LeetCode 194: Transpose File

Problem Restatement

We are given a text file named file.txt.

Each line contains several words separated by spaces.

We need to transpose the file.

That means:

• The first column becomes the first row.
• The second column becomes the second row.
• The third column becomes the third row.
• And so on.

Example input:

name age
alice 21
ryan 30

Expected output:

name alice ryan
age 21 30

The original rows become columns.

Input and Output

Item	Meaning
Input	A file named file.txt
Output	The transposed version of the file
Separator	Words are separated by spaces
Language	Bash shell script

Examples

Input:

name age
alice 21
ryan 30

Visual form:

Column 1	Column 2
name	age
alice	21
ryan	30

After transposing:

Row 1	name alice ryan
Row 2	age 21 30

Output:

name alice ryan
age 21 30

Another example:

Input:

a b c
d e f
g h i

Output:

a d g
b e h
c f i

First Thought

We need to rebuild the file column by column.

While reading each row, we can collect values belonging to the same column.

For example:

a b c
d e f
g h i

Column 1 becomes:

a d g

Column 2 becomes:

b e h

Column 3 becomes:

c f i

This suggests storing data indexed by column number.

Key Insight

awk automatically splits each line into fields.

For example:

a b c

Inside awk:

Variable	Value
$1	a
$2	b
$3	c
NF	3

NF means "number of fields".

So while processing each line, we can append every field to a column buffer.

Example:

Column Buffer	Value
col[1]	a d g
col[2]	b e h
col[3]	c f i

After reading the whole file, we print each buffer.

Algorithm

For every line:

1. Loop through all fields from 1 to NF.
2. If the column already has content, append a space.
3. Append the current field to that column.
4. Track the maximum number of columns.

After processing all lines:

1. Print every column buffer from 1 to max_columns.

Correctness

Each line is split into fields automatically by awk.

For every field position i, the algorithm appends the current field $i to the buffer corresponding to column i.

Therefore:

• All first-column values are stored together.
• All second-column values are stored together.
• And so on.

The order of insertion matches the original row order.

So each column buffer becomes exactly one row of the transposed matrix.

Finally, printing the buffers from left to right outputs the correct transpose of the original file.

Complexity

Suppose the file contains r rows and c columns.

Metric	Value	Why
Time	O(r × c)	Every field is visited once
Space	O(r × c)	The transposed content is stored in memory

Implementation

awk '
{
    for (i = 1; i <= NF; i++) {
        if (NR == 1) {
            col[i] = $i
        } else {
            col[i] = col[i] " " $i
        }
    }
}
END {
    for (i = 1; i <= NF; i++) {
        print col[i]
    }
}
' file.txt

Code Explanation

The main loop processes each line:

{
    for (i = 1; i <= NF; i++) {

NF is the number of fields in the current line.

So the loop visits every column.

For the first row:

if (NR == 1)

we initialize the column buffer directly.

NR means "number of records", which is the current line number.

For later rows:

col[i] = col[i] " " $i

we append the new value with a space separator.

At the end:

END {

runs after the entire file has been processed.

Then:

print col[i]

prints each transposed row.

Testing

Create a sample file:

cat > file.txt << 'EOF'
name age
alice 21
ryan 30
EOF

Run the solution:

awk '
{
    for (i = 1; i <= NF; i++) {
        if (NR == 1) {
            col[i] = $i
        } else {
            col[i] = col[i] " " $i
        }
    }
}
END {
    for (i = 1; i <= NF; i++) {
        print col[i]
    }
}
' file.txt

Expected output:

name alice ryan
age 21 30

Another test:

Input:

a b c
d e f
g h i

Expected output:

a d g
b e h
c f i

Single-row test:

Input:

x y z

Expected output:

x
y
z