从制表符分隔文件移除指定字段/列的技术实现需求

Got it, let's tackle these two common tab-separated value (TSV) editing tasks using awk—it's the go-to tool for this kind of field-level wrangling, and it keeps things clean and efficient.

Case 1: Remove All Fields Containing a Specific String #

Goal #

Take a TSV file, and for every line, remove any field that includes your target string (e.g., "four"). Output the remaining fields as a valid single-tab-separated TSV.

Solution Command #

awk -F'\t' -v OFS='\t' -v target="four" '
{
    out = ""
    for (i = 1; i <= NF; i++) {
        # Skip fields that contain the target string (use $i != target for exact matches)
        if ($i !~ target) {
            out = out OFS $i
        }
    }
    # Remove leading tab added by the first OFS
    sub(/^\t/, "", out)
    print out
}' your_input.tsv

Example #

Input line:

one	two	three	four	five	six	four

Output line:

one	two	three	five	six

Breakdown #

• -F'\t' tells awk to split input lines by tabs.
• -v OFS='\t' sets the output field separator to a single tab, so we don't get messy spacing.
• -v target="four" passes your target string into the script (easy to change for different values).
• The loop checks each field: if it doesn't contain the target, we add it to our output string.
• sub(/^\t/, "", out) cleans up the leading tab that gets added when we start building the output.

Case 2: Remove Specific Columns (Avoid Multiple Tabs) #

Goal #

Remove columns by their numeric position (e.g., column 4 and 7) from a TSV, ensuring the output stays properly formatted with only single tabs between fields (no empty tab gaps from deleted columns).

Solution Command #

awk -F'\t' -v OFS='\t' -v cols_to_remove="4,7" '
{
    split(cols_to_remove, del_cols, ",")
    # Create a lookup for columns to delete
    delete del_map
    for (i in del_cols) {
        del_map[del_cols[i]] = 1
    }
    out = ""
    for (i = 1; i <= NF; i++) {
        # Keep columns not in the delete list
        if (!(i in del_map)) {
            out = out OFS $i
        }
    }
    sub(/^\t/, "", out)
    print out
}' your_input.tsv

Example #

Input line:

one	two	three	four	five	six	four

Output line (removing columns 4 and 7):

one	two	three	five	six

Breakdown #

• split(cols_to_remove, del_cols, ",") turns your comma-separated column list into an array.
• del_map acts as a quick lookup to check if a column should be deleted.
• The loop builds the output by only including columns not in del_map, so we never create empty tab gaps in the first place.
• Again, sub(/^\t/, "", out) cleans up the leading tab from the output string.

Quick One-Liner for Single Column Removal #

If you only need to delete one column (e.g., column 4), you can simplify things:

awk -F'\t' -v OFS='\t' '{for(i=4;i<NF;i++) $i=$(i+1); NF--; print}' your_input.tsv

This shifts all columns after the deleted one left by one, then reduces the field count (NF--) to drop the now-duplicate last field—no extra tabs to clean up!

内容的提问来源于stack exchange，提问作者Chris