Gear that distributes files to target projects.
The ADCIDs are mapped to the Flywheel group ID using the custom info found in the NACC admin metadata project.
Ideally we only distribute to centers for which the file is applicable. To figure out which are applicable, we have to associate the distributed file with another one, usually a CSV run through csv-center-splitter, to pull an associated ADCID list.
The associated_csv_regex basically defines a capture regex that runs on the input filename, and searches for another CSV file in the project that matches the capture. If found, the gear will then attempt to read that CSV file for ADCIDs.
For example, say our associated_csv_regex is ^(.*?)-reference\.csv.
If our input file to be distributed is called my-special-file-reference.csv, the regex will capture the string my-special-file and search for a CSV file whose name matches that string exactly.
In other words, it will grab
my-special-file.csv
but not
my-special-file-other.csv
0-my-special-file.csv
If not found, an error will be thrown. Otherwise, the gear will open my-special-file.csv and look for a column with adcid (or field provided with the adcid_key config) and determine the ADCID list from there. It will then only write to projects in this found ADCID list.
Note this configuration is ignored if staging_project_id is provided, as there is only one definitive project we are writing to in that scenario.