Tabular: prepend header

Connect a text or tabular dataset to tp_awk_tool when a downstream step needs a fixed first line. An awk BEGIN action prints that line before any input row. Choose the row action according to the input: print every row when it has no header, or use NR > 1 when the new line replaces an existing first-row header. The tool’s infile input receives the dataset, code holds the awk program, and the cited workflows leave variables empty.

Add a header to headerless rows

The VGP purge-duplicates workflow uses a one-line program to label metric rows. A gxformat2 step’s relevant state is:

tool_id: toolshed.g2.bx.psu.edu/repos/bgruening/text_processing/tp_awk_tool/9.5+galaxy3
tool_state:
  code: 'BEGIN {print "Metric\tAlternate"} {print}'
  infile: { __class__: ConnectedValue }
  variables: []

With input N50<TAB>4200, the output is Metric<TAB>Alternate followed by the unchanged input row. In awk, \t in the double-quoted string becomes a tab. {print} prints the original line without rebuilding its fields, so its existing separators remain intact. Connect the step’s infile port to the upstream dataset and declare the awk output in the complete workflow.

Replace a header and normalize rows

The MAGs generation workflow replaces the incoming header and adds .fasta to genome names that lack that suffix:

tool_id: toolshed.g2.bx.psu.edu/repos/bgruening/text_processing/tp_awk_tool/9.5+galaxy2
tool_state:
  code: |-
    BEGIN {OFS="\t"; print "genome\tcompleteness\tcontamination"}
    NR > 1 {
        if ($1 !~ /\.fasta$/)
            $1 = $1 ".fasta"
        print $1, $2, $3
    }
  infile: { __class__: ConnectedValue }
  variables: []

For an input headed bin<TAB>complete<TAB>contam with a data row bin_1<TAB>95<TAB>2, the output is genome<TAB>completeness<TAB>contamination and bin_1.fasta<TAB>95<TAB>2. NR > 1 suppresses the old first line. OFS="\t" makes the rebuilt data rows tab-separated when print receives three fields. The awk wrapper passes FS and OFS as tabs. Check that the input is tab-separated, especially if a field contains spaces or an empty column.

Check the input and result

• If the input is already headed, plain {print} leaves that old header as the first data row. Use NR > 1 only when the first physical line should be discarded. On a headerless input, it would discard data.
• The new header is emitted even when the input has no rows. Decide whether that header-only result is useful to downstream tools.
• print $1, $2, $3 writes only those three parsed fields. Use {print} when columns and original separators must pass through unchanged.
• Verify the output’s first two lines, tab separators, field count, and datatype expected by the next tool. The wrapper copies the input datatype and metadata to the output, so set or convert the datatype if the next tool expects a different one.

The IWC survey records pins 9.3+galaxy1, 9.5+galaxy0, 9.5+galaxy2, and 9.5+galaxy3 for tp_awk_tool. These snippets reproduce observed program shapes, not a promise that every pin is interchangeable. Keep the version and Tool Shed repository metadata consistent with the workflow you are editing, then validate the complete workflow. The examples above are step excerpts, not standalone gxformat2 workflows.

For header removal without a replacement, use a first-line removal operation. For stacking a collection while retaining only one header, use tabular-concatenate-collection-to-table. A separate constant-text dataset followed by tp_cat can also prefix text when that dataset is already part of the workflow design, but the cited header-injection steps use awk.

Evidence and related patterns

The iwc-tabular-operations-survey records the MAGs awk program and identifies the VGP header-injection steps. See tabular-compute-new-column when rows need a computed field, and tabular-filter-by-regex when the operation is line filtering. The checked-in test under verification/workflows/tabular-prepend-header/ compares both complete outputs against expected tab-separated files.