基于列号在输入文件行追加值的Shell脚本问题排查与修正

Let's break down what's going wrong with your current code, then build a robust, general solution that meets your requirements.

What's Wrong with Your Current Sed Command? #

Your existing sed logic appends the value after the Nth character, rather than overwriting the preceding characters starting from your target column. Here's why:

sed -i 's/\(\.\{'$1'\}\)/\1'$2'/' "input.txt"

This regex matches exactly $1 arbitrary characters, then keeps those characters (\1) and adds your value ($2) right after them. For your example with column_num=1 and append_val=45, it matches the first character (1), then turns it into 145—hence the output 145200 ... instead of the desired 4500 ....

General Solution with Awk (Recommended) #

Awk is more intuitive for character-position manipulation and handles edge cases better than sed. This function will:

• Take a 0-based target column, your value, and the input file
• Calculate the range of characters to overwrite (starting from target_col - value_length + 1 up to target_col)
• Safely split the string, insert your value, and write back to the file

function insert_val_at_col {
    local target_col="$1"  # 0-based column number where the value ends
    local val="$2"
    local file="$3"
    local val_len="${#val}"
    local start_col=$(( target_col - val_len + 1 ))

    # Validate we don't try to overwrite characters before the start of the string
    if (( start_col < 0 )); then
        echo "Error: Target column $target_col is too small for a value of length $val_len. Minimum required target column is $((val_len - 1))" >&2
        return 1
    fi

    # Convert 0-based positions to awk's 1-based string indexing
    local prefix_length="$start_col"
    local suffix_start=$(( target_col + 2 ))

    # Use awk to split, modify, and rebuild each line
    awk -v pre_len="$prefix_length" -v suf_start="$suffix_start" -v new_val="$val" '{
        prefix = substr($0, 1, pre_len)
        suffix = substr($0, suf_start)
        print prefix new_val suffix
    }' "$file" > temp_modified.txt && mv temp_modified.txt "$file"
}

How to Use It #

# Your example parameters
column_num=1
append_val=45
insert_val_at_col "$column_num" "$append_val" Input.txt

Before:

1200 5600 775000 34555

After:

4500 5600 775000 34555

Key Details #

• 0-based Columns: This matches your original requirement (column 1 refers to the second character in the string)
• Automatic Length Calculation: The function uses ${#val} to get the length of your input value, so it works for any length of value
• Safety Check: It prevents invalid operations where the target column is too small to fit your value
• Handles Spaces: Since we're working with raw character positions, spaces are treated as valid columns (as you noted you won't insert values in these positions anyway)

Alternative Solution with Sed #

If you prefer using sed, you can construct a regex that replaces the exact range of characters. Note: This requires extra care if your value contains special regex characters (like /, &, or *).

function insert_val_at_col_sed {
    local target_col="$1"
    local val="$2"
    local file="$3"
    local val_len="${#val}"
    local start_col=$(( target_col - val_len + 1 ))

    if (( start_col < 0 )); then
        echo "Error: Target column $target_col is too small for a value of length $val_len" >&2
        return 1
    fi

    # Escape special characters in the value for sed (handles /, &, etc.)
    local escaped_val=$(printf '%s' "$val" | sed 's/[\/&]/\\&/g')

    # Replace the range from start_col to target_col with the escaped value
    sed -i -E "s/^(.{$start_col}).{$val_len}/\1$escaped_val/" "$file"
}

This works similarly to the awk version, but we first escape special characters in the value to avoid breaking the sed regex.

内容的提问来源于stack exchange，提问作者Anna