Skip to contents

Source Pre-Pipeline Input Builder Scripts

Source: R/execute.R
sourceInputBuilderScripts.Rd

Auto-discovers and sources builder scripts from pre-pipeline directories in a MANDATORY dependency order. Designed to be called from main.R before the production pipeline. This ensures concept sets are available before cohorts, and dependent cohorts can reference already-loaded cohort definitions.

Usage

sourceInputBuilderScripts(
  projectPath = here::here(),
  verbose = TRUE,
  warnMissing = TRUE
)

Arguments

projectPath

Character. Path to the project root. Defaults to current project.

verbose

Logical. If TRUE (default), displays which scripts are being sourced.

warnMissing

Logical. If TRUE (default), warns when directories don't exist.

Value

Invisibly returns a list with:

  • sourced_files: Character vector of sourced files (absolute paths)

  • directories_checked: Character vector of directories checked

  • error_summary: List of any errors encountered (always empty on successful return; a non-empty list raises an error instead)

Details

Scripts are sourced in the following FIXED order (skipping any that don't exist):

  1. inputs/conceptSets/R/import_atlas_concept_set.R

  2. inputs/conceptSets/R/import_capr_concept_set.R

  3. inputs/cohorts/R/import_atlas_cohort.R

  4. inputs/cohorts/R/import_capr_cohort.R

  5. inputs/cohorts/R/import_sql_cohort.R

  6. inputs/cohorts/R/build_dependent_cohorts.R

This order is enforced to guarantee dependencies are satisfied:

  • Concept sets load first (may be needed by cohort definitions)

  • Base cohorts load next (may be needed by dependent cohorts)

  • Dependent cohorts load last (can reference base cohorts)

Use makeInputBuilderScript to create scripts with the correct naming convention. Missing scripts are silently skipped, allowing flexible configurations.

Errors inside builder scripts are collected across all scripts (so they can be fixed in one pass) and then raised as a single error: input building is a hard precondition of pipeline execution, so a failed script stops the run before execStudyPipeline() can be reached.