from submitting the same issue if the title already exists. (Yes, I know this would never happen because we have Github.)

The following configuration ensures that each package issue report has a unique title:

dynamic_config = list(
  list(
    config_type = "unique",           # Type of configuration
    config_col = "issue_title",       # Column to check for uniqueness
    result = "stop",                  # Action on duplicate ("warn" or "stop")
    result_field = "title_error"      # Field for custom error message
  )
)

This configuration will:

  1. Check the entered issue title against existing records in the database
  2. If a duplicate is found, either:
    • Display a warning message (result = “warn”)
    • Prevent form submission and display an error message (result = “stop”)

The validation normalizes text by:

  • Converting to lowercase
  • Trimming whitespace
  • Removing special characters