Widen the URL query parameters into columns

The url_query column of a metadata tibble holds, per study result, the query parameters the worker arrived with, which for a Prolific study link are PROLIFIC_PID, STUDY_ID and SESSION_ID. This turns them into one character column per parameter, named in snake case with prefix in front (query_prolific_pid, query_study_id, query_session_id), so that the tibble joins to a Prolific export without unnesting. Study results without a parameter get NA; a tibble without any parameter is returned unchanged. Two parameters that differ only in case or separators (PROLIFIC_PID and prolific_pid) would share a column name; the second gets a numeric suffix and a warning names both keys.

Usage

jatos_url_query(metadata, prefix = "query_")

Arguments

metadata

A metadata tibble with a url_query list column.

prefix

Prefix of the new column names. The default keeps STUDY_ID from colliding with the study_id column.

Value

metadata with one column per parameter appended; url_query stays.

Examples

meta <- jatos_flatten_metadata(
  system.file("extdata", "metadata.json", package = "jatosr")
)
jatos_url_query(meta)[, c("study_result_id", "query_prolific_pid", "query_session_id")]
#> # A tibble: 6 × 3
#>   study_result_id query_prolific_pid query_session_id
#>             <int> <chr>              <chr>           
#> 1            9001 NA                 NA              
#> 2            9002 NA                 NA              
#> 3            9002 NA                 NA              
#> 4            9003 NA                 NA              
#> 5            9004 p-0004             s-0004          
#> 6            9004 p-0004             s-0004