Tabular: to collection by row

Tool

Use toolshed.g2.bx.psu.edu/repos/bgruening/split_file_to_collection/split_file_to_collection/0.5.2.

The IWC-attested shape is split_by: col: split a tabular file into a dataset collection, one element per row or key. id_col chooses the column that becomes the element identifier. match_regex and sub_regex clean or extract that identifier.

This is the inverse of tabular-concatenate-collection-to-table, where collapse_dataset turns a collection into one tabular dataset.

When to reach for it

Use this when a single manifest, sample sheet, accession list, or combined results table must become a collection so the next tool can map over each row/key.

Do not use this when the input is already a collection. Do not use this for collection-to-table row-binding; use tabular-concatenate-collection-to-table.

Parameters

tool_id: toolshed.g2.bx.psu.edu/repos/bgruening/split_file_to_collection/split_file_to_collection/0.5.2
tool_state:
  input: { __class__: ConnectedValue }
  split_parms:
    split_by: col
    id_col: "1"
    match_regex: (.*)
    sub_regex: \1

• split_parms.split_by: col: split by a tabular column.
• split_parms.id_col: 1-indexed column whose value becomes the collection element identifier.
• split_parms.match_regex / sub_regex: regex and replacement used to produce the final identifier.

Pitfalls

• id_col is 1-indexed.
• Pick a stable, unique identifier column; duplicate values produce ambiguous collection elements.
• Regex cleanup is downstream metadata cleanup, not cosmetic only.
• Headers matter. Ensure each split element gets the header behavior the downstream tool expects.

See also

• iwc-transformations-survey — Recipe J and candidate boundary.
• tabular-concatenate-collection-to-table — inverse operation using collapse_dataset.
• sync-collections-by-identifier — downstream collection alignment by element identifiers.